The Case of Light Verb Constructions in Persian - Relais d

Explorations of Semantic Space:

The Case of Light Verb Constructions in Persian

Neiloufar Family

A thesis submitted to Ecole des Hautes Etudes en Sciences Sociales

in fulfillment of the requirements for the degree of Doctor of Philosophy in Cognitive Science

Paris, France April 2006

© Copyright Neiloufar Family 2006

i

ACKNOWLEDGEMENTS .......................................................................................................................... III ABBREVIATIONS .........................................................................................................................................V 0 INTRODUCTION .........................................................................................................................................1

0.1 LANGUAGE AND THE STUDY OF MIND.......................................................................................................1 0.2 ELEMENTS UNDER STUDY .........................................................................................................................2 0.3 THE INTEREST OF STUDYING PERSIAN.......................................................................................................3 0.4 OUTLINE OF THIS STUDY .........................................................................................................................4

1 LIGHT VERB CONSTRUCTIONS .............................................................................................................5 1.1 LIGHT VERBS ...........................................................................................................................................5

1.1.1 Definitions........................................................................................................................................5 1.1.2 Open and Closed Class Words..........................................................................................................6

1.2 PERSIAN LVCS.........................................................................................................................................7 1.3 PREVIOUS STUDIES ON PERSIAN................................................................................................................9

1.3.1 Some Preliminary Grammatical Facts ..............................................................................................9 1.3.2 Contribution of Each Element and Event Structure .........................................................................16 1.3.3 Noun Incorporation ........................................................................................................................19 1.3.4 Other Studies..................................................................................................................................20

1.4 OTHER LANGUAGES................................................................................................................................21 1.5 CHAPTER REVIEW...................................................................................................................................23

2 FUNDAMENTAL THEORIES AND STUDIES........................................................................................25 2.1 PRODUCTIVITY .......................................................................................................................................25

2.1.1 Definition .......................................................................................................................................25 2.1.2 A Study on Productivity ..................................................................................................................26 2.1.3 Productivity in Persian LVCs .........................................................................................................27 2.1.4 Conclusion .....................................................................................................................................30

2.2 COMPOSITIONALITY................................................................................................................................30 2.2.1 Definition .......................................................................................................................................30 2.2.2 Compositionality in Persian............................................................................................................31

2.3 POLYSEMY .............................................................................................................................................32 2.3.1 Definition .......................................................................................................................................32 2.3.2 Theories and Studies on Polysemy ..................................................................................................33 2.3.3 Polysemy and Meaning in Persian LVs...........................................................................................37 2.3.4 Conclusion .....................................................................................................................................39

2.4 CONSTRUCTIONS ....................................................................................................................................40 2.4.1 Construction Grammar...................................................................................................................40 2.4.2 Previous Studies on Persian ...........................................................................................................42 2.4.3 Persian Constructions ....................................................................................................................43 2.4.4 Conclusion .....................................................................................................................................45

2.5 CHAPTER REVIEW...................................................................................................................................45 3 ISLANDS .....................................................................................................................................................47

3.1 CORE MEANINGS AND ISLANDS...............................................................................................................47 3.1.1 Core Meanings and Islands ............................................................................................................47 3.1.2 Defining Constructions Through Complements...............................................................................48 3.1.3 Persian...........................................................................................................................................49 3.1.3 Example Islands .............................................................................................................................50

3.2 MORE ON ISLANDS AND CONSTRUCTIONS ...............................................................................................53 3.3 PREVIOUS RESEARCH PROJECT ...............................................................................................................55 3.4 THE LIGHT VERBS ..................................................................................................................................57

3.4.1 ZÆDÆN........................................................................................................................................59 3.4.2 XORDÆN......................................................................................................................................78 3.4.3 GEREFTÆN .................................................................................................................................88 3.4.4 DADÆN ......................................................................................................................................105 3.4.5 Conclusion ...................................................................................................................................124

3.5 CHAPTER REVIEW.................................................................................................................................125

ii

4 ISLAND GENERATIVITY ......................................................................................................................127 4.1 PAST STUDIES OF ALTERNATING CONSTRUCTIONS ................................................................................127

4.1.1 Alternations in English .................................................................................................................127 4.1.2 Alternations in Persian .................................................................................................................129

4.2 ALTERNATIONS: A CASE OF SHARED ISLANDS ......................................................................................130 4.2.1 Alternations between kærdæn and ∫odæn ......................................................................................131 4.2.2 Alternations between zædæn and xordæn......................................................................................136 4.2.3 More Nuanced Alternations..........................................................................................................138 4.2.4 Is It Just Transitivity? ...................................................................................................................141 4.2.5 Alternations between dadæn and gereftæn....................................................................................146 4.2.6 Conclusion ...................................................................................................................................148

4.4 CHAPTER REVIEW.................................................................................................................................148 5 CONSTRUCTIONS AND DISCUSSION.................................................................................................149

5.1 TRANSPARENCY ...................................................................................................................................149 5.1.1 Transparent LVCs ........................................................................................................................149 5.1.2 Idiomatic LVCs.............................................................................................................................150

5.2 CONSTRUCTIONS ..................................................................................................................................150 5.2.1 Idiomatically Combining Expressions...........................................................................................151 5.2.2 Dual Nature of Constructions .......................................................................................................153 5.2.2 Motivations...................................................................................................................................156 5.2.4 Relations between Constructions ..................................................................................................158

5.3 THE PERSIAN VERBAL SYSTEM .............................................................................................................158 5.3.1 What’s the Difference? .................................................................................................................158 5.3.2 Why So Many LVs?.......................................................................................................................161 5.3.3 Productivity..................................................................................................................................162

5.4 A NEUROLINGUISTIC STUDY.................................................................................................................163 5.5 CHAPTER REVIEW.................................................................................................................................163

6 THESIS SUMMARY.................................................................................................................................165 REFERENCES.............................................................................................................................................167 APPENDIX...................................................................................................................................................173

iii

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude to Pierre Cadiot for being a valuable advisor. Pierre overlooked my work, provided constructive comments on linguistic issues, while he allowed me to pursue my own ideas freely. He also endured the tedious task of helping me edit the French version of this thesis to the level of academic standard for which I thank him again.

I am highly indebted to Dick Carter who generously offered me his time and energy, studying my work at each step and arguing with me over controversial issues. Every discussion with him was motivating and insightful, the completion of my work was only possible with his help.

A third pivotal person who greatly inspired my work is my father, Nader Family, whose insatiable curiosity and intuitive questions helped drive this project forward. My warmest gratitudes for his assistance while I was producing my corpus and helping me find the software for the visual analysis.

A special thanks goes to Victor Rosenthal for his timely assistance and advice, especially during the important initial phases of my research; Bernard Victorri for finding time in his busy schedule for very insightful and open discussions at crucial points during my research; and Laleh Ghadakpour who in addition to personal guidance, studied and commented every chapter of my thesis.

I am truly grateful for Gilbert Lazard’s intuitive observations and incisive questions as a member of my thesis committee. I would also like to express my thanks to Alain Peyraube for presiding my committee and for his contributions in drawing important comparative parallels on the focus of my work in other languages.

Thanks also to Jean-Louis Dessalles who, as my master’s studies advisor, supported and encouraged me to pursue my ideas further and then continued with his help afterward. And also to colleagues in Iran who helped setup a productive exchange with Iranian researchers, Mohamad Vahedi-Langarudi and Reza Nilipour.

Ecole doctorale Cerveau-Cognition-Comportement (ED 3C) which granted the funding for this research deserves recognition for its generosity and I thank its administration for making this effort possible.

And finally, thanks to friends and family for their love and support: Simmin Moghaddam, Hadi Danaee, Farhod Family, Fereshteh Family, Heidi Bishay, Madjid Khaladj, Arden Schweich, Nassim Hashemnia, and Narly Golestani.

v

ABBREVIATIONS

The following is a list of abbreviations used in the morpheme glosses of this thesis:

ACC accusative (marked by the ra- morpheme) GEN ezafe morpheme (similar to genitive, used to express possession/relation) IMP imperative INDEF indefinite INF infinitive NEG negative OBJ direct object marker PL plural PROG progressive PRS present PST past PTCP participle SG singular SUBJ subjunctive

1

0 INTRODUCTION

In this thesis, I address the problem of the construction of meaning of compound expresions, which is at the core of the Persian verbal system. I investigate the semantic properties of light verb constructions in Persian in order to address some fundamental properties of human languages: productivity, compositionality, polysemy, and building constructions (especially linguistic). These properties are present in all human languages. This thesis sheds new light on the organization of the Persian verbal system. Using notions from existing theories, I give a new treatment of LVCs in Persian. The methods used in this work allow us to deal with issues not dealt with in earlier work.

0.1 Language and the Study of Mind

Linguistics has long been recognized important to the study of cognition. We use language as a tool to communicate our views of the world as well as our personal experiences and mental states. It is reasonable to assume that the structure of this highly elaborate tool should reflect some general properties of the mechanisms related to categorization and perception used by the brain. Therefore, linguistics is a crucial component of cognitive science and requires sustained scientific analysis.

In the study of the language faculty, we try to find evidence for the existence of more general cognitive mechanisms through the analysis of structure and meaning as expressed through utterances. Although brain studies have not to date discovered the biological “mechanism” of language, or how its functions are distributed in the brain, we can already support or disprove certain theories of language through imaging and other types of neuropsychological findings.

But in tandem with biological studies, it is essential that we understand better what it is that we are studying. We need better definitions of issues and problems involved in cognitive processing as they relate to language, in order to efficiently narrow the direction of our investigations.

In my research I am trying to understand, or at least better define, certain aspects of a fascinating phenomenon in human language, and use its apparent and somewhat clearly defined parameters to create a well directed road sign for future research.

One of the fundamental characteristics of natural language is productivity. This is the ability to form an infinite number of novel grammatical utterances within the bounds of the linguistic system. Since this cognitive ability exists in higher reasoning as well, and since language can provide us with solid empirical data, the study of productivity in language should allow us to relate our findings to other areas of cognitive science.

Another defining characteristic of natural language is compositionality, or, more generally, the principles through which the meaning of compound expressions is calculated. The meaning of compound expressions can result from the exact combination of the meanings of its constituent elements, or it can have a completely opaque meaning, apparently unrelated to its constituents. Most compound expressions in language fall somewhere between these two extremes, where the meaning of the whole does not result merely from the sum of the meanings of its parts. This property of language is directly related to properties of the human conceptual system, especially to the construction and interpretation of compound concepts.

2

Yet a third characteristic of natural language is polysemy, or the use of a single form to express a set of related meanings. It presumably forces us to use our inferential processing capacities more than we would if all linguistic elements had a one-to-one form to meaning pairing. This expansion of the meaning of a form allows much flexibility in expression and comprehension. The study of this phenomenon could also shed light on the dynamics of language and perhaps, on other cognitive processes.

To study these phenomena, it is useful to study a language that provides an abundance of clear cases of productivity, compositionality, and polysemy. As we will see below, the Persian verbal system seems to be an excellent candidate to serve these requirements.

The most adequate framework available to study the type of data present in the Persian verbal system is Construction Grammar. The methods in this analysis stem from the basic tenets of this theory of grammar. Constructions are form-meaning pairs where the meaning of the whole is more than the meaning of the parts. This framework allows for non-linguistic knowledge to be tied to particular elements or compositions.

0.2 Elements under Study

Before I begin discussing the linguistic data, it is necessary to specify the type of elements we are studying. The same utterance can have different meanings or interpretations in different contexts. In this study, I am interested not in the elements that might modify the meaning in specific contexts, but rather in the basic meaning common to most contexts.

In his Course in General Linguistics (1915), Ferdinand de Saussure differentiates la parole (speech) from la langue (language). Langue denotes the abstract principles or ideal capacity to use language knowledge, whereas parole denotes individual occurrences of language. La parole is not possible without la langue.

This langue/parole differentiation was echoed in modern linguistics as a distinction between performance and competence (Chomsky, 1965, p.4). Performance denotes the use of language competence, and provides evidence of that internal knowledge of language. Importantly, in performance, factors other than linguistic competence can affect the interpretation of the uttered form. Competence is the practical knowledge a speaker has of his or her language.

In line with a long tradition, Fuchs & Victorri (1996) make a distinction between “énoncé-type” (utterance-type) and “énoncé-occurrence” (utterance-occurrence) (p. 27). The former is defined by the linguistic properties present in the utterance-occurrence. The utterance-occurrence is an element that is used concretely in a social interaction where psychological, referential, and pragmatic parameters come into play. The authors provide the example It is cold as an utterance-type that has different utterance-occurrences depending on the context of its utterance. If it is in response to Will you come to the market?, the response would be referring to the outside temperature and insinuate a negative answer. Whereas if it is uttered at a party, it might refer to the temperature of the room and perhaps provides a reason for the person keeping his or her coat on. This shows different utterance-occurrences that can proliferate from a single utterance-type. The definition of an utterance-type is therefore the constant contribution of semantic elements in all possible utterance-occurrences.

3

In this study I study utterance-types, as specific elements of competence. I am interested in purely linguistic parameters (e.g. semantics), and am not concerned with psychological and pragmatic parameters that might interact in the interpretation of specific utterances.

0.3 The Interest of Studying Persian

One of the most striking features of the Persian language is its deceptively small repertoire of approximately 160 simple verbs. This small number conceals a complex system of verbal constructs. In this limited set of simple verbs, there are around 15 verbs, the “light verbs” (LV) from which a vast array of verbal notions is constructed. These LVs combine with preverbs (PV) to produce new verbal expressions, sometimes with meanings seemingly unrelated, or non-transparently related, to the LV when used alone. These PVs can be nominal, adjectival, adverbial, or prepositional words or phrases. The two components of these light verb constructions (LVCs), also known as complex predicates or compound verbs, interact to form new verbal notions. This productive process is at the core of a system that has evolved according to certain principles that help organize verbal semantic space in Persian. An analysis of the verbal system must crucially account for the productivity of these forms. Examples of these verbs include:

be zæban aværdæn to tongue bring say, utter xun ændaxtæn blood throw cause to bleed kæm ∫odæn few become be diminished, become less qæbul da∫tæn assent have maintain as true arame∫ dadæn calmness give calm someone jæ∫n gereftæn party obtain celebrate t∫opoq ke∫idæn pipe pull smoke a pipe qorbæt ke∫idæn remoteness pull suffer from longing for home Similar constructions are common in many languages, namely in English in forms like take a walk, give tongue, pull on a pipe, make a speech, or in French faire le ménage or prendre une douche. However, unlike Persian, most languages often have simple verbs that can roughly replace these compound forms (e.g. to walk, to speak, se doucher), which usually make up a rather small set compared to simple verbs. In Persian, on the other hand, as in certain other languages, LVCs are often the only option for expressing verbal notions.

Due to the quasi-compositional nature of these forms, Persian LVCs provide an interesting window into the dynamics of meaning construction from compound expressions. The expression of such diverse meanings through the use of the same LV results in high levels of meaning extensions at the core of this verbal system. This manner of verb formation constitutes the basis for productivity in the system. Thus, Persian is a language within which these linguistic properties can be jointly investigated.

Our endeavor is not to describe the large set of LVs in detail, but to investigate the dynamics of this relatively closed system. A large variety of actions are expressed using a small set of LVs allowing for extensions across cognitive domains (e.g. space and time, emotional processes, intentional and non-intentional actions). My goal includes finding the patterns of productivity in the system and describing the mechanism through which new verbs are constructed and understood. I also hope to find contrasts and similarities in LV meanings that might point to fundamental properties of the human conceptual system. This study may suggest answers to certain problems in current theories of concepts and conceptual structure.

4

0.4 Outline of This Study

In the first chapter, I will discuss the notion of light verb as it relates to Persian and I also give an overview of previous studies done on light verbs in Persian and in other languages. In the second chapter, I present and attempt to define, in relation to our study, key linguistic terms that will be employed in the analysis of this system. These include the notions of productivity, compositionality, and polysemy. I will also discuss relevant grammatical frameworks such as Construction Grammar. In the third chapter, I define the notion of “islands” and provide a description of some islands centered on chosen verbs. In the fourth chapter, I discuss how the LVs work together in the system and how forms are interrelated. In the fifth chapter, I discuss consequences of such a system on the production of productive utterances and on the structure of the conceptual system compared to languages with a different structure. And finally, in the sixth chapter, I offer a conclusion and suggest possible topics for future research.

5

1 LIGHT VERB CONSTRUCTIONS

1.1 Light Verbs

1.1.1 Definitions

Light verb constructions (LVCs), also known as complex predicates or compound verbs, consist of a semantically weak verb (the light verb) and a non-verbal constituent. The resulting meaning of this multi-word expression is often opaque and it is not easy to discern what types of semantic and syntactic information each element contributes. An LV often contributes less of its lexical meaning than in its use as a full verb. For example, English LVs include take and give when used in LVCs such as take a walk or take a seat, and give a talk or give a break. The meanings of the verbs seem to become reduced, or bleached in LVC contexts. The amount or type of information portrayed by the LV in these constructions is subject to much debate in linguistics.

The first use of the term ‘light verb’ (Jespersen 1965, Volume VI : p 117) refers to the process in Modern English of placing a semantically empty verb which takes marks of person and tense, before the main notion of the expression. In these compounds, the non-verbal element provides most of the semantic information and the light verb contributes little or nothing to the semantics of the clause. In English, for V NP constructions, such as take a drive, take a walk, have a drink, have a laugh, the largest part of the argument structure of the LVC seems to be derived from the non-verbal element of the phrase. This led Jespersen to conclude that the light verb does not contribute much information to the construction (Jespersen, 1965). Unlike Persian, LVCs in English are much less frequent than simple verbs, and often have simple counterparts that express similar meanings, as with walk vs. take a walk.

Cattell (1984) considers the set of light verbs to be part of the verbal grammatical category. The light verb provides a predication license to the preverb, merely expressing that the action expressed by the preverb has occurred. In his book, Composite Predicates in English, he gives a comprehensive overview of previous studies on English LVs. Most of his own analysis pertains to delineating the syntactic and semantic constraints that differ between LVCs and their lexical counterparts. He describes the environments in which LVCs occur, but concludes that it is not possible to provide an algorithm that predicts which LV a particular complement will combine with (Cattell, 1984, p. 91).

In the generative framework, Grimshaw and Mester (1988) used the term in their discussion of the suru construction in Japanese, a verb with the general meaning “to do” often compounded with nominals. This verb is semantically light and purportedly plays only a functional role in the sentence.

Chomsky (1995) following Keyser and Hale (1993), Larson (1988) and others, uses ‘light verb’ for a verb that serves as a syntactic place holder, and typically has an incomplete event structure. According to these theories the LV holds the place of AGR (where the agreement of the clause is marked) in a sentence, acting as the head of a transitive. In most languages it is also phonetically weak and doesn’t have an explicit phonological form.

These theories tend to underestimate the amount of semantic information the LV contributes to the LVC. Brugman (2001) defines LVs as “lexical items which provide simultaneously semantic contributions and discourse functions to their constructions.” She claims that light

6

verb constructions are headed by a “verb which typically has a very wide distribution – it is polysemous and may be associated with a wide variety of patterns of argument selection.” An LV can be viewed as expressing a subset of more general semantic features present in the full verb counterpart.

1.1.2 Open and Closed Class Words

Many linguists make a fundamental distinction between the grammatical and lexical classes in a language. The lexical subsystem, sometimes referred to as open class, includes all verbal, nominal, adjectival roots, and any word class that is productive and doesn’t resist new additions. On the contrary, the grammatical, or closed (functional) class, subsystem only includes grammatical words, relations, word order patterns and compositional syntactic structures. Words from this class serve to express grammatical and semantic relations and are generally very high in frequency and low in semantic content.

In such a system, for a listener to construct a cognitive representation from a linguistic utterance, the lexicon must provide content, while the grammatical subsystem provides cues for their structure. The structure and content interact with respect to elaborate constraints including topology, perspective, attention, viewpoint, location and motion, force and causation. This interaction of conceptual content and structure creates a unified cognitive representation from disparate sources of information. This also allows for an infinite range of conceptual representations, constructed in context dynamically where elements contribute specific content depending on the context.

Differentiating closed class and open class words can be difficult in certain cases, partly due to the fact that some closed class words are historically derived from open class words. Some linguists, like Hopper (1993, p 137-155) have posited that words in a language span a continuum between open and closed class elements. Others find there to be two distinct categories.

In Toward a Cognitive Semantics, Talmy discusses the lexical and grammatical subsystems from a cognitive perspective. A cognitive representation of an utterance is the imagistic experiential complex it triggers (Talmy, 2000). The closed grammatical class provides the ‘skeletal’ structure of the image, and the lexical items provide the content. Closed class words refer to what Talmy calls “schematic categories.” These categories interact with each other to form a system abiding to overarching principles which help organize the grammatical and lexical subsystems. These principles incorporate dimensions such as force, causation, or configuration. Since these grammatically specified notions “specify a crucial set of concepts”, they determine conceptual structure (Talmy, 2000).

LVs seem to be one of the linguistic phenomena that cannot be properly categorized as being part of either extremes of this continuum. On the one hand, the set of LVs scarcely accepts new additions, similarly to grammatical markers like auxiliaries. On the other hand, LVs are definitely not semantically empty, or devoid of lexical content, which is not necessarily true for auxiliaries, which may serve purely grammatical functions.

In her study on LVCs in Urdu, Butt & Geuder (2004) describe how auxiliaries differ from LVs. As in Urdu, auxiliaries in Persian behave differently than LVs. Verbs like budæn (to be), da∫tæn (to have), and xastæn (to want) have specific functions in the grammar when used in their auxiliary form (e.g. to express different verb tenses, modals, or the future). Unlike these auxiliaries, LVs appear in diverse contexts and not exclusively in a single tense or

7

aspectual context. They also display different patterns in regard to word order, or topicalisation. Furthermore, grammaticalisation tends to change lexical items phonetically and syntactically, as can be seen in contexts where the auxiliary verb budæn is expressed as a suffix. Persian LVs never undergo these types of morphological or phonetic changes. LVs show more stability than auxiliaries.

Thus, as has been shown by Butt & Geuder, LVs must be clearly distinguished from auxiliaries. Auxiliaries are verbs that serve a purely grammatical function. Though historically related to full lexical verbs, their diachronic development branches away from the original meaning. Though the semantic properties of LV are hard to pin down, they are not necessarily the result of consistent bleaching of an originally full lexical verb by a process of grammaticalisation. Hook (1991) originally proposed this explanation for the shift from full verb to auxiliary. He proposes that the more frequent an LV in a language, the weaker the semantic connection between such an LV and its full counterpart. As lexical items of a language go through a grammaticalisation process, they gradually lose their original meaning and eventually only serve as functional elements in the grammar. A gradient might exist with full verbs at one extreme and auxiliaries at the other. Butt & Gueder (2004) argue that the LV stage of a verb does not fall between these two extremes and that LVs are not the result of an evolving process of grammaticalisation.

Butt & Geuder (2004, p 35), conclude that

light verbs serve the purpose of structuring event concepts. To this end, they are reduced to an entirely schematic meaning. The crucial difference between light verbs and auxiliaries is that, with light verbs, this schematic information is applied in order to structure a lexical content (namely the full verb), not in order to express a grammatical meaning as is the case with auxiliaries. So even if there were a gradual development by which light verbs are becoming semantically weaker over time, this would not mean that they are becoming more grammatical: We would just arrive at a more coarse-grained structuring of event concepts.

I follow a similar argument as Butt and Geuder’s proposal that LV meaning is “linked to the main full verb reading via analysis under which light verb usages are considered to be an instance of polysemy” (2004, p25). For the purposes of our study, we will assume that LVs in Persian have a similar amount of lexical content as those of Urdu. One main difference of LVCs in Persian is that they do not combine with other verbs to form complex predicates as they do in Urdu.

1.2 Persian LVCs

One of the unique characteristics of the Persian verbal system is its predominant use of light verb constructions (LVC) to express verbal notions. Unlike English, or French, where each have over 5000 simple verbs, Persian has less than 200. All other verbal notions are expressed through complex expressions. LVCs in English or French, which occur much less frequently than simple verbs, often have a simple counterpart that expresses a similar meaning. For example, take a stroll (faire une promenade) can be replaced by stroll (promenade), or give a shout (pousser un cri) can be replaced by shout (crier). In Persian, however, the LVC used to express a verbal notion is almost always the only option and can only be translated by a simple verb in a different

8

language such as English. LVCs in Persian are formed by an LV in conjunction with a nominal element, an adjective, a preposition, or a prepositional phrase. These forms are discussed in any grammar of the Persian language (Lazard, 1992; Lambton, 1961; Rastorgueva, 1964; Beeman, 1986; Windfuhr, 1987; Mahootian, 1997). Some examples appear below (with the LV in upper case, the preverb in lower case).

dæst ZÆDÆN hand hit applaud, touch xæmiyaze KE∫IDÆN yawn pull yawn dust DA∫TÆN friend have like/love gu∫ DADÆN ear give listen hærf ZÆDÆN sound hit talk negah DA∫TÆN sight have keep email ZÆDÆN email hit send an email tayp KÆRDÆN type do type qose XORDÆN grief eat grieve

Out of the limited set of simple verbs in Persian, around 20 are used as a basis to form LVCs. In this study, I concentrate on 14 of the most common LVs. These verbs and the basic English translation of their full verb meanings appear below:

amædæn [come] kærdæn [do] ændaxtæn [throw] gereftæn [obtain, hold] aværdæn [bring] ke∫idæn [pull] bordæn [take] oftadæn [fall] dadæn [give] xordæn [eat, collide] da∫tæn [have] zædæn [hit] ∫odæn [become] ræftæn [go]

LVC forms have been gradually replacing simple verbs since the beginning of Modern Persian in the 13th century (Khanlari, 1973). For this reason, Persian can provide us with a large amount of data that has been in continuous usage over several centuries, giving us access to a robust system allowing us to search for evolutionary steps as new words were introduced from other cultures. The following series of LVCs, correspond either to previously simple verbs (given first in each line), now expressed only through LVCs, or to simple verbs that are becoming archaic. When used in a clause or sentence, such forms are equivalent, semantically and syntactically. geristæn ==> gerye KÆRDÆN cry do cry agahidan ==> agah KÆRDÆN aware do inform lafidæn ==> laf ZÆDÆN brag hit brag qæltidæn ==> qælt XORDÆN flip eat flip over LVCs in Persian often have non-compositional meanings, where the meaning of the whole is not a function of the meaning of its parts. In other words, the two or more elements that combine to form these LVCs have a partially idiosyncratic (unpredictable) form meaning-pairing. Due to this quasi-compositional (since the forms are not fully idiomatic, but motivated from the meaning of the constituents) structure of the LVC, the system presents an appropriate architecture to explore theories that attempt to link natural language semantics and fundamental cognitive processes.

9

LVs in Persian have developed diffuse ranges of implied or extended meanings, usually derived from their core meanings. The core meanings are noted through general English translations. We can expect to find central meanings to be the base for the multivaried set of extended or modified meanings available in the set of LVCs of each LV. Thus, when I present an LV within an LVC, it appears in capital letters, as not to confuse it with the full verb counterpart. Furthermore, Persian speakers unanimously provide this full verb meaning, and not the extended meanings, as the definition for the LV. For example, practically all Persian speakers would immediately say that the word xordæn means ‘to eat,’ while in fact, as we will see in the analysis of the data, there are over ten different extended or modified meanings which can be triggered from the same subjects when presented with LVCs of this LV. Similarly, zædæn will be translated as ‘to hit’ by all subjects, although it also occurs in certain constructions to express notions such as injection, playing a musical instrument, cutting off, calling on the telephone, ringing a bell, or rusting, to name a few. Before presenting the cognitive and linguistic interest of this study of LVs in Persian, I will give a brief overview of previous studies of Persian LVCs, which will provide a better understanding of the syntactic properties of Persian verbs, followed by an overview of previous studies done on LVCs in other languages.

1.3 Previous Studies on Persian

The classification of Persian compound verbs has been a subject of considerable interest, especially in the areas of syntax and semantics. All linguists who have studied Persian verbs have recognized that there is a problem with categorizing LVC forms as purely lexical or purely compositional. Most attempts at analyzing the contributions of each element in these constructions have been within strict linguistic frameworks, where the form and function of each entity remain distinct and no account of cognitive parameters is given. Though helpful in highlighting some interesting aspects of Persian LVs, these analyses fail to explain certain properties that play a critical role in language, namely productivity and predictability. These properties are important for language acquisition and production, and will be discussed in subsequent chapters. In the following subsections, I will give a brief outline of the main problems that linguists have highlighted regarding LVCs in Persian. Since most linguists come from slightly different backgrounds and have different methods for analyzing the data, giving a coherent synopsis of the studies done thusfar is not easy. I will present some grammatical facts that will familiarize the reader with the data, and I will mention linguistic dilemmas that may or may not be solved in the framework I will adopt in the analysis section of this thesis.

1.3.1 Some Preliminary Grammatical Facts

One of the prevalent questions that repeatedly appears in studies on Persian LVCs is whether they are lexical units or are constructed post-syntactically. In this section, I provide some examples from Megerdoomian (2002, chapter 3) and Karimi-Doostan (1997), but also provide insight and evidence for certain facts from other works as cited. Some of the facts do

10

not apply to all of the LVCs in the system, but can serve to show the heterogeneity of the phenomenon. LVCs can undergo all derivational processes that can be applied to simple verbs, while abiding by the same semantic or grammatical constraints. These include gerundive nominalization (a), agentive nominalization (b), adjective formation (c), participle formation (d), and adverb formation (e) (Megerdoomian, 2002; Karimi-Doostan, 1997; Goldberg, 1996; Vahedi-Langrudi, 1996). These derivational processes are productive, considering certain grammatical restrictions (e.g. only agentive LVCs can undergo the agentive nominalization). (a) gerundive nominalization: past stem + æn

sigar ke∫id-æn dar in saxteman mamnu æst. cigarette pull-æn in this building forbidden be.PRS.3SG. Smoking is not allowed in this building. æz hærf zæd-æn-e∫ mælum bud ke ne-mi-ay-æd. from speech hit-æn-3SG evident was that NEG-PROG-come.PRS-3SG. It was evident from his manner of talking that he wasn’t coming.

(b) agentive nominalization: present stem + ænde (or +ø)

gozare∫ dæh-ænde news give- ænde reporter bazi kon-ø play do player

(c) adjective formation: infinitive + i

in kelid peyda ∫odæn-i nist this key found become-i NEG.be.3SG This key is not to be found.

(d) participle formation: past stem +e

miz-e xak gereft-e table-GEN dust take-e dust covered table

(e) manner adverb formation: present stem + an

bæt∫e-ye ∫eytun jiq zæn-an varede otaq ∫od. child-GEN naughty scream hit-an entered room become.PST.3SG. The naughty child entered the room screaming.

Furthermore, only the LVC can undergo such transformations, and not the LV alone, though a valid form may be possible with the simple verb counterpart of the LV. Here, the agentive nominalization is not possible with the verb kærdæn ‘to do’ alone. This shows that the stem indeed attaches to the whole LVCs and not just the LV. There are other examples, such as the one below with gereftæn ‘to get,’ where the LV can take an agentive nominal suffix when it appears as a full verb. It can also be nominalized in the LVC form. It is important to note that the simple verb counterparts of some LVs do allow these transformations (e.g. ræftæn,

11

ræftæni), resulting in identical forms as in the LVCs they appear in. However, the unacceptable cases show that when the verbs appear as LVs, it is the LVC that undergoes the transformation and not the LV.

kærdæn to do, make *kon-ænde doer bazi KÆRDÆN play bazi konænde player

gereftæn to get gir-ænde receiver mahi GEREFTÆN to fish mahi gir-ø fisherman LVCs resist separation in a sentence. Adverbs cannot occur between the PV and LV, as seen in (a), but must precede LVCs as shown in the (b) sentences. (a) *aparteman ejare zud ræft.

apartment rent quickly go.PST.3SG. *the apartment was rented out quickly. *ali ranændegi yæva∫ kærd ali driving slowly do.PST.3SG. *Ali drove slowly.

(b) apartman zud ejare ræft. apartment quickly rent go.PST.3SG. The apartment was rented out quickly. ali yæva∫ ranændegi kærd. ali slowly driving do.PST.3SG. Ali drove slowly.

Compare this to non-LVC forms where the verb and the direct object can be separated with adverbs.

Ali ketab ziyad mi-xun-æd. Ali book much PROG-read.PRS-3SG. Ali reads books a lot. Ali ruye nun kære kæm mi-mal-æd. Ali on bread butter little PROG-spread.PRS-3SG. Ali spreads too little butter on the bread.

12

Similarly, the direct object of a sentence cannot occur between the two elements of the LVC. (a) * Færhad nejat ∫irin ra dad.

Farhad save Shirin-ACC give.PRS.3SG. * Farhad saved Shirin.

*fer muhayæ∫ ra zæde æst. curl her hair-ACC hit-PTCP be.PRS.3SG. *She has curled her hair.

(b) Færhad ∫irin ra nejat dad. Farhod Shirin-ACC save give.PST.3SG. Farhod saved Shirin. muhayæ∫ ra fer zæde æst. her hair-ACC curl hit-PCPT be.PRS.3SG. She has curled her hair.

Many linguists have put forth the fact that an LVC has a single main stress accent as evidence for its being a single unit in the lexicon (Family, 1998; Vahedi-Langrudi, 1996; Megerdoomian, 2002; Goldberg, 1996). In simple forms like (a) below, the stress falls on the final syllable of the verbal stem by default, whereas in LVCs like (b), the stress falls on the final syllable of the PV (example from Goldberg, 1996). (stressed syllable capitalized) (a) Ali mærd ra ZÆD. (simple verb)

Ali man ACC hit.PST.3SG. Ali hit the man.

(b) Ali ba Babæk HÆRF zæd. (LVC) Ali with Babak word hit.PST.3SG. Ali talked with Babak.

LVCs can also display phrasal behavior. Individual components of the construction can undergo certain morphological derivations. For example, certain verbal inflections are prefixed to the verb. In LVC contexts, this means that the inflection appears between the PV and the LV. Examples of these inflections, which express the imperative, the progressive, and negation markers, appear below in bold type.

kæbabha ra bad be-zæn! kebobs ACC wind IMP-hit.PRS! Fan the kabobs! salon ra jaru mi-ke∫-æd. salon ACC broom PROG-pull.PRS-3SG. He is sweeping the room. diruz baran næ-amæd. yesterday rain NEG-come.PST. It didn’t rain yesterday.

13

Modals, like tævanestæn ‘to be able’ or da∫tæn ‘to have’, or the auxiliary of the future tense xastæn ‘to want’ also appear between the PV and the LV. When an LVC takes xastæn, the future marker (with full meaning want) this verb carries the agreement markers.

morad zæmin da∫t mi-xord ke man gereft-æm-æ∫. morad ground had PROG-eat that I obtain.PST-1SG-3SG. Morad was falling when I caught him. in gonje∫k pærvaz ne-mi-tun-æd bo-kon-æd. this sparrow flight NEG-PROG-able.PRS-3SG SUBJ-do PRS-3SG. This sparrow cannot fly. in zobale ra dur xah-æm ændaxt. this trash ra far want.PRS-1SG throw.PST. I will throw this trash away.

In certain cases, namely when the PV is a process noun (Karimi-Doustan, 1997) or an abstract noun that can bear a-structure, the PV can be modified within an LVC. The components of an LVC can sometimes be relativized or coordinated. For instance, PVs can appear in the plural and be modified by PPs or other elements. For example, the sentence with the basic LVC e∫tebah kærdæn is shown in (a), and the version with the modified PV e∫tebah-at-e færavan-i appears in (b). (a) hæsæn dær næql-e in dastan e∫tebah kærd.

Hasan in telling-GEN this story mistake do.PST.3SG. In telling the story, Hasan made a mistake.

(b) hæsæn dær næql-e in dastan e∫tebah-at-e færavani kærd. Hasan in telling-GEN this story mistake-PL-GEN numerous do. PST.3SG. Hasan made many errors in telling the story.

Compare these examples to cases where the PV is not a process noun, but a non-predicative (or concrete, non-argument bearing noun). In these cases, the PV cannot be relativized.

Ali be radiyo gu∫ kærd. Ali to radio ear do.PST.3SG. Ali listened to the radio. *Ali gu∫-e xubi be radio kærd. Ali ear-GEN good to radio do.PST.3SG. Ali listened well to the radio.

Emphatic elements might appear after the PV element and before the LV (Megerdoomian, 2002, p63). In these examples, the emphatic particle ke appears between the two elements of the LVC.

gerye ke mi-kærd mæn hæm ba u mi-nalid-æm cry that PROG-do.PST.3SG I also with him PROG-groan. PST -1SG. Whenever he cried, I also groaned with him.

14

deraz ke mi-ke∫id dærd-e kæmær-e∫ kæm mi-∫od. long that PROG-pull.PST.3SG pain-GEN back-POSS little PROG-became.PST.3SG. When he lay down, his back pain was releived.

As mentioned earlier, adverbs cannot intervene between the PV and the LV. However, the adjectival form of the adverb can modify the PV, making the PV a nominal phrase. The ungrammatical sentence in (a) is rendered acceptable by changing the adverb into an adjective modifying the PV, shown in (b). However, this seems only to occur in cases where the LVC is nearly or completely compositional in meaning. (a) *ali ranændegi yæva∫ kærd

ali driving slowly do. PST.3SG. *Ali drove slowly.

(b) ali ranændegiy-e yæva∫i kærd.

ali driving-GEN slow do. PST.3SG. Ali drove slowly

Similarly, constituents corresponding to direct objects in other languages can sometimes appear as an ezafe (roughly a possessive marker) complement to the PV element, in which case they appear between the PV and the LV components. The ungrammatical sentence in (a) becomes acceptable if the direct object becomes the object of the PV, as in (b). Again, this occurs in cases where the LVC in question is semantically transparent. (a) *setaye∫ æli ra kærd-æm.

praise Ali ACC do.PST-1SG. *I praised Ali.

(b) setaye∫-e æli ra kærd-æm.

praise-GEN Ali ACC do. PST-1SG. I praised Ali.

In certain cases, the PV can be gapped from the LV, without causing any ungrammaticality (see Karimi-Doostan, 1997: p 67)

sal-ha sasan ra ∫ekænje va æzab dad-ænd. year-PL Sasan ACC torture and torment give PST-3PL. They tortured and tormented Sasan for years.

Sometimes, the PV can be a noun phrase or grow to be a determiner phrase. In these cases the size of the PV can be limitless, and in some frameworks this lessens the plausibility that LVCs of this type are stored individually in the lexicon. In the following example (from Karimi-Doostan, 1997) the basic LVC of interest here is lætme zædæn ‘damage hit = to damage.’

lætme-i ra ke tægærg be baq-e mæn zæd bavær næ-kærdæn-i bud. damage-INDEF ACC that hail to garden-GEN I hit.PST.3SG.belief NEG-do PST-NOM be. PST.3SG. The damage that the hail caused to I garden was unbelievable.

15

In transitive LVCs, the direct object clitic can attach to the PV element of the LVC. This clitic can also attach to the whole LVC, appearing as a suffix to the LV. This property most clearly emphasizes the dual nature of LVCs with respect to lexicon and syntax. The following examples are both acceptable.

kotæk-e∫ zæd. beating-3SG(clitic) hit.PST.3SG. He hit him. kotæk zæd-e∫. beating hit PST.3SG-3SG(clitic) He hit him.

Goldberg (2004) uses this as a piece of evidence that supports the analysis of Persian LVCs in a constructional framework. Goldberg posits that there is a construction that treats LVCs as Vo (or basic verbal element, or lexical unit) (the following diagrams are visual instantiations of constructions): The above diagram shows that a Vo consists of two elements and the first receives phonetic stress. She shows that other constructions in the language can override the Vo status of the LVC, and allow its unit-status to be flexible. For example, the future marker and verbal markers that occur preverbally can override the above construction. This occurs because of the dominance of the following constructions (here the future marker) over the one above (p10). She claims that for the case of the clitic (see examples above), where the clitic can occur in both internal and external positions, there are two constructions, neither of which preempts the other.

Cat: V0 Χ0 < V0

Future xâstan < V0past

Future-CP X0 < xâstan <V0 past

V’ predicate / V0-CL

V’ predicate / X0-CL < V0

CP+clitic

Simple verb + clitic

16

One construction calls for the clitic to occur post verbally and the second one above calls for it to occur after the primary stress of an LVC (which falls on the PV element) (Goldberg 2004, p 14). In her analysis, LVCs result from a particular construction which occurs as a unit in the lexicon. This doesn’t entail atomic syntactic status. Constructions occur on a hierarchy allowing more specific constructions possibly blocking the more general constructions. The basic tenets of this framework will be discussed in further detail in the following two chapters.

1.3.2 Contribution of Each Element and Event Structure

One prominent question that arises when investigating the Persian verbal system is the semantic and syntactic content contributed by each element of the LVC. Karimi & Mohammad (1992) propose that LVs in Persian are semantically empty. They claim that the entire semantic content comes from the PV and the LV serves as a verbal marker. They present examples where the change of LV, while retaining the same PV, does not affect the resulting meaning. However, there are many similarly constructed pairs which refute this claim, and Karimi (1997) and Goldberg (1996) discuss further problems with this view. In addition, examination of sets of LVCs constructed with a single PV and different LVs shows drastic differences in meaning. One example of such a set appears below, and there are many other similar cases.

ab gereftæn water GEREFTÆN juice ab ke∫idæn water KE∫IDÆN rinse ab ræftæn water RÆFTÆN shrink ab ∫odæn water ∫ODÆN melt ab dadæn water DADÆN water (i.e. a plant) ab oftadæn water OFTADÆN become watery Karimi-Doostan (1997) and Vahedi-Langrudi (1996) also claim that LVs are semantically bare. All the semantic and argument structure is said to be provided by the PV. However, other evidence shows that switching the LV in a particular LVC can change the argument structure. Vahedi-Langrudi deals mostly with the syntactic structure of LVCs, while Karimi-Doostan discusses semantic features in detail. Bashiri (1981) shows that aspectual1 differences emerge upon substituting one LV for another. Furthermore, such aspectual alternations usually involve large sets of LVCs. One such alternation, which will be discussed in chapter 5, involves the LVs zædæn ‘to hit’ and ke∫idæn ‘to pull.’ Forms constructed with zædæn express an instantaneous or repetitive action, whereas those expressed with ke∫idæn imply a prolonged or durational action. This semantic difference is subtle and is often translated with the same simple verb in English or French. Some examples are listed below.

jar zædæn jar ke∫idæn yell (short or repetitive, durative) dad zædæn dad ke∫idæn shout

1 Aspect, a purely linguistic concept, has certain psychological realisations. Comrie (1976) defines aspect as a type of temporal description of an action. For example, imperfect action is an expanded view of the action, seen from within as if the action is ongoing. Perfective action, on the other hand, is action seen as a complete whole. These qualities are not present in the actions themselves, but in the linguistic encoding of the perspective taken.

17

sær zædæn sær ke∫idæn pay a short visit / long visit næfæs zædæn næfæs ke∫idæn pant / breathe dar zædæn dar ke∫idæn hang / string up on the gallows Karimi-Doostan (1997) deals with the aspectual contributions of LVs in some detail. He claims that there are two classes of LVs which he names Dynamic and Stative. Dynamic LVs can be either Transitory or Initiatory. Transitory LVs usually express the change of state of the subject argument, whereas Initiatory LVs take an agent who initiates an action. Stative LVs express non-changing states. The following table shows how he categorizes the 16 most frequent LVs.

DYNAMIC STATIVE Transition LV Initiatory LV xordæn ‘eat’ zædæn ‘hit’ da∫tæn ‘have’ yaftæn ‘find’ dadæn ‘give’ ∫odæn ‘become’ bæx∫idæn ‘offer’ amædæn ‘come aværdæn ‘bring’ gereftæn ‘take’ ke∫idæn ‘pull’ ræftæn ‘go’ bordæn ‘carry’ didæn ‘see’ goza∫tæn ‘put’ kærdæn ‘do’ kærdæn ‘do’

Only da∫tæn expresses stative actions, all other LVs are dynamic. Transition LVs are always unaccusative, or intransitive verbs where the subject does not initiate the action. Initiatory LVs can be either transitive (require a direct object) or unergative (an intransitive verb where the subject is responsible for initiating the action). In his thesis, he provides some grammatical tests to support this categorization. He also discusses accusativity and transitivity in detail. Karimi-Doostan (1997) also divides nominal PVs into three different groups in an attempt to distinguish different types of behavior for the resulting LVC. Non-predicative nouns are concrete nouns that can never be separated from the LV in a sentence. Process nouns are nouns that don’t occur in non-LVC contexts, but they can be modified, relativized, and separated from the LV. Verbal nouns cannot occur independently, but they lack certain nominal and verbal properties that would allow them to be modified. Karimi-Doostan (1997) applies a widely used method to distinguish bounded and unbounded verbs2, namely testing the acceptability of temporal adverbs within the clause. This method consists of inserting adverbials of the type en une heure and pendant une heure in French, or in an hour and for an hour in English. Bounded predicates can combine with in an hour, but cannot occur with for an hour, which expresses an unbounded time span. Karimi-Doostan shows that in certain cases, the choice of LV dictates the boundedness of the LVC. Examples from his thesis appear below. 2 Also known as the perfective/imperfective distinction. Perfective actions have a beginning and an end (the action is viewed from the outside), whereas imperfective actions continue over a span of time (the action is viewed from the inside).

18

Dæst-e daryu∫ dær yek saniye / *sa’æt-ha dærd gereft. Hand-GEN Dariush in one second / hour-PL pain obtain. PST.3SG. Dariush’s hand (started to) hurt within one second / *for hours. Daryu∫ * dær yek saniye / sa’æt-ha dærd ke∫id. Dariush in one second / hour-PL pain pull.PST.3SG. Dariush hurt * in one second / for hours. In the first sentence, the combination of dærd ‘pain’ with gereftæn results in a bounded LVC. When combined with ke∫idæn, the resulting LVC is unbounded. Hence, the aspectual distinction comes from the choice of LV. Megerdoomian (2002, p 69) shows that changing the PV can also change the aspectual interpretation of an LVC. For example, in the following sentences, the change of PV changes the boundedness of the resulting LVC. Hale *dær ærz-e nim sa’æt / sa’æt-ha gerye kærd. Haleh in length-GEN half hour / hour-PL cry do.PST.3SG. Haleh cried *in half an hour / for hours. Hale dær ærz-e nim sa’æt / ? sa’æt-ha qofl-e dær ra baz kærd. Haleh in length-GEN half hour / hour-PL lock-GEN door ra open do.PST.3SG. Haleh opened the door lock in half an hour / ?for hours. In the first case, the LVC accepts the durative adverbial, while the bounded reading is unacceptable. Whereas in the second case, the same LV, with a different PV accepts a bounded meaning. With an unbounded adverbial, the sentence is marginally possible, but would mean that Haleh locked and unlocked the door repeatedly for hours. In her thesis, Megerdoomian suggests that the PV contributes substantive information (the core lexical meaning), internal arguments, and aspectual properties. The LV contributes external arguments and other aspectual properties (causation, change of state, duration, inception). In her analysis, she finds the decomposition of LVCs into two components inadequate, and proposes a method of further breaking up the components of the construction. For further discussion, see Megerdoomian (2002, p 71-82). In attempts to describe LV transitivity alternations, there has been an interest in the event structure of LVs. Megerdoomian (2002) argues that “the transitive predicate is formed when a CAUSE event is added on top of the intransitive structure” (p74). Haji-Abdolhosseini (2000) takes a generative approach to the study of the event structure of LVCs. He attempts to predict the types of LVs a given PV occurs with, as well as predict the meaning of the resulting compound. He briefly explores these issues in the framework of the Generative Lexicon, but at the end fails to give a coherent account of the data. In a recent paper, Folli et al. (2004) investigate the elements that determine event structure in Persian LVCs. They show that the event structure of the LV can differ from that of its heavy counterpart, and also that, though it determines certain attributes of the LVC (e.g. agentivity), the LV does not fully determine the event structure of the LVC. This also depends on the type of PV occuring with the LV. In relation to nominal PVs, they claim that when the PV is

19

eventive, the LVC may be telic (accomplishment), otherwise the LVC is atelic (activity or semelfactive)3. They further show to some limited extents that the semantics of the PV determines whether it can combine with certain LVs. For example, they signal the importance of the difference between alternating inchoative/unaccusatives (marking the beginning of a process/ not accepting external cause) and non-alternating ones. Their example of the kærdæn/∫odæn alternation limitation appears below (Folli et al., 2004: p19). (a) Kimea sorx ∫od.

Kimea red became.PST.3SG. Kimea blushed.

(b) *Papar Kimea ra sorx kærd.

Papar Kimea ra red do.PST.3SG. *Papar made Kimea blush.

Here, the concept of blushing can only be caused internally and the PV ‘red’ cannot combine with kærdæn to mean ‘cause to blush.’ A similar restriction comes about with PVs forming inchoatives with xordæn, whose LVCs tend to have an agentive alternation with zædæn. These types of alternations will be discussed further in Chapter 4.

1.3.3 Noun Incorporation

Another issue that has been dealt with by several linguists is noun incorporation, a type of noun-verb compounding. In these analyses (Karimi, 1996; Dabir-Moghadam, 1997; Samvelian, 2001), LVCs are grouped into two different classes depending on whether or not these forms result from the incorporation of the direct object into the predicate. Most of these studies deal solely with LVCs that have nominal PVs. Noun incorporation occurs when a direct object loses its grammatical function and becomes incorporated into the verb to create an intransitive verbal complex. Dabir-Moghadam (1997, p 43) differentiates noun-incorporated forms from combined forms. He uses the disappearance of grammatical object markers, structural differences which inhibit the appearance of pronominal reference, nominal ellipsis, and gapping, as evidence for noun incorporation. Compared to combined forms, incorporated forms remain fairly transparent (semantically), have an un-incorporated counterpart, and are more productive than the other forms. He also focuses on the many similarities between the two types of LVCs and maintains that they are conceptual wholes, whether they are incorporated or combined forms. Karimi & Mohammad (1996) had previously argued against noun incorporation in Persian. They claim that LVs in Persian are semantically empty, and all the information are carried by the PV elements. In a review of the their paper, Dabir-Moghadam (1997) invalidates certain examples, though he fails to respond to evidence involving cliticization (cf. Karimi & Mohammad, 1997) and the incorporation of indirect objects. Samvelian (2001) returns to this theoretical issue involving Persian LVCs. She provides an overview of the two previous studies mentioned here, and introduces her own syntactic analysis of noun-incorporation in Persian. However, in a later paper, Samvelian (2003), she 3 Activities are duratives while semelfactive actions are punctual atelic actions.

20

claims that in fact both types give rise to new LVCs, and each of these might display different syntactic behavior. For the purposes of this thesis, the important fact to be retained from the studies on noun incorporation is that all types of LVCs, whether from the incorporation of an object in the verb or just a combination of concepts, the semantics can range from opaque to transparent. Since all LVCs undergo the same morphological derivations, it is not necessary to separate them in the above ways when investigating meaning (Vahedi-Langrudi, 1996). In the present study, I am interested in the semi-compositional semantics present in most LVCs in Persian. Previous studies highlight important properties of Persian LVCs: differences in argument structure, differences in syntactic behavior, seperability, and non-compositional semantics. Although the processes through which these forms became constructions may be rooted in different grammatical phenomena, the resulting forms share common semantic and syntactic properties. These draw attention to a structured organization of the concepts that are expressed through specific Persian LVs.

1.3.4 Other Studies

The study of idioms and the amount of semantic transparency of LVCs is also relevant to the present study and I will return to it in Chapter 5. Here, I will describe briefly a study which focused on this issue. In her 1997 article, Karimi discusses two different types of LVCs. She claims that there are a handful that are pure idioms, as in (a). These forms are completely opaque and non-compositional. However, most Persian LVCs are idiomatic combining expressions as in (b), in the spirit of Nunberg et al. (1994).

(a) xær kærdæn donkey KÆRDÆN fool dæst ændaxtæn hand ÆNDAXTÆN insult bala ke∫idæn up KE∫IDÆN steal sæbok kærdæn light KÆRDÆN belittle

(b) nærm ∫odæn soft ∫ODÆN soften

pi∫ ræftæn ahead RÆFTÆN progress qærar dadæn position DADÆN put fekr kærdæn thought KÆRDÆN think

In many cases, the meaning of an LVC cannot be attributed to the sum of its components, but rather to the combination. This often results in an idiosyncratic meaning. However, we will see that this idiosyncratic meaning can give rise to productivity. Because of this, and depending on how the term idiom is defined, these forms cannot be considered as idiomatic, as idioms are usually greatly restricted in productivity.

Dabir-Moghadam (1996) argues that certain LVCs considered to be compounds are in fact “frozen sentences” (p46). These forms include the special construction where the PV seems to serve as the subject of the verbal component of a sentence. These are peculiar constructions that occur with the LVs gereftæn, amædæn, and ∫odæn. They will be formally presented in Chapter 5, a couple of examples appear below:

næfæs-æm gereft. breath-1SG obtain.PST.3SG. I was out of breath.

21

xo∫-et amæd. good-2SG come.PST.3SG. You liked it.

Dabir-Moghadam’s dismissal of these forms is unfounded. Despite their particular syntactic construction, they only occur with LVs in Persian and they often have non-compositional meaning. I treat these forms as LVCs and include them in my analysis.

In an older article, Sharifi explores the LVCs in Persian within a Chafean framework (1975a). He gives an analysis of the verbs gereftæn (1975b) and ke∫idæn (1975c) which I will return to in the discussion of these verbs in Chapter 4. In his more general article, he provides a descriptive analysis of the LVC phenomenon. He says that LVCs are “used to introduce new semantic concepts, to replace verbs of Arabic origin, or to provide new ways of expressing an already existing semantic concept” (p 460). He claims that LVs are in fact surface realizations of “underlying semantic units and processes involved in the structure of simple verbs.” Hence, he argues that the core meaning of the heavy verb is present in LV usage and he applies Chafean (1970) verbal selectional units such as process, action, state, state-ambient, action-ambient and process-action, to LVCs in Persian. He notes, however, that the theory must be developed for a complete analysis to be possible.

Tabaian (1978) claims that there are no such thing as LVCs in Persian, and that most of what has been referred to as compound verbs in previous literature can be categorized as verbal phrases. He differentiates between three classes of verbs: prefixed (LVCs with a preposition as PV), complement and verb, and nominal and auxiliary. However, he fails to account for the non-compositional semantics and the productivity of most of these forms. Also, he does not define what he means by transparency, and does not discuss noun-incorporation in his treatment of the groups of verbs he labels as “complement and verb”.

1.4 Other Languages

In this section, I will briefly sketch several studies that have been done on LVCs in other languages. By no means is this a complete overview of the abundant literature, but only a scattering of the works most closely related to the subject of this thesis.

Many South Asian languages also display a similar type of phenomenon. Mohanan (1994) explores LVCs with nominal PVs in Hindi. She claims that these constructions are monoclausal (they have only one syntactic predicate), and that they constitute phrasal categories constructed post-lexically. Also, she claims that the PV, though it contributes to argument structure, can also be an argument of the LV. She also tries to show how argument types like agent, theme, and experiencer should be represented with finer-grained parameters.

Similar to Persian, LVs in Hindi have heavy counterparts. For example, the verb kar ‘to do’ and de ‘to give’ can occur as a heavy verb (a) or as an LV (b).

(a) raam-ne apnaa homwark kiyaa. Ram self’s homework do.

22

Ram did his homework. niinaa-ne raam-ko kitaab dii. Nina Ram book give. Nina gave Ram a book.

(b) raam-ne mohan-par barosaa kiyaa.

Ram Mohan reliance/trust do. Ram trusted/relied on Mohan. niinaa-ne raam-kii kahaanii-par dyaan diyaa. Nina Ram story attention give Nina paid attention to Ram’s story.

Case markings in Hindi provide an extra tool with which to explore the grammatical function of LVCs in the language. Using these clues, and also by comparing argument structure of the heavy verb and LV counterparts, Mohanan devises a formalism to capture the semantic structure of the LVCs. She separates semantic structure from grammatical function structure, argument structure, and grammatical category structure (Mohanan, 1994, p 438). Thus, she shows that there are two different types of LVCs in Hindi: those where the PV is an argument, and those where it is part of the construction. As evidence, she uses patterns in passivisation, agreement of the PV with the LV, the placement of NEG, and constraints in gapping.

In Persian, as seen in the previous sections, the LVCs all act the same with respect to agreement and negativization. The passive construction in Persian is obscure, and the inchoative construction which usually replaces it, never uses the PV as subject. Also, as described by Karimi-Doustan (1998), separability of the PV and the LV (gapping), is dictated by the type of nominal PV, and not necessarily its relation to the LV.

Extensive work has also been done on Urdu, a language highly related to Hindi. Butt (1997) mostly focuses on V + V predicates in Urdu, where the PV of the construction is another verb. This type of construction does not exist in Persian, however, the analyses of LVs in Urdu can provide some insight to the aspectual properties expressed by the LV. She shows that the LV of the construction contributes aspectual information such as inception, completion, and volition. The example below displays this phenomenon. The LV changes while the PV stays the same, and the only thing that changes is the notion of conscious choice (Butt, 1997, p 122).

vo ciik par-aa pronoun scream fall He began screaming suddenly (despite himself). us-ne ciik daal-aa pronoun scream put He screamed violently (on purpose).

The verb par ‘to fall’ expresses lack of conscious choice, whereas daal ‘to put’ implies volition. Similar aspectual alternations exist in Persian as was discussed above and as we will see in Chapter 4.

23

Japanese and Korean have LV-like verbal elements (suru and hata, respectively) that occur with nominal PVs. Grimshaw and Mester (1988) analyze suru as an LV. They show that it is the PV that contributes all the lexical content of the construction, and suru serves only as a verb marker. It does not affect the argument structure of the clause it appears in.

1.5 Chapter Review

The working definition of light verb in my study is the same as that used in most linguistic literature on the subject. That is, an LV is a semantically weak verb that occurs with a non-verbal element to express a verbal notion. The resulting meaning is often non-compositional, but not fully frozen either.

Following Butt & Geuder (2004), I distinguish between LVs and auxiliaries. In this study, I will not explicitly define any criteria that isolate LVCs in Persian, nor will I focus on their inherent syntactic properties. Rather, I consider an LVC to be where a verbal and non-verbal element combine to produce a novel meaning that is usually not totally predictable from the meaning of the parts. I presented some grammatical and lexical facts about Persian LVCs, discussed some issues that have been looked at previously and provided an overview of these studies. And finally, I briefly sketched a summary of a handful of studies that investigate LVCs in other languages. In the next chapter, I will highlight the main linguistic issues that pertain to my analysis of Persian LVCs.

25

2 FUNDAMENTAL THEORIES AND STUDIES

In this chapter, I will discuss a few linguistic and cognitive issues that are relevant to this thesis. Productivity, compositionality, polysemy, and constructions each come from long traditions in linguistic and cognitive theories. Here, I cannot delve into a comprehensive overview of the details of the different debates concerning these subjects, but merely provide some definitions and present older work as it pertains to this thesis.

2.1 Productivity

2.1.1 Definition

Coining new words or expressions on the basis of other words or established expressions allows for a potentially infinite number of expressions in natural language. The mechanism that allows for productivity4 is of immediate interest in the study of language and cognition. The predictability and systematicity of form or meaning that result from this mechanism are highly valuable for the language user.

The notion of productivity is one of the major contributions of generative grammar. A speaker who masters patterns and rules for combining words in a language can form novel multi-word structures. When this occurs, established patterns and rules allow speakers to generate new patterns. Different labels have been used to describe this phenomenon. Chomsky first referred to this notion as linguistic creativity. During the time when behaviorist doctrine was popular, which stated that human cognition was based on a stimuli-response interaction, Chomsky highlighted the capacity of a speaker to make use of “the intricate structure of specific rules and guiding principles to convey his thoughts and feelings to others, arousing in them novel ideas and subtle perceptions and judgments” (Chomsky, 1975, p 4).

Adams (1973, p 197) defines productivity in terms of a pattern that may be used as a model for new items if the occasion demands. Bauer (1983, p 18) posits the condition that a form can be used synchronically in word formation processes. Spencer (1991, p49) considers a rule to be productive when it is used actively in the creation of new words.

In his book Words and Rules, Pinker points out that “specifying kinds of words rather than a string of actual words, the rules allow us to assemble new sentences on the fly and not regurgitate preassembled clichés—and that allows us to convey unprecedented combinations of ideas” (Pinker, 1987: p 6). When forms can be built according to rules or other forms, a recognizable pattern emerges. These patterns can provide hints as to how productivity functions.

Patterns can occur at different levels of the language: the sentence structure (e.g. argument structure and word order), the clause or phrase level (e.g. verbal constructions and idiomatic constructions), or at the word level (e.g. morphological derivations).

Although no specific mechanism has been discovered for this powerful generative capability, we can expect that such mechanisms include analysis and synthesis based on predictable or established links between form and meaning. This systematicity would play an important role

4 Here, I do not refer to productivity and use of syntactic recursivity, which also falls under the label of productivity in certain works (e.g. Fodor, 2001). The present study pertains to morphological productivity.

26

in allowing children to bootstrap in language acquisition. The mechanism would likely depend on first generalizing and then refining the analysis and synthesis loop.

Experimental evidence shows that speakers extend existing patterns for new verbs. Only the combinations of known meanings can be used to build new concepts. Clark (1996) discusses the acquisition of certain grammar rules by children. She observes that general-purpose verbs, such as go, do, make, give, and put, are amongst the first verbs children learn crosslinguistically. These verbs coincide with the LVs of Persian, and seem to be cognitively salient across languages. Hence, these verbs might provide certain semantic and syntactic patterns that consequently proliferate throughout the grammar as it develops in a speaker. The constraints imposed by the meanings of these words will direct the productivity of related forms.

Certain phenomena in language are totally productive, and some are not productive at all. An example of a highly productive process is addition of the plural suffix, –s, in English. The addition of this suffix to any noun in English renders the form into its plural form, though a handful of exceptions exist. For instance, the addition of –ren to form the plural of child is totally non-productive. Other non-productive processes include the introduction of loan words into the language, along with many other isolated processes not carried out according to any patterns. Between these two extremes (total productivity and non-productivity), semi-productive patterns can be found. Semi-productivity occurs when a process displays systematic behavior, though the amount of productivity of the forms is constrained by semantic or other restrictions. The Persian verbal system is rich in this type of phenomenon. We will see examples of this in the next chapter and how it relates to Persian in section 2.2.3.

Most studies on productivity provide a quantitative method for measuring the productivity of a process in a language. These studies require massive amounts of corpus data, historical data, or both, and mathematical definitions of the process under study. The equations that calculate the level of productivity of a process involve the ratio of actual word forms and possible word forms, and provide a probability of the occurrence of a certain word form. A review of problems in the measurements of productivity can be found in Plag (2004).

Though the measurement of productivity is highly controversial, there is little doubt that speakers are equipped with some associative mechanism that allows them to generalize patterns and use these in linguistic expression. Productivity is a universal linguistic phenomenon. Since we do not have much quantitative data for Persian, I will rather focus on a qualitative description of the phenomenon as it relates to Persian. In the next section, I will discuss a recent statistical study done on LVCs in English. This study is of particular interest, since the goal of the researchers is similar to the goals of this study, though their access to abundant quantitative data allows them a different type of analysis.

2.1.2 A Study on Productivity

Productivity allows avoidance of redundant entries in the lexicon, or the stock of forms in memory. Using the knowledge that [-ed] added to a verbal root in English provides us with the past form in most cases, for example, is more efficient than storing the word ‘walk’ as well as the word ‘walked’ in the lexicon as separate entries. In fact, one needs only to store just those verb forms that are exceptions to the general rule (e.g. go/went). Apart from these morphological examples of productivity, where affixes make new forms, productivity can occur at the level of multi-word expressions. For example, the expression go + N where N is a word signifying ‘crazy,’ is a semi-productive expression in English (Goldberg, 1995). The

27

whole expression means ‘to become crazy,’ and the noun in this expression is constrained by semantic restrictions. Namely, only words with the meaning ‘crazy’ can occur in this expression. This set is potentially infinite.

In a statistical corpus study on LVCs in English, Stevenson et al. (2004) investigate the degree of acceptability of LVCs based on linguistic properties. They focus on “LV a V” constructions (e.g. take a walk, give a groan). Here, the V corresponds to the nominal complement in the construction. The goal of their study is to determine the degree that “words form a valid complement to a given light verb” on the basis of corpus statistics (p 2). One of the goals of this study is to eventually support the development of accurate computational lexicons. In English, the verbs take, give, and made are examples of LVs. These form LVCs when they occur in conjunction with particular types of nominal verbs. Examples of these forms include take a walk, give a talk, give a speech, and give a howl.

The authors hypothesize that the LVCs are not fully idiosyncratic, but exhibit systematic behavior; especially “class-based behavior—i.e., that the same light verb will show distinct patterns of acceptability with complements across different verb classes” (p 3). Through a statistical analysis of results from numerous Google searches on the internet and comparison to human judgment tests, they find that the acceptability of LVCs indeed depends on the semantic properties of the complement. The categorization of the complements according to these properties allows creativity and predictability in the language to a certain degree. The existence of these forms and the syntactic and semantic constraints that dictate the formation of patterns provides an example of semi-productivity in language. They find that the potential complements in LVCs are separated into semantically similar classes which are defined by English verb classes in Levin (1987)5.

This study focused on nominal complements derived from verbs, which Persian LVCs don’t allow. Thus, Levin-like classes cannot be used for an analysis of Persian. Even with Levin’s classes, the authors found that the classification is too coarse-grained, and other syntactic and semantic features need to be sought to be able to give a more precise account of this phenomenon. Fine-grained features are necessary for this type of analysis. In the next sub-section, I provide an overview of how LVCs are productive in Persian; a more detailed qualitative analysis will be provided in the next chapter.

2.1.3 Productivity in Persian LVCs

LVC formation in Persian is a semi-productive process. Not only have LVCs gradually replaced previously simple verb forms, but any novel verbal notion can only be expressed through this compounding process. The productivity of Persian LVCs is semantically based.

Persian LVCs are constructed by combining a PV with one of a set of 12-20 light verbs (LV). Each one of the most common of the LVs can be used to form around 400 verbal notions. As new concepts and terms have been invented and imported, the number of LVCs constructed by each LV has expanded. This continuous expansion and evolution has caused the “meaning” or semantic contribution of the LVs to become gradually dispersed and vague. Nonetheless, this expansion has not been random, but has been guided by semantic constraints on possible combinations. Importantly, the system continues to work effectively

5 Levin (1987) published a highly detailed descriptive work on English verbs. Rich in semantic and syntactic data, this book is often used as a reference of English verb classification. Relevant to studies like Stevenson et al., Levin shows that the semantics of a verb determines its syntactic behavior.

28

to maintain complete coherence for production and allows creation of new terms while preserving the old. Though historical facts about the evolution of the system are scarce, and there are no dictionaries or other sources that can provide accurate statistical data on LVCs. LVC formation has been a productive process in Persian for at least 10 centuries.

Unlike English, where nouns can often be used as verbs, as in ‘I will phone you,’ (the noun ‘phone’ appears as a verb to mean ‘contact by phone’), Persian speakers must always produce verbs by combining a non-verbal element with an LV.

Consider the following examples, all of which are recently created LVCs based on nouns (PVs) introduced into Persian from other languages:

email zædæn email ZÆDÆN send off an email telefon zædæn phone ZÆDÆN call on the phone losion zædæn lotion ZÆDÆN apply lotion bat mitzva gereftæn bat mitzva GEREFTÆN arrange a bat mitzvah hepatit gereftæn hepatitis GEREFTÆN catch hepatitis tayp kærdæn type KÆRDÆN type (on a computer) tayp zædæn type ZÆDÆN type (on a manual typewriter) mesaj dadæn message DADÆN send a message It is important to note that there are some non-LVC verbs in Persian. But many of these are scarcely used in everyday spoken language, even in formal settings, and are gradually becoming obscure and used solely in literary and poetic registers. Though there are a hundred or so very common verbs that are used frequently as heavy verbs (e.g. budæn ‘to be,’ xabidæn ‘to sleep,’ didan ‘to see,’ xændidæn ‘to laugh,’ etc). The following are examples of more obscure simple verbs which have been replaced by LVCs.

ku∫idæn ku∫e∫ kærdæn to try geristæn gerye kærdæn to cry (sob) avixtæn avizan kærdæn to hang (something) When a noun appears in the language through the introduction of a new idea, material, object, or method, it must be combined with an LV to be expressed as a verb. For example, when spray fixative was first introduced to the market, it was verbalized with zædæn ‘to hit’ rather than pa∫idæn ‘spray’. This was a natural selection as one of the constructions zædæn occurs in serves to express “controlled application of something onto another.” This and other similar examples provide evidence for the productivity of LVCs in Persian.

Initially, adult Persian learners find most of Persian LVCs idiosyncratic. To them, these constructions sound like idioms. However, as we will see, these forms are at least partially predictable and thus cannot be called idioms, or at least not of the frozen kind, which, by definition, have unpredictable meaning and are not productive.

While there are many LVs to choose from, the LV that appears in a novel LVC is never random. The predictability of the LV that appears in the expression of a particular notion is crucial to understanding how the LVs interact with the PVs. In order to express specific notions, PVs call on specific LVs.

Productivity is systematic. We observe that if one understands a whole predicate, then one can understand its constituents. The notion of systematicity of predicates and how it relates to

29

productivity is outlined in Ghadakpour (2002, p 185). However, as we have seen above with the LVCs, systematicity in productivity occurs at lower levels of linguistic expression than whole predicates. Since systematic phenomenon is potentially infinite (in this case, knowing the properties of certain constructions allows the production of an infinite set of new LVCs), there must be a productive mechanism giving rise to this human capability.

The search for the phenomena which brings out the contextual meaning of the LV in any LVC is our initial task. In other words, how is it that when the word telephone was introduced, telefon zædæn ‘telephone hit’ and telefon kærdæn ‘telephone do’ became acceptable, but telefon xordæn ‘telephone eat’ or telefon ændaxtan ‘telephone throw’ did not. Meanwhile, lazer ‘laser ‘can be used with zædæn, and xordæn, meaning ‘to shoot a laser at’ and ‘to be shot with a laser’, respectively.

In a study on event types in Persian LVCs, Haji-Abdolhosseini (2000) tries to “see to what extent one can predict what verb(s) a given noun combines with and what literal meaning is derived as the result of this combination” (p 8). He approaches this problem in a generative framework. He finds that only transparent LVCs are analyzable, and others are metaphorical and difficult to classify.

A problem in studying productivity in Persian is that corpora or data are not easily accessible. LVCs are not systematically listed in dictionaries. Searching the web for frequency poses problems since bracketed searches are not functional with Persian fonts and, more importantly, an LV can appear separated from its PV (as described in Chapter 1) in a given corpus.

Another problem with investigating productivity in Persian is that there are pragmatic and real world knowledge restrictions on the type of LVCs that can be constructed. Both these parameters are beyond the scope of most theories. However, the processes involved are systematic enough to allow some predictability and in some cases total predictability. Further discussion of the integration of this type of knowledge follows in the descriptions in chapter 3.

The patterns revealed by the LVCs in Persian will not necessarily be systematic or purely compositional, like the [-er] noun suffix of English. In Introduction to Cognitive Grammar, Langacker discusses this English morpheme in the word stapler. He explains that this English word does not simply mean ‘something that staples’ (which is what we would derive by computing staple + [-er]). The meaning of this composite structure cannot be totally rule-derived, but the word’s “organization and meaning are determined in large measure by the rule, even though it has properties above and beyond those the rule itself specifies” (Langacker, V.1: p. 15, 1990).

While classification is difficult, Persian can provide us with a large amount of data that has been in continuous usage for over fourteen centuries, giving us access to a robust system and allowing us diachronic comparisons and access to evolutionary steps as new words were introduced from other languages. Furthermore, the productivity of these constructions is neither purely syntactic nor morphological.

In the next chapter, I will first try to establish a strong and clear phenomenological description of the system, then I will discuss how the concepts which appear fit within current linguistic and cognitive theories, and how they support or disprove parts of such theories.

30

2.1.4 Conclusion

In this section I presented the notion of productivity, namely the systematic ability speakers have of constructing novel utterances based on established forms. I gave a summary of a recent study done on LVCs in English, and then an introduction to the type of productivity manifested in the Persian verbal system.

2.2 Compositionality

2.2.1 Definition

The meaning of expressions in natural language result from the interaction and sum of the meanings of the constituent elements. An unresolved issue in linguistics has been to define how the meaning of expressions is derived from the meaningful expressions within it. This type of systematicity is called systematicity of extraction, namely that if a whole expression is understood, then the elements that make up the expression are also understood (Ghadakpour, 2002, p 185). Searching for an explanation for systematicity gives rise to theories of compositionality.

Traditionally, philosophers like Frege (1892), one of the pioneers of the study of compositionality, are of the opinion that the meaning of an expression emerges directly from the meaning and structure of its constituents. Frege proposed the principle of compositionality which states that in a language L:

For every complex expression e in L, the reference of e in L is determined by the structure of e in L and the references of the constituents of e in L.

Montague (1970) uses this principle to formulate a related condition that there must be a strict mapping from syntax to semantics. The meaning of an expression results from a semantic function, parallel to the syntactic function of that expression. In other words, the semantic units (words) of a sentence are combined according to the operations dictated by the syntactic structure of the sentence.

I will not further discuss the extrapolations of this view, since there are many and it remains a controversial issue. Rather, I will present another view of compositionality before briefly citing examples from Persian.

One of the contested problems with the compositionality principle is that it assumes that each word in a complex expression has a stable and unambiguous meaning. As we will see in the next section of this chapter, on polysemy, word meaning cannot be defined exactly. Meanings change in different contexts and gain content from extralinguistic input.

Because of the inadquacies of describing the system as solely compositional, researchers have developed an alternate view, called gestalt compositionality . The following discussion on this type of compositionality comes from Fuchs & Victorri (1996, p 178-180).

Inspired by previous studies on visual perception and the role of gestalt figures in the visual domain, it can be deduced that the language faculty shares certain cognitive mechanisms with other cognitive domains. While maintaining that language has many domain specific properties, linguistic theories can be inspired by work in other domains.

31

Traditional compositionality extracts the meaning of an expression by first calculating the meaning of the constituents and then combining them (with a recursive method of disambiguation). One of the presuppositions of gestalt compositionality is that not all composite expressions are deconstructable. Other, global parameters come into play to disambiguate the meaning of the constituents. Traditional compositionality can be thought of as a sub-principle to this compositionality principle. Namely, traditional compositionality refers to cases where the constituents happen to be readily decomposable into separate units which contribute in a linear fashion to the expressions they are in.

In our study on the Persian verbal system, the shortcomings of the principle of compositionality emerge explicitly. An adequate framework is needed to account for the non-compositional meaning of the LVCs, the abundance of meanings that might be due to the LV and the productivity of the whole. In the next subsection, I provide some examples to illustrate how meaning surfaces from the interaction of elements and their respective construction.

2.2.2 Compositionality in Persian

The notion of compositionality pertains directly to the study of Persian LVCs: these constructions are comprised of two or more elements that combine and often give rise to non-compositional (in the traditional sense) meaning. Here, I’ll give a few examples which I will return to in the next chapter.

The complications of positing LVCs as compositional corresponds highly to the complications of positing that the LVs have a large number of meanings that aren’t necessarily related. For example, the verb zædæn can be used in a construction that means to transfer a substance into an object by means of a nozzle-like or sharp object. Here are some examples:

benzin zædæn gas ZÆDÆN pump with gas gazoil zædæn diesel ZÆDÆN pump with diezel bad zædæn wind ZÆDÆN fill with air (e.g. a tire)

Taken as compositional LVCs, these constructions would be the combination of the PV and an LV having a meaning similar to “transfering [PV] into [the direct object] through a pointy or nozzle-like object.” These meaning specifications are necessary because zædæn could not mean “to transfer,” but is more nuanced. The only acceptable PVs are ones that express substances that can be added to an object through a nozzle that goes into the receptor. In this vain, forms like *roqæn zædæn “oil hit” will not mean add oil to the car6, since oil is not added through a nozzle.

Positing such an idiosyncratic meaning is superfluous, no other known language has such nuanced verbs. Furthermore, the meaning of these constructions, as I will show in the next chapter, is related to the meaning of other constructions in the system. And also, not any PV can combine with this LV. Real-world knowledge plays a role in choosing the right PV for the construction. These restrictions aren’t captured in either the PV or the LV of these forms.

6 This form does exist in Persian, but it belongs to a different set of meanings, namely one of spreading a viscous material on a surface. This and other uses of zædæn will be discussed in the next chapter.

32

Hence, unless we posit detailed and highly nuanced definitions for each of the 40+ forms constructed with zædæn, the above constructions cannot undergo a compositional analysis. Instead, the meaning of zædæn emerges from the interaction between the construction and the type of PV that occurs in it. This is where ideas from gestalt compositionality can work to avoid either positing an abundant amount of meanings to zædæn or from having high levels of ambiguity.

To capture these linguistic and cognitive traits, the framework needed must accomodate idiosyncratic meaning and elements of gestalt compositionality. And, as seen in this chapter, the chosen framework must allow us to discover a basis from which LVC meaning emerges and is disambiguated, as well as a basis for productivity in the system.

In the remainder of this chapter, I present the idea of polysemy and how it is a problem for traditional compositionality. I will then present the basic ideas of construction grammar, a framework that helps in understanding comparable linguistic phenomena. On compositionality, Goldberg (1995, p 16) discusses similar ideas to gestalt compositionality:

By recognizing the existence of contentful constructions, we can save compositionality in a weakened form: the meaning of an expression is the result of integrating the meanings of the lexical items into the meanings of constructions. In this way, we do not need to claim that the syntax and semantics of the clause is projected exclusively from the specifications of the main verb.

2.3 Polysemy

2.3.1 Definition

Polysemy refers to instances where a word has two or more systematically related but different meanings. These related meanings are sometimes extended from a main, or central meaning of an expression. Simply listing separate definitions of a word does not shed light on the polysemy relations of the word. The study of polysemy involves exploring the relations between different meanings of a word. It involves seeking the explanation for motivations of extended or related meanings of a word. The understanding of these motivations will help in understanding how semantic information is organized and accessed in the lexicon.

Polysemy is the normal case, i.e. most words have more than one related senses. It occurs when there are several related, but distinct, meanings associated with a single form. For example, the word foot in English is multiply polysemous. It can be used in expressions like my left foot the foot of the mountain, the foot of the bed, on foot, two feet long, etc. In the first expression, foot means the bottom part of a leg, in the second, the bottom part of the mountain, in the third it refers to the floor at the end of a bed, the fourth to walking, and the fifth to units of measurement. These uses are not the same, but are clearly related.

The French counterpart pied has a similar, but not identical set of related senses. Compare au pied de la montagne, le pied de la table, etc. These are similar to the English (foot of a mountain, leg of a table). These meanings are also clearly related.

To take another example: the English verb walk is used with different but related senses in he walked all day, he walked the dog, the pitcher walked the batter, etc. Again, the French marcher is also polysemous, with related expressions such as rien ne marche, marcher sur la lune, marcher sur ses mains, etc.

33

Polysemy is not to be confused with homonymy or homophony, where two words happen to share the same form but have unrelated meanings. For example, the words piece and peace are homophones in English, as are voler (in the air) and voler (take what is not yours) in French. Their phonetic form is the same, while the meanings are totally unrelated.

2.3.2 Theories and Studies on Polysemy

In this section, I will discuss certain studies and views on multiple lexical meanings within different frameworks. This overview displays the different types of parameters that have been evoked in the past when studying polysemy. The main question these studies investigate is how to delineate the relations between multiple senses of a word. Though stemming from diverse theories, all the studies presented consider language to be related to and perhaps inseparable from other cognitive faculties. These theories belong to the category of theories that fall under the name Cognitive Linguistics.

The study of polysemy is crucial for determining the overarching principles that explain the formation and evolution of meaning in language. Extensions of the meaning of lexical items must be motivated for such a system to be robust. When extensions are motivated, meanings and different senses form patterns and a systematic structure. This structure allows for prediction of meaning and results from the non-arbitrary relations between meanings of a word. Also, with respect to acquisition, Lakoff (1987) argues that “it is easier to learn something that is motivated than something that is arbitrary” (p. 346). He further argues that it is easier to remember and use motivated as opposed to arbitrary knowledge.

As we will see in our investigation of Persian LVCs, the idea of motivated extensions of meaning is very important in the organization of semantic space. Most LVC uses of an LV can be defined using elements inherent to the core meaning of the LV as a heavy verb. In other words, LVCs constructed with zædæn ‘to hit’ can have meanings like applying, injecting, emitting, and wounding. I will show that there are certain parameters, linguistic and cognitive or perceptual, that underlie the various uses of this verb as an LV.

Some semanticists (e.g. Lakoff, 1989; Rosch, 1988) consider polysemy to be related to prototype based categorization, where each meaning of a word is part of a category that has a prototypical example. Wittgenstein (1953) suggests that concepts (lexical) form “a complicated network of similarities overlapping and crisscrossing: sometimes overall similarities, sometimes similarities of detail.” As we will see in the next section, different linguists have different approaches to extracting the relevant parameters inherent to a word’s meaning.

One of the assumptions of the cognitive linguistics treatment of polysemy is that language uses fundamental notional categories to organize meaning. In Women, Fire, and Dangerous Things, for example, George Lakoff speaks of idealized cognitive models where an ideal or prototypical instance of a concept is placed at the center of a set of related meanings. The related senses branch off from this core instance. Such a radial structure helps us characterize our knowledge of a particular word or category. In Lakoff’s models,

polysemy arises from the fact that there are systematic relationships between different cognitive models and between elements of the same model. The same word is often used for elements that stand in such cognitive relations to one another (p13)

34

Lakoff gives an example of such radial categorization with the Japanese affix hon (Lakoff, 1987) which is used to demarcate long thin objects, such as pencils, canes, sticks, etc. This initial description is considered the central case. From this core, other extended meanings radiate, forming a network of non-central cases. For this affix, these radial cases include trajectories (e.g. kick of a ball, throwing an object), activities done with long thing objects (e.g. injections, martial arts), activities similar to those done with long thin objects (e.g. judo, which doesn’t involve swords or staffs), and also communication (e.g. Japanese scrolls, telephone wires, TV programs, letters).

Lakoff provides several other examples from different languages of how items are organized in a radial diagram, where the central case unifies all other meanings. Brugman (1981), revisited in Lakoff (1987) investigates the English preposition over. The following sentences (from Lakoff, 1987, p 418-419), display the many meanings this preposition can take.

1. The painting is over the mantle. 2. The plane is flying over the hill. 3. Sam is walking over the hill. 4. Sam lives over the hill. 5. The wall fell over 6. Sam turned the page over 7. Sam turned over. 8. She spread the tablecloth over the table. 9. The guards were posted all over the hill. 10. The play is over. 11. Do it over, but don’t overdo it. 12. Look over my corrections, and don’t overlook any of them. 13. You made over a hundred errors.

Lakoff and Brugman show that the many uses of over can be related with respect to physical conditions to which the variants refer. Using the notions of ‘trajector’ (the object) and ‘landmark’ (the reference point, or what the object is over), Lakoff presents different variations corresponding to whether these physically touch each other, physically move, cover one another, etc. These different configurations form a relational diagram.

The physical instantiations of the senses of over shown in this study result in a network of meanings. In the study of polysemy, Lakoff (1987) and Brugman (1981) contribute the idea that the physical domain can serve as a basis from which relations between word meanings can be structured. Many different types of perceptual and cognitive parameters have been posited to represent the core or prototypical cases which lie at the center of these theories. Like Lakoff, Langacker (1990) and others cognitive linguists (Talmy, 2000; Fauconnier, 1993), claim that the basic semantic features of a word involve physical, dynamical, or force related parameters. Secondary meanings of these words result from metaphorical or abstract extensions of these primitive features. The existence of these features does not preclude traditional linguistic parameters, such as aspectual parameters or thematic roles, which occur in conjunction with these more cognitive parameters.

Followers of this sort of cognitive linguistics argue that regularities in the semantics of natural language can only be accounted for in terms of image schemata or schematic mental imagery. This trend rejects a formalist approach to semantics where language is treated as an independent system, in favor of an 'embodied' approach, according to which the meanings of

35

linguistic items are metaphorical extensions determined by our embodied experience in the world. In a different domain, Dehaene discusses the mathematical capacities of humans and remarks that an associative system like the brain needs images to function properly at an intuitive level (Dehaene, 1997). To ignore the imagery evoked by two syntactically different utterances would be to overlook an essential aspect of the data (Langacker, 1987).

Talmy (2000) posits force-dynamic parameters to account for different meaning structures. These parameters involve abstract notions such as causativity7 that can be extended from the physical domain to the psychological/social domain. The schemas in Talmy’s theory involve an actor and undergoer, and their default tendency (a state of rest or of action). The balance or the interaction between the two entities results in a force dynamic configuration, where there is usually a struggle between the states of two entities.

Brugman (2001) examines the relationship between light verbs and their full verb counterparts in English. She posits that light verbs retain the force-dynamic configurations of their full verb counterparts, though these often apply to the psychological, as opposed to physical, domain. Apart from differences in force dynamics, she uses semantic role assignment and syntactic structure to differentiate different instances of polysemous LVs. She suggests that the different meanings of light verbs in English can be characterized in terms of the distinctions of force-dynamic action or interaction. For example, in terms of where the energy for the action begins and ends, the classes of arguments for take include (Brugman, 2001, p 568):

Sensation and perception (subject referent = energy source and energy sink) take a whiff / sniff / taste / sip take a look / glance / glimpse

Self-oriented action (subject referent = energy source and energy sink) take a walk / shower / spill / fall / rest

Other-oriented physical action (subject referent = agent/energy source) take a stab / punch / swipe

Other-originated physical action (subject referent = patient/energy sink) take a hit / beating

Brugman’s examples display one way to categorize the instances of the verb take in English. She uses force-dynamic and also linguistic parameters to define the basic or core characteristics of the prototypical cases of take. Thusfar, the theories and studies presented have put forth different cognitive and linguistic parameters to explain the elements present in the core instance of an item. In other words, these studies all assume the existence of a core from which a network of related meanings are constructed (through metaphorical or other types of extensions). In their Theorie des Formes Sémantiques, Cadiot & Visetti present a theory in which language, and especially lexical semantics involves three different levels: themes, profiles, and motifs. Words are not based on a prototype instance. Each word has a series of motifs, or traits, which convey a potential for all the different possible uses of that element in context. These motifs are inherently unstable and are only stabilized when an expression appears in

7 Causativity, where an agent causes or forces a patient to do something or be in a particular state, is discussed in Talmy (1988, p 67). He claims that the traditional primitive notion of causation can be captured in a force-dynamic framework as being a set of patterns, or different combinations of the primitive features.

36

context. In context, the stabilized motif serves to express one of the profiles of the word. The theme component becomes relevant at the level of discourse and pragmatics.

Analyzing prepositions and some other polysemic fully lexical words in French, Cadiot & Visetti argue that the motifs of a word are not necessarily physical in nature. Rather, the features emerge from perceptual attributes and possible resulting action on the physical world, such as apprehending, displaying, anticipating. In other words, motifs do not belong to a specific domain, but concern many domains related to qualitative and modal ways of perceiving.

For example, in their analysis of the word école ‘school’ in French, they posit that there are three different axes on which the motifs can be placed. First, related to telicity, their parameters include program, project, and perspective of transmission. Then, related to configuration, école expresses an agglomeration, or an assembly. Also included in the motifs of école are the notions of seriality (time), and conformity. These motifs encompass physical and temporal factors as well as social and psychological ones.

Cadiot uses the same type of analysis in his articles on the French verbs jouer ‘to play’ (Cadiot, 1999) and monter ‘to climb’ (Lebas & Cadiot, 2003). Though at first glance, the variety of meanings associated with each of these verbs suggests a case of homonymy, the intuition of the native speaker solicits a more sophisticated analysis. For jouer, Cadiot’s method consists in classifying the different meanings of the verb primarily (but not exclusively) with respect to the type of preposition it occurs with. Motifs such as orientation towards an object, compositionality, intention, and controlled activity are amongst those he puts forth in explaining the differences in meanings.

For monter, Lebas & Cadiot (2003) analyze differences in meaning that can be extracted over linguistic tests. For example (p 11), when monter occurs with the auxiliary AVOIR in the past, it expresses a dynamic action of ‘climbing’. Whereas, when the verb occurs with the auxiliary ETRE, it expresses a static view of the action, which is difficult to translate into English, but might be equivalent to ‘had mounted.’ Motifs that he allots to this verb include ascending movement, elevation, inchoation, augmentation, assemblage, and combination. By completing this classification, Cadiot wishes to show that the essence of semantic traits of the notion expressed by the verb monter or any other linguistic utterance is qualitative in nature and cannot be based on purely linguistic or purely physical parameters.

Cadiot & Visetti are interested in the dynamic interaction between motifs of a word and the context in which they appear. However, their work is still work in progress, and they present no systematic way of applying their theory.

Regardless of the approach taken, it is a well recorded fact that polysemy lies at the core of any linguistic system. Whether the network of meanings results from an extension from a prototypical instance or from the instantiation of selected inherent properties, the disambiguation of lexical items is a fundamental process of language cognition. In this thesis, I do not confine the data within the bounds of either of the above approaches. Rather, I am inspired by both types of theories and attempt to provide an analysis that will efficiently predict the meaning of an expression as well as extract relevant cognitive parameters that emerge.

Fuchs and Victorri (1996) present a novel computational approach to the disambiguation of polysemic units which they name gestalt compositionality (see discussion in the section on

37

compositionality). The polysemic unit interacts dynamically with other units in the sentence in order to lessen the amount of possible ambiguity of a sentence. They base their computational model on ideas from dynamical systems theory and describe the French adjective sec as an example, amongst others. Using a synonym dictionary, they find relations of proximity between synonyms of this word, and project the results onto a two dimensional geometric space that represents all the different possible meanings.

Different types of interpretation (i.e. ambiguous, precise, and indeterminate) can be distinguished by the shape and size of the function that results from the contextual cues co-occurring with the word. Their methods provide intriguing results for computational linguistics and modeling natural language. Their mathematical analysis allows a clear graphical display of the interconnected networks of meaning constructed around words in language and provides a preliminary sketch of a method to extract meaning in context. Unfortunately, similar methods for Persian do not exist due to the scarcity of analysable electronic material available, and the lack of synonym dictionaries that include LVCs.

To create an interconnected network of meanings allotted to a form requires particular cognitive abilities. Namely, it requires the speaker to extend the meaning of a form in an organized and predictable manner. The speaker or hearer must construct or reconstruct the intended meaning of a form by inferring its position in a semantic space. The Persian verbal system seems to intensively solicit such a capacity.

2.3.3 Polysemy and Meaning in Persian LVs

Since almost all verbal notions are constructed on the basis of a limited set of LVs, meanings for these LVs have become vague and difficult to define. Intuitively, the large sets of LVCs constructed with a given LV share certain properties. However, the semantic contribution of the LV is not concrete, but very general. The original heavy meaning of the verb is backgrounded in LV instances and serves as the core meaning.

If the LVs in Persian are bleached instances of their heavy counterparts, then polysemy should not be assigned to the LV itself. On the other hand, polysemy cannot be attributed to the LVC as a whole, unless it is to express the type of polysemy present in English for words like bite or French mordre. In Persian, gaz zædæn ‘to bite,’ is an LVC constructed with the LV zædæn. This LVC can be used, as in English and French, to express notions like fish biting bait, a dog biting the mailman, or someone biting his lip. However, this type of polysemy (or polyreferentiality) occurs at the level of the LVC, not the LV. I am interested in the dynamics of the meaning of the LV.

The full verb zædæn can be translated most concisely by the English verb ‘to hit.’ However, when it is used as an LV, the meaning of the construction often deviates, sometimes strongly, from this original meaning. A sample of the many possible LVCs constructed with this LV appears below.

tar zædæn guitar ZÆDÆN play the guitar dad zædæn yell ZÆDÆN yell qædæm zædæn step ZÆDÆN stroll ni∫ zædæn sting ZÆDÆN sting felfel zædæn pepper ZÆDÆN add pepper (to food) maye zædæn mix ZÆDÆN prepare ræng zædæn paint ZÆDÆN paint

38

∫æm∫ir zædæn sword ZÆDÆN strike with a sword ampul zædæn syringe ZÆDÆN administer a shot foru∫gah zædæn store ZÆDÆN opening a store tohmæt zædæn accusation ZÆDÆN accuse gaz zædæn bite ZÆDÆN bite jib zædæn pocket ZÆDÆN pick a pocket mævad zædæn drugs ZÆDÆN take drugs jævane zædæn sprout ZÆDÆN sprout t∫æmæn zædæn grass ZÆDÆN trim grass hærf zædæn word/sound ZÆDÆN talk As can be seen in the above set of LVCs, the LV zædæn can lose most of the semantic features related to the verb ‘to hit.’ For example, there is no hitting (in a non-metaphorical sense) when you stroll or add pepper. Here, each of the LVCs can be considered as polysemous, in the fashion of gaz zædæn discussed above, but this is not due to the meaning of zædæn. In Persian, this verb is used in LVCs that express notions such as applying, trimming, attacking, build, play, prepare, do a repetitive movement, and emit sounds.

Hence, these notions are not assigned to the LV zædæn or to a particular LVC constructed with zædæn. If zædæn was assigned the vast number of different meanings that can be extracted in this way, it would mean that the LV can be used in any expression that evokes that context. However, novel uses of zædæn in contexts referring to applying, trimming, attacking, etc, are constrained by the type of PV uttered and also by grammatical restrictions. These senses emerge from the interaction of zædæn with a particular type of PV.

My main goal in this study is to provide a quasi-formal method to differentiate these meanings. Each LV produces families of constructions that portray different but related meanings. The details of the analysis will be discussed in the next chapter.

Another common LV in Persian is gereftæn ‘to get, obtain, catch.’ The notions expressed with the help of this LV include hiring, removing, choosing, attaining a quality or trait, understanding, hunting, and setting up.

kargær gereftæn worker GEREFTÆN hire a worker zæn gereftæn woman GEREFTÆN get a wife mahi gereftæn fish GEREFTÆN fish naxon gereftæn nail GEREFTÆN clip the nails taxi gereftæn taxi GEREFTÆN take a taxi æks gereftæn photo GEREFTÆN take a picture du∫ gereftæn shower GEREFTÆN shower sor’æt gereftæn speed GEREFTÆN speed up ræng gereftæn color GEREFTÆN get colored, tanned ævæzi gereftæn wrong GEREFTÆN mistake a thing for another donbal gereftæn follow GEREFTÆN follow up gæt∫ gereftæn cast GEREFTÆN set in a cast tæmas gereftæn contact GEREFTÆN contact ∫ohræt gereftæn fame GEREFTÆN become famous bu gereftæn smell GEREFTÆN get smelly xak gereftæn dust GEREFTÆN get covered with dust

39

Some of the LVCs clearly display a derivation of the original meanings of the LV. For example, in mahi gereftæn, there is the idea of catching a fish. Or less concretely, ∫ohræt gereftæn, there is the idea of receiving a trait or an attribute.

Yet another list of LVCs demonstrates the polysemy of the LV ændaxtæn whose full meaning is ‘to throw.’ Here, the notions expressed include installing, setting, and throwing.

gir ændaxtæn stuck ÆNDAXTÆN cause problems rah ændaxtæn path ÆNDAXTÆN start up, ignite ja ændaxtæn place ÆNDAXTÆN set into place, install dur ændaxtæn far ÆNDAXTÆN throw away, discard jelo ændaxtæn front ÆNDAXTÆN set forward serke ændaxtæn vinegar ÆNDAXTÆN set vinegar t∫eft ændaxtæn lock ÆNDAXTÆN hook the bolt of the lock ∫i∫e ændaxtæn glass ÆNDAXTÆN install a window fær∫ ændaxtæn rug ÆNDAXTÆN spread out or unfold a rug tir ændaxtæn shot ÆNDAXTÆN shoot a bullet nur ændaxtæn light ÆNDAXTÆN shine light on something Some of these forms clearly show motivation for the use of this particular LV. For example, dur ændaxtæn means ‘to discard.’ This LVC can be used to express discarding something that doesn’t involve throwing. However, the original use of the PV dur ‘far,’ is probably due to the fact that when you throw something far, it is often entails that you are getting rid of it permanently. In fact, the same is true of the English expression throw away and the French jeter which can have the meaning “discard.” In English, the meaning of throw is not compositionally part of the meaning of throw away although it clearly motivated the expression (i.e. he threw away his letters doesn’t entail he threw anything, maybe he just marked them for disposal, or picked them up and gently placed them in a rubbish bin). The meaning of the whole expression is more general and can be considered in some sense as a metaphorical extension of the original meaning. This relation allows speakers to learn the meaning of this form with more ease.

The relation between the two elements of these constructions is not strictly compositional, but stems from metaphorical or motivational reasons (extensions). Since the LVCs can be non-compositional in many cases, the meaning of the LV need not be “bleached.” If the meaning of the LV were bleached, each form could be derived compositionally, which is not the case.

As mentioned earlier, native speaker intuition suggests that these diverse meanings are not accidental. In the next chapter, I will show that the different meanings associated with the LVCs constructed with a given LV are organized in such a way as to avoid most ambiguities that might arise, despite the myriad of possible meanings allotted to the LV. Furthermore, this organization can give rise to productivity in the system. The framework which seems adequate enough to incorporate natural language phenomena, like LVCs, is Construction Grammar, the topic of the next section.

2.3.4 Conclusion

In this section I presented the notion of polysemy, or instances where an expression has a series of systematically related meanings. I provided a sketch of some different cognitive approaches to the study of word meaning. And finally, I presented lists of LVCs with a

40

common LV to display the type of extensions of meaning that are possible within each set of LVCs. I briefly discussed and rejected the idea that LVs in Persian are polysemous, since this would entail many different definitions allotted to each LV. These LVs are not bleached, either, since many of the meanings of the resulting LVCs are motivated and not derived compositionally from the LV.

2.4 Constructions

In the three previous sections, I presented three issues relevant to the study of Persian LVCs: productivity, compositionality, and polysemy. Several questions emerge immediately when faced with data from the Persian verbal system. How do Persian speakers organize semantic space of each LV in order to resolve the ambiguity that would result from their wide usage? Is there an algorithm that can be applied to the system to predict the meaning of a new verb form? How are new verb forms constructed and understood?

In search of the most useful means available to help in investigating answers to such questions, cognitive linguistic theories provide us with valuable tools. Theories of this type focus on the cognitive processes involved in language processing and understanding. Discussed briefly in section 2.1 in the studies on polysemy, theories that fall into this category consider the role that language plays in the conceptual and cognitive system. Rather than focusing on abstract and formal structures in language, cognitive linguists incorporate other areas of cognition into the study of language. These perceptual and conceptual parameters include space, time, and force-dynamics. In this view, language is a means through which humans, confined to physical bodies in particular environments, are able to express a range of phenomena, both perceived and conceived. In other words, language must be explained through other capacities other than just purely linguistic ones.

Language comprehension requires a detailed mental model of the linguistic information provided or uttered. In other words, linguistic as well as real-world knowledge play a role in constructing an accurate interpretation of an expression.

Amongst the variety of cognitive linguistic theories, Construction Grammar (Fillmore and Kay, 1996; Goldberg, 1995) seems most appropriate to account for the data of Persian. The fundamental units of language are pairings of form and meaning. Formal theories only take into account the abstract structure of linguistic units and do not necessarily include all the semantic and pragmatic information associated with the structure. The pairings between form and meaning are often called constructions. These can be as small as words or affixes, or they can be whole sentence structures. Crucially, constructions include both universal and general knowledge as well as idiosyncratic and language-specific information.

As will be evident in the following sections and in the next chapter, the present analysis, while inspired by this approach, does not strictly adhere to the types of constructions discussed so far in the literature. Before I discuss how this theory relates to Persian, I will give an overview of the basic tenets of the original formulation of this theory.

2.4.1 Construction Grammar

Construction Grammar (CG) is a linguistic theory that assumes that language consists of a repertoire of complex patterns, or constructions, that integrate form and meaning in conventionalized and often non-compositional ways. In a construction, form is the combination of syntactic, morphological, or prosodic patterns, and meaning, in a broad sense,

41

includes lexical semantics, pragmatics, and discourse structure. A grammar in this view consists of intricate networks of overlapping and complementary patterns that serve as ‘blueprints’ for encoding and decoding linguistic expressions of all types (Goldberg, 1995).

The basic tenets of CG include the following (Goldberg & Jackendoff, 2004):

1- There is a cline of grammatical phenomena from the totally general to the totally idiosyncratic.

2- Everything on this cline is to be stated in a common format, from the most particular, such as individual words to the most general, such as principles for verb position, with many subregularities in between. That is, there is no principled divide between “lexicon” and “rules.”

3- At the level of phrasal syntax, pieces of syntax connected to meaning in a conventionalized and partially idiosyncratic way are captured by constructions.

According to CG, a construction exists if one or more of its properties cannot be strictly predicted from the inherent properties of the lexical elements within it. The description of a construction specifies the types of elements that can occur within it, as well as how they will interact with the construction. In relation to idiosyncrasy, constructions are similar to idioms. Idioms are listed in the lexicon with a syntactic structure, a meaning, and often phonological information.

Some constructions are easy to spot because of their odd syntax. For example, the term let alone appears as part of a particular construction (e.g. She can't boil potatoes, let alone cook a meal). Other constructions, however, do not have marked elements, but impose strict constraints on the syntax of an expression (e.g. the resultative construction, the ditransitive construction). Most aspects represented by constructions are language-specific and cannot be explained by universal linguistic parameters.

Importantly for the present study, in verbal constructions, the verb alone does not determine the argument structure of the sentence. Rather, the argument structure results from the composite effects of the verb and the construction. This property of constructions allows for great reduction of verbal polysemy in the lexicon.

Most properties of constructions are based on semantic and syntactic information. However, CG also leaves room for real world knowledge and its effect on construction meaning. Real world knowledge includes information a human has of how the world functions and what types of actions are pragmatically plausible.

Constructions can occur in “families,” where a group of constructions share some syntactic and/or semantic properties. These families can be partially productive, based on semantic or pragmatic parameters. No claim is made that constructions or the parameters used to construct them are innate or universal, although presumably there are strong universal constraints.

CG, like other forms of cognitive linguistics, considers meaning construction and comprehension as cognitive mechanisms and formalisms to be grounded in other brain processes. The cognitive parameters implicated in these constructions include embodiment, cultural or social issues, and other cognitive traits of the agents who use them.

42

CG provides concise tools to describe polysemy, productivity, and non-compositional or idiosyncratic semantics. These properties are central to the Persian verbal system. In the next section, I give an overview of the two studies done on Persian within a CG framework, and section 2.4.3, I explain further reasons why this framework provides interesting tools for the current analysis.

2.4.2 Previous Studies on Persian

In two of her studies, Goldberg investigates different aspects of Persian LVCs within a CG framework. In both, she discusses how Persian LVCs act in some ways as single words and in other ways as multi-word expressions. The assumption that Persian LVCs are constructions allows a certain flexibility in the statement of the syntactic restrictions they undergo.

In her discussion of the future marker (Goldberg, 1996), she uses tools from Optimality Theory (Prince & Smolensky, 1993) to account for the patterns observed in the data. She discusses the future marker, the auxiliary xastæn ‘to want,’ which necessarily occurs between the PV and the LV, as do some other auxiliaries (as discussed in Chapter 1).

Ali telefon xah-æd kærd. Ali telephone want.PRS-3SG do.PST. Ali will call. * Ali xahæd telefon kærd. Ali want.PRS-3SG telephone do.PST.

She first posits a constraint CPvo, which states that the Xo (lexical elment, here the PV) and the Vo (LV) of an LVC are expressed as a Vo (a single verbal unit). This constraint requires the LVC to be considered as a single unit. Then, to account for the future, there must be a higher ranking constraint, FUT, which requires the future auxiliary to appear directly before the verbal root. The first constraint serves to disallow free separation of the PV and the LV, since they are considered to be a single unit. And the FUT constraint, which requires the future marker to appear between the two elements, must be ranked higher (be more important and override any lower constraints) than the first constraint. This results in the correct positioning of the future auxiliary and the non-grammaticality of the second sentence above.

Another piece of data considered in this work is the post-PV positioning of the DO8 clitic (this property is discussed in section 1.3). She posits a constraint which says that DO clitics are suffixed after the PV. However, this suffix can also occur post-LVC, depending on stylistic differences. Both of the following forms are grammatical.

mæsxære-æ∫ kærd-ænd. joke-3SG do.PST-3SG. They mocked him. mæsxære kærd-ænd-æ∫ joke do.PST-3SG-3SG. They mocked him.

8 DO = Direct Object

43

To account for this, Goldberg suggests that the DO constraint is not ranked (given necessary priority) over the CPvo constraint which calls for the inseparability of the two elements of the construction. This free floating constraint will yield the first sentence above if placed before the CPvo constraint, and after to yield the second.

Furthermore, to restrict the type and number of elements that can possibly separate the LV from its PV (in cases where the CPvo constraint is violated), another constraint, CPv, is introduced. This constraint states that the LVC prefers to remain under a v-bar (in other words, not separated by non-morphemic elements).

In conclusion, she suggests that Persian LVCs are stored as lexical items, though they can later be separated. CG analyses do not entail that a stored entity has atomic syntactic status. Rather, the unit-nature of the LVCs is determined on the basis of what other constructions and constraints it interacts with.

In a later article (2004), Goldberg pursues the idea that Persian LVCs are words by default. She revisits the previous data, but provides a dependence hierarchy analysis instead of an Optimality Theory analysis. This analysis more efficiently captures the existing patterns, without having to posit a separate constraint each time an element intervenes between the LV and the PV. (See Goldberg (2004) for further details.)

Goldberg is interested in the word or phrasal status of Persian LVCs in general. The present study focuses on constructions at a finer-grained level. The constructions depicted in this study are specific instantiations of the more generic constructions presented by Goldberg. These constructions include specific overtly expressed LVs and particular types of PVs. The actual LV is included in the formal description of the construction. Different patterns in LVCs have different types of meaning, or semantic specificities, as will be seen in the next section and next chapter. In CG, “a hierarchical network of constructions clearly enables the theory to be in principle fully descriptively adequate” (Goldberg, 2004, p3). Constructions higher in the hierarchy represent general patterns, and lower level constructions account for idiosyncratic and language-specific information.

Inasfar as the LVCs are constructions, I consider them to be units. Their semantics can range from transparent to opaque and some are more separable (syntactically) than others. These differences between particular LVCs do not effect their participation in the system. The LVCs in this study might have different linguistics behaviors, but they are all treated as the same type of phenomenon in regards to meaning construction and disambiguation. Our goal is to understand how the LVs function within the context of these constructions.

2.4.3 Persian Constructions

In this study, we are interested in how Persian speakers disambiguate LVs when they appear in different LVCs. From this, we might also begin to stipulate how new verbs are produced and understood by these speakers. The tools made available through CG allow us to capture these and the idiosyncratic and often non-compositional semantic properties of LVCs in Persian. Since this study is concerned with specific verb meanings that result from the combination of a specific LV with a type of PV, the constructions we deal with occur at a very language-specific level. This doesn’t mean that the same type of patterns won’t be found in other languages.

44

Here, I will briefly present a couple of examples to illustrate the type of phenomenon I will explore in the next chapter. Taking a specific LV, ke∫idæn who’s heavy verb meaning is ‘to pull,’ a variety of different meanings can be observed from its combination with different PVs.

dad ke∫idæn shout KE∫IDÆN shout qælyan ke∫idæn hooka KE∫IDÆN smoke a hooka feraqæt ke∫idæn separation KE∫IDÆN suffer from longing ælæm ke∫idæn banner KE∫IDÆN hoist a banner æbru ke∫idæn eyebrow KE∫IDÆN draw in eyebrows The above meanings seem to be unrelated. However, when faced with all the different possible LVCs constructed with ke∫idæn, patterns in the data emerge. Some of the different meanings seem to be highly related and form clusters that are defined by the type of PV in the construction. For example, if the PV is any smokable substance or an instrument that can be smoked out of, the resulting meaning will mean to smoke that substance or out of that instrument, respectively.

hæ∫i∫ ke∫idæn hash KE∫IDÆN smoke hash sigar ke∫idæn cigarette KE∫IDÆN smoke cigarettes qælyan ke∫idæn hooka KE∫IDÆN smoke a hooka pip ke∫idæn pipe KE∫IDÆN smoke a pipe On the other hand, if the notion expressed involves building or setting something up across a significant length in space, the PV will be the object that is drawn out and the resulting meaning stabilizes.

jade ke∫idæn road KE∫IDÆN make a road divar ke∫idæn wall KE∫IDÆN build a wall nærde ke∫idæn fence KE∫IDÆN put up a fence Ke∫idæn has about twenty different constructions in which it can appear each with a specified type of PV. Likewise, each LV gives rise to a series of constructions. These constructions form a network of different meanings which allow some predictability and some basis for productivity.

Furthermore, the networks built around a single LV share nodes with networks built around other LVs. For example, any PV expressing a lesson or type of advice can be combined with gereftæn ‘to get’ to express receiving advice or a lesson.

dærs gereftæn lesson GEREFTÆN be taught in class sæboq gereftæn precedence GEREFTÆN learn a lesson pænd gereftæn advice GEREFTÆN get advise The same PVs can be used with dadæn ‘to give’ to express giving the lesson or advice.

dærs dadæn lesson DADÆN teach in class sæboq dadæn precedence DADÆN give a lesson pænd dadæn advice DADÆN advise

45

These two clusters display an agent/patient alternation that can be predicted from the full verb meanings of the LVs. Other clusters with these and other LVs also display this sort of general alternation that can be explained through linguistic terms. And there are yet other alternations that require a more detailed idiosyncratic description. These alternations allow for further predictability in the system and are revisited in Chapter 4.

As we will see in the next chapter, I use the basic ideas of CG, but propose that the constructions that occur at the level of LVCs have much more idiosyncratic information in them than the constructions usually presented in CG studies. This information includes the explicit LV and a specification of semantic properties relevant to the PV.

2.4.4 Conclusion

In this section, I presented the fundamental ideas of Construction Grammar. I also presented two studies by Goldberg (1995, 2004) relevant to LVCs in Persian. I then gave a quick sketch of the type of patterns that need to be accounted for in the LVC data.

2.5 Chapter Review

In this chapter, I presented some basic theoretical notions pertaining to the data to be treated in this study, most importantly, productivity, compositionality, and polysemy. The theories that seem most adequate in describing the data are those which consider language in relation to other cognitive processes, namely those of cognitive linguistics. From the many theories available, Construction Grammar seems to provide the best tools to account for this data. In CG, there is no strict line between lexical and phrasal elements. Furthermore, CG permits the use of real-world knowledge in the construction of linguistic utterances. Polysemy, productivity, and non-compositional aspects of semantics can be described adequately within this framework. In the next chapter, I introduce “islands” of constructions form the basis of this verbal system.

47

3 ISLANDS

3.1 Core Meanings and Islands

3.1.1 Core Meanings and Islands

At this point I introduce the idea of a core meaning for the LVs. Thus far, the LVs have been translated into English with respect to their core meanings. For lack of a better way to translate the meaning, the core meaning will sometimes be used, even in LVC contexts. However, in most of the glosses, it will be standard to write the LV in small capitals to show that the full verb, or core meaning, does not emerge fully. By the core meaning I simply refer to the meaning which a Persian speaker will utter as the “meaning” of a verb when asked out of context. This meaning occurs in non-LVC contexts. However, as an LV, this meaning has developed diffuse ranges usually motivated by the core meaning.

We can expect to find this core meaning to be the base for the extended and modified meanings of the LVs. The core meaning of the full verb provides certain skeletal features which reemerge partly or fully in most of its LV instantiations. This study focuses on the fourteen verbs listed below, in order of usage frequency9 along with their original, full verb meanings10.

kærdæn to do zædæn to hit dadæn to give gereftæn to get da∫tæn to have amædæn to come aværdæn to bring xordæn to eat ke∫idæn to pull ∫odæn to become bordæn to take ræftæn to go ændaxtæn to throw oftadæn to fall The non-core meanings of the LVs cannot necessarily be directly predicted from the core meaning, none the less, we can almost be certain that they are not arbitrary. Similarly, idiomatic LVCs don’t have predictable meaning, but a relation between the idiosyncratic meaning and the core meaning exists (see chapter 5 for a discussion on idioms). In the descriptive section of this chapter, we will see the limits to which the core meanings are extended in LVC contexts.

9 Karimi-Doustan, 1998, p 83. 10 These English translations also serve the purpose of expressing (in an approximative manner) general linguistic traits that are shared cross-linguistically. Though the LV version of these verbs might deviate greatly from the core meaning, these translations provide a base from which these extensions can be observed.

48

3.1.2 Defining Constructions Through Complements

In Why can you have a drink when you can’t *have an eat? (1982), Wierzbicka examines the have a V construction in English. This construction consists of the LV have and verbal noun. For example, have a walk, have a swim, have a taste, have a bite, have a look, etc. She suggests that the grammaticality of constructions like have a drink or have a try as opposed to *have an eat or *have a manage is due to the repeatability of the actions, i.e. enumerative vs continual.

One of the major differences between the English and the Persian cases is that in English, the LVC usually has an analogous simple counterpart. For example, have a walk has the counterpart walk or have a swim has the counterpart swim. In her study, Wierzbicka tries to explain the semantic differences of these two analogous forms. This study investigates the difference in the usage of the particular Persian LV (and not the complement as in English) and it’s heavy counterpart.

Wierzbicka’s analysis provides some interesting insight into the different possible semantic variants of the have a V construction and her methods of description will be useful in describing the variants of LVCs in Persian. She also looks at semantic invariants of the have a V, providing valuable insight to the types of extensions that are possible, delineating what can be considered as a type of core construction.

She begins by roughly defining the type of constructions she is studying and by giving an initial idea of the semantic invariant of the construction. The essence of her study consists of examining the different sub-types of the have a V construction. According to her study, there is a minimum of ten different constructions associated with the LV have. She posits the following structure as the semantic invariant, found in all possible have constructions. It is the core construction.

X had a V => for some time, not a long time, X was doing something (V) it could cause something (good) to happen in X

that nobody else would know about X was doing it not because X wanted anything to happen to

anything other than himself X could do it more than one time

Wierzbicka does not use a conventional meta-language to describe this construction. Rather, she uses a language that is “in principle intelligible and intuitively verifiable” (p 300). This alleviates the burden of defining linguistic notions in minute detail, and provides a concise, though at times fuzzy definition of terms.

The ten constructions presented share the above general structure, but each have one or several construction specific nuances. These constructions are partly defined by the type of complement they receive. For example, two separate constructions are consumption of small parts of objects which could cause one to feel pleasure and action aiming at perception which could cause one to know something and which would not cause one to feel bad if it didn’t (a. and b. below, respectively).

49

a. (e.g. have a bite, a lick, a suck, a chew, a nibble) X had a lick (of Y). => for some time, not a long time, X was doing something (V) taking through his mouth into his body small parts of Y it could cause X to feel something good (when X was doing it) X was doing it not because X wanted to cause anything to

happen to anything other than himself X could do it more than one time. b. (e.g. have a look (at), a listen (to), a smell (of), a feel (of), a taste (of)) X had a listen (to Y). => for some time, not a long time, X was doing something (V) it could cause X to come to know something about Y X would not feel anything bad if X did not come to know

anything about Y because of doing it X was doing it not because X wanted anything to happen to anything other than himself X could do it more than one time.

As seen in a and b, the elements of the core construction emerge in both constructions. The existence of each construction, with its specific nuances, allows for fine grained prediction of the meaning of particular have a V forms in English. If the complement of the LV belongs to one of the categories defined in a construction, the meaning of the whole can be predicted from the construction. The existence of these types of constructions also allows for productivity, since the type of complement that occurs with have in each of the constructions is highly specific and defined within the construction. One can imagine making new forms based on this structure, using similar complements.

Wierzbicka provides a detailed sketch of the different have a V constructions and then compares them with each other. At the end of her study, she also compares these constructions with those of take (take a V). Certain complements can occur with both of these verbs. A similar phenomenon exists in Persian, which I will present and discuss in the next chapter.

Another study related to this issue deals with the productivity and acceptability of certain complements with a given LV (Stevenson et al., 2004). They found that the complements that occur in particular constructions belong to a particular semantic class. In other words, the constructions are again found to partially depend on and partially determine the type of complement that combine with the LV. The details of this study were presented in section 2.2.2. The important point to retain from these studies is that constructions can include lexical items (e.g. LVs), and they can also call for very specific, sometimes idiosyncratically defined, types of complements (e.g. type of PV).

3.1.3 Persian

In analyzing the list of LVCs of a given LV in the corpus of Persian, my first task has been to discover semantic resemblances and sets of constructions, as with the different have constructions in English in the above study. At first glance, the LVCs of a given LV might seem to be intuitively related or maybe totally unrelated. Variation of the type of PV, as defined by their common properties, results in a variation of the meaning of the LVC.

50

Conversely, variation of the LV also results in the variation of the meaning of the LVC. Detailed investigation shows that sets of LVCs actually share certain attributes. It is essential to look both for PVs that accept multiple LVs and single LVs and their possible PVs to obtain clues as to how the core meanings evolve into peripheral meaning.

First focusing on individual LVs, we can more easily isolate groups of LVCs in what I call islands. These islands are clusters of LVCs which express similar verbal notions, based on the same LV and a specific type of PV. The islands seem to form based on certain, but not all, inherent attributes of the PV. In other words each PV has attributes which activate certain meanings of the LV and the LV then in turn contributes relevant features inherent to it, creating a meaning different from the core meaning of either component. Each island of LVCs has an underlying construction that encodes this information. The result is an LVC with a meaning that’s not necessarily predictable from the meaning of its parts. The LVC will belong to an island with other LVCs that share similar PVs and that serve to express similar verbal concepts.

To better illustrate these ideas, I will discuss the patterns that allow for island formation through concrete examples from Persian similar to the ones given at the end of the previous chapter.

3.1.3 Example Islands

If the verb gereftæn ‘to get’ appears as an LV with a PV that expresses a type of liquid that has the property of hardening or setting, then the LVCs will mean to cover with such a liquid. This transitive construction takes a subject who does the covering, and an object that is covered. The following forms belong to this island

gæt∫ gereftæn plaster GEREFTÆN set in plaster tæla gereftæn gold GEREFTÆN gold-leaf siman gereftæn cement GEREFTÆN cover in cement doktor dæst-e bæt∫æ-ro gæt∫ gereft. doctor hand-GEN child-ACC plaster get.PST.3SG. The doctor put the child’s hand in a cast. It is interesting to note that this sense of gereftæn also appears in the more general English verb ‘set’ as applied to both plaster and cement as well as to gelatin, yogurt, pudding, etc. where the liquid mix transforms into a thicker form or becomes completely solid. The notion of ‘hardening’ exists in both constructions in Persian, and is defined by this property of the PV, but the action expressed in the two differs. The Persian construction above expresses the act of covering with such a ‘setting’ mix, rather than the more simple phenomenon of the liquid hardening or thickening, as in the English ‘set.’ This construction is also similar to the French verb ‘prendre,’ as in ‘le ciment a pris.’

A different construction expresses the setting of yogurt or pudding (hardening that doesn’t necessarily cover anything and occurs at room temperature), also with the LV gereftæn (in

51

this case the construction calls for a different syntactic structure). This intransitive construction is expressed with the substance as PV11.

mast gereft yogurt GEREFTÆN the yogurt set jele gereft jello GEREFTÆN the jello set bæd æz t∫ænd sa’æt, mast gereft. after from few hour, yogurt get.PST.3SG. The yogurt set after a few hours. A specific LV and a few known PVs which have certain attributes, comprising of the physical, chemical, visual, tactile, and other common attributes define these islands. From these, other LVCs with similar meanings can be constructed. Each island’s underlying construction serve as a basis for productivity, it provides rules for making new forms. Real world knowledge about the functional attributes of the PV in these constructions plays a crucial role in judging the acceptability of the form. The former case expresses the notion of covering with different materials that are liquids that can be applied over an object and which then set permanently into a hardened state. And in the latter case, liquids congeal or set, permanently (which excludes freezing processes (e.g. water to ice) which are usually reversible).

Another island formed in gereftæn’s semantic space involves PVs that refer to liquids again. However, these liquids are always within an object and don’t flow out naturally. The LVCs in which these PVs occur with gereftæn express the extraction of such essences. Unlike the first set above, this construction yields an intransitive LVC, with only an agentive subject who does the extracting.

golab gereftæn rose-water GEREFTÆN extract liquid ∫ire gereftæn syrup GEREFTÆN extract syrup xun gereftæn blood GEREFTÆN pull blood baqeban æz botteha ∫ire gereft. gardner from plants syrup get.PST.3SG. The gardner extracted syrup from the plants. This is a well defined island, which requires the PV to have the property of a liquid naturally present in an entity which can be extracted. In other words, when taking water from the ocean, we cannot say ab gereftæn (water GEREFTÆN), while we could use that to express squeezing juice out of a fruit or even taking water from a deep well using a bucket, but not a pump. There is a semantic unpredictability in relation to the full verb gereftæn, but also a syntactic difference that marks the existence of a different construction.

As with the previous islands, the structure of this LVC can be generalized with regards to the PV: any liquid essence combined with gereftæn in an LVC will mean the extraction of that liquid from its source. For example, xun gereftæn specifically means extracting blood from a

11 This construction has a special syntactic structure in that it only appears in the third person singular. It seems that the PV is somehow the subject of the LVC. Certain linguists (e.g. Dabir-Moghadam 1997, Bashiri 1996, etc) do not consider these forms actual LVCs. They claim that they are full frozen sentences. However, others (Langarudi 1996, Ghomeishi 1991, etc) refer to them as impersonal constructions. We will return to these in Chapter 5.

52

vein, and not grabbing a testube of blood from a lab or covering something with blood and letting it dry (as this is not the function of blood).

Islands can be found for every LV. The more frequent the LV, the more islands it is likely to form. For example, at least thirty different islands can be distinguished for the verb gereftæn, which is the second most frequent verb in the system after kærdæn ‘to do.’ Xordæn, a less common LV, has at least fifteen identifiable islands.

Xordæn has some islands that express more abstract or emotional notions, or notions that cannot be defined solely through physical or perceptual attributes. For example, one can loosely categorize the common attributes of the PVs in one of xordæn’s islands as being a continuous, irrepressible, negative feeling that has to be suffered as the result of ones personal actions or state of mind. The subject is affected. Note that in the sentence below, the PV can be replaced with one of the others in the list below, with the corresponding change in meaning.

æfsus xordæn regret XORDÆN regret ænduh xordæn sorrow XORDÆN grieve a lost chance /opportunity æzab xordæn torture XORDÆN suffer heavily sohrab æz æz dæst dadæn-e rostæm xejli æfsus xord. Sohrab from from hand give-GEN Rostæm much regret eat.PST.3SG. Sohrab heavily regretted the loss of Rostæm. It is problematic to conclude that xordæn must have multiple meanings completely independent of its core meaning. For instance, if it actually meant to suffer, we could plausibly assign an abundant number of other meanings to xordæn to account for the data. Further, we would expect the verb xordæn to be usable for any instance of suffering, which is not the case (see example below). It is clear that xordæn is associated with a special meaning or meanings that only show up in specific constructions. Meanings associated with xordæn are triggered by certain properties inherent in the PVs with which it combines and the construction in which they occur. Also, as can be seen from the description above, the meaning of the construction is more nuanced than to suffer.

An example of a context in which suffering is expressed and where xordæn does not appear is an island formed with some LVCs of ke∫idæn ‘to pull.’ This island expresses continuous sufferance without necessarily being the result of an action, but of injustice. We cannot substitute the LV xordæn in these cases.

æzab ke∫idæn torture KE∫IDÆN grieve, be tortured12 rænj ke∫idæn rage KE∫IDÆN suffer entezar ke∫idæn waiting KE∫IDÆN long for

æz duri-e dust-æm æzab ke∫id-æm. from distance-GEN friend-1SG torture pull.PST-1SG. I suffered from being away from my friend.

12 Notice that the PV æzab can be used both with xordæn and with ke∫idæn resulting in a slight change of meaning. This seems to be accidental, since no other PVs are shared between these two LVs. Furthermore, the construction with xordæn is much less frequent and even sounds odd for some Persian speakers. So, there isn’t much interference between these two constructions in the grammar.

53

Another piece of evidence, which supports the claim that it is not the numerous different meanings of the LVs that surface in each of these LVCs, is that Persian speakers will only utter the core meaning of these words when asked for them out of context. Thus, if asked for the word for suffer, xordæn would not be uttered, but rather different LVCs probably based on xordæn.

A conceivable objection to this argument might be that Persian speakers don’t have access to the meaning of the LV out of context, it surely exists in their minds and is associated with the LV. However, a closer look at the type of meanings that would be portrayed by the LVs shows that this hypothesis is not plausible. For instance, other islands of xordæn include one defined by a type of weapon. This construction expresses being wounded by the weapon or projectile signaled by the PV (the subject is the target of the action). Some examples include:

∫æm∫ir xordæn sword XORDÆN be hit by a sword gule xordæn bullet XORDÆN get shot kard xordæn knife XORDÆN be stabbed pa-æ∫ dær jæng gule xord. foot-3SG in war bullet eat.PST.3SG. His leg was shot in the war.

This construction is reminiscent of the English I took a bullet (but in English, only bullet is allowed as a weapon). In Persian, if the weight of the meaning were placed on xordæn, i.e. an attempt is made to make this set of meanings compositional by assigning a meaning to xordæn, it would have to be something like “be wounded by a weapon of type X ...”. The existence of verbs with such a meaning is not plausible in natural language. It is too nuanced and there is perhaps no other language in which you can express such a notion with a simple verb. I argue that it is not xordæn or other LVs that express meanings, but the constructions themselves. One of the key issues I wish to explore is how to get a finer understanding or characterization of constructions.

3.2 More on Islands and Constructions

Islands are groups of LVCs where one type of PV combines with a particular LV to produce different LVCs with highly related meanings. These patterns can be captured in constructions. Each construction includes a specific LV, a type of PV (defined by common attributes, including physical, perceptual, semantic, and real world knowledge), and the meaning of the whole construction. This meaning portrays general aspectual and syntactic information, as well as idiosyncratic semantic information associated with it (not linearly predictable from the meaning of its constituents).

It is important to note that not all the LVCs constructed with a specific LV are valid members of one of the islands of the LV. There are many LVCs with opaque meanings that can be considered idiomatic, as well as some transparent LVCs that don’t seem to fall into any of the island patterns. However, in most cases a motivation for the use of a particular LV in the LVC can be found. Idiomatic LVCs include:

pæhlu gereftæn side GEREFTÆN bring to port (ship) iman da∫tæn faith DA∫TÆN believe (religion) kutah amædæn short AMÆDÆN accept a compromise

54

It is more precise to say that some of the constructions are semi-productive, and not fully productive. As discussed in section 2.2.1, semi-productivity occurs when a process displays systematic behavior, though the amount of productivity is constrained by semantic or other restrictions. In Persian, the fact that the constructions that define the islands involve particular semantic information is one sign of semi-productivity. Furthermore, there are certain syntactic constraints on the possibilities of PVs. For example, anaphora or pronominal elements cannot replace the PV in context. This shows that the semantics and syntax of the constructions is rigid, to a certain extent.

Moreover, the existence of an island in a set of LVCs doesn’t mean that the notion expressed by that island is necessarily productive. Some islands are more productive than others. The more tokens we find of a specific island, the more certain it is that the island is generative and actually exists in the system. In this way, instances of a particular construction and the construction itself mutually motivate each other.

Consider the two sets of LVCs constructed with zædæn ‘to hit’ below. The first set is much more limited in the number of new PVs or imaginable PVs it can take, whereas the second seems to be quite open.

hærf zædæn letter ZÆDÆN speak, talk laf zædæn speech ZÆDÆN vaunt gæp zædæn chat ZÆDÆN chat Ma tæmam ruz rajebe gozæ∫te gap zæd-im We all day about past chat hit.PST-2PL. We chatted all day about the past. gitar zædæn guitar ZÆDÆN play the guitar piano zædæn piano ZÆDÆN play the piano violon zædæn violin ZÆDÆN play the violin Homayun dær konsert violon-e kelasik zæd. Homayoon in concert violin-GEN classical hit.PST.3SG. Homayoon played a classical violin piece in the concert. The first set expresses the emission of speech and accepts any PV expressing a type of speech element or act. The number of possible PVs that can be accepted by this construction is rather limited. Intuitively, there are not many new ways of expressing speech acts in natural language. Consequently, the productivity of this set of constructions is low. On the other hand, the second set of LVCs accepts any new noun expressing a musical instrument to express the action of playing that instrument. The productivity of this island is high, many new forms can be created based on its structure.

Goldberg (1995, p134) defines token frequency as the number of times a particular word occurs in a construction, and type frequency as the number of words that occur with a particular construction. A construction with high type frequency will have a higher productivity. Following these definitions, the first LVC above has a high token frequency (speak, talk), and the second set represents a construction with high type frequency and is more productive than the former.

55

It should be noted that some LVCs have meanings which can be expressed by simple verbs in other languages. For instance, the verb zædæn ‘to hit’ occurs in over 400 LVCs. In this large set of LVCs, many islands form, and the constructions that result are very varied. For example, in the first set above, the LVCs can be translated by single lexemes in English or French (namely, speak, vaunt, chat in English, or parler, se vanter, causer in French). These cases are also, not coincidentally, those where it is plausible to speak of the LV as more bleached.

On the other hand, in the second set above, the meaning is expressed in English or French by a combination. These are constructions in these languages as well. In English the construction is play a Ninstrument, and in French jouer un Ninstrument. These latter cases resemble semi-idiomatic forms, where the PV retains it’s full meaning and the LV contributes an idiosyncratic element that must be learned. Once learned, the construction is highly productive. LVCs and their relation with idioms will be further discussed in chapter 5.

Between these two types of extreme cases, namely transparent and non-transparent, there are many intermediate cases where the contribution of each element is not clear. We will include all types of islands in our analysis, whether they are productive or not, and whether they are transparent or not. Although some islands do not provide direct clues for the production of new forms, they do represent interesting extensions of the meaning of the LVs that might be useful in the study of other, more productive, islands. Also, they avoid redundancy in the lexicon and help children as they acquire new vocabulary.

3.3 Previous Research Project

Before presenting the islands found in each of the LVs under study, I will briefly summarize the findings of Family (2001). In this study, the LVs in Persian were evaluated according to certain established cognitive parameters. These include those pertaining to time, movement, and space and are represented schematically. Each parameter has between four and six possible values. For example, a movement-related parameter can be represented by the schema below, where one entity enters the domain of another entity as time goes on. Analogously, other values show an entity leaving a domain, remaining constant, or just appearing in a domain.

Other parameters include aspect (the internal temporal structure of the action), force (how the energy flows to and from the participants), and thematic structure (the thematic role of the subject of the LV). Evaluating the LVs produced the following table, which already shows many contrasting values.

56

zædæn [hit] 1 2 1 1 oftadæn [fall] 1 3 1 3 dadæn [give] 3 2 4 1 gereftæn [get] 2 2 1 1 bordæn [carry] 4 1 4 2 aværdæn [bring] 3 2 1 1 xordæn [eat] 2 3 1 3 amædæn [come] 3 1 1 3 da∫tæn [have] 5 4 2 2 ∫odæn [become] 2 1 3 3 ræftæn [go] 3 1 4 2 ke∫idæn [pull] 4 2 1 1 kærdæn [do] 2 2 3 1

1 punctual internal in agent 2 accomplish spec. steady experiencer 3 process unspec. something patient 4 durative no force out 5 state

Ideally, each LV has a value for each parameter, and if these were projected onto a two-dimensional plot of the multi-dimensional space formed by the parameters, the LVs would be as far away from each other as possible. This would mean that each has its own sub-space in the semantic space and each expresses a meaning that contrasts maximally with any other LV. The computation results can be summarized as in the graph below, and indeed shows that the LVs spread out in the multi-dimensional space mapped here onto two dimensions.

57

The results support the hypothesis of the study. However, the values given to each LV in the previous study are not absolute. In other words, the values of each parameter might vary from LVC to LVC within a single group produced by a particular LV. The results are too coarse-grained to have any detailed predictive value.

This study was a first look at how cognitive parameters might play a role in organizing the semantic space of verbal notions in Persian. In the next sections, I present further developments on the structure of the semantic space produced by each LV, and in the chapter following that, I present how these structures can be interconnected across LVs.

3.4 The Light Verbs

In this section, I describe the structure of the semantic space of a few LVs. As a visual aid, I propose to present the space in a two-dimensional diagram. This should not be taken as a suggestion that the actual structure of each verb adheres to this arrangement rigidly. The diagrams are merely an exploratory tool. The core meaning of each verb is given in the center of the diagrams, and the islands occur at the periphery.

Between the core and the periphery, branching occurs. The resulting divisions and groupings of islands closely represent the rough approximations of broad semantic divisions. In other words, islands that seem to express semantically related notions occur closer to each other in this space (though there is currently no metric).

It is also important to note that though the islands lie at the end nodes of these diagrams, the traits that are expressed by the branches are not mutually exclusive. Certain islands could just as well be at the end of a different branch than the one showed. The patterns of the branches

58

displayed were chosen for visual convenience, and other structures (though restricted semantically) are possible, theoretically.

The highest nodes (closest to the core) do not represent islands themselves. Each of these nodes can branch and take many types of PVs to express different types of related actions loosely categorized under the label of the previous node. Any node divides into much more specific actions for which a specific type of PV can be described that occurs in each construction.

The islands lie at the end nodes of each branch. Defined by a type of action and specific semantic properties of the PV involved, these islands serve as bases for new verbs that illicit a particular LV.

It is important to note that the divisions, groupings, and even the islands themselves are the product of empirical selection informed by other native Persian speakers. Due to the lack of experimental or statistical data, there is no metric presented in the semantic spaces at this time. These diagrams represent an initial guess as to a possible way to organize most of the constructions known for a give LV.

In this chapter, I will present four LVs that are relatively rich in structure. I chose to present these four in detail since they will be further discussed in chapter 4. Other LVs, which will be mentioned in later chapters, are included in Appendix I.

In the following pages, each LV is presented with its full island distribution diagram, followed by its meaning as a full verb in Modern Persian, and its argument structure as a full verb. I will then briefly discuss general issues related to its use as an LV before giving a detailed description of the islands and possible core properties, shared by all islands. The description of the islands are systematically organized as follows: a) label of the island, b) meaning expressed, c) type of PV required, d)remarks, and e) examples.

In the descriptions of the role of the subject in full and light verb forms are expressed using Dowty’s proto-roles (Dowty, 1991). In his article, Dowty suggests that traditional thematic roles (agent, patient, etc) are too rigid to be applicable to certain empirical data. He proposes proto-roles (proto-agent and proto-patient). Most proto-agents have the property of having a volitional involvement in an action, having sentience or perception, causing an event or change of state in another participant, or moving (relative to the position of something external). Proto-patients, on the other hand, undergo changes of state, are causally affected by another participant, or are stationary relative to the movement of another participant. Arguments might have traits that correspond to both these role types, but they are assigned the proto-role from which they take most of their traits. I adopt this type of role assignment when discussing the arguments of the verbal elements under study.

The exploration of possible overarching properties of each LV will allow us to compare the LVs and perhaps lead to a better understanding of how this minimal set of verbal notions can give rise to such a rich repertoire of verbal notions. The results might help reveal how new or borrowed notions are produced in Persian, given the structure of each LV. The structure of the semantic space of the LVs will also allow us to analyze how they interact in the system, which leads to higher scale generativity.

59

3.4.1 ZÆDÆN

The original meaning of zædæn is to hit or engulf with a force. In its full verb use, zædæn is a transitive verb with a meaning close to the English verbs hit and beat or the French frapper. The syntactic and semantic argument structure of the full verb are as follows:

ZÆDÆN Syntax: NP1 NP2 ZÆDÆN The subject of this verb is always a proto-agent. If there is an instrument involved in the action, it is expressed with the addition of a prepositional phrase (ba NP3, where ba is a preposition meaning with). The verb can also be used to express one entity (NP1) hitting another entity (NP2) against a third entity (NP3). In this case, the third entity would be expressed in a prepositional phrase headed by another preposition (be NP3, where be roughly means to).

LVCs constructed with zædæn express a vast array of actions. Zædæn is one of the most common LVs in Persian. LVCs constructed with this LV can be either transitive or intransitive (around 2 out of 3 are intransitive).

A similar LV exists in Arabic. Some compound verbs in Arabic have similar lexical constituents as those in Persian, but there is no direct correspondence. As mentioned by Ibrahim (2002, p337), “le recoupement lexical des verbes supports d’une language à l’autre est quasi nul13.” This single LV in Arabic maps onto compound verbs in French with donner, faire, fixer, dresser, mettre, tenir, ouvrir, and tirer, but neither frapper or battre. The meaning of this Arabic LV always has a certain focus on “l’intention motrice au détriment de l’instrument don’t elle se sert14.” The Arabic LV has kept the notion of violence and speed of movement from the original meaning of the full verb.

The islands formed by zædæn fall into one of six broad categories. In other words, the core meaning of zædæn diffuses into roughly six extended meanings: affecting, auto-motion, emitting, quick action, piercing and transfering, and stealing. These meanings project from the core:

13 There is almost no exact lexical correspondence of LVs between languages. 14 Intentional motor movement that is at the detriment of the instrument it involves.

60

ZÆDÆN[to hit]

auto motion

emitting

piercing and

transfer-ing

stealing

quick action

affecting

Of the characteristics that all the LVCs constructed with zædæn seem to share, there are some that are most prominent (though some exceptions exist). These include the fact that the subject of the LVC is usually an agent, who does something. Sometimes this action causes a change of state in another entity, but other times it only concerns himself.

The action undertaken is usually a spontaneous or instantaneous action that has the potential to be iterated. It is a relatively quick and effective action. For example, grooming is faster than plastic surgery, hence ∫ane zædæn ‘comb’ or jaru zædæn are possible, but only æmæl pelastik kærdæn (plastic surgery) is possible, and not *æmal pelastik zædæn. Or, when taking a nap, t∫ort zædæn is possible, but to express sleeping a longer period, an alternative verb form must be used. Furthermore, ‘to dive’ can be expressed with zædæn (∫irje zædæn), but ‘to swim’ is expressed with kærdæn (∫ena kærdæn).

Though zædæn, as a full verb, is always transitive (takes a direct object), the majority of LVCs, with the exception of most LVCs occuring in the cluster labelled “affecting,” are intransitive.

ZÆDÆN: AFFECTING Zædæn expresses an action carried out by a proto-agent entity on another entity. It is a directed, intentional action that changes the state of another entity. It is usually abrupt or swift in time or space (like cutting or hitting). Similar to many other actions expressed with zædæn, these are actions that are potentially iterative, but cannot be durative. In other words, they cannot be used in sentences with durative clauses, an aspectual property discussed in the previous section. There are three islands that don’t cluster with the other islands due mostly

61

to their idiosyncratic syntactic structure (hang, cut, and steal, below). They could be grouped together, but each one has a meaning close to that of the cluster of islands immediately to its right.

hand held

ZÆDÆN[to hit]

protecting

enhancing

appearance texture

liquid

taste!ane [brush]

mesvak [toothbrush]

lif [loufa]

kære [butter]

roqæn [oil]

losion [lotion]

næmæk [salt]

golab[rose-water]

ab [water]

ætr [perfume]

direct

trimming

topology

repairing

damaging

projectile

weapon

ri! [beard]

t!æmæn [grass]

mu [hair]

gereh [knot]

ta [fold]

fer [curl]

bæxiye [stitch]

peyvænd [graft]

lætme [damage]

qat! [chop]

sædæme [damage]

trickery

bambul[cheat]

gul [fool]

hoqe [trick]

tir [bullet]

mu!æk [missile]

toop [canon]

kard [knife]

!æm!ir [sword]

t!aqu [knife]

ju! [acne/boil]

kæpæk [mold/ fungus]

tæbxal [cold soar]

t!ador[tent]

tæxte [board]

pærde [curtain]

type of hit

blunt

zærbe [punch]

lægæd[kick]

kotæk[beating]

t!æko! [hammer]

potk [sledgehammer]

surfacepa [foot]

dæst [hand]

suzæn [needle]

indirect

ædvie' [spice]

affecting

hurting

bænd [connection]

physical attack

modifying

ZÆDÆN: Affecting: Protecting Meaning: Hang an object that serves as a divider or that serves to block or protect something. PV: a type of board or cloth. Remarks: Intransitive or transitive. If it means to set up a divide between two things, the entities divided are expressed in a clause headed by beyn (between), and if it expresses hanging something on a surface, the surface is expressed in an indirect object. The construction that represents this island has a special syntactic structure. Namely, the PV can easily take a direct object marker. In transitive cases, the object can be the periphery of

62

something (e.g. dor-e otaq-ra pærde zædænd. around-of room-ra curtain hit-3SG. They hung curtains around the room – i.e. on the walls). pærde zædæn curtain ZÆDÆN hang a curtain t∫ador zædæn tent ZÆDÆN set up a tent tæxte zædæn board ZÆDÆN hang a board up Koli-ha væsæte rah-e Esfæhan t∫ador zæd-ænd. Nomad-PL middle path-GEN Isphanan tent hit.PST-3PL. The nomads setup tent on the way to Isphahan. ZÆDÆN: Affecting: Enhancing: Appearance Meaning: groom or clean with a tool (carrying out its function). PV: grooming tool. Remarks: Transitive (object is surface or entity being groomed). This islands especially takes PVs that are grooming or cleaning tools that require repetitive strokes on the same area. For example, it would be difficult to use the word æbr (sponge) as a PV, since the cleaning action does not necessarily entail a repetitive movement but dragging the sponge across a surface (see LV ke∫idæn ‘to pull’). kise zædæn sac ZÆDÆN rub with exfoliating glove boros zædæn brush ZÆDÆN brush sohan zædæn file ZÆDÆN file qabl æz lak alcol zædæn ruy-e miz ra sombate zæd. before from laquer hit.INF top-GEN table ACC sandpaper hit.PST.3SG. He sanded the table top before applying laquer.

ZÆDÆN: Affecting: Enhancing: Texture Meaning: spread a material across a surface. PV: a spreadable semi-liquid, viscous material. Remarks: Intransitive, surface expressed in an indirect object. moræba zædæn jam ZÆDÆN spread jam ræng zædæn paint ZÆDÆN paint (apply or spray) sorme zædæn eyeliner ZÆDÆN apply eyeliner nædjar paye-ha-ye miz ra ræng zæd. carpenter legs-PL-GEN table ACC paint hit.PST.3SG. The carpenter painted the table legs. ZÆDÆN: Affecting: Enhancing: Taste Meaning: add spice or another element to a compound (usually food). PV: substance added. Remarks: This island can be either transitive or intransitive with the same meaning. When transitive, the food that the spice is added to is expressed as a direct object, whereas in intransitive cases, the same entity is expressed in an indirect object.

63

ædviye zædæn spice ZÆDÆN add spices næmæk zædæn salt ZÆDÆN add salt felfel zædæn pepper ZÆDÆN add pepper a∫pæz æz inke mo∫tæri qæza ra bi∫tær næmæk zæd narahæt ∫od. cook from that client food ACC more salt hit.PST.3SG upset become.PST.3SG. The cook became upset because the patron added extra salt to the food. ZÆDÆN: Affecting: Enhancing: Liquid Meaning: spray or splash a liquid across a surface. PV: a liquid. Remarks: Intransitive, surface expressed in an indirect object. This action is necessarily a controlled application of something onto another. Compare to pa∫idæn ‘to spray’ which is the general verb for spray. ab zædæn water ZÆDÆN spray/splash water ætr zædæn perfume ZÆDÆN apply perfume bejaye ∫ost-o-∫uye ∫æxsi mærdom dær æsre vosta ætr mi-zæd-ænd. instead washing personal people in age middle perfume PROG-hit.PST-3PL. During the middle ages, instead of personal hygiene, people wore perfume. Comment: This island doesn’t necessarily mean a random spraying as with pa∫idæn, such as watering flowers. It generally refers to applying a liquid in a controlled manner onto a very limited area. ZÆDÆN: Affecting: Trimming Meaning: trim overgrown organic matter. PV: organic material that usually grows in innumerable quantities. Remarks: Intransitive, but has a special syntactic structure: the PV acts as a direct object and often receives an object marker. This special syntax makes this construction idiom-like, but an island exists due to the unpredictable semantics these forms have in common. mu zædæn hair ZÆDÆN trim hair ri∫ zædæn beard ZÆDÆN shave a beard off t∫æmæn zædæn grass ZÆDÆN mow the grass qæbl az inke t∫æmænha ra be-zæn-æd, usta karim muhay-æ∫ ra zæd. Before that grass-PL SUBJ-hit-3SG, Usta Karim hair-3SG ACC hit.PST.3SG. Usta Karim had his hair trimmed before he mowed the lawn. ZÆDÆN: Affecting: Modifying: Direct: Surface Meaning: touch a surface with a hand or foot or an instrument, usually leaving a mark or imprint. PV: instrument doing the touching. Remarks: Intransitive, surface touched expressed in an indirect object. The instrument used for this action is usually sharp unless it is a body part (or it can at least leave a non-negligible effect on the surface), but the action doesn’t entail hurting the object.

64

dæst zædæn hand ZÆDÆN touch t∫ængal zædæn fork ZÆDÆN pierce/touch with a fork suzæn zædæn needle ZÆDÆN touch with a needle kæfa∫ ruy-e kæf∫ ra suzæn-e riz zæd væ næq∫i ziba dad. Shoemaker top-GEN shoe ACC needle-GEN tiny hit.PST.3SG and pattern nice give.PST.3SG. The shoemaker put a nice needlework pattern on the shoe with a fine needle. ZÆDÆN: Affecting: Modifying: Direct: Topology Meaning: cause an organized topological change in an entity. PV: type of topological transformation. Remarks: Transitive, it is important that the type of action carried out on the entity does not necessarily deform or damage the entity. In other words, the result of the action can’t be random in shape. It is in sympathy with the nature of the entity acted upon. However, the change must be direct and intentional. The form *t∫uruk zædæn (wrinkle hit) does not exist. Wrinkles are random in geometry and are not usually intentional. ta zædæn fold ZÆDÆN fold gere zædæn knot ZÆDÆN tie a knot fer zædæn curl ZÆDÆN curl garson rumizi-ha ra tæmiz kærd va ta zæd. waiter table-PL cloths ACC clean make.PST.3SG and fold hit.PST.3SG. The waiter cleaned and folded the tablecloths. ZÆDÆN: Affecting: Modifying: Direct: Damaging Meaning: inflict damage on an entity. PV: general noun meaning damage or wound. Remarks: Intransitive: the damaged entity occurs in a prepositional phrase headed by be (to). asib zædæn injury ZÆDÆN injure, damage zæxm zædæn wound ZÆDÆN damage, wound lætme zædæn damage ZÆDÆN damage (e.g.reputation, progress) seil dehkæde ra nætavanest ziyad asib be-zæn-æd. flood village ACC unable.PST.3SG much damage SUBJ-hit.PST-3SG. The flood couldn’t cause much damage to the village. ZÆDÆN: Affecting: Modifying: Direct: Repairing Meaning: fuse or connect entities or parts of a single entity. PV: type of connection or instrument/material used for fusing or connecting. Remarks: Transitive when the subject acts upon a single entity whose parts might need to be connected (e.g. hem of a skirt, two parts of a pipe). If, however, the first entity (the object) is fused to a second entity, the second occurs in a prepositional phrase headed by be (to). kuk zædæn stitch ZÆDÆN stitch a seam peyvænd zædæn graft ZÆDÆN graft plants, organs væsle pine zædæn patch ZÆDÆN patch over a hole

65

mærdom raz-e peyvænd zædæn-e derækht-an ham no ra zud yad gereft-ænd. people secret-GEN graft hit.INF-GEN tree-PL same species ACC early memory get.PST-3SG. People learned the secrets of grafting trees of same species early on. ZÆDÆN: Affecting: Modifying: Indirect Meaning: acquire marks or irritation as evidence of some internal or external biological or chemical reaction on a surface. PV: type of mark or reaction. Remarks: Intransitive, the LV seems to agree in number and person with the PV (3SG), though the actual subject of the sentence, or the entity or person undergoing the reaction, might be plural. Impersonal constructions will be further discussed in chapter 5. The subject is not an agent, but undergoes a reaction that causes the appearance of the thing expressed by the PV. pine zædæn callus ZÆDÆN develop a callus zæng zædæn rust ZÆDÆN rust tavæl zædæn blister ZÆDÆN develop a blister teke-ye abtæla dad-e hit∫gah zæng ne-mi-zæn-æd. piece-GEN goldplate give-PTCP never rust NEG-PROG-hit.PRS -3SG. Goldplated pieces will never rust. ZÆDÆN: Affecting: Hurting: Tricking Meaning: trick someone. PV: trick. Remarks: Transitive, this LV is used as opposed to the more generic kærdæn, since it is in a cluster of islands that pertain to negatively effecting an entity. However, it is interesting to note that if the English word trick was borrowed into Persian, the resulting verb would be trick kærdæn, and not *trick zædæn, perhaps because kærdæn is more general and used as the default. hoqqe zædæn trick ZÆDÆN slight, trick kælæk zædæn trick ZÆDÆN play a trick, fool, cheat naro zædæn double-cross ZÆDÆN double-cross u dær væræq-bazi kælæk mi-zæn-æd he in card-playing trick PROG-hit.PRS-3SG. He plays tricks in card games. ZÆDÆN: Affecting: Hurting: Physical Attack: Weapon: Hand Held Meaning: wound or penetrate another entity with a weapon. PV: a sharp, penetrating weapon. Remarks: Transitive, the weapon must be sharp and directly penetrate the patient of the action. For example, one can be wounded by a gun, but the form *tofæng zædæn (gun hit) does not exist, because it is not the weapon that penetrates the patient. It is the bullets that penetrate the patient (see next island). xænjær zædæn scythe ZÆDÆN hit with a scythe t∫aqu zædæn knife ZÆDÆN stab with a knife ∫æm∫ir zædæn sword ZÆDÆN stab with a sword

66

tarix mi-gu-yæd ke Ali ra hengam-e namaz-æ∫ ∫æm∫ir zæd-ænd history PROG-say-3SG that Ali ACC time-GEN prayer-3SG sword hit.PST-3PL History says that Ali was hit by a sword during his prayer. ZÆDÆN: Affecting: Hurting: Physical Attack: Weapon: Projectile Weapon Meaning: attack with a projectile weapon. PV: projectile weapon. Remarks: Intransitive. Similar limitations as previous island. mu∫æk zædæn missile ZÆDÆN hit a missile tir zædæn bullet ZÆDÆN shoot with a bullet sæt∫me zædæn pellet ZÆDÆN shoot with pellets polis-ha ælkapon ra dær ∫ikago tir zæd-ænd væ ko∫tænd police-PL Al Capone ACC in Chicago bullet hit.PST-3SG and kill.PST-3SG. The police shot and killed Al Capone in Chicago. ZÆDÆN: Affecting: Hurting: Physical Attack: Blunt Instrument Meaning: strike some entity with a heavy or blunt instrument. PV: blunt instrument used for striking. Remarks: Transitive. t∫æko∫ zædæn hammer ZÆDÆN hammer potk zædæn sledge hammer ZÆDÆN hit with a sledge hammer t∫omaq zædæn club ZÆDÆN beat with a club gærdæn koloft-ha daneshjuy-an ra t∫omaq zæd-ænd. neck thick-PL student-PL ACC club hit.PST-3SG. The rednecks beat the students with clubs. ZÆDÆN: Affecting: Hurting: Physical Attack: Type of Hit Meaning: hit an entity with ones hands, feet, or head. PV: type of hit. Remarks: Transitive, this action must directly affect the patient in a hurtful manner. For example, one cannot say *hol zædæn (push hit), because the notion of push implies moving an entity without necessarily hurting it. mo∫t zædæn punch ZÆDÆN punch lægæd zædæn kick ZÆDÆN kick sili zædæn slap ZÆDÆN slap u mærdi ra ke be-u biehterami kærd-e bud yek sili mohkam zæd. She man ACC which to her disrespect do.PST-PTCP be.PST.3SG one slap hard hit. She slapped the man who had offended her very hard.

67

ZÆDÆN: AUTO-MOTION

Two types of reflexive, dynamic, movements are expressed with zædæn. The types of movement expressed with this LV are usually instantaneous, either carried out once or iterated in succession.

ZÆDÆN[to hit]

rotation

rhythmic

auto motion

position pit! [turn]

qælt [sommersault]

varune [underwater

sommersault]

bal [wing]

æsa [cane]

tolombe [pump]

sær pa [on foot]

zanu [knee]

agitated

ZÆDÆN: Auto Motion: Rhythmic Meaning: carry out a rhymic action requiring an instrument. PV: the instrument. Remarks: Intransitive, the action described by these verbs is inherent to the instrument. The functional properties of the instrument, which are part of real world knowledge, are evoked. æsa zædæn cane ZÆDÆN walk w/ a cane (hitting it on the ground) tolombe zædæn pump ZÆDÆN pump water out of a well bal zædæn wing ZÆDÆN flap wings (used for winged animals) parænde-ye azad ∫od-e bal zæd o ræft. bird-GEN free become.PST-PTCP wing hit.PST.3SG and leave.PST.3SG. The freed bird flew away.

68

ZÆDÆN: Auto Motion: Agitated: Position Meaning: position one’s body in a certain configuration. PV: configuration. Remarks: Intransitive, similar to the English “striking a pose.” zanu zædæn kneel ZÆDÆN kneel t∫onbæk zædæn squat ZÆDÆN squat qamæt zædæn erect posture ZÆDÆN stand in a position to pray emperatur-e ∫ekæst xord-e jelo-ye do∫mæn zanu zæd. emperor-GEN defeat eat.PST-PTCP front-GEN enemy knee hit.PST.3SG. The vanquished emperor kneeled in front of the enemy. ZÆDÆN: Auto Motion: Agitated: Rotation Meaning: Flipping or rotating of one’s own accord. PV: type of rotational movement. Remarks: Intransitive, volitional (as opposed to the same PV with xordæn in next section). pit∫ zædæn turn ZÆDÆN turn varune zædæn flip ZÆDÆN do a somersault (esp. in water) mæl’æq zædæn somersault ZÆDÆN do a somersault u æz xo∫hali mæl’æq mi-zæd. he from joy somersault PROG-hit.PST.3SG. He did somersaults in joy.

ZÆDÆN: EMITTING

The type of emission expressed by zædæn pertains generally to auditory, though there is one island whose LVCs express visual emission. Something is emitted by the subject whether or not there is a receiver. Consequently, all these forms are intransitive. An important trait common to these islands is the segmented nature of the emission - either the sound or visual emission is contingently instantaneous or instantaneous by nature. There is no change of state: the subject carries out an action that doesn’t directly affect his state or the state of any other entity. The subject is the emitter, and not the agent.

I will show later that some of the islands also appear with ke∫idæn ‘to pull,’ whose aspectual characteristics differ from zædæn in this respect.

69

coherent speech

ZÆDÆN[to hit]emitting

aural

visual

instru-ment

humansounds

musical

type of music

jaz [jazz]

se zærbi [three beat]

muzik [music]

tombæk [tombak]

sæntur [santur]

gitar [guitar]

bærq [shine]

jæræqe [spark]

su su [light]vocal

expres-sive

sounds

be!kæn[snap]

dæst [clap]

kæf [applause]

annoying utterance

hærf [speech]

gæp [word/sound]

laf [vaunt]

veng[infan'ts cry]

qor [mumble]

emotionalnoises zar [sob]

qæhqæh[laughter]

dad [yell]

jiq [shout]

ærbade [yell]

noise

jez[sizzle]

vez [fizzle]

ZÆDÆN: Emitting: Visual Meaning: emit a bright visible stimulus. PV: bright shimmer, or flash. Remarks: Intransitive. It is the only island that expresses non-auditory emission. It is usually a spontaneous and potentially repeating visual phenomenon. The subject is never human, and never an agent. The use of zædæn is probably due to the focus on the spontaneity of the event. jæræqe zædæn spark ZÆDÆN spark t∫e∫mæk zædæn wink ZÆDÆN blink (as in a light) zæbane zædæn flame ZÆDÆN flare up (as in a fire) hengami ke væsæt-e xiyaban resid t∫eraq qermez obur-e piade ∫oru kærd be t∫e∫mæk zædæn. while that middle-GEN street arrive.PST.3SG red light cross-GEN pedestrian start do.PST.3SG blink hit.PST-INF. The pedestrian red signal started blinking when he got to the middle of the street. Comment: In colloquial registers, this form can be used to say something has a hint of a specific visual property, or a tint of a color. Here, the PV is in an indirect object. For example: t∫ehre-æ∫ be afqani mi-zæn-æd. face-3SG to Afghan PROG-hit.PRS-3SG. His face has Afghan features.

70

in pirahæn be qermez mi-zæn-æd. this shirt to red PROG-hit.PRS-3SG. This shirt has a red tint.

ZÆDÆN: Emitting: Aural: Music: Instrument Meaning: play a musical instrument to emit “musical” sounds. PV: any musical instrument. Remarks: Intransitive, highly productive. This island is semantically equivalent to the English use of the word play in play an instrument, or the French jouer un instrument. folut zædæn flute ZÆDÆN play the flute violon zædæn violin ZÆDÆN play the violin piano zædæn piano ZÆDÆN play the piano hengami ke juki folut zæd mar æz sabæd sær birun kærd. while that juki flute hit.PST.3SG, snake from basket head outside do.PST.3SG. When the juki played his flute the snake poked its head out of the basket. Comments: It is fairly easy to formulate a compositional account of this island, positing a separate definition for the full verb zædæn. In this case, where productivity is maximal, and the syntax is much freer, this analysis is equivalent to an analysis where each island represents a different and separate lexical meaning for zædæn. However, it is important to note that one cannot conjoin across islands. In other words, one cannot say *Hæsæn folut va hæft zærbi zæd. *Hasan flute and seven beat hit.PST.3SG. *Hasan played the flute and a seven beat rhythm (see next island). ZÆDÆN: Emitting: Aural: Music: Type of Music Meaning: play a genre or style of music. PV: type or sytle of music or rhythm. Remarks: Intransitive, highly productive. This island is very similar to the previous island, the only difference is that the PV is of a different class. muzik-e kelaasik zædæn classical music ZÆDÆN play classical music haft-zærbi zædæn seven-beat ZÆDÆN play a seven-beat rhythm jaz zædæn jazz ZÆDÆN play jazz music tæqi ba folut-e irani mi-xast jaz be-zæn-æd. Taqi with flut-GEN Iranian PROG-want.PST.3SG jazz SUBJ-hit.PRS-3SG. Taqi wanted to play jazz with the Iranian flute. ZÆDÆN: Emitting: Aural: Noise Meaning: make sounds of sizzling. PV: onomatopoeia sizzling sound. Remark: Intransitive, probably zædæn is used here as opposed to other more generic LVs, such as kærdæn, because emission is in the semantic space of zædæn, and it expresses the spontenaity of the noise, as with the visual stimulus island at the beginning. The sound emitted by meat on a hot grill conveys that the meat is being sizzled, and similarly the crackling noises emitted by a fire tell of its continuous burning.

71

jez zædæn sizzle ZÆDÆN sizzle jelez velez zædæn crackling ZÆDÆN crackling sound (burning,grilling) kæbab-e jigær ru-ye mænqæl jez mi-zæd. liver-GEN kabob on-GEN barbecue-pit sizzling PROG-hit.PST.3SG. The liver kabob was sizzling on the barbecue pit. ZÆDÆN: Emitting: Aural: Expressive Sounds Meaning: make clapping or clicking sounds with the hands. PV: palm or hand, or configuration making sound (snap, etc). Remarks: Intransitive, these forms are used solely to express applause or to express keeping a rhythm. It involves only socially typical ways of doing so, which includes hands, and maybe in some contexts, feet (pa), but no other body parts can be included. Carries a lot of the hitting notion of zædæn, especially repetitive and rhythmic hitting. kæf zædæn palm ZÆDÆN clap, ovation be∫kæn zædæn snap ZÆDÆN snap dæst zædæn hand ZÆDÆN clap, applaud tæma∫agær-an bæraye taqi kæf-e istade zæd-ænd. spectator-PL for Taqi palm-GEN standing hit.PST-3PL. The spectators gave Taqi a standing ovation. Comment: This is similar to the English put your hands together and the French battre les mains. ZÆDÆN: Emitting: Aural: Vocal: Coherent Speech Meaning: speak, talk in prose. PV: words that mean sound or speech. Remarks: Intransitive, the entity being spoken to is expressed in a prepositional phrase headed by ba (with), similar to the English speak or talk to/with. Here, the notion of speaking entails communicating in a dialogue (though not always), compared to the island below, this island is a two sided communication. Since no PV can actually be added to the set of PVs and the existing forms are highly uncompositional, these forms can be considered as idiomatic. The sole reason that they are grouped into an island is because they all use zædæn and express very similar notions. Not that zædæn can’t be used to express poetry recitation, as in she’r zædæn, although poetry is definitely comprehensible utterances. Poems are sang, not emitted by zædæn as normal speech is, and they don’t entail a dialogue. hærf zædæn utterance ZÆDÆN speak, talk laf zædæn speech ZÆDÆN vaunt gæp zædæn chat ZÆDÆN chat bæd æz sal-ha ne∫æst-im væ xeyli gæp zæd-im. after from year-PL sit.PST-1PL and much chat hit.PST-1PL. After many years we sat and chatted a lot. Comment: A similar notion is used in the English expression shoot the breeze or strike up a conversation or beat your gums.

72

ZÆDÆN: Emitting: Aural: Vocal: Annoying Utterance Meaning: speak in a specific manner as to communicate negative emotions such as stress, displeasure, boredom, etc. PV: manner of speech. Remarks: Intransitive, there is communicative intent, though the verb expresses an action that is carried out by a single agent. veng suggests a onomatopeic term that would occur in these actions. veng zædæn whine ZÆDÆN whine (speak in a whinning way) qor zædæn grumble ZÆDÆN complain zer zædæn drivel ZÆDÆN talk nonsense an mærd amæd væ dær barey-e xod-e∫ kolli zer zæd. That man come.PST.3SG and about-GEN self-3SG lots drivel hit.PST.3SG. That man came and driveled much about himself. ZÆDÆN: Emitting: Aural: Vocal: Emotional Sounds Meaning: make vocalizations that convey emotions (i.e. yelling and crying sounds). PV: the type of vocalized sound. Remarks: Intransitive, this noise is usually sharp and emitted once or repetitively. As opposed to the previous island, the PV express the type of production, and not the manner. dad zædæn yell ZÆDÆN yell zar zædæn cry ZÆDÆN cry, sob t∫ah t∫ah zædæn chirping ZÆDÆN sing like a nightingale (staccato) xanænde xeyli ali t∫ah-t∫ah mi-zæd. singer very great chirping PROG- hit.PST.3SG. The singer sang great staccato.

ZÆDÆN: QUICK ACTION

Creating or producing something or obtaining quick measurements can be expressed with zædæn. Since zædæn often implies a swift and rapid action, these LVCs focus on expressing a trivial act. These forms are usually used in the colloquial language.

73

ZÆDÆN[to hit]

prepara-tion

maye [starter]

qaleb [mold]

xame [cream]

quick action

creation

measure-ment

ændaze [size]

!i!e [glass]

livan [cup]

ZÆDÆN: Quick Action: Creation Meaning: produce any item or establishing an institution or store. PV: object produced or manufactured. Remarks: Intransitive, process is expressed in a trivial manner, the speaker gives the impression that the process is an instantaneous, and thus effortless task, it is productive since it occurs in colloquial registers. livan zædæn cup ZÆDÆN produce cups mæqaze zædæn store ZÆDÆN open a store pol zædæn bridge ZÆDÆN build a bridge komæk-kar-an særi yek pol-e pantun zæd-ænd. rescuer-PL quick one bridge-GEN pantoon hit.PST.-3PL. The rescuers quickly built a pontoon bridge. Comment:As the verb monter in French (studied by Cadiot and Lebas, 2000), zædæn can be used to describe an organization or mounting of something, like restoran zædæn ‘restaurant hit = open a restaurant’ or livan zædæn ‘cup hit = manufacture cups’. Here, as with monter, ‘les caractéristiques qu’on prête à ces entitiés ressenties comme extérieures à leur procès de gestation ne sont en fait que des réalisations d’un rapport qualifié: le sujet présente le produit de son activité dans sa dimension de projet soutenu, création et/ou artifice15’ (Cadiot and 15 these caracteristics, that we assign and feel as exterieur to the action are actually just a skilled accomplishments: the subject produces a creation or a project.

74

Lebas, 2000: p14). The PV holds a profile which is carried out with the cinetic characteristics of the LV. ZÆDÆN: Quick Action: Measure Meaning: undertake any short, trivial action. PV: action. Remarks: Intransitive or transitive, depending on the PV or action that is expressed. t∫ort zædæn nap ZÆDÆN take a nap ændaze zædæn measurement ZÆDÆN make a quick measurement tæxmin zædæn estimate ZÆDÆN make an estimation pire mard hami∫e bæd-e nahar yek t∫orti mi-zæn-æd. old man always after-GEN lunch one nap PROG-hit.PRS-3SG. The old man always takes a short nap after lunch. ZÆDÆN: Quick Action: Preparation Meaning: prepare a starter or making a mold. PV: starter, mold. Remarks: Intransitive. The action is usually one that starts the transition of a material from one form into another, such as milk into yogurt, hops into beer, cake into a baked form, mud or cement into brick form, liquid cream into whipped, etc. xame zædæn cream ZÆDÆN whip cream maye zædæn starter ZÆDÆN add a starter qaleb zædæn mold ZÆDÆN make a mold ∫agerd-e ∫irini-pæz ævvæl bayæd xame zæd-æn ra yad be-gir-æd. student-GEN pastry-cook first must cream hit-INF ACC memory IMP-get-3SG. Pastry chef students must first learn how to whip cream.

ZÆDÆN: PIERCING AND TRANSFERING

These islands in this cluster have in common the use of an instrument that pierces or goes through the surface of something in order to transfer a substance into the pierced object. In other words, there is an outside to inside transfer action.

75

ZÆDÆN[to hit]

piercing and

transfer-ing

hæ!i! [hash]

tæriyak [opium]

injecting

drug

ampool [shot]

vaksæn [vaccination]

mævad [substance]

fueling

benzin [gasoline]

gazoil [deisel]

ZÆDÆN: Piercing and Transfering: Drug Meaning: voluntarily ingest or inject some substance into the body. PV: drug. Remarks: Intransitive, highly productive: any drug can occur in this construction. The subject is always a human acting on himself, unless it is medication, in which case another person can administer the drug (as in the example sentence below). The quickness with which the drug takes effect could be a reason for the choice of zædæn as LV, since taking drugs orally cannot occur in a construction with zædæn. hæ∫i∫ zædæn hash ZÆDÆN smoke hash penesilin zædæn penicilin ZÆDÆN administer penicilin kokain zædæn cocaine ZÆDÆN take cocaine enqædr penesilin zæd-æn-e∫ ke defa-e badan-e∫ hit∫ ∫od-e. so much penicillin hit-3SG-3SG that defense-GEN body nothing become-PTCP. They have injected him with so much penicillin that his body’s defense has dropped to nothing. Comments: The core, or prototypical, case of this island involves a drug that is injected with a needle. However, the meaning has extended to include any ingestion of a psychoactive drug (usually with trivial connotations). This island also provides the basis for a common colloquial utterance expressing the notion of eating in a trivial way or it suggests that it puts you in a good or euphoric mood (c.f. French: se taper du shit, se taper un hamburger, etc). For example:

76

ye t∫el o kæbab be-zan-im? one rice and kebab SUBJ-hit.PRS-1PL? Are you up for some rice and kabab? (Lit. Shall we inject ourselves with rice and kebab?) ZÆDÆN: Piercing and Transfering: Injecting Meaning: Pierce the skin with an object that will inject something into the vein. PV: instrument that pierces, usually depositing a substance though the skin. Remarks: Transitive, the semantic relations are different from the previous island (this one specifies the instrument used, whereas the previous specifies the substance), and in these cases, the subject is not necessarily the person undergoing the effects of the action. If the subject administers a shot to someone else, the undergoer will be expressed as a direct object. ampool zædæn shot ZÆDÆN administer a shot soræng zædæn syringe ZÆDÆN administer a shot with a syringe æz bæs soræng zæd-æm bazu-æm ∫od-e abke∫. so much syringe hit.PST-1SG arm-1SG become-PTCP sieve. I taken so many shots that my arm has become like a sieve. (the subject administers his own shots) ZÆDÆN: Piercing and Transfering: Fueling Meaning: put fuel to a machine through a nozzle shaped object. PV: fuel substance. Remarks: Transitive, the subject never undergoes the effects, but adds fuel to a machine (usually used for a car), expressed in an indirect object. It is essential that the instrument with which the substance is added have a nozzle at the end that enters the receptor. The type of PV that is acceptable in this case is ontologically limited, though can be potentially productive (as new fuel-like substances replace the more typical petroleum-based fuel). The difference between this island and the first island in this cluster is mainly that the subject does not undergo the effects, and it is interesting to note that the only way to differentiate the two is through real-world knowledge (the properties of the PV). benzin zædæn gas ZÆDÆN pump with gas gazoil zædæn diesel ZÆDÆN pump with diesel geris zædæn grease ZÆDÆN add grease ma∫in xeyli seda mi-dæh-æd, bayæd hæme ja∫ o geris be-zæn-æm. car much noise PROG-give.PRS-3SG, must all place ACC grease SUBJ-hit.PRS-1SG. The car makes a lot of noise, I must lubricate it all over.

77

ZÆDÆN: STEALING

ZÆDÆN[to hit]

stealing

jib [pocket]

bank [bank]

gænj [treasure]

Meaning: steal money from inside a container-like place. PV: container being stolen from. Remarks: Intransitive, but like the cut and hang islands above, the PV easily accepts a direct object marker. This island is not very productive and is used in the colloquial language. These are idiom-like expressions, where the “containers” must be known places where money is kept. For example, one cannot say *kase zædæn (bowl hit) to mean steal money out of a bowl, since it is not a typical place one would keep money. This island has corresponding nouns that describe specific types of robbers, such as jib zan, bank zan, rah zan, and so on. (Also, consider the English hit a bank, kock off) kæf zædæn palm ZÆDÆN steal (from someone) bank zædæn bank ZÆDÆN rob a bank jib zædæn pocket ZÆDÆN pickpocket da∫t mi-ræft bank be-zæn-æd, to rah mask-e∫ ra æz jib-e∫ zæd-ænd. have.PRS.3SG PROG-go.3SG bank SUBJ-hit.PRS-3SG, on way mask-3SG ACC from pocket-3SG hit.PST-3PL. He was on his way to rob a bank, but they picked his mask from his pocket.

78

3.4.2 XORDÆN

In Modern Persian, the full verb xordæn has two different, unrelated meanings and two respective argument structures. It can either be a transitive verb meaning to eat (French: manger), or a verb taking only an indirect object to mean to collide, or hit (French: entrer en collision). The two core meanings of xordæn don’t seem to be related, and none of the etymological sources available shed any light on the origins of the second meaning. However, we will find traces of both of these meanings in the LVCs constructed using xordæn as an LV.

XORDÆN (EAT) Syntax: NP1 NP2 XORDÆN XORDÆN (HIT) Syntax: NP1 be NP2 XORDÆN When expressing the notion of eating, the full verb xordæn takes a volitional subject, or in Dowty’s terms, a proto-agent argument. When xordæn expresses the notion of hitting, there is no volition. In these cases, the subject is a proto-patient and collides with another object accidentally or non-intentionally.

Xordæn produces mostly intransitive LVCs where the subject is a proto-patient, undergoing a change of state, or experiencing a state. It is an inchoative verb, that can be either telic or atelic, depending on the type of action (undergoing an action or experiencing a state).

The meaning of the actions expressed by this LV is most generally one of being effected, and usually has a negative connotation. The LV doesn’t seem to have retained any of the lexical content of xordæn, but rather a figurative or metaphorical extension of the original meaning. The eat meaning of xordæn serves as a basis for a metaphorical extension resulting in the idea of undergoing an atelic action. In other cases, the notion of collide gives rise to undergoing telic or sudden actions.

Except for a handful of prepositions and prepositional phrases that form more idiomatic LVCs, only nominal PVs combine with xordæn. In our corpus, xordæn forms around 125 LVCs. Though not comparable to zædæn in frequency, this verb is interesting partly because of the islands it shares with zædæn, a phenomenon I return to in the next chapter.

The LVC islands of xordæn fall into four broad categories. These encompass meanings related to being affected, suffering, exploiting, and being agitated. These further branch off into finer-grained classifications whose end nodes are islands.

79

XORDÆN[to eat]

suffering

motion

usurping

affected

XORDÆN: AFFECTED

The richest branch, measured in terms of the number of associated islands, expresses undergoing or being affected by an action. The subject of these LVCs usually undergoes the action expressed. It is interesting to note that the group of islands that branch off to the right in the diagram are mostly used for inanimate object, whereas those to the left are animate. Some of these LVCs can be considered as inchoative alternants of analogous LVCs constructed with zædæn. As we will see in the next chapter, several xordæn islands correlate highly and alternate with a cluster of islands formed with zædæn.

80

modifying

XORDÆN[to eat]

hurt-ing

weapon

type of hit

touching

fusing

fer [curl]

gere [knot]

pit! [twist]zærbe

[shock]

væsle [connection]

bæxiye [stich]

ju! [weld]

ta [fold]

!ælaq [whip]tir

[bullet]

golule [bullet]

projec-tile

hand held

kard [knife]

sæt!me [pellet]

lægæd [kick]

sili [slap]

kotæk [beat]

trickgul [trick]

færib [trick]

qæbn [cheat]

instrument

t!æko! [hammer]

næzær [opinion]

potk [mallet]

ængo!t [finger]

suzæn [needle]

naxon [nail]

affected

damaginglætme [damage]

qat! [chop]

sædæme [damage]

reshaping

XORDÆN: AFFECTED

The richest branch, measured in terms of the number of associated islands, expresses undergoing or being affected by an action. The subject of these LVCs usually undergoes the action expressed. It is interesting to note that the group of islands that branch off to the right in the diagram are mostly used for inanimate objects, whereas those to the left are animate. Some of these LVCs can be considered as inchoative alternants of analogous LVCs constructed with zædæn16. As we will see in the next section, several xordæn islands correlate highly and alternate with a cluster of islands formed with zædæn.

16 Each LV in the system has a similarly structured semantic space (networked islands). Many of the LVCs constructed with the LV xordæn have alternants with the LV zædæn ‘hit’ (by alternants, I refer to LVCs that have a single PV occuring with different LVs to express similar but related meanings). Alternations are systematic and can be considered as cases of shared islands, connecting two or more different LVs, and in this way inter-connecting all the verbs in the system. It is beyond the scope of this article to explore the dynamics between LV spaces.

81

Diagram 3: XORDÆN: AFFECTED

XORDÆN: Affected: Modified: Fused Meaning: become fused or connected to parts of itself or to other entities usually through a natural process. PV: type of connection or instrument/material used for fusing or connecting. Remarks: Intransitive. These forms are used when the subject becomes fused or mended as a consequence of a natural process (rust, humidity, organic growth) and generally not the consequence of the actions of a conscience being. For example, the term kuk xordæn is rare, because stitching can only be done by a volitional external entity. Or, ju∫ xordæn ‘weld or fuse’ can be used for a material when the fusion is the result of heat or rust or other environmental factors, but not directly when an entity has welded the items together (though if the speaker doesn’t know, care, or remember who welded it, but only assumes the action has taken place, this form can be used). In the LVCs expressing the fusion of two different entities, the second entity occurs as an indirect object. kuk xordæn stitch XORDÆN be closed up by stitches peyvænd xordæn graft XORDÆN be grafted (plants, organs) væsle pine xordæn patch XORDÆN be patched up in lebas qæ∫æng æst hært∫ænd besyar væsle pine xord-e æst. this dress beautiful be.PRE.3SG despite much patch eat-PTCP be.PRE.3SG. This dress is beautiful even though it has been patched up quite a bit. XORDÆN: Affected: Modified: Damaged Meaning: be damaged or deteriorated. PV: type of damage or wound.

82

Remark: Intransitive. These forms express substantial physical damage sustained by the subject. This damage is usually incurred by effects of the physical environment and doesn’t necessarily involve an external, conscious agent. The damage usually diminishes the value and usefulness of the subject. asib xordæn injury XORDÆN be injured, be damaged zæxm xordæn wound XORDÆN be damaged, wounded lætme xordæn setback XORDÆN sustain setback(e.g. progress) saltænæt pæræst-an dar enqelab lætmehaj-e ziyad xord-ænd. royalty worshiper-PL in revolution setback-GEN much eat.PST-3PL. The royalists sustained much setback in the revolution. XORDÆN: Affected: Modified: Topology Meaning: undergo an organized, topological change. PV: type of topological transformation. Remarks: Intransitive. The topological change expressed by these verbs is not imposed or directly inflicted by an external entity, but rather by the environment (natural process) or unintended consequence of an action. For example, one cannot use the term fer xordæn to refer to someone’s hair after a visit to the hair salon, though it could be used if the curls result from humidity in the air. The change usually damages the subject or at least results in an unwanted state. For example, one cannot say gere xordæn for a string that has been purposely tied into a knot, though the same form can be used to express a wire having gotten tangled from too much motion (a consequence of another action, e.g. a tangled telephone wire when one walks around while using the phone). ta xordæn fold XORDÆN get curled gere xordæn knot XORDÆN get tied in a knot fer xordæn curl XORDÆN get curled hengami ke kenar-e dærya resid-im mu-hay-e ham-e-man fer xord. when that side-GEN sea arrive.PST-2PL hair-PL-GEN all-GEN-2PL curl eat.PST.3SG. When we arrived at the beach, all our hair got curled. XORDÆN: Affected: Modified: Surface Meaning: be touched with a hand or foot or an instrument, usually leaving a mark or imprint. PV: instrument doing the touching. Remarks: Intransitive. The instrument used for this action is usually sharp unless it is a body part (it can leave a non-negligible effect on the surface of the subject), and though the action doesn’t entail hurting, it might have negative or damaging effects on the subject. This effect is usually not the direct intent of an action, but a consequential result. In other words, someone might touch a surface, not meaning to leave an imprint, and so the imprint is not the intent of the action: the surface can be said to have been dæst xorde. dæst xordæn hand XORDÆN be touched, altered suzæn xordæn needle XORDÆN be touched/pierced with a needle pa xordæn foot XORDÆN get hit with a foot in æks æsl nist, dast xord-e æst. this picture original NEG.be.PRS.3SG, hand eat-PTCP be.PST.3SG.

83

This picture is not an original it has been altered. XORDÆN: Affected: Hurting: Weapon: Type of Hit Meaning: be hit with another entity’s hands, feet, or head. PV: type of hit. Remarks: Intransitive. This action must directly affect the subject in a hurtful manner. For example, one cannot say *hol xordæn (push eat), because the notion of push can imply acting on an entity without necessarily hurting it. An agent carries out the action, though not explicitly expressed. sili xordæn slap XORDÆN be slapped lægæd xordæn kick XORDÆN get kicked mo∫t xordæn fist XORDÆN get punched tæræf anqædr xode∫-o lus kærd ke yek sili xord. guy so himself-ACC pest do.PST.3SG that one slap eat.PST.3SG. The guy made such a pest of himself that he got slapped in the face17. XORDÆN: Affected: Hurting: Weapon: Hand Held Meaning: be wounded or penetrated by a weapon. PV: a sharp, penetrating weapon, usually hand held. Remarks: Intransitive. The weapon must be sharp and directly penetrate the subject. For example, one can be wounded by a gun, but the form *tofæng xordæn (gun eat) does not exist, because it is not the gun that penetrates, but the bullets (see next island). xænjær xordæn scie XORDÆN be hit with a scie t∫aqu xordæn knife XORDÆN be stabbed with a knife ∫æm∫ir xordæn sword XORDÆN be stabbed with a sword gozærkon væsæt-e mahlæke t∫aqu xord. passerby middle-GEN melee knife eat.PST.3SG. The passerby was stabbed in the middle of the melee. XORDÆN: Affected: Hurting: Weapon: Projectile Meaning: be attacked by a projectile weapon. PV: a projectile weapon. Remarks: Intransitive. This island is similar to the previous island, but only differs in the type of weapon used. mu∫æk xordæn missile XORDÆN get hit by a missile tir xordæn bullet XORDÆN get shot with a bullet sæt∫me xordæn pellet XORDÆN get shot with pellets mohafez-e ræis jomhur tir xord. guard-GEN boss republic bullet eat.PST.3SG. The president’s guard got shot.

17 In French slang (argot), one can say “il s’est mangé un pain” (lit. he ate himself some bread) to express someone getting punched (Vanhove, personal communication)

84

XORDÆN: Affected: Hurting: Weapon: Blunt Meaning: be struck with a heavy or blunt instrument. PV: a blunt instrument used for striking. Remarks: Intransitive. The subject usually undergoes a quick blow or repetitive quick hits by the instrument. t∫æko∫ xordæn hammer XORDÆN be hammered potk xordæn mallet XORDÆN be hit with a mallet gu∫tkub xordæn meat-hammer XORDÆN be hit with a meat-hammer in ma∫in xeyli t∫æko∫ xorde æst. this automobile much hammer eat-PTCP be.PRS.3SG. (lit.)This automobile has been hammered quite a bit. This automobile has been repaired often. XORDÆN: Affected: Hurting: Trick Meaning: be tricked. PV: trick. Remarks: Intransitive. The subject of these forms goes through a negative process. These islands are not productive, but represent an island since there are several forms that express a similar idea. hoqqe xordæn trick XORDÆN slight, trick kælæk xordæn trick XORDÆN trick naro xordæn double-cross XORDÆN double-cross bit∫are hæmi∫e sær-e bazi hoqqe mi-xor-æd. helpless always head-GEN game trick PROG-eat.PRS-3SG. The poor guy always gets tricked in games.

XORDÆN: SUFFERING

The following LVCs express suffering caused by a process or condition affecting a person physically or mentally. The cause of the suffering is usually an unintended result of an action. This is one of the only sets in the system that expresses abstract notions which otherwise mostly occur with the generic LV kærdæn ‘to do’. The LVCs in these islands are all atelic, activity verbs: they express durational conditions.

85

XORDÆN[to eat]

suffering

condition

emotion

qose [grief]

æsæf [regret]

æzab [torment]

gorosnegi[hunger]

gij [dizziness]

særma [cold]

XORDÆN: Suffering: Emotional Meaning: suffer from a negative emotion. PV: emotion of regret, sorrow or grief. Remarks: Intransitive. The LVCs express the durational suffering from an emotional burden. This emotion is a continuous, irrepressible, negative feeling that has to be suffered as the result of ones personal actions or experience. nedamæt xordæn regret XORDÆN regret qose xordæn concern XORDÆN worry, be concerned ænduh xordæn sorrow XORDÆN grieve hæmi∫e qosey-e færda ra mi-xor-æd. always concern-GEN tomorrow ACC PROG-eat.PRS-3SG. She always worries about the future. XORDÆN: Suffering: Physical Meaning: suffer from a physical condition that could cause bodily damage. PV: a natural but uncomfortable condition that causes suffering or might entail more serious ailment. Remarks: Intransitive. These LVCs specifically express the condition or the process that causes the suffering, and not the symptoms. One cannot say *deldærd xordæn (stomach ache eat) since this is a symptom (e.g. of hunger) and not a condition that causes suffering. Nor can one say *særgije xordæn (vertigo eat) which is a symptom of dizziness gij xordæn. gorosnegi xordæn hunger XORDÆN suffer from hunger særma xordæn cold XORDÆN catch cold (from the cold) giji xordæn dizziness XORDÆN get dizzy

86

u fekr mi-kon-æd æz særma særma xord-e æst. he thought PROG-do.PRS-3SG from cold cold eat-PCTP be.PRS.3SG. He thinks he has caught a cold from the cold weather.

XORDÆN: USURPING

XORDÆN[to eat]

usurping

nozul [interest]

re!ve [bribe]

sævari [riding]

Meaning: Exploit service or property. PV: the type of good that is being taken advantage of. Remarks: Intransitive. The LVCs in this island express the notion of taking advantage of another person’s labor or property. Here, the original meaning of xordæn, ‘eat,’ emerges in a metaphorical expression denoting gluttony. re∫ve xordæn bribe XORDÆN accept a bribe nozul xordæn interest XORDÆN charge interest pul xordæn money XORDÆN embezzle or extract money hoquq-e kæm ba’es-e re∫ve xord-æn ziyad ∫od-e æst. salary-GEN small cause-GEN bribe eat-INF much become-PCTP be.PRS.3SG. Low salaries have become the cause of much bribery.

XORDÆN: AGITATED

In these LVCs, the subject undergoes certain types of motion. The motion is usually unintentional on the part of the subject and often repetitive.

87

XORDÆN[to eat]

motion

rotation

general

qælt

[sommersault]

vul

[slither]

pit!

[twist]

sekændari

[stumble]

tækan

[motion]

XORDÆN: Agitated: General Meaning: move. PV: type of movement. Remarks: Intransitive. These LVCs express non-goal oriented movement, usually non-volitional. The movement results from an internal, uncontrollable condition or an external agent, such as twitching from muscle spasms (internal) or being shaken by someone to be woken up (external). tækan xordæn movement XORDÆN jerk, shake, wag telo telo xordæn sway XORDÆN sway vul xordæn fidget XORDÆN fidget bæt∫e æz bihoselegi hæmæ∫ vul mi-xord. child from boredom constantly fidget PROG-eat.PRS-3SG. The kid constantly fidgeted from boredom. XORDÆN: Agitated: Rotation Meaning: rotate. PV: type of rotational movement. Remarks: Intransitive. Similar to the previous island, this set of LVCs express uncontrollable motions, but involve the rotation or turning of the subject. pit∫ xordæn roll XORDÆN be rolled qælt xordæn flip XORDÆN get flipped mæl’æq xordæn somersault XORDÆN go into a somersault, flip over ma∫in-e mosabeqe seta mæl’æq xord. car-GEN race three somersault eat.PST.3SG. The race car flipped over three times.

88

3.4.3 GEREFTÆN

This verb has always had two different, but related meanings that persist in Modern Persian. The first meaning can be characterized globally by verbs like hold and grab (also block), and the second by verbs like get, receive, and obtain. The two possible argument structures appear below.

GEREFTÆN (HOLD) Syntax: NP1 NP2 GEREFTÆN

GEREFTÆN (GET)

Syntax: NP1 NP2 (FROM NP3) GEREFTÆN The subject of the full verb is a proto-agent in all cases, except for the case where the meaning of gereftæn is closest to receive, where the subject is a proto-patient. Given the argument structure of the full verb gereftæn above, we see that it takes at most three arguments, one grammatical subject, one grammatical object, and, in certain cases, an optional grammatical indirect object (the source from which the object is obtained). When gereftæn means block, the subject can be either the one blocking the object, or the verb can have a single argument in which case the subject would express a passage that has been blocked (with no overt object).

As an LV, gereftæn is telic, a goal-oriented or resultative action where an argument goes through a change of state or brings into possession. Karimi-Doustan (1997) classifies this verb as a Transition dynamic light verb. He claims that like other transition LVs, gereftæn provides a slot for an internal argument who’s change of state results in the telicity of the verb. Since gereftæn denotes momentaneous and telic achievements, it is not durative, though it can be a temporally extended achievement in certain contexts. Karimi-Doustan claims that gereftæn is the only LV that can be either Initiatory or Transitory (the first case accepts proto-agent as subject, the second proto-patient). The analogous verb “get” in English also has similar grammatical properties where the “active meaning is hard to disentangle from the non-volitional ones” (Dowty, 1991, p581)

In LV cases, gereftæn can take one or two arguments. Most LVCs constructed with gereftæn are intransitive though a couple of islands produce transitive LVCs. The subject, almost always human, can be either a proto-agent or a proto-patient. Gereftæn is one of the most frequent in the system, and consequently has many islands.

Sharifi (1975, p 469) signals that “most of the derived semantic units [of the LV] seem to share somewhat the basic meaning of the simple verb root.” Only the context, the PV, and arguments, can determine the particular meaning expressed by the verb gereftæn. In all forms, there is the idea of being ‘held together’, sometimes the idea of ‘seizure, overtaking’. Cramping, plugging, grasping, choking, seizing, or stumbling are all processes where ‘seizure’ or ‘taking’ changes the state of a patient or experiencer by a force which creates a closing or cramping situation (Sharifi, 1975: 473). This patient is usually the direct object of the LVC, though it can also be the subject as we will observe in the examples below.

Except for a handful of adjectives and prepositions, gereftæn combines almost exclusively with nominal PVs. These nouns usually express the entity that is obtained through the action. A first classification of the different meanings portrayed with the contribution of gereftæn in

89

LVCs yields six main categories: removal, grasping, ritual, covering, getting (from someone), and taking upon oneself.

take upononeself

GEREFTÆN

grasp

remove/extract

ritual

get fromsomeone

cover

GEREFTÆN: GRASP

This group of islands express notions related to catching or grabbing hold of. The subject carries out an action that results in him grasping something that was out of his control, or it was fleeting and about to disappear. In other words, the object is captured or restrained within the grip or constraints imposed by the subject. The ideas of domination or regulation emerge in these forms. As seen above, the several meanings of the full verb gereftæn include grasping. However, in the LVC instances, the original meaning can extend to figurative or more nuanced senses.

90

GEREFTÆN

grasp

capture

outlaw

wild animal

film [film]

video [video]

æks [photo]

mahi [fish]

!ekar [hunt]

dozd [thief]

æsir [hostage]

zendani [prisonner]

grip

consider

nædide [unseen]

geran [dear]

ævæzi [error]

enan [reins]

tormoz [breaks]

ni!gun [pinch]

catch

pærvane [butterfly]

GEREFTÆN: Grasp: Consider Meaning: understand or grasp something in the manner expressed by the PV. PV: an adjective expressing an evaluation. Remarks: Transitive. This is one of those rare islands that accepts adjectives as PV. The construction expresses how the subject perceives, interprets, or judges the object. This construction can be compared to the English “do you get it” or “take it easy, take it seriously, etc.” The judgement always occurs subjectively, the subject seizes the situtation in a particular way. In some cases, he is consciously making the decision to understand something in a particular way (e.g. nædide). In others, he is not conscious (e.g ævæzi). geran gereftæn dear GEREFTÆN consider dear nædide gereftæn unseen GEREFTÆN disregard, pass off as unseen ævæzi gereftæn erroneous GEREFTÆN mistake smthg for smthg else mærdæk-e mæst mæn o ba dust-æ∫ ævæzi gereft. guy-GEN drunk me-ACC with friend-3SG erroneous take.PST.3SG. The drunk guy took me for his friend. GEREFTÆN: Grasp: Record Meaning: record an image or sound on media. PV: any type of media. Remarks: Intransitive, entity being recorded appears as indirect object. This construction is analogous to the English catch on film or take a picture (also French prendre une photo). The use of the type of media to verbalize recording on that media is also used in English and French, in to tape or filmer, respectively. A fleeting scene (composition, lighting, action, and also sound) change quickly, but are brought into the grasp of the subject through recording. Photographing a scene is different from painting it because of the instantaneous and exact nature of the former. Hence, painting cannot occur as a construction in this island.

91

film gereftæn film GEREFTÆN record on film æks gereftæn photo GEREFTÆN photograph vidio gereftæn video GEREFTÆN record on video æz hæme kar-e bæt∫e vidio mi-gereft. from all action-GEN child video PROG-take.PST.3SG. He shot videos of the kid’s every act. GEREFTÆN: Grasp: Grip Meaning: grip with hands or mechanical means. PV: physical object that is grasped to serve its function, usually of control. Remarks: Intransitive. The subject gains more control over something by gripping a functional part of it (e.g. the speed by clutching the break, the reins to control the horse). The means of control are grasped. This can be compared to the English or French “hold a child’s hand” or “tenir la main d’un enfant” where the action of holding or clutching insinuates controling (as opposed to just holding). tormoz gereftæn brakes GEREFTÆN hit the breaks enan gereftæn reins GEREFTÆN pull a horse’s reins kelat∫ gereftæn clutch GEREFTÆN press on the clutch (manual car) sævarkar dorost be moqe enan gereft. rider right in time rein take.PST.3SG. The rider took the rein just in time. GEREFTÆN: Grasp: Catch: Criminal Meaning: capture. PV: a captive or an outlaw. Remarks: Intransitive. Again, the entity apprehended is fleeing or free and must be brought under the control of some person or authority. dozd gereftæn thief GEREFTÆN apprehend a thief æsir gereftæn hostage GEREFTÆN take a hostage zendani gereftæn prisoner GEREFTÆN obtain prisoners be omid-e jælb-e næzær-e donya t∫ænd æsir gereftænd. in hope-GEN attraction-GEN view- GEN world several hostage take.PST-3PL. They took several hostages in the hope of attracting the world’s attention. GEREFTÆN: Grasp: Catch: Wild Animal Meaning: hunt. PV: a wild animal. Remarks: Intransitive. The productivity of this island is restricted to wild animals, or animals that are difficult to get into the subjects grasp. Wild animals are free, not domestic ones. Therefore, the word dog can be used in this construction if it refers to a wild dog or a runnaway dog, not the subject’s own dog. mahi gereftæn fish GEREFTÆN fish ∫ekar gereftæn prey GEREFTÆN hunt pælæng gereftæn leopard GEREFTÆN hunt a leopard

92

adæm-e zeræng æz abe gelalud mahi mi-gir-æd. person-GEN agile from water muddy fish PROG-take.PRS-3SG. Opportunists catch fish in muddy waters.

GEREFTÆN: RITUAL

The following set of islands expresses carrying out societal or personal rituals or routines.

GEREFTÆN

performritual

social

personal

ærusi [wedding]

jæ!n [celebration]

tævælod [birthday]

du! [shower]

ruze [fast]

rejim [diet]

zæn [wife]

doust pesar [boyfriend]

marital

taharæt[hygene]

GEREFTÆN: Ritual: Marital Meaning: get a sexual/life partner. PV: opposite sex partner. Remarks: Intransitive. The PV usually expresses a woman partner, where the subject is a male. The only case where the partner expressed is a man appears below (boyfriend). To express a women getting a husband, the generic verb kærdæn is used in ∫ohær kærdæn ‘husband do.’ A similar expression exists in English: take a wife. The idea of possession is clear in this island, especially considering the male dominated nature of traditional societies. Accordingly, the usage of these forms are less prevalent in cosmopolitan cities. siqe gereftæn temporary wife GEREFTÆN get married temporarily zæn gereftæn woman GEREFTÆN take a wife doustpesar gereftæn boyfriend GEREFTÆN get a boyfriend u dær jævani-ha-æ∫ zæn gereft. he in youth-PL-3SG wife take.PST.3SG. He took a wife in his youth. Comment: Though not a very productive LVC island (since there is only a limited number of marriage or partnership possibilities), gereftæn has come to mean ‘to take as a wife’ in certain contexts. This meaning, allotted to the full verb gereftæn, could be due to the existence and high usage (or social importance) of this island. An example of this usage:

93

Hassan doxtær-e Hossein ra gereft Hassan girl-GEN Hossein ACC get.PST.3SG. Hassan married Hossein’s daughter. GEREFTÆN: Ritual: Societal Meaning: have a gathering to celebrate or commemorate. PV: type of festive/socially important occasion. Remarks: Intransitive. These verb forms are similar to the English throw a wedding/party/etc or have a wedding/party/etc. xætm gereftæn memorial GEREFTÆN have a memorial service ærusi gereftæn wedding GEREFTÆN throw a wedding tævælod gereftæn birthday GEREFTÆN have a birthday party næmorde bæra-æ∫ xætm gereft-ænd. undead for-3SG memorial take.PST-3PL. They celebrated his demise prematurely. GEREFTÆN: Ritual: Personal Meaning: partaking in a ritual or action that has set form. PV: ritual or habit that involves the cleansing of the body. Remarks: Intransitive. These forms all involve a conscious partaking in a controlled or calculated process that results in the physical cleansing of the body, though sometimes involving spiritual or religious beliefs. These rituals bring “something” of value to a person who partakes in them. The English and French terms take a shower/bath or prendre une douche/un bain might be related to this island. The equivalent form also exists in Persian: du∫ gereftæn. ruze gereftæn fast GEREFTÆN fast vozu gereftæn ablution GEREFTÆN cleanse oneself before prayer rejim gereftæn diet GEREFTÆN diet ba ruze gereft-æn anha ehsas-e næsdiki mi-kon-ænd. with fast take-INF they feeling-GEN closeness PROG-do.PRS-3PL. The sense a closeness by fasting.

GEREFTÆN: COVER

These islands express the idea of covering or blocking. This action can either be oriented towards some object or towards the subject. In other words, in some islands, the subject is a proto-agent, in others, it is a proto-patient.

94

GEREFTÆN[to obtain]

cover

enrobe

seizecramp

næfæs [breath]

xak [dust]

tæla[gold-leaf]

qey [blear]

oq [gag]

vir [craving]

!a! [urine]

become covered

cover

qælb [heart]

gat! [plaster]

bærf[snow]

a!qal [trash]

atæ![fire]

GEREFTÆN: Cover: Enrobe: Cover Meaning: wrap or cast. PV: a soft and often viscous material that then hardens to cover the object. Remarks: Transitive. The subject is a volitional entity who uses the material expressed by the PV to wrap or cast the object completely. The form gel gereftæn can also be used idiomatically, meaning to close up indefinetly (e.g. a store). This form expresses the idea of blocking passage, or covering up as not to let anything through to the entity covered. Also, it is interesting to note that the transitive nature of this construction is due to the PV not representing the object that is attained, obtained or possessed, but the instrument used to carry out the action. tæla gereftæn gold GEREFTÆN gold plate gæt∫ gereftæn plaster GEREFTÆN cover with plaster gel gereftæn mud GEREFTÆN cover with mud dærz-e divar ra gæt∫ gereft-ænd. crack-GEN wall ACC plaster take.PST-3SG. They plastered over the crack in the wall. GEREFTÆN: Cover: Enrobe: Become Covered Meaning: natural accumulation of a substance on a surface. PV: organic material (usually degrading). Remarks: Transitive. This construction has an interesting argument structure in that the semantic subject (the entity that is covered) is expressed as the object, and there is no overt subject. The PV seems to take the place of the grammatical subject. This construction expresses a a gradual and natural process of a build-up forming on the surface of something.

95

This process usually results from neglect, or the lack of maintenance of the object by an external entity. læjæn gereftæn slime GEREFTÆN get covered with slime qey gereftæn blear GEREFTÆN become bleary xak gereftæn dust GEREFTÆN get covered in dust dær ærz-e t∫ænd dæqiqe hæmet∫iz ra xak gereft. in width-GEN few minute everything ACC dust take.PST.3SG. Everything was covered by dust in a few minutes. GEREFTÆN: Cover: Be Overwhelmed Meaning: be overwhelmed with an urge or condition. PV: an involuntary psychological or biological urge that must be manifested or repressed. Remarks: This LVC has a special structure. The semantic subject is always a conscious being expressed through a pronomial suffix attached to the PV (here it appears in the 1st person singular). This construction and other similar constructions are presented with further detail in chapter 5. The construction expresses a hidden force or a natural force that is out of the control of the subject. This force seizes the subject. oq-æm gereft gag GEREFTÆN to feel like vomiting xænde-æm gereft laughter GEREFTÆN to be seized with laughter bazi kærdæn-æm gereft play do GEREFTÆN to become playful æz hærf-ha-ye an mærd-e mohem xænde-æm gereft. from word-PL-GEN that man-GEN important laughter-1SG take.PST.3SG. I couldn’t but laugh at that important man’s pronouncements. GEREFTÆN: Cover: Cramping Meaning: cramp. PV: body part. Remarks: This island also has a special kind of structure, because the PV must always be followed by a pronominal possessive suffix. Below, breath is not a body part, but the meaning of the corresponding construction is related to the other forms in this island. Here, my breath ceases to be smoothly functioning, as with the body parts (perhaps because of the cramping feeling one senses in the lungs during this process). As with the previous island, a natural or biological force seizes the subject. næfæs-æm gereft my breath GEREFTÆN I am short of breath pa-m gereft my foot GEREFTÆN I have a cramp in my foot qælb-æm gereft my heart GEREFTÆN I have chest pains væsæte mosabeqe pa-∫ gereft. middle-GEN race leg-3SG take.PST.3SG. He got a leg cramp in the middle of the race.

GEREFTÆN: OBTAIN

This clusters of islands expresses meanings closest to that of obtain. The subject is always an agent accepting or obtaining something from some external source. Unlike the first island

96

above, labelled grasp, this group of islands express actions that require a greater involvement of human mental capacities. While they express the transfer of an entity or a service (from a source to the subject), these often include acts of persuasion, agreement, knowledge, etc.


service

represent

karmænd[worker]

næqa! [painter]

pæræstar [nurse]

vækil [lawyer]

gævah [witness]

vaset[mediator]

obtain

lessonresponse

money

help

ejare [rent]

nozul [interest]

vam [loan]

yari [help]

komæk [help]

re!ve [bribe]

dærs [lesson]

sæboq [precedence]

pænd [advice]æhval

[health]

jævab [answer] xæbær

[news]

info

employ

safety/cure

permission

ælaj [cure]

pænah [refuge] !æfa

[cure]

roxsæt [permission]

ejaze [permission]

reserve

bind

qol [promise]

væ'ede [promise]

ja [place]

væqt [time]

GEREFTÆN: Obtain: Reserve Meaning: reserve a place or time. PV: place or time. Remarks: Intransitive. The subject reserves a time or a place. A similar expression exists in English (e.g. get an appointment). ja gereftæn place GEREFTÆN reserve a seat/place væqt gereftæn time GEREFTÆN get an appointment nobæt gereftæn turn GEREFTÆN take a turn bæra-ye didæn-e ræis-æ∫ væqt gereft. for see.INF-GEN boss-3SG time get.PST.3SG. He got an appointment to see his boss. GEREFTÆN: Obtain: Permission Meaning: get permission or allocation. PV: permission or allocation. Remarks: Intransitive.

97

æzn gereftæn leave GEREFTÆN get permission roxsæt gereftæn permission GEREFTÆN get permission ejaze gereftæn permission GEREFTÆN get permission u qæbl æz residæn ejaze næ-gereft-e bud. he before from arrive.INF permission NEG-take.PST.PTCP be.PST.3SG. He hadn’t gotten permission before arriving. GEREFTÆN: Obtain: Money Meaning: charge money. PV: a fee or bribe. Remarks: Intransitive. An optional indirect object can express the source from which the subject obtains the money. The money transfers possession in return for a service, fee, interest, or bribe. ejare gereftæn rent GEREFTÆN charge rent re∫ve gereftæn bribe GEREFTÆN accept a bribe nozul gereftæn interest GEREFTÆN charge interest re∫ve gereftæn yek t∫iz-e addi ∫od-e æst. bribe take. INF one thing-GEN habitual become-PTCP be.PRS.3SG. Taking a bribe has become a normal thing. GEREFTÆN: Obtain: Info: Lesson Meaning: learn a lesson or getting advice. PV: lesson or advice. Remarks: Intransitive. The indirect object expresses the event or the conscious being who bestows the new knowledge or wisdom upon the subject. Not only does a external source make available the knowledge, but the subject often extracts some knowledge from the event, especially when the source is not a conscious being, but an event. dærs gereftæn lesson GEREFTÆN learn a lesson sæboq gereftæn precedence GEREFTÆN understand the moral of... pænd gereftæn advice GEREFTÆN take a lesson pænd gereftæn yek bar khub æst, dobar tænbih. lesson take.INF one time good be.PRS.3SG, twice punishment. Taking a lesson the first time is good, the second time it’s for punishment. GEREFTÆN: Obtain: Info: Response Meaning: obtain information or an answer. PV: information that must be extracted from an external source. Remarks: Intransitive. The productivity of this island is limited probably due to the abstract nature of the PV. There are only a limited number of lexemes that mean answer or information. jævab gereftæn answer GEREFTÆN get an answer

98

pasox gereftæn response GEREFTÆN get a response xæbær gereftæn news GEREFTÆN get news of ba telefon æz anha xæbær gereft-æm. by telephone from them news take.PST-1SG. I called for some news about them. GEREFTÆN: Obtain: Service: Represent Meaning: assign a representative. PV: person that acts as official representative or witness. Remarks: Intransitive. The subject hires or assigns a representative for himself. The service transfered is not of physical labor, but of social or legal expertise. vaset gereftæn intermediary GEREFTÆN hire an intermediary vækil gereftæn lawyer GEREFTÆN engage a lawyer gævah gereftæn witness GEREFTÆN take as witness bæraye feysæle-ye kar vaset gereft-æm. for conclusion-GEN task intermediary take.PST-1SG. I introduced an intermediary in order to resolve the issue. GEREFTÆN: Obtain: Service: Employ Meaning: hire. PV: a type of worker. Remarks: Intransitive. This LVC is usually used when a single person (not an institution) employs someone. Furthermore, the less physical work a job requires, the less probable it is that this construction is used. In other cases, the type of worker appears as an object to the more general LVC estexdam kærdæn ‘employ do.’ kargær gereftæn worker GEREFTÆN employ a worker næqa∫ gereftæn painter GEREFTÆN employ a painter pæræstar gereftæn nurse GEREFTÆN employ a nurse bæraye ræng kardæn-e otaq næqa∫ gereft-i? for paint do.INF-EZ room painter take.PST-2SG? Did you hire a painter to paint the room? GEREFTÆN: Obtain: Service: Help Meaning: receive help. PV: physical, moral, or financial help. Remarks: Intransitive. yari gereftæn aid GEREFTÆN get help komæk gereftæn help GEREFTÆN get help vam gereftæn loan GEREFTÆN obtain a loan bæraye xæridæn-e xane-æ∫ vam gereft. for buy.INF-GEN house-3SG loan take.PST.3SG. He took out a loan to purchase his house.

99

GEREFTÆN: Obtain: Safety Meaning: get refuge or be cured. PV: refuge or cure. Remarks: Intransitive. The subject is the entity being saved, intentionally or by happenstance. ∫æfa gereftæn cure GEREFTÆN be cured pænah gereftæn refuge GEREFTÆN be provided refuge nejat gereftæn saved GEREFTÆN be save ba kæmi zæhmæt æz nabesamani nejat gereft. with little work from unsettledness save take.PST.3SG. With a bit of hard work he was saved from his unsettled situation. GEREFTÆN: Obtain: Bind Meaning: bind someone to perform an act or a deed. PV: word expressing a binding. Remarks: The subject is the entity that binds someone (expressed by an indirect object). If the deed is implied in the context of the conversation it is not required in a sentence produced with this LVC. qol gereftæn promise GEREFTÆN make promise væ’de gereftæn promise GEREFTÆN make promise madær-æm æz-æm qol gereft ke kare badi næ-kon-æm. mother-1SG from-1SG promise get.PST.3SG that deed-GEN bad NEG-do.PRS-1SG. My mother made me promise her that I wouldn’t do anything bad.

GEREFTÆN: RECEIVE

These islands are similar in nature to the previous set, but in these cases, the subject does not exert any effort to attain the things that he receives. The subject is a proto-patient, and unlike the previous islands, the goods or services received are not usually to his benefit.

GEREFTÆN

naxo!i [unwell]

jan[life]

bu[smell]

færman [order]

sefare! [order]

dæstur [order]

order

abele [pox]

særætan [cancer]

disease

physical

receive

GEREFTÆN: Receive: Illness Meaning: catch an illness.

100

PV: a type of illness. Remarks: Intransitive, similar to catching a cold or getting cancer in English.. særætan gereftæn cancer GEREFTÆN get cancer abele gereftæn pox GEREFTÆN catch the pox særgije gereftæn vertigo GEREFTÆN get dizzy adæm æz tæma∫a-ye film-e hit∫kak særgije mi-gir-æd. person from watching-GEN film-GEN Hitchcock vertigo PROG-take.PRS-3SG. One gets vertigo from watching Hitchcock’s movie. GEREFTÆN: Receive: Order Meaning: receive an order. PV: order or request. Remarks: Intransitive. sefare∫ gereftæn order GEREFTÆN take an order dæstur gereftæn directions GEREFTÆN be directed færman gereftæn decree GEREFTÆN be commanded garson ba ehteram væ sæbr æz anha sefare∫ gereft. waiter with respect and patience from them order take.PST.3SG. The waiter took their order with respect and patience. GEREFTÆN: Receive: Physical Meaning: absorb. PV: something that can be imbued into things and generally permiate to fully engulf it. Remarks: Intransitive. atæ∫ gereftæn fire GEREFTÆN catch on fire jan gereftæn life GEREFTÆN get energized bu gereftæn smell GEREFTÆN get smelly lebas-æm tu-ye kafeye por æz sigari bu gereft clothes-1SG in-GEN café- GEN full of cigarette-smokers smell take.PST.3SG. My clothes became smelly in the café full of smokers.

GEREFTÆN: ASSUME

The subject of these islands assumes a position or a shape, he takes a configuration upon itself. Depending on the construction, the subject may or may not be actively involved in attaining the configuration or trait acquired.

101

assume


settlement

look

position

pose

shape

position/ place

relative

arame! [calm]

næzm [order]

!ohræt [fame]

qærar [position]

ja[place]

væqt [time]

tæræf [side]

donbal [after]

æz sær[from head]

!ekl [form]

suræt [face]

næx! [design]

jest [gesture]

qollab [hook]

qiyafe [face]

trait

saman [settled]

GEREFTÆN: Assume: Position: Relative Meaning: take a position relative to something else. PV: a physical position. Remarks: Transitive. The object is the entity relative to which the subject positions himself. Though the PV expresses a physical position, the construction never occurs in contexts of physical configuration. Rather, these constructions are used to express the subject’s position with respect to more intangible entities. For example, singing a song from the beginning can be expressed with æz sær gereftæn (below). Or, following up on a legal case can be expressed with donbal gereftæn (below). The action refers to a process of rerunning or developing on a process from a certain point. This is probably why it wouldn’t make sense to use the noun end in these constructions, since there is no continuation after that point. æz sær gereftæn from head GEREFTÆN take it from the start (restart) donbal gereftæn after GEREFTÆN follow up violonzæn ahæng ra æz sær gereft. violinist song ACC from head take.PST.3SG. The violinist started the piece over. GEREFTÆN: Assume: Position: Position/Place Meaning: take up time or place. PV: a time/space nominal.

102

Remarks: Intransitive. These LVCs are in fact polysemous (see reserve island above). Each of these can be used to mean waste or take up room/time/place, in the sense of this computer takes up space on this table. væqt gereftæn time GEREFTÆN take up time ja gereftæn place GEREFTÆN take up room qærar gereftæn position GEREFTÆN be placed neve∫tæn-e mæ∫q væqt mi-gir-æd. write.INF-GEN homework time PROG-get.PRS-3SG. Writing homework takes time. GEREFTÆN: Assume: Condition: Shape Meaning: develop into a coherent or recognizable shape. PV: a general word for shape or form. Remarks: Intransitive. The subject of this construction doesn’t partake in the process described. Rather, this usually gradual process occurs autonomously, though usually in consequence of the direct actions of one or more external agents (e.g. a sculpture takes shape because the sculptor sculpts it or the dance choreography acquired a coherent form as all the dancers danced their part). ∫ekl gereftæn form GEREFTÆN take shape næq∫ gereftæn design GEREFTÆN be outlined suræt gereftæn face GEREFTÆN get a form zære zære mojæsæme-ye ostad ∫ekl gereft. little little sculpture-GEN master shape take.PST.3SG. Little by little the master’s sculpture took shape. GEREFTÆN: Assume: Condition: Pose Meaning: put one’s body in a specific and recognizable configuration. PV: shape or configuration. Remarks: Intransitive. The subject volitionally configures parts of his body to attain the required position expressed by the PV. qollab gereftæn hook GEREFTÆN make a hook with the hands jest gereftæn gesture GEREFTÆN strike a pose qiyafe gereftæn face GEREFTÆN make a face ali bæraye mohsen qollab gereft ta æz divar bala ræ-væd Ali for Mohsen hook take.PST.3SG until from wall up go.PRS-3SG. Ali made a hook for Mohsen so that he could climb the wall. GEREFTÆN: Assume: Trait Meaning: acquire a trait. PV: a psychological or abstract trait. Remarks: Intransitive. The subject does not directly partake in his own acquisition of the trait. This construction expresses the end of a potentially gradual process through which the subject acquires this trait.

103

∫ohræt gereftæn fame GEREFTÆN become famous ræng gereftæn color GEREFTÆN get tanned/colored aram gereftæn calmness GEREFTÆN become calm bæt∫e ba didæn-e madær-æ∫ aram gereft. child with see.INF-GEN mother-3SG calmness take.PST.3SG. The child calmed down at the sight of his mother. GEREFTÆN: Assume: Settlement Meaning: become settled or organised. PV: settlement or organisation. Remarks: Transitive. The subject falls into place or is arranged (usually without an explicit external force). These LVCs can be used for abstract notions (e.g. life) as well as concrete ones (e.g. books on a bookshelf).

næzm gereftæn organization GEREFTÆN become organized, ordered tærtib gereftæn hierarchy GEREFTÆN be arranged saman gereftæn welfare GEREFTÆN become settled bæt∫e-ha dær kelas ba vorude nazem næzm gereft-ænd. kid-PL in class with entrance-GEN monitor organization take.PST-3PL. When the monitor entered the class the kids became orderly.

GEREFTÆN: REMOVE

This group of LVCs express removing something from an object. The item removed is part of the object as a whole.

GEREFTÆN

remove

growth(external)

surface

æbru[eye brow]

naxon [nail]

gærd [dust]

læke [stain]

t!erk [dirt]

golab[rose-water]

!ire [syrup]

ab mive [fruit juice]

extract(internal)

dæmaq [nose]

104

GEREFTÆN: Remove: Extract Meaning: extracting an essence or liquid from its original source. PV: an essence or liquid. Remarks: Intransitive. (repeated from section 3.1.3) The liquids expressed by the PV are always within an object and don’t flow out naturally, but can be extracted. In other words, when taking water from the ocean, we cannot say ab gereftæn (water GEREFTÆN), while we could use that to express squeezing juice out of a fruit or even taking water from a deep well using a bucket, but not a pump. Xun gereftæn specifically means extracting blood from a vein, and not grabbing a testube of blood from a lab or covering something with blood and letting it dry (as this is not the function of blood).

ab mive gereftæn fruit juice GEREFTÆN juice golab gereftæn rose-water GEREFTÆN extract rose-water ∫ire gereftæn syrup GEREFTÆN extract syrup sorur ba metod-e sonæti golab mi-gir-æd Sorur with method-GEN traditional rose-water PROG-take.PRS-3SG. Sorur extracts rose-water the traditional way. GEREFTÆN: Remove: Growth Meaning: remove of human tissue that is in excess. PV: human tissue that grows external to the body. Remarks : Intransitive. This island expresses the removal of biological matter that tends to grow in excess (e.g. sheep hair). The final state of this action is the absence of any excess of the matter in question, which is different than the idea of trimming. mu gereftæn hair GEREFTÆN remove hair pæ∫m gereftæn fur GEREFTÆN remove fur naxon gereftæn nail GEREFTÆN clip nails æz xers mu gereftæn qænimæt æst. from bear hair take.INF contentment be.PRS.3SG. Be content with removing just one hair from a bear. GEREFTÆN: Remove: Residue (surface) Meaning: remove something that has become encrusted on another entity. PV: residual matter that collects on the surface of objects. Remarks: Intransitive. This construction rarely occurs in a sentence, but is used as a base for a common nominal derivation. t∫erk gereftæn dead skin GEREFTÆN exfoliate gærd gereftæn dust GEREFTÆN dust off qey gereftæn blear GEREFTÆN remove bleariness nezafætt∫i hæme ja ra gærd gereft cleaner all place ACC dust take.PRS.3SG. The cleaner dusted off the whole place.

105

3.4.4 DADÆN

Dadæn, in Modern Persian, expresses the notion of giving, offering, or making available. The subject gives an obect to a second entity. The argument structure is as follows:

DADÆN (GAVE) Syntax: NP1 NP2 (be NP3) DADÆN

The subject, NP1, is always an agent and the object, NP2, is always the entity that is being transferred to NP3 which always occurs as an indirect object headed by be ‘to.’ As in English and French, most of the lexically simple occurrences of this verb (as a simple verb) express a ditransitive action where the central case is when one entity causes another entity to receive a third entity.

The verb give is polysemous in many languages, including English, and French donner. Viberg (2002) demonstrates that the Swedish verb ge ‘give’ is highly polysemous, with some forms similar to the English verb give. In that paper, the simple verb ge was studied alongside its occurrences in complex constructions (LVCs or particle and verb). The different categories of definitions extracted include concrete possession, yield, surrender, existence, production, emergence of sound, emotion, departure, and power and possibility (p.5). Examples of these occurrences of ge are presented diagrammatically, where the prototypical case occurs in the middle (meaning transfer of something from someone to someone else).

The LVCs of dadæn express a number of loosely related notions. Consequently, many islands emerge in the semantic space of this LV. The continuum is from the very tangible and concrete exchange of material to rather abstract events. The latter types include actions such as altering of states of mind, transfer of information and knowledge, yielding of social or political right, power, or permission, and providing emotional or spiritual aid. Imposing physical changes to states of thing, and emanating tangible signals, especially visible or audible, are examples of the former type. The “physical” or tangible part of the LVC continuum created with dadæn is closer to the main meaning of the full meaning of the verb and as the PVs becomes more abstract, the LVCs become more semantically opaque.

The resulting LVCs can be transitive or intransitive and the PV is usually a noun.

The LVCs constructed with dadæn fall under five main categories: offer, relinquish, impose, conduct, and produce.

106

DADÆN[to give]

produce

offer

impose

relinquish

conduct

Most of the LVCs formed with dadæn take a proto-agent. In some, especially those under the produce category, the subject is non-volitional, although not necessarily a proto-patient. Bringing forth and imposing are two verbal actions that prevail throughout many of the LVCs.

DADÆN: OFFER The meanings expressed in these islands are closest to the full meaning of dadæn. In all these LVCs, a receiver undergoes or receives the transfer, emission, or change of state initiated by the subject. The receiver usually gains something from the interaction.

107

promotion

DADÆN[to give]offer

express

explain

say

knowledge

norishment

provide

help

safety/cure

permit

presentation

render

trait

settlement

pleasure

favor

lesson

response

reserve

social

opportunity

næzær [opinion]

pasox [response]

jævab [answer] neda

[message]

xæbær [news]

e'elam[announcement]

tozih [explanation]

!æhr [delineation]

tæfsil [interpretation]

sæboq [precedence]

dærs [lesson]

ændærz [advice]

næmaye! [show] konferans

[conference]

soxænrani [lecture]

ejare [rent]

pestan [breast]

qors [pill]

nan [bread]

mædæd [help]

komæk [aid]

!æfa [cure]

pænah [refuge]

ælaj [cure]

saman [settlement]

næzm [organization]

tærtib [order]

mehmani [get-together]

!am [dinner]

!irini [sweets]

!ohræt [fame]

sor'æt [speed]

hal [joy]

mæze [taste]

ejaze [permission]


nobæt [turn]

ja [place] væqt [time]

forsæt [chance]

emkan [possibility]

jai [place]

money

product/ service

kerye[let]

nozul[usury]

maliat [taxes]

monetarytæxfif

[discount]

ærzan[cheap]

justice

ensaf[honesty]

hæq[right]

jayezeh[prize]

dæreje[rank]

erteqah[promotion]

okazyon[possibility]

DADÆN: Offer: Express: Explanation Meaning: explain. PV: any lexeme meaning explanation, or clarification, especially one that adds to or brings out unknown details of a subject. Remarks: Transitive. The object is the entity that being clarified or explained. An optional prepositional clause headed by bæraye ‘for’ explicitly expresses the person to which the explanation is offered. These constructions especially expresses the bringing about of unknown details of a subject. This verb form can be compared to the English give an explanation, or give an account.

108

tæfsil dadæn details DADÆN tell with excruciating detail tozih dadæn explanation DADÆN explain, elucidate ∫æhr dadæn description DADÆN give an account of, report guyænde væziyat-e pænahænde-ha ra dæqiq ∫æhr dad. newsperson condition-GEN refugee-PL ACC exact description give.PST.3SG. The news-reader gave a detailed account of the refugees’ conditions. DADÆN: Offer: Express: Knowledge: Announce Meaning: deliver a message or give news of something. PV: any type of unknown message or news. Remarks: Intransitive. The content of the message or news delivered occurs in a relative clause following the LVC. The message can include simple facts to spiritual proclamations and religious edicts. xæbær dadæn news DADÆN give news neda dadæn proclamation DADÆN proclaim e’elam dadæn announcement DADÆN announce radio jayeze gereftæn-e ostad ra xæbær dad. radio prize get.INF-GEN master ACC news give.PST.3SG. The radio gave the news of the master receiving the prize. DADÆN: Offer: Express: Knowledge: Response Meaning: provide one’s answer or reaction to something. PV: any general expression that can be considered a “reply.” Remarks: Intransitive. The response that occurs as PV can be an answer to a question or inquiry, a decision relating to a request, or an expert opinion expressed regarding a subject. The questions need not be explicit. As in English and to a certain degree French (répondre), answering the door and answering the telephone/telegraph/letter, are valid constructions. jævab dadæn answer DADÆN answer pasox dadæn response DADÆN respond næzær dadæn opinion DADÆN give one’s opinion jærah æz næzær dadæn xoddari kærd. surgeon from opinion give.INF self-control do.PST.3SG. The surgeon refused to give an opinion. DADÆN: Offer: Express: Knowledge: Lesson Meaning: give a lesson or advice. PV: lesson or advice. Remarks: Intransitive. The subject makes available some knowledge or directly teaches the indirect object. The subject may not be aware of its role in the process, such as a drunkard being a lesson to the youth. sæboq dadæn precedence DADÆN teach a moral pænd dadæn advice DADÆN advise dærs dadæn lesson DADÆN give a lesson

109

mærde movæffæq pænd ne-mi-dæh-æd, dærs mi-dæh-æd man-GEN success advice NEG-PROG-give.PRS-3SG, lesson PROG-give.PRS-3SG. A successful man doesn’t give advice, he teaches. DADÆN: Offer: Express: Presentation Meaning: give a public performance or presentation. PV: a one-to-many presentation. Remarks: Intransitive. The one-to-many attribute of the presentation type is essential to the PVs of this island. This island is analogous to the English forms give a concert, give a talk/lecture, give a presentation, etc. konferans dadæn conference DADÆN give a talk at a conference næmaye∫ dadæn performance DADÆN perform soxænrani dadæn lecture DADÆN lecture memar-e saxteman dær bareye tarh-æ∫ konferans dad. architect-GEN building in about design-3SG conference give.PST.3SG. The building’s architect gave a conference about its design. DADÆN: Offer: Money Meaning: pay dues. PV: certain types of monetary transfer. Remarks: Intransitive. The money transfer is in return for a service, usage fee, loan interest, or for an illicit favor. Generally valid PVs are considered necessary, albeit burdensome, payments and not voluntary or payments for luxury, such as paying for a concert ticket. ejare dadæn rent DADÆN pay rent re∫ve dadæn bribe DADÆN bribe nozul dadæn interest DADÆN pay interest u ba nozul dadæn movafeq nist. he with interest give.INF agreeable be.PRS.NEG.3SG. He doesn’t agree with paying interest. DADÆN: Offer: Provide: Nourishment Meaning: feed or make food available, to administer medicine (for medicine: prescribe). PV: a type of food/medicine. Remarks: Transitive. The object of this construction is the person receiving the nourishment or medicine. The subject provides the food or medicine as alimentation, not just as an offering. The medicine that is administered must be orally taken. Injected medicines are not valid PVs. But all types of medicine, including the method of administration, are valid PVs when they are prescribed. These distinctions go beyond the English where LVCs with“give + medication” encompass both prescription and administration. ∫ir dadæn milk DADÆN breastfeed qæza dadæn food DADÆN feed nan dadæn bread DADÆN provide bread

110

hær ankæs ke dændæn dæh-æd nan dæh-æd each who that tooth give.PRS-3SG, bread give.PRS-3SG. That who gives one teeth provides one bread. (Verse of a poem: whoever creates life, sustains it) DADÆN: Offer: Provide: Product Meaning: offer a product for temporary use (sometimes against a fee). PV: rent or lend. Remarks: Transitive. The subject rents out, or loans, the object to another person or entity (expressed as an indirect object). For example: Homayoun xane ra be Hassan ejare dad, ‘Homayoun rented the house to Hassan.’ ejare dadæn rent DADÆN rent keraye dadæn rent DADÆN rent qærz dadæn loan DADÆN lend bank-ha bæraye komæk dær tose-e eqtesad asan qærz mi-dah-ænd. bank-PL for help in expansion-GEN economy easy loan PROG-give.PRS-3PL. Banks give easy loans in order to help expand the economy. DADÆN: Offer: Provide: Help Meaning: provide help: personal, physical, or financial. PV: a type of aid, or the notion of assistance. Remarks: Intransitive, the receiver of the help occurs as an indirect object or as a direct object: Ali Hassan ra yari dad, or Ali be Hassan yari dad, are both valid. There is a subtle difference, in that the second case suggests that Hassan had asked for help whereas in the first case it is not clear if help had been solicited or not. The meaning of these constructions is reminiscent of the English give a hand. The PVs meaning help can also occur with the LV kærdæn to express the same meaning. yari dadæn aid DADÆN help komæk dadæn help DADÆN help vam dadæn loan DADÆN give a loan dolæt dane∫ju-yan ra yari mi-dæh-æd. government student-PL ACC help PROG-give.PRS-3SG. The government helps the students. DADÆN: Offer: Provide: Safety Meaning: give refuge or cure someone. PV: refuge or cure. Remarks: Transitive. The object is the entity being saved, intentionally or by happenstance. Pezeshk anha ra pænah dad, ‘the doctor gave them refuge,’ suggests that the object (them) was given refuge by the doctor intentionally, while kar ma ra nejaat dad, ‘work saved us’ is a happenstance (e.g. having stayed longer at the office saved us from having to go to a boring assembly). ∫æfa dadæn cure DADÆN cure pænah dadæn refuge DADÆN provide refuge nejat dadæn saved DADÆN save

111

moælem-e ∫ena jan-e anha ra nejat dad. instructor- GEN swimming life-GEN them ACC save give.PST.3SG. The swimming instructor saved their lives. DADÆN: Offer: Provide: Social Occasion Meaning: provide by way of social entertainment, usually involving food. PV: type of ritual that involves providing food and entertainment for a festive occassion. Remarks: Intransitive. The subject usually assembles some people to provide food or drink. This get-together is usually to celebrate a positive outcome of some event. Analogous English forms include give a party. ∫am dadæn dinner DADÆN give a dinner ∫irini dadæn sweets DADÆN provide sweets to celebrate mehmani dadæn get-together DADÆN have a get-together be eftexar-e piruzi tenisbaz dust-an-æ∫ ra ∫am dad. in pride-GEN victory, tennis-player friend-PL-3SG ACC dinner give.PST.3SG. In celebration of his victory, the tennis player gave a dinner (for all his friends). DADÆN: Offer: Render: Settlement Meaning: put in order, give organization to, or arrange. PV: settlement or organisation. Remarks: Transitive. The subject acts directly, or indirectly and causes the object to be put in order or become organized. These LVCs can be used for abstract (e.g. life) as well as concrete (e.g. books on a bookshelf). næzm dadæn organization DADÆN organize tærtib dadæn hierarchy DADÆN arrange, put in order saman dadæn welfare DADÆN equip with furniture, prepare modir-e jædid be edare næzm dad. director-GEN new to office order give.PST.3SG. The new director organized the office. DADÆN: Offer: Render: Trait Meaning: bring about a desired state to another person or entity. PV: a positive or desired psychological condition or abstract trait. Remarks: Transitive. The trait is usually a positive one, deserved or not, and generally it is acquired in shorter time than expected. The adjectives derived from the nouns for this island are generally valid in LVCs in kærdæn to suggest a similar type of action. arame∫ dadæn calm DADÆN calm niru dadæn power DADÆN strengthen ∫ohræt dadæn fame DADÆN make famous bi ∫æk qæzay-e tæbiyi særitær be hærkæs niru mi-dæh-æd. without doubt food-GEN natural faster to anyone power PROG-give.PRS-3SG. Without doubt, natural foods strengthen one faster.

112

DADÆN: Offer: Render: Pleasure Meaning: trigger a pleasant state on another person. PV: the notion of pleasure. Remarks: Intransitive. The subject of these forms is never human, but rather an event, action, or an inanimate entity that gives pleasure to another entity that can be, but not necessarily, expressed as an indirect object. The pleasure attained by the object might be a consequence of a person’s actions, and not from the person himself. It can be implied that a person’s actions or the ambiance he creates gives pleasure. Such usage is rather colloquial. For example the expression an khanande hal mide can be translated as ‘that singer gives good vibes/is a good entertainer,’ and is understood as referring to the singer’s singing. keyf dadæn pleasure DADÆN please kam dadæn satisfaction DADÆN give satisfaction hal dadæn health DADÆN give a good feeling t∫ayi ba abe dorost xeyli keif mi-dæh-æd. tea with water right much pleasure PROG-give.PRS-3SG. Tea made with proper water is very pleasurable. DADÆN: Offer: Opportunity Meaning: make something possible or give a person an opportunity or chance. PV: chance, opportunity, possibility, or probability. Remarks: Intransitive. The thing or person being given a chance occurs as an indirect object. emkan dadæn possibility DADÆN allow for something forsæt dadæn chance DADÆN give a chance jai dadæn place DADÆN give a place for someone mærd-e porhærf be hit∫ kæs forsæt næ-dad jævæb dah-æd man-GEN loquacious to noone chance NEG-give.PST.3SG answer give.PRS-3SG. The loquacious man didn’t give anyone the chance to answer. DADÆN: Offer: Favor: Incentive Meaning: offer a monetary incentive or favor to make a sale happen. PV: type of incentive. Remarks: Any type of incentive to make a product attractive and affordable is usually a valid PV, including extending credit. tæxfif dadæn discount DADÆN give a discount ærzan dadæn cheap DADÆN offer cheaper kahe∫ dadæn reduction DADÆN reduce bæd æz eid-e sale no qeymæt-ha ra kahe∫ dadænd. after holiday-GEN year new price-PL ACC reduction gave-3PL. They reduced the prices after the New Year holiday. DADÆN: Offer: Favor: Justice Meaning: offer a favor by consideration of what is right. PV: a just opinion or judgment.

113

Remarks: Intransitive. The person to whom the justice is given occurs as indirect object. This construction appears in two similar situations. In the first, an action occurs where the PV is a decision that is just and in favor of the indirect object. In the second situation, the subject requests a just review or decision. hæq dadæn right DADÆN find a person in the right ensaf dadæn honesty DADÆN make an honest judgment hakem be mo∫tari hæq dad. arbitrator to client right give.PST.3SG. The arbitrator sided with the client. DADÆN: Offer: Favor: Permission Meaning: giving permission or allocating something to someone. PV: permission or allocation. Remarks: Intransitive. æzn dadæn leave DADÆN permit roxsæt dadæn permission DADÆN permit ejaze dadæn permission DADÆN allow moælem ejaze dad anha mæ∫qe∫-an ra dær kelas be-nevis-ænd teacher permission give.PST.3SG they homework-3SG in class SUBJ-write-3PL. The teacher permitted them to do their writing exercises in class. DADÆN: Offer: Favor: Promotion Meaning: promote or ascend someone in rank. PV: measure of social, official, or military standing. Remarks: Transitive. dæreje dadæn degree DADÆN promote to higher rank erteqa dadæn promotion DADÆN promote to the next job grade rotbe dadæn position DADÆN promote to an official position bæd æz salha xedmæt be u rotbe dad-ænd. after from year-PL service to him position give.PST-3PL. They gave him a graded position after years of service. DADÆN: Offer: Favor: Reserve Meaning: appoint a spatial or temporal spot. PV: general word for place in time or space. Remarks: Intransitive. The subject allots a time or a place to the person that occurs as an indirect object. A similar expression exists in English (e.g. give an appointment). ja dadæn place DADÆN give a placement væqt dadæn time DADÆN give an appointment nobæt dadæn turn DADÆN assign a turn

114

be xastdaran-e molaqat dir væqt mi-dæh-ænd. to applicants-GEN meeting late time PROG-give.PRS-3PL. They give appointments to those requesting meetings far in the future.

DADÆN: RELINQUISH

These two islands occur separately in dadæn’s semantic space because they represent the only two constructions where the subject loses something as the action is carried out. The subject can voluntarily or involuntarily relinquish something. The English equivalent of these notions can usually be expressed in the verbs give up or give away.

DADÆN[to give]

extiyar [control]

relinquish

voluntarily

in-voluntary

rezayat[agree]

del[heart]

jan[life]

vekalæt[attorney]

DADÆN: Relinquish: Voluntary Meaning: give up control or right. PV: an expression of control or right. Remarks: Intransitive. The subject lets the person expressed as an indirect object to take control of a situation or take over the subject’s personal rights. rezayæt dadæn agreement DADÆN agree extiyar dadæn control DADÆN give control vekalæt dadæn proxy DADÆN give power of attorney u be farzænd-an-æ∫ bæraye hæme kar extiyar dad-e æst. he to children-PL-3SG for all task control give-PTCP be.PRS.3SG. He has given his children control of everything DADÆN: Relinquish: Involuntary Meaning: lose something valuable. PV: cherished personal attribute or symbols of dignity, control, and independence. Remarks: The PVs are always symbolic rather and truly physical. This island’s LVCs are highly idiomatic, but since they express similar notions, they cluster to form an island.

115

jan dadæn life DADÆN perish, sacrifice ones life del dadæn heart DADÆN fall in love, lose one’s heart ba negah-e ævvæl xosro be ∫irin del dad. with look-GEN first Xosro to Shirin heart give.PST.3SG. Xosro lost his heart to Shirin at first sight.

DADÆN: IMPOSE

The islands in this cluster are similar in meaning to those in the first cluster, offer, with one main difference that the receiver of the action in this case cannot choose whether or not to accept what is imposed upon him. The grammatical subject imposes or provides a position, state, or topological configuration upon the object of the LVC. Depending on the type of PV, the construction may expresses a change of state in the object caused by the subject of an action. Although the subject causes this change of state, it need not be an agent.

regiment

pærvære! [raise]

uponothers

a bind

DADÆN[to give]

impose

embelish

rotation

order

rank

shape

exposurediscomfort

topology

physicalchange

motion

general

!ekænje [torture]

connection

azar [torment]

gorosnegi [hunger]

ræbt [relation]

ju! [weld]

etesal [join]

pit! [twist] fer [curl]

xæm [bend]

!ekl[form]

læbe[edge]

næq! [pattern]

aftab [sun]

gærma [heat]

bu [smell] qælt [flip]

pit! [turn]

t!ærx [roll]

qeymæt [price]

færman [order]

sefare! [order]

dæstur [order]

tab [swing]

tækan [jerk]

gærde! [movement]

fasele [distance]

nam[name]

displacement

hol[push off]

kut![migration]

ab tæla [gold]

lo'ab [polish]

ab roqæn [oil]

halæt[character]

physical

personallinkage

vefq[coordination]

ræbt [relation]

a!ti[reunited]

uponself

sogænd[promise]

væ'ede[promise]

qowl[assurance]

qæsæm[swear]

saze![reunited]

værze! [training]

xak [dust]

rejim[diet]

abuse

misusezæhmæt[hardship]

handle

fe!ar [squeeze]

male! [massage]

qelqelæk [tickle]

position

ja [place]

qærar [position]

DADÆN: Impose: Discomfort: Abuse Meaning: triggering or imposing an unpleasant state on another person. PV: a word expressing the notion of suffering or torment. Remarks: Transitive. The PV may be physical or psychological torment or torture, indirect or direct. PVs expressing indirect torment can be feelings of greed, envy, jealousy. The LVC

116

then implies these as torments caused by the subject’s actions, including flaunting of wealth, social position, good fortune, etc. zæjr dadæn torment DADÆN torment hers dadæn envy DADÆN cause envy ∫ekænje dadæn torture DADÆN torture u sæy mi-kærd ba xærj-e ziyad-e pul doxtær-æmuyæ∫ ra hers be-dah-æd. she try PROG-do.PST.3SG with spending-GEN much-GEN money cousin-3SG ACC envy SUBJ-give-3SG. She tried to make her cousin envious by spending a lot of money. DADÆN: Impose: Discomfort: Misuse Meaning: abuse a friendship for personal gain. PV: trouble or inconvenience. Remarks: Transitive. The LVCs are often used to express ones gratitude for a friend’s help by exaggerating the trouble they went through, or to express the fact that someone is abusing a relationship thusly. zæhmæt dadæn effort DADÆN inconvenience someone dærde-sær dadæn trouble DADÆN be a headache u ba vorud-e bimoq-e be dust-æ∫ dærde-sær dad. he with entrance-GEN untimely-GEN to friend-3SG headache give.PST.3SG. He caused a headache for his friend with his untimely arrival. DADÆN: Impose: Connection: Physical Meaning: connect something, physically. PV: a type of physical connection (irreversible). Remarks: Transitive. A permanent connection is made between rigid items using a process that uses the material of the items only, without a binding agent. This includes physical or chemical welding or using tight compressive connections such as twisting two wires together or a clamp. Weld is general and includes bonds which are provided by melting or the two parts, and then cooling the joint after putting the molten sides together. Melting may be by heat, such as in glass, plastics, or metals, or by chemical solvents used on acrylic type material. Glue, adhesives, tapes, solder are not a valid PVs. Junctions are made between electrical wires by twisting them together or holding them against each other by physical compressive holders such as clamps. Thus tying two pieces of rope, thread, belts, or connecting lakes by canals are not valid contexts for the forms in this island. Lakes are not physically solid entities, neither are ropes or threads considered rigid or “solid”. etesal dadæn junction DADÆN connect ju∫ dadæn welding DADÆN weld teknisien sim-ha-ye qæt ∫od-e ra etesal dad. technician wire-PL-GEN disconnected become-PTCP ACC connection give.PST.3SG. The technician connected the disconnected wires.

117

DADÆN: Impose: Connection: Personal Meaning: cause two parties to get along. PV: state of peaceful coexistence or friendship. Remarks: Transitive. The object of this construction is usually a phrase containing two entities, linked with the preposition ba ‘with.’ saze∫ dadæn collusion DADÆN make two people agree a∫ti dadæn reconciliation DADÆN make 2 people reconcile solh dadæn peace DADÆN make peace miyani do tæræf ra a∫ti dad. mediator two side ACC reconciliation give.PST.3SG. The mediator reconciled the two sides. DADÆN: Impose: Connection: Linkage Meaning: relate. PV: phrase or word expressing some type of relation or correlation. Remarks: Transitive. This construction is abstract: the subject relates two usually abstract entities. ræbt dadæn connection DADÆN relate, connect nesbæt dadæn relation DADÆN attribute, ascribe tarix kæ∫f-e qove jazebe ra be niyuton nesbæt mi-dæh-æd. history discovery-GEN power-GEN gravity ACC to Newton attribute PROG-give.PRS-3SG. History attributes the discovery of gravitational force to Newton. DADÆN: Impose: Regimen Meaning: place under specific program for developing proper physical condition. PV: : type of course or regimen used for physical conditioning or rehabilitation. Remarks: Transitive. The subject acts as the teacher who oversees and maintains the object’s adherence to the prescribed regimen. pærvare∫ dadæn breed DADÆN raise værze∫ dadæn exercise DADÆN exercise (something) rejim dadæn diet DADÆN put on a diet ashoq sæg-æ∫ ra rejim mi-dad. Issac dog-3SG diet PROG-give.PST.3SG. Issac would put his dog on a diet. DADÆN: Impose: Physical Change: Topology Meaning: reshape the external form of something PV: word expressing a geometric manipulation. Remarks: The change inflicted by the subject usually reshapes the subject temporarily. The change requires force, heat, or a special device. The resulting state is usually not the natural state the object should be in, but nothing is physically added to the object to accomplish the task. The object may be abstract, such as a story or report, which is made convoluted by the subject’s deviations from the main line of the narrative.

118

xæm dadæn bend DADÆN bend pit∫ dadæn twist DADÆN twist fer dadæn curl DADÆN curl u tebqe ræsm-e hengam mu-ha-yæ∫ ra fer dad. he according-GEN tradition-GEN time hair-PL-3SG ACC curl give.PST.3SG. He curled his hair in fashion of his days. DADÆN: Impose: Physical Change: Shape Meaning: give something form. PV: a general word for sculptural form. Remarks: Transitive. The subject is like a sculptor who gives a shape to something or part of something. The reshaping can include adding to the object, such as a handle. This construction can also be used for abstract objects, such as concepts or reports. ∫ekl dadæn shape DADÆN shape form dadæn form DADÆN shape læbe dadæn edge DADÆN add an edge ostad ru-ye t∫ærx be goldan form dad. master face-GEN wheel to vase shape give.PST.3SG. The master shaped the vase on the turntable. DADÆN: Impose: Physical Change: Handle Meaning: imposing a physical pressure or acting on something physically. PV: a type of physical action usually capable of reshaping malleable objects. Remarks: Transitive. The object is generally a malleable entity that the subject can reshape, or one that has parts that respond to the specific PV, such as buttons, pedals, levers. fe∫ar dadæn squeeze DADÆN squeeze male∫ dadæn massage DADÆN massage qelqelæk dadæn tickle DADÆN tickle madær bozorg hæmi∫e bæt∫e-ha ra qelqelæk mi-dad. grandmother always children-PL ACC tickle PROG-give.PST.3SG. The grandmother would always be tickling the children. DADÆN: Impose: Physical Change: Embellish Meaning: cover with. PV: coating. Remarks: Transitive. The subject coats the surface of an object, usually covering it fully with the substance expressed by the PV. loab dadæn enamel DADÆN enamel seyqæl dadæn polish DADÆN polish ab noqre dadæn silver-plating DADÆN coat with silver kuzegær jam ra loab dad. potter cup ACC enamel give.PST.3SG. The potter enameled the cup.

119

DADÆN: Impose: Bind: Upon Self Meaning: bind oneself to perform an act or a deed. PV: word expressing a binding. Remarks: The subject is the entity that binds itself and there is no explicit object required other than the optional inclusion of the deed. If the deed is implied in the context of the conversation it is not required in a sentence produced with this LVC. qol dadæn promise DADÆN promise væ’de dadæn promise DADÆN promise kandid-e ræyis jomhuri xeyli væde dad. candidate-GEN president republic much promise give.PST.3SG. The candidate for the presidency made a lot of promises. DADÆN: Impose: Bind: Upon Others Meaning: bind another person to perform an act or a deed. PV: a word expressing the execution of a promise. Remarks: Transitive. qæsæm dadæn oath DADÆN put someone under oath sogænd dadæn oath DADÆN put someone under oath contora dadæn contract DADÆN put someone under contract qazi-an-e jadid ra sogænd dad-ænd judge-PL-GEN new ACC oath give.PST-3PL. The new judges were given the oath. DADÆN: Impose: Exposure Meaning: expose to an environmental element. PV: natural environmental element. Remarks: Transitive. The subject exposes the object to a type of environment. The object consequently undergoes a change of state because of this exposure. The exposure is either curative, pleasant or damaging, causes illness or physical damage to the object. The element the object is exposed to often carries the connotation of being positive or negative, such as sun vs. cold, air vs. smoke, and so on. The implication is always that the exposure is forced and under the control of the subject. aftab dadæn sun DADÆN expose to the sun bad dadæn wind DADÆN expose to fresh air særma dadæn cold DADÆN expose to the cold anha lebas-ha-ye næftalini ra bad dad-ænd. they clothe-PL-GEN moth-balled ACC wind give.PST-3PL. They aired the moth balled clothes. DADÆN: Impose: Motion: Rotation Meaning: cause an entity to rotate. PV: a type of rotational movement.

120

Remarks:Transitive. The subject makes the object turn or flip. When the object is a person, the imposition of the motion is considered an irritation by placing the object in an uncomfortable position that debases him. t∫ærx dadæn turn DADÆN make something turn qælt dadæn roll DADÆN roll something pit∫ dadæn twist DADÆN twist around dust-an-æ∫ u ra ru-ye sahel qælt dad-ænd. friend-PL-3SG him ACC on-GEN shore roll give.PST-3PL. His friends rolled him on the beach. DADÆN: Impose: Motion: Displacement Meaning: forcefully move or make movement possible PV: a type of displacement or condition requiring displacement. Remarks: Transitive. The subject either imposes or facilitates the displacement of the object. In the latter case, the PV expresses a condition that the object was not able to attain by itself. The context determines if the displacement is voluntary or facilitated. For example, in kut∫ dadæn the subject facilitates the migration of a herd to move them into safety or more lush location, while if it is related to forcing a tribe to migrate it is an imposition. Similarly in færar dadæn, a person is allowed to escape to freedom as he wishes, while a domestic animal is frightened into escaping and put in danger of unfamiliar surrounding. In both cases, the notion of facilitating or forcing a displacement is present. hol dadæn push DADÆN push kut∫ dadæn migration DADÆN force to migrate færar dadæn escape DADÆN allow to escape cauboy-ha gælle gav-ha ra sad-ha kilometr kut∫ mi-dad-ænd. cowboy-PL herd cow-PL ACC hundred-PL kilometers migration PROG-give.PST-3PL. The cowboys move the cow herd hundreds of kilometers. DADÆN: Impose: Motion: General Meaning: make something move physically. PV: a type of movement. Remarks: Transitive. The movement of the object is fully controlled by the subject, from beginning to end: it may be a pleasant action, such as showing someone around, or a disturbing one, such as tækan ‘shaking’. gærde∫ dadæn prommenade DADÆN show someone around tækan dadæn shaking DADÆN shake tab dadæn swing DADÆN swing u ra tækan dad ta bidar-æ∫ kon-æd. him ACC shake give.PST.3SG until awaken-3SG do.PRS-3SG. He shook him to wake him. DADÆN: Impose: Rank Meaning: assign a name, rank, value, or title to someone or something. PV: a general word that distinguishes or ranks an entity (in its class).

121

Remarks: Intransitive. The ranked entity occurs as an indirect object. The subject is usually an authority that can affirm such an assessment. Name here is a title and not a personal name and thus, “naming a child”, is not a possible meaning. læqæb dadæn pseudonym/title DADÆN give title nam dadæn name DADÆN name qeymæt dadæn value DADÆN give a value dær gozæ∫te-ha adæm-ha-ye mohem ra læqæb mi-dad-ænd. in past-GEN person-PL-GEN importance ACC pseudonym PROG-give.PST-3PL. In the past they would give a pseudonym title to important persons. DADÆN: Impose: Position Meaning: provide or make a place or position for something. PV: a general word meaning spatial position. Remarks: Transitive. The position may be literal, such as a physical placement, or figurative, such as placing data in the proper order or sets, or to find a way to put a person’s name on a list. Place may also be in time, such as finding a time for an appointment for someone, or putting time between beats (in music). ja dadæn place DADÆN find a place for something qærar dadæn position DADÆN place something somewhere fasele dadæn separation DADÆN put space or time between polis bæraye contorole terafik ma∫inha ra fasele mi-dad. police for control-GEN traffic car-PL ACC separation PROG- give.PST.3SG. The police was spacing out the cars to control the traffic flow. DADÆN: Impose: Order Meaning: place an order, make a request. PV: order or request. Remarks: Transitive. The direct object is the entity requested by the subject. sefare∫ dadæn order DADÆN specify an order dæstur dadæn order DADÆN place an order færman dadæn order DADÆN command næqi bæraye hæme qæza sefare∫ dad. Naqi for all food order give.PST.3SG. Naqi ordered food for everyone.

122

DADÆN: CONDUCT

DADÆN[to give]

conduct

progress

event

completion

ænjam[final]

ænjam[perform]

edame[continuation]

suræt[face]

feysæle[closure]

xateme[ending]

tælaq[divorce]

DADÆN: Conduct: Event Meaning: effect the termination of an act in progress. PV: end, closure. Remarks: Transitive. The subject causes an act in progress to come to an end, or defines the end of the process. The PV can generally be any concept that defines a clear termination. The direct object is an ongoing phenomenon, such as a story, a course, a marriage, a partnership, etc. Imported word meaning the same are also valid, such as in script writing it is ending dadæn. xatæme dadæn end DADÆN bring to an end feysæle dadæn closure DADÆN bring closure to tælaq dadæn divorce DADÆN grant a divorce jæm∫id kar-ha ra feysæle dad. Jamshid work-PL ACC closure give.PST.3SG. Jamshid brought all tasks to close. DADÆN: Conduct: Progress Meaning: affect the progress of a phenomenon. PV: a chronological event describing an action related to completion or pursuing of a task. Remarks: Transitive. edame dadæn continuation DADÆN continue ænjam dadæn completion DADÆN complete peygir dadæn follow-up DADÆN follow up on something mærd-e porhærf zerzer-æ∫ ra edame dad. man-GEN loquacity haranguing-3SG ACC continuation give.PST.3SG. The talkative man continued with his harangue.

123

DADÆN: Conduct: Completion Meaning: execute a task. PV: a completed task as a metaphor of a physical form. Remarks: Transitive. The general sense is of accomplishing a task with clear and undisputable success. suræt dadæn shape DADÆN give a final shape ænjam dadæn ending DADÆN complete a task dars-æ∫ ra be moqe ænjam dad. studies-3SG ACC to time ending give.PST.3SG. He completed his studies on time.

DADÆN: PRODUCE

Dadæn can also be used to express yielding a result or yielding fruits or flowers. Here, the subject makes something apparent. Again, the notion of providing or yielding prevails.

DADÆN[to give]

produce

product mive [fruit]

gol [flower]

!ire [syrup]

result

emanation

nur [light]bu [smell]

nætije [conclusion]

sud [profit]

zærær [loss]

condition linæt[softness]

yobosæt [constipation]

del dærd [stomach ache]

aramesh[calmness]

seda[noise]

DADÆN: Produce: Product Meaning: blossoming of flowers or the yielding of fruit. PV: something yielded by a plant or tree. Remarks: Intransitive. The sense is that of expressing a plant offering signs of natural progress, such as sprouting leaves, blossoming, and giving fruit. gol dadæn flower DADÆN blossom

124

mive dadæn fruit DADÆN give fruit (a tree) ∫ire dadæn syrup DADÆN give syrup (a tree) DADÆN: Produce: Emanation Meaning: emanate. PV: any emanation or spreading that can be seen, heard, or smelled. Remarks: Intransitive. The essense of the object is in its intangible quality, such as heat, light, sound, and smell. In this sense, taste is not valid PV as tasting requires touching the object with the tongue. This construction is similar to give off in English. bu dadæn smell DADÆN smell nur dadæn light DADÆN give off light seda dadæn sound DADÆN make noise dær-e ahæni xeili seda mi-dah-æd. door-GEN steel much sound PROG-give.PRS-3SG. The steel door makes a lot of noise. DADÆN: Produce: Result Meaning: yielding results. PV: a type of result, positive or negative, from an act or an investment. Remarks: Intransitive. The subject is generally an act, such as a trade, a business, an endeavor, or a financial instrument or transaction, such as stocks, real estate, loans, etc. zærær dadæn loss DADÆN attain a loss sud dadæn gain DADÆN give a monetary benefit nætije dadæn result DADÆN give results komæk be mohtaj-an be ejtema sud mi-dah-æd. help to needy-PL to society profit PROG-give.PRS-3SG. Helping the needy benefits the society. DADÆN: Produce: Condition Meaning: cause a condition. PV: physical ailment. Remarks: Intransitive. The subject of this construction is usually non-human and has the inherent property of causing the state expressed by the PV to come about in humans. deldærd dadæn stomachache DADÆN cause a stomach ache linæt dadæn laxity DADÆN act as laxative yobusæt dadæn constipation DADÆN cause constipation xordæne ziyadi qar∫t be adæm deldærd mi-dæh-æd. eating-GEN much-GEN mushroom to person stomachache PROG-give.PRS-3SG. Eating too much mushrooms gives one a stomachache.

3.4.5 Conclusion

The data presented in this chapter display the types of varied patterns of meaning formation in Persian LVs. It is a daunting task to list overarching semantic traits for any of the LVs. No

125

single list would be able to capture and summarize all the notions expressed by a particular LV. The corpus here was meant to give a solid overview of the issue.

Almost any conjecture proposed for one or more islands will fail to hold for at least one other. For example, at the beginning of the section on zædæn, it is suggested that LVCs constructed with zædæn are instantaneous actions that are usually quick in nature. This trait emerges in most of the LVCs of zædæn, but a non-negligible number of constructions there do not abide by that rule. The island Affecting: Enhancing: Texture and those under the label Emitting: Aural: Music are not necessarily quick, whole actions that are possibly iterative, as the discussion at the beginning of the section suggested. Furthermore, constructions with zædæn can be either unergative or unaccusative, resulting in very different verb types.

For the LV xordæn, most of the islands express a change of state or an action undergone by the proto-patient. However, one of its islands, Exploiting, takes a proto-agent.

A survey of the variety of LVCs shows that it is highly unlikely that traits inherent to the LVs zædæn, xordæn, gereftæn, and dadæn can be explicitly delineated in list-form based on simple formulae. However, some reasonably close traits do emerge in large clusters of islands. This clustering of islands might explain why there are no LV-wide alternations in the system. Alternations can at most occur between many, but not all, LVCs of pairs of LVs. As we will see in the next chapter, islands can be shared in the system, where a different LV results in slightly different meanings. Clusters of shared islands usually have overarching features in common with each other.

3.5 Chapter Review

In this chapter, I introduced the notion of islands of LVCs and how they pertain to the Persian verbal system. These islands provide a blueprint for different verbal notions. Before presenting the full description of the four most common verbs, I addressed some issues relevant to the idea of islands and also provided an overview of a past study. Islands are represented in the semantic space of an LV by a construction. These constructions contain rich semantic and syntactic information. The LVs zædæn, xordæn, gereftæn, and dadæn were discussed in detail and it was shown that it is not possible to extract any overarching traits from the distinct islands forming the semantic space. For a list of LVCs of the other LVs and a preliminary diagram for each, I refer the reader to Appendix I. In the next chapter, I will explore the relations between LVs thorough connections between their alternating island.

127

4 ISLAND GENERATIVITY

In the previous chapter I presented the notional islands that form the foundation of verbal semantic space in Persian. Each LV is comprised of clusters of islands organized in such a way as to provide Persian speakers with an efficient mapping system with which to comprehend and produce new utterances despite the unclear contribution of the LV. The organization of these spaces reduces redundancy in the lexicon and structures the semantic space of each LV.

Now that we have an idea of this organization, I will discuss another important property of the system: the set of LVs can alternate, that is, given a certain PV, its occurrence with different LVs will result in different but highly related meanings. This occurs at the island level involving constructions that serve as a basis for productivity. These alternations are therefore productive. Pairs or sets of pairs of LVCs with the same PV and different LV often display aspectual or grammatical alternations18 (examples will be given below). As in most languages, Persian LVCs don’t always alternate according to purely formal parameters, but can take into account more idiosyncratic semantic information. This generative property of the system allows further efficiency in the retrieval and expression of new LVCs.

In 4.1.1, I will present previous studies that have highlighted this phenomenon in other languages. In section 4.1.2, I will provide a few examples from Persian and describe the language specific patterns of the system. In section 4.2, I will discuss how these productive alternations can be understood within the framework of this study, and I will give examples of such alternations.

4.1 Past Studies of Alternating Constructions

4.1.1 Alternations in English

Alternations, which display generativity, exist in English LVCs. In the article presented in section 3.1, Wierzbicka (1982) briefly discusses the alternations between the have a V and the take a V constructions in English. She posits that the have a V construction “belongs to a family of constructions which includes at least two other members: take a V and give a V” (p337). There are systematic alternations between the have a V and take a V constructions, but the alternations do not encompass all possible instances of either construction. She gives the following list of examples of alternating and non-alternating pairs (p338):

have a look at take a look at have a bath take a bath have a walk take a walk have a lick take a lick have a sip take a sip have a play *take a play have a read *take a read have a cry *take a cry have a cough *take a cough have a pee take a pee have a try *take a try have a look for *take a look for

18 An alternation occurs when a single PV combines with more than one LV to express different meanings.

128

have a think *take a think have a suck ?take a suck have a chew *take a chew have a listen *take a listen have a feel *take a feel have a chat *take a chat have a cuddle *take a cuddle *have a dive take a dive *have a leap take a leap *have a step take a step

According to Wierzbicka, one of the main semantic traits that doesn’t allow a full-scale alternation between these two constructions is that the semantic core, or invariant, of the take a V construction requires that the action start at a definite moment in time. An initial, momentary impulse, which usually involves a deliberate physical motion, characterizes the action expressed through this construction. Unlike the have a V construction, which allows repetitive or aimless movements, the take a V construction requires a voluntary subject who knows what he wants and takes an initiative. The take a V action is seen as unitary and quick. Therefore, the acceptability of the alternations lies in the inherent properties of the verbal complement. Recall that in their study on productivity, Stevenson et al. (2004, p5) also showed that some classes of verbal complements occur more with some LVs than with others.

In her study on LVs, Brugman (2001) focuses on discerning the contribution of each LV, not how they alternate. However, she mentions the alternation between the LVs take and give. She says that the former LV refers to subjective experiences, whereas the latter refers to other oriented actions or to the potential utility of the action towards an object (p569).

As we saw in the last chapter, LVCs often do not have purely compositional semantics. Their meaning lies in their proper constructions, and is highly dependant on the type of complement the LV occurs with. It is important to note that the semantic opacity of the construction does not affect the alternation possibilities of the related forms. Consider the construction go N, where N is any expression meaning crazy, and the construction meaning is ‘go crazy’ (e.g. go bananas, go bonkers, go nuts). This semi-productive construction is highly idiomatic, but alternates systematically with drive N to mean ‘cause to go crazy.’ Knowing that this pattern exists allows a specific form (e.g. go cuckoo) to predict the analogous form (e.g. drive cuckoo), without having to store each explicit example in the lexicon.

Goldberg states “the repertoire of constructions is not an unstructured set. There are systematic generalizations across constructions,” and “constructions form networks and are linked by inheritance relations which motivate many of the properties of a particular construction” (p67). In other words, the relations between constructions further delineate individual constructions since certain traits are particularly highlighted in each relation. Construction Grammar also predicts the type of alternations dealt with in this chapter. The relation between “two constructions is quite analogous to a rule in that the existence of one form can be used to predict the existence of the other form” (p138). As will be illustrated by the examples from Persian, aspectual, grammatical, or other types of traits can describe the rules or relations that allows two constructions to be linked.

129

In a study on Urdu, Butt (1997) shows that LVs like de ‘give,’ le ‘take’ and daal ‘put’ require ergative subjects while par ‘fall’ and bait ‘sit’ cannot occur with PVs that require a subject with conscious choice. She concludes that the use of different LVs results in these types of case restrictions. It is, in fact, a widespread property of LVs to have selectional restrictions on the type of complement with which they combine.

4.1.2 Alternations in Persian

In the previous chapter I extracted patterns from lists of LVCs that shared a common LV. The next step in my investigation is to analyze how the LVs potentially interact with each other. This will give us a more global view of the system as well as provide us with further clues as to how the system generates new forms. In order to study these interactions, patterns must be extracted from sets of LVCs based on different LVs that share a PV. Some examples are:

ab gereftæn water GEREFTÆN juice ab ke∫idæn water KE∫IDÆN rinse ab ræftæn water RÆFTÆN shrink ab ∫odæn water ∫ODÆN melt, dissolve ab dadæn water DADÆN give a drink of water, irrigate ab oftadæn water OFTADÆN become watery, become flooded sær amædæn head AMÆDÆN excel, attain mastery sær aværdæn head AVÆRDÆN bring to an end sær ∫odæn head ∫ODÆN come to an end sær dadæn head DADÆN free, let go sær ræftæn head RÆFTÆN overflow, boil over sær xordæn head XORDÆN be discouraged, disappointed sær zædæn head ZÆDÆN inspect briefly, impromptu visit sær kærdæn head KÆRDÆN put or wear on the head, to begin sær gereftæn head GEREFTÆN take a good turn, come into effect sær ke∫idæn head KE∫IDÆN drink up, supervise zæmin xordæn earth XORDÆN fall down (negative connotation) zæmin oftadæn earth OFTADÆN drop down zæmin ændaxtæn earth ÆNDAXTÆN throw down zæmin zædæn earth ZÆDÆN hit against the ground Previous studies on Persian have dealt with similar minimal pairs or sets in the system, usually in the contexts of studying the grammatical traits of individual LVs. Grammatical and aspectual properties (e.g. event-type, duration) have been put forth as reasons for these alternations. Though these traits might capture large scale alternations where many PVs are shared by two LVs, more idiosyncratic traits involving perceptual or other cognitive features are needed to account for finer-grained alternations (or, in some cases, lack of alternations where they are expected).

None of the studies thus far have presented a coherent image of how the system operates as a whole. Why don’t alternations occur LV-wide, and how are alternant LVs chosen? In the following section, I present the approach taken in this thesis. The notion of islands allows for a global view of how different parts of the system interconnect.

130

4.2 Alternations: A Case of Shared Islands

We have seen that inherent properties of a PV partly determine the meaning of a construction. However, apart from purely linguistic parameters, features related to perception and real-world knowledge are necessary for the prediction of some forms. In other words, the structure of the semantic space and the islands to which the LVC belong play a large role in determining its meaning and alternations with other LVCs.

When a certain LVC island exists in one LV’s space, LVCs with the same PVs, can sometimes be expected to be present in the space of one or more other LVs. In such cases, two or more LVs share islands populated with the same PVs, but to express different notions. Hence, any island found in the space of a specific LV, regardless of its level of productivity, might signal the existence of other forms in the language. This again allows for an alleviation of redundant LVC entries in the lexicon. We also see here the reason for the separate notions of islands and constructions. Islands can be shared between LVs, and each island can have more than one construction related to it (e.g. if two LVs share an island, that island will host two separate constructions).

Not all constructions with the same LV display the same type of behavior. No LV allows all of its LVCs to alternate with all of the LVCs of another LV. In other words, most LVs share some of their islands with another LV, but not all, or they might share some islands with several other LVs. Different alternation patterns represent different types of meaning differences between two constructions. Large alternation sets, where many islands are shared, are usually due to general aspectual properties; and local alternations, where only one or a limited number of islands are shared, are more semantically nuanced. As we saw in Chapter 3, LVs can occur in numerous LVCs to express a large variety of meanings. Different meaning attributes of the LV are implicated in different constructions. This could be one reason why alternations don’t occur LV-wide (too much meaning variety within a single LV).

Karimi-Doostan (1997) examines aspect in LVCs. In section 1.3.2, I presented his division of LVs into two classes: Transition and Initiatory. He claims that for every Transition LV, there is one or more Initiatory counterpart, though he doesn’t indicate how this counterpart is chosen from the available LVs. Through several linguistic tests, such as adding a temporal for/in X time phrase to a sentence, he shows that Transition LVs (usually unaccusative19) like xordæn, ∫odæn, gereftæn, amædæn, and ræftæn cannot occur with for X time, but can occur with in X time. On the other hand, their Initiatory counterparts (usually unergative or transitive), zædæn, dadæn, kærdæn, and aværdæn, the inverse pattern is observed. This shows that Transition LVs resist atelic readings whereas Initiatory LVs resist telic readings. He also shows that the choice of LV determines the duration of an event. His example, repeated from section 1.3.2, involves the LVs gereftæn and ke∫idæn.

Dæst-e daryu∫ dær yek saniye / *sa’æt-ha dærd gereft. Hand-GEN Dariush in one second / hour-PL pain get.PST.3SG. Dariush’s hand (started to) hurt within one second / *for hours.

19 I will remind the reader the working definition of unaccusative/unergative distinction in intransitive verbs (from Cadiot, Lebas, and Visetti, 2005) . Unergative verbs have a deep agentive subject as single argument: [NP[VPV]], whereas unaccusative verbs have a single argument, a direct object that is promoted at the surface to subject position: [VPVNP]. In other words, the unaccusative verb takes a subject that plays the role of a patient.

131

Daryu∫ * dær yek saniye / sa’æt-ha dærd ke∫id. Dariush in one second / hour-PL pain pull.PST.3SG. Dariush hurt (had pain) * in one second / for hours. Megerdoomian (2002) also discusses the aspectual alternation between gereftæn and ke∫idæn. She claims that gereftæn is a bounded action, representing the culmination of an event where the subject is affected by an action (unaccusative), whereas ke∫idæn expresses an unbounded, durational event where the subject is the agent or the initiator (unergative)20.

Understanding the process through which these shared islands emerge may allow us to predict and judge verb forms in the system. To do this, I will focus on a few alternations present in the system. First, I will describe a general alternation pattern between two LVs not discussed previously (kærdæn and ∫odæn), since they portray the basic dynamics of alternating pairs in the system. I will then focus on alternations involving the LVs presented in Chapter 3.

4.2.1 Alternations between Kærdæn and ∫odæn

The two most generic and most common LVs are kærdæn ‘to do’ and ∫odæn ‘to become.’ They don’t appear in the previous chapter because, compared to the other LVs in the system, they retain the least amount of semantic information as LVs. Their semantic spaces are sparse, although they are very productive. I present a rough sketch of their islands below to show how they alternate, as their dynamics are representative of those seen with LVs with more sustained core meaning. I do not rule out the possibility that there will be other layers of data to analyze pertaining to these two common LVs. These LVs might have a slightly richer semantic space, but for the purposes of this study, and showing how islands can be shared, the current analysis suffices.

Karimi-Doostan (1997) describes kærdæn as the most common and “lightest” LV in the system (p 124) and Vahedi-Langrudi (1996) says that these two LVs (kærdæn and ∫odæn) are “definitely the most frequent LVs used to form” LVCs in Persian (p 151). Kærdæn is analogous to the English LV ‘to do’ or ‘to make,’ or the French LV ‘faire.’

20 Though, in this case, the unaccusative/unergative distinction does not play a role in the alternation.

132

KÆRDÆN [DO]

state

undergo

xo!k [dry]kæj [bent]

pærdaxt [payment]

action

qæ! [faint]qolqol

[bubble]

montæqel [transfer]

xæbær [news]

tæb [fever]

ræha [free] geriye

[cry]

unergative

zendegi [life]

!uxi [joke]

KÆRDÆN: STATE Meaning: cause something to be in a new state. PV: adjective expressing any physical, mental, or emotional state. Remarks: Transitive. The subject causes the object to undergo a change of state. The resulting LVC is atelic.

kæj kærdæn bent KÆRDÆN bend xo∫k kærdæn dry KÆRDÆN dry ræha kærdæn free KÆRDÆN free kar-ha-yæ∫ hameh ra ∫ad kærd. actions-PL-3SG everyone ACC joyful do.PST.3SG. His actions made everyone joyful. KÆRDÆN: ACTION Meaning: cause an action to take place. PV: noun expressing an action. Remarks: Transitive. The subject acts in such a way that something results from this action, but the object doesn’t necessarily go through a change of state. The action itself is highlighted and not the change of state that it might entail for the object. mota’ele kærdæn investigation KÆRDÆN investigate pærdaxt kærdæn payment KÆRDÆN transfer payment montæqel kærdæn transfer KÆRDÆN transfer

133

danesju mæqale ra mota’ele kærd. student article ACC investigation do.PST.3SG. The student studied the article. KÆRDÆN: UNDERGO Meaning: undergo or show symptoms of a change of state. PV: noun expressing an internal process that causes a state change. Remarks: Intransitive. The LVCs of this island are unergative and express auto-processes. The subject undergoes an action that affects itself directly or indirectly. The resulting LVC is telic and the state results from an internal cause and is usually temporary. tæb kærdæn fever KÆRDÆN come down with a fever qæ∫ kærdæn faint KÆRDÆN faint qolqol kærdæn bubble KÆRDÆN boil u mariz ∫od væ tæb kærd. he sick become.PST.3SG and fever do.PST.3SG. He became sick and came down with a fever. KÆRDÆN: UNERGATIVE Meaning: do an action that involves oneself. PV: a noun expressing a sound or an action. Remarks: Intransitive. This island includes unergative LVCs and LVCs of emission. geriye kærdæn cry KÆRDÆN cry ∫uxi kærdæn joke KÆRDÆN joke zendegi kærdæn life KÆRDÆN live u dær birun-e ∫æhr zendegi mi-kærd. she in out-GEN city life PROG-did.PST.3SG. She lived outside of the city. ∫odæn ‘to become’ is usually described as a typical inchoative verb. It doesn’t allow for aspectual variation, and is inherently telic. The PV is always an adjective describing a state or a noun describing a process. This LV is analogous to the English LVs ‘to get’ or the more general verb ‘to become,’ and the French ‘devenir.’

134

!ODÆN[BECOME]

state

internal


pærdaxt [payment]

undergo


ræha [free]

sorx [blush]

zeræng [clever]

atæ!i [enraged]

∫ODÆN: STATE Meaning: undergo a change of state. PV: adjective expressing an externally caused state change. Remarks: Intransitive. The subject undergoes a change of state that can be caused by some external force. In other words, an implicit agent can control the action. The state may be physical, mental, or emotional. kæj ∫odæn bent ∫ODÆN bend xo∫k ∫odæn dry ∫ODÆN dry (reflexive) ræha ∫odæn free ∫ODÆN become free bæd æz t∫ænd sa’æt ræxt-ha xo∫k ∫od-ænd. after from few hour laundry-PL dry become.PST-3PL. The laundry dried after a few hours. ∫ODÆN: UNDERGO Meaning: be acted upon. PV: noun expressing an action. Remarks: Intransitive. The subject doesn’t necessarily go through a change of state, but an action is done involving it. pærdaxt ∫odæn payment ∫ODÆN be paid montæqel ∫odæn transfer ∫ODÆN be transferred mota’ele ∫odæn investigation ∫ODÆN be investigated an marizi che joori be u montaqel ∫od? this illness which how to him transferred become.PST.3SG? How did he catch this illness? (Lit. How was this illness transferred to him?)

135

∫ODÆN: INTERNAL Meaning: undergo a change of state through an internal auto-action. PV: adjective expressing an internally caused state. Remarks: Intransitive. The subject undergoes a change of state that cannot be the direct result of an external force, but must be a result of a process internal to the subject, often instigated by an external force or event. sorx ∫odæn red ∫ODÆN blush zeræng ∫odæn clever ∫ODÆN become clever atæ∫i ∫odæn enraged ∫ODÆN get angry/feisty hær dæfe bah-e∫ sohbæt mi-kon-am, sorx mi-∫æv-æd. each time with-3SG speak PROG-do.PRS-1SG, red PROG-become PRS -3SG. She blushes each time I speak to her. Many LVCs made with this LV are inchoative/unaccusative alternants to certain transitive forms with kærdæn. The alternating LVCs of kærdæn are transitive and their corresponding ∫odæn LVCs are intransitive. LVCs formed with ∫odæn are telic, whereas those formed with kærdæn are generally atelic (see Karimi-Doostan, 1998, p115 for linguistic tests displaying this property). However, as can be seen in the diagram below, not all islands in kærdæn alternate with those in ∫odæn, and vice versa.

!ODÆN[BECOME]internal


pærdaxt [payment]

ræha [free]

sorx [blush]

zeræng [clever]

atæ!i [enraged]

KÆRDÆN [DO]

undergo

qæ! [faint]qolqol

[bubble]


tæb [fever]

geriye [cry]

unergativezendegi

[life]

!uxi [joke]

state

action

mota'ele [investigate]

With the following examples, Folli et al. (2004) propose that certain ∫odæn forms, namely those in the Internal island described above, do not have corresponding LVCs with kærdæn (p 19).

136

a. Kimea sorx ∫od. Kimea red become.PST.3SG. ‘Kimea blushed’

b. Papar Kimea-ro sorx kærd.

Papar Kimea ACC red do.PST.3SG.. *Papar made Kimea blush.

The LVC in a. is unaccusative, but the LVC in b. cannot have a causative reading because blushing is an internally caused reaction and cannot occur with an agent21. Vahedi-Langrudi (1996) points out that certain cases necessitate real-world knowledge in order to block alternations with kærdæn (p192)).

Conversely, kærdæn forms inherently unergative actions, which cannot alternate with ∫odæn (Megerdoomian, 2002; Folli et al., 2004). A reasonable conjecture is that these LVCs are already intransitive and thus cannot partake in the transitivity alternations formed with ∫odæn.

The alternation between these two LVs serves an as example of how this phenomenon is captured in the present framework. In the next section, I explore LVs with more complicated semantic spaces. I focus on zædæn and xordæn.

4.2.2 Alternations between zædæn and xordæn

One group of islands formed with zædæn expresses roughly the notion of “acting on something”. Parallel to this group of islands, the LV xordæn combines with the same set of PVs to express the non-agentive action in the group of islands labeled as “undergoing an action or change of state.” Karimi (1996) labels these LVs causative/unaccusative pairs. Some examples of these forms appear below.

t∫aqu zædæn knife ZÆDÆN injure with a knife t∫aqu xordæn knife XORDÆN get wounded by a knife gul zædæn trick ZÆDÆN trick gul xordæn trick XORDÆN be tricked gereh zædæn knot ZÆDÆN tie a knot gereh xordæn knot XORDÆN get tied into a knot bæxiye zædæn stitch ZÆDÆN stitch bæxiye xordæn stitch XORDÆN get stitches Xordæn forms the intransitive, unaccusative, inchoative LVCs of a subset of zædæn LVCs. This is similar to the pattern seen above with kærdæn and ∫odæn. Folli et al. (2004) describe zædæn as a semelfactive22 activity verb and xordæn as its inchoative counterpart. Vahedi-Langrudi (1996) claims that zædæn functions primarily as a simple activity verb, and that the 21 It is interesting to note here that in English, *she blushed me is unacceptable, as is *he fiestied me (meaning he made me fiesty). It seems that internal states of a certain kind cannot go through this derivation. The Persian data mirrors the phenomenon in English, though we would expect the problem in English to be due to the lack of transparency in the verb form. However, it seems that kærdæn loses part of its transparency in this case. The meaning of the LV is not merely to cause, which would have rendered the unacceptable forms grammatical. 22 A single act out of a possible series.

137

PVs that occur with it qualify as instruments, handled by an agent to change the state of another entity. Xordæn, on the other hand, expresses externally caused eventualities, spontaneous change in state through repeated instances of coming into contact of an instrument with the subject. Though these characterizations do not hold true for all the forms, some of the alternations can be explained thusly.

It is apparent that there have to be systematic rules that are intuitive to all Persian speakers about the relationship between these two LVs, namely, expressing causing a change of state, or undergoing a change of state. The above cases can be viewed as spaces. There are two constructions associated to each island, one with xordæn and the other with zædæn. With xordæn, the subject is a proto-patient (in the terms of Dowty, 1991), and with zædæn, the subject is proto-agent of the same action. The link between the two LVs can be visualized as follows:

offensive

ZÆDÆN[to hit]

act on something

offensive

state change

topology

fusion

damage

gereh [knot]

ta [fold]

fer [curl]

bæxiye [stitch]

bænd [connection]

peyvænd [graft]

lætme [damage]

qat! [chop]

sædæme [damage]

trick

bambul [trick]

gul [trick]

hoqe [trick]

type of hit

instru-ment

zærbe [hit]

lægæd [kick]

kotæk [beating]

t!æko! [hammer]

næzær [opinion]

potk [sledgehammer]

XORDÆN[to eat]

undergo

hand held

projectile

weapon

tir [bullet]

mu!æk [missile]

toop [canon]

kard [knife]

!æm!ir [sword]

t!aqu [knife]

state change

weapon

If we go back to the LV diagrams for zædæn and xordæn, we will see that the above diagram is the left and bottom branches of the other two diagrams, respectively. The islands that are shared have been highlighted.

Megerdoomian also investigates other alternations. She uses minimal LVC pairs to illustrate transitive and intransitive alternations. Namely, she shows that “transitive clauses formed with light verbs dadæn ‘give’ and zædæn ‘hit’” can be replaced by their “unaccusative counterparts” to provide the intransitive version of the same clause (p. 82)

This suggests the existence of a generative system in which one construction predicts another. For example, given “sustaining injuring from” and “successfully injuring with” and

138

a newly introduced weapon (e.g. cruise missile, dagger), one verbal form in one island would presuppose the other. This notion is expressed with zædæn and xordæn.

cruise XORDÆN cruise ZADÆN dagger XORDÆN dagger ZADÆN Specific information pertaining to the PV semantics and real-world knowlegde constrain the meaning differences between alternating forms. An important point regarding the alternations between zædæn and xordæn is that not all PVs that combine with xordæn also combine with zædæn, and vice versa. The systematicity occurs at the level of islands only and cannot encompass all possible islands or LVCs of a given LV. It is not the case that all the branches extending from xordæn interconnect with those extending from zædæn. In this case, the islands shared between the two LVs tend to form a cluster around the general extensions “act on something” and “undergo an action.” However, not all branches extending from even these two more general nodes connect with the other LV (see full diagrams in previous chapter). This pattern was also observed for the kærdæn and ∫odæn alternations. Consider the following unacceptable alternations between zædæn and xordæn:

∫ekæst XORDÆN (to be defeated) <≠> *∫ekæst ZADÆN tombæk ZADÆN (to play the drum) <≠> *tombæk XORDÆN The specifications of the PV cannot allow for these meanings. A tombæk is not a sharp instrument with which one can attack and pierce or penetrate another entity. This fact reinforces the assumption that the meaning of the construction is attributed to the interaction of the constituents. In other words, zædæn doesn’t mean “to attack,” but it occurs in a construction that means “attack and penetrate an entity with a sharp instrument” and calls for a PV with particular perceptual and functional attributes.

Karimi (1996) says that a set of unaccusative LVs corresponds to a set of transitive LVs in Persian. However, she points out that the combination of LVs and PVs is restricted and therefore LVs cannot be generally interchangeable with their unaccusative counterparts (1996, p18). She suggests that the above form with xordæn is not acceptable because the PV tombæk can only combine with accomplishment events, and xordæn cannot occur in these types of events. In a more recent article (Folli et al., 2004), Karimi and colleagues explain that certain events expressed by zædæn (e.g. emission verbs) are inherently unergative and require an agent.

Vahedi-Langrudi (1996) postulates that since zædæn has a wider range of aspectual properties, not all of the LVCs of the two verbs can alternate. Also, xordæn expresses a certain independence from an external agent, whereas forms with zædæn require external control. He stresses that the “aspectual properties seem to have an effect on their combinability” (p 239). However, as seen above, xordæn occurs with weapons to express being hit by a weapon, which necessarily involves an external agent. However, his claim supports my proposal that alternations occur at the island level, and are not all-encompassing.

4.2.3 More Nuanced Alternations

A crucial question arises at this point: since pairs of LVs seem to display transitivity alternations, why can’t this be captured by only two bleached LVs? Why do we need so many LVs? This issue will be taken up more comprehensively in the next chapter. However, we

139

can begin to answer the question here by looking at isolated alternations, where LVs share only a single island.

One such instance can be observed in the previous diagram, by adding some connections. ke∫idæn ‘to pull’ and ændaxtæn ‘to throw,’ alternate with zædæn and xordæn in two of the islands that involve weapons. Consider the following forms:

1c) t∫aqu ke∫idæn knife KE∫IDÆN pull a knife on someone 1d) *t∫aqu ændaxtæn knife ÆNDAXTÆN (not a valid construction to express attacking with a knife) 2c) *tir ke∫idæn bullet KE∫IDÆN (not a valid construction)23 2d) tir ændaxtan bullet ÆNDAXTÆN shoot a bullet 1c shows that any gun or knife-like PV (hand-held weapon) combines with ke∫idæn to mean roughly ‘pulling a knife/gun on someone or threatening or attacking with a knife/gun’ (without necessarily shooting or stabbing the person like with zædæn). 1d, however, is unacceptable, because only weapons that are projectiles combine with ændaxtæn (hence the acceptability of 2d). Also, tir ke∫idæn ‘bullet PULL’ is unacceptable, since a bullet is not a “weapon” which can be used manually by itself. The two alternations occur at the most fine-grained level: they only share one island with the other LVs. These alternations can be captured in the following diagram:

23 In fact, tir ke∫idæn is a valid construction only with a homonym of tir which means searing pain, and not a bullet. Tir has many homonyms, such as pole, the name of the 4th Persian calendar, etc. Only tir, meaning searing pain combines with ke∫idæn, all other homonyms are excluded from combining with this LV.

140

offensive

ZÆDÆN[to hit]

act on something

offensive

state change

topology

fusion

damage

gereh [knot]

ta [fold]

fer [curl]

bæxiye [stitch]

bænd [connection]

peyvænd [graft]

lætme [damage]

qat! [chop]

sædæme [damage]

trick

bambul [trick]

gul [trick]

hoqe [trick]

type of hit

instru-ment

zærbe [hit]

lægæd [kick]

kotæk [beating]

t!æko! [hammer]

næzær [opinion]

potk [sledgehammer]

XORDÆN[to eat]

undergo

hand held

projectile

weapon

tir [bullet]

mu!æk [missile]

toop [canon]

kard [knife]

!æm!ir [sword]

t!aqu [knife]

state change

weapon

ÆNDAXTÆN[to throw]

KE!IDÆN[to pull]

This phenomenon allow us to predict and preclude a third form from the connections between the LVs. If we introduce the word cruise to Persian, the system will predict a verb form with ændaxtæn and precludes one with ke∫idæn, whereas a PV like dagger shows the opposite pattern:

cruise XORDÆN cruise ZADÆN cruise ÆNDAXTÆN ≠ *cruise KE∫IDÆN dagger XORDÆN dagger ZADÆN dagger KE∫IDÆN ≠ *dagger ÆNDAXTÆN Importantly, these examples show that very specific attributes of the weapon-type PV allow or disallow combination with a third LV. The LVs don’t alternate on purely grammatical or aspectual features. In the case above, the dynamics of the action determine the acceptability of the alternation. Ke∫idæn means ‘to pull or to draw,’ an action that comes to mind when thinking of the motion of extending or drawing a weapon through the air towards a target in a controlled manner. The dynamic configuration of this action plays a large role in shaping the meaning of the corresponding construction, along with features such as agentivity, control,

141

and the fact that the action doesn’t cause a change of state24. Similarly, the LV ændaxtæn has the ideas of an initial impulse, a projection, and a trajectory.

4.2.4 Is it just transitivity?

Let us return to what seems at first glance to be a transitivity alternation between zædæn and xordæn. Xordæn is not the only intransitive alternate available for zædæn. Some transitive/intransitive pairs exist with LVs other than xordæn, excluding LVCs with xordæn. One of these is constructed with the LV gereftæn ‘to get.’ This island expresses a process involving the absorption of something.

næm zædæn moisture ZADÆN wet, moisten næm gereftæn moisture GEREFTÆN get moist or become damp atæ∫ zædæn fire ZADÆN set on fire atæ∫ gereftæn fire GEREFTÆN catch fire mærd-e divane xune-ye xod-æ∫ ra atæ∫ zæd. man-GEN crazy home-GEN himself-3SG ACC fire hit.PST.3SG.. The crazy man set fire to his own house. astin-æm ke næzdik-e ∫ole ∫od, atæ∫ gereft. sleeve-1SG that near-GEN flame become.PST.3SG., fire get.PST.3SG.. As my sleeve got close to the flame, it caught on fire. There aren’t many examples from this particular island (not very populated, but the existence of these few forms remains very informative. Vahedi-Langrudi (1996, p240) is the only linguist who looks at this alternation. His explanation for this alternation is that the PVs in 24 Configuration is not the only parameter that’s evoked in alternations between ke∫idæn and zædæn. Alternations between these two LVs can involve durational aspect in other islands. These forms belong to the emission island seen in zædæn, namely the island labeled “emotional noises.” dad zædæn shout ZÆDÆN shout (short) dad ke∫idæn shout KE∫IDÆN shout (long) Though the duration explanation for these alternations hold true globally, other islands that alternate with ke∫idæn have more nuanced explanations. Consider the following forms (from the appearance, weapon, touch, and drugs islands of zædæn): 1. ∫ane ke∫idæn brush KE∫IDÆN brush 2. ∫æm∫ir ke∫idæn sword KE∫IDÆN pull a sword 3. mævad ke∫idæn drugs KE∫IDÆN smoke drugs 4. dæst ke∫idæn hand KE∫IDÆN slide a hand across These islands are scattered in the semantic space of both these LVs. The form in 1 highlights the pulling action of brushing (whereas the form with zædæn would either express a quick brushing of the hair or describe the action as a whole). 2 depends on the physical actions involved, as explained above. 3 highlights the pulling, or inhaling, of a substance into the body, whereas a form with zædæn describes injecting or piercing. Finally, 4 describes an action of pulling across a surface, rather than simply touching it (as expressed with the analogous LVC with zædæn). These LVCs display the fact that deciphering the meaning or choosing an LVC requires both real-world knowledge (about the function of the PV), aspectual properties, and knowledge about the physical qualities of the action involved (usually from the core meaning of the LV. These forms show that the LVs are not bleached, but add substantial information to add to the meaning of the whole expression.

142

these cases do not count as instruments and there is no physical or abstract repetitive movement or manipulation targeting the theme of the action which he claims is representative of the zædæn/xordæn alternation. However, though there aren’t instruments involved, there is no reason to suppose that there is no manipulation affecting the theme of the action in the above pair. Setting fire to something requires some kind of contact with it, as does wetting something. Furthermore, some alternations between zædæn/xordæn, as seen above, don’t involve instruments. Perhaps one of the reason for the non-occurrence of xordæn in this pair, at least in the case of “catching on fire” is, as Vahedi-Langrudi also mentions, that there is no inception with xordæn. Gereftæn expresses inception. Xordæn, on the other hand, expresses the beginning and the end of the action, and it expresses impact.

These types of alternations help further describe the characteristics of each LV and its constructions. Finding restrictions and limits to alternating features allows us to develop a more fine-grained description of the core meaning of the LVs.

Another alternation difficult to explain is one between zædæn, xordæn, and dadæn. This alternation further shows the limits of the transitivity argument for the zædæn/xordæn alternation. These three LVs share the island of rotational movement. Consider the following examples:

qælt zædæn somersault ZADÆN do a somersault qælt xordæn somersault XORDÆN ‘undergo’ a somersault qælt dadæn somersault DADÆN make something flip over t∫ærx zædæn turn ZADÆN turn t∫ærx xordæn turn XORDÆN ‘undergo’ a turn t∫ærx dadæn turn DADÆN make turn Dadæn, and not zædæn, is valid for expressing an agentive action in this island. The difference between zædæn and xordæn concerns intentionality, and both are reflexive. With zædæn, the action suggests intentionality whereas with xordæn, the LVC expresses the same action without intention.

143

DADÆN

XORDÆN

movem

ent

ZÆDÆNrotation

pit! [twist] qælt

[flip]

t!ærx [turn]

As can be seen in the following examples, the subject of the LVC with zædæn is volitional and propels himself in a rotational movement, whereas the subject of the LVC with xordæn is inanimate and merely undergoes the movement. With dadæn, the subject imposes a movement on the object. Consider the following sentences (dadæn expresses a transitive, causative action; zædæn expresses a conscious unergative action25; and xordæn expresses an unconscious action: e.g. inanimate).

ræqqas ba ritm-e musik t∫ærx mi-zæd. dancer with rhythm-GEN music turn PROG-hit.PST.3SG.. The dancer spun to the rhythm of the music. hævapeimaye kaqæzi dær hæva t∫ærx xord. airplane-GEN paper in air turn eat.PST.3SG.. The paper airplane turned in the air. mot∫-æ∫ ra gereft o do bar t∫ærx dad. wrist-3SG ACC get.PST.3SG. and two times turn give.PST.3SG.. He grabbed his wrist and twisted it around twice. From the patterns of LVC formation we saw with zædæn, we would expect the transitive version of the verb to be expressed with zædæn and not dadæn. However, it seems that it is 25 Interestingly, zædæn can also be used for inanimate objects in certain contexts (e.g. ma∫in t∫ænd bar t∫ærx zæd: the car flipped over several times). This could be because of the momentum already present before the rotational action is undertaken.

144

more important, in these forms, to express the unergative aspect through zædæn. On the other hand, dædæn expresses mostly causative actions26.

Furthermore, dadæn and xordæn share islands that don’t connect to any in zædæn. Xordæn occurs in causative/inchoative alternations with dadæn as well. Consider the following LVCs:

tækan xordæn shake XORDÆN shake (no causation) tækan dadæn shake DADÆN cause something to shake ∫ekænje xordæn torture XORDÆN be tortured ∫ekænje dadæn torture DADÆN torture qose xordæn grief XORDÆN grieve qose dadæn grief DADÆN cause to grieve nozul xordæn interest XORDÆN charge interest nozul dadæn interest DADÆN pay interest The above forms each come from different islands shared by xordæn and dadæn. The reason for the choice of dadæn over the expected LV, zædæn, is not clear in these cases.

26 In certain limited cases, even dadæn occurs in unergative LVCs (e.g. læm dadæn ‘lean dadæn,’ to lean back on, relax)

145

XORDÆN[to eat]

suffering

motion

emotion

qose [grief]

æsæf [regret]

æzab [torment]

qælt [sommersault]

pit! [twist]

tækan [motion]money

nozul [interest]

re!ve [bribe]

DADÆN[to give]

offer impose

rotation

motion

general

ZÆDÆN[to hit]

Vahedi-Langrudi (1996, p242) suggests that it might just be idiosyncratic. He suggests that one of the differences between dadæn and zædæn is that the former causes events that occur with a continuous or permanent contact of two objects. Also its PVs are never instruments of contact, and the action expressed does not involve any sort of manipulation. It seems indeed that the LVCs that alternate with dadæn express a “longer” activity than those with zædæn which are usually short or abrupt.

This explanation does not suffice. There is an island shared between zædæn and xordæn that expresses “damaging something” or “being damaged” (see islands in Chapter 3). According to the above mentioned suggestion, the LVCs of this island should alternate between dadæn and xordæn because of manipulation of the object, but they don’t. Also, the last example above is not a causative/inchoative alternation. They are both agentive. In the first case, the subject obtains money (note that this island in xordæn has negative connotations, expressing the notion of usury or of being unjust). In the second case (with dadæn), the subject gives or pays money. Importantly, because of the negative connation of the first LVCs, these constructions cannot alternate in all contexts. Namely when the LVC with xordæn only occurs in contexts where the transaction is unjust. The more neutral receiving of money would be expressed by an analogous construction with gereftæn whose alternation with dadæn will be explored in the next section.

146

4.2.5 Alternations between dadæn and gereftæn

Of the LVs studied in the previous chapter, dadæn and gereftæn also possess large numbers of shared islands. Many of the alternations between these two LVs can be grossly characterized as unaccusative/inchoative alternations as discussed in the previous section with zædæn and xordæn. This can be expected given the core meanings of the two LVs. Dadæn ‘to give’ expresses causing a transfer, and gereftæn ‘to obtain’ expresses receiving. Around fifteen islands are shared between these two LVs. Some example LVCs are:

ejaze dadæn permission DADÆN grant permission ejaze gereftæn permission GEREFTÆN get permission næzær dadæn opinion DADÆN give an opinion næzær gereftæn opinion GEREFTÆN get an opinion pænah dadæn refuge DADÆN give refuge pænah gereftæn refuge GEREFTÆN take refuge sefare∫ dadæn order DADÆN order sefare∫ gereftæn order GEREFTÆN take an order dærs dadæn lesson DADÆN teach dærs gereftæn lesson GEREFTÆN learn ∫ekl dadæn shape DADÆN give shape ∫ekl gereftæn shape GEREFTÆN take shape tærtib dadæn order DADÆN put some order in something tærtib gereftæn order GEREFTÆN get organized Unlike the zædæn/xordæn alternation, this alternation does not involve islands that are in close proximity. In other words, where in zædæn we had a cluster of related islands alternating with their analogous forms with xordæn, gereftæn and dadæn share islands that are dispersed in their semantic spaces. For this reason, it is difficult to give a concise two dimensional visual description. If we warp the two spaces, we can get a figure that shows most, but not all of the relations between the two LVs.

147

a bind

DADÆN[to give]

offer

impose

express knowledge

provide

render

favor

physicalchange

assume


service

obtain

settlement

look

position

shape

position/ place

lesson

response money

help

ejare [rent]

nozul [interest]

vam [loan]

yari [help]

komæk [help]

re!ve [bribe]

dærs [lesson]

sæboq [precedence]

pænd [advice]

æhval [health]

xæbær [news]

arame! [calm]

næzm [order]

!ohræt [fame]

qærar [position]

ja[place]

væqt [time]

!ekl [form]

suræt [face]

næx! [design]

færman [order] sefare!

[order]

dæstur [order]

order

info

receive

safety/cure

permission

trait

saman [settled]

ælaj [cure]

pænah [refuge] !æfa

[cure]


ejaze [permission]

reserve

bind

qol [promise]

væ'ede [promise]

ja [place]

væqt [time]

Though many of the alternations do indeed fit the pattern of transitivity alternations, this cannot be the only overarching generalization. We already saw that dadæn alternates with other LVs, as shown in section 4.2.4 (e.g. xordæn ‘to eat’ and ke∫idæn ‘to pull), as does gereftæn, as shown in the same section (e.g. zædæn ‘to hit’).

Gereftæn and dadæn also share islands that don’t alternate according to the subject’s role in the sentence. Consider the following LVCs:

bu dadæn smell DADÆN be smelly (emanate a smell) bu gereftæn smell GEREFTÆN get smelly jan dadæn life DADÆN die jan gereftæn life GEREFTÆN be invigorated In all these cases, the subject is a proto-patient. In the first cases, with dadæn, the subject smells, involuntarily giving off an odor; and with gereftæn the subject absorbs an odor. In the second pair, the subject gives up his life, or receives life energy, respectively with dadæn and gereftæn. Though these forms don’t alternate according to transitivity, they highlight the crucial parameter of directed movement, or transfer, which plays an important role in the meanings of these LVCs. Perhaps, one of the reasons the alternations do not occur LV-wide

148

is because each LV expresses many different parameters, and each parameter is foregrounded or backgrounded differently depending on the notion expressed. Thus, sometimes transfer is not the priority.

4.2.6 Conclusion

In the last three sections, we explored the alternations between the LVs presented in Chapter 3. We found that apart from linguistic parameters, like inchoative, transitivity, unergativity, other factors also play important roles to differentiate and validate the LVs in the system. These parameters include physical parameters such as the type of movement involved in the action expressed, and the direction of the movement.

4.4 Chapter Review

From the facts surveyed in this chapter, we can make certain predictions about the dynamics of the verbal system. First, as we saw with zædæn and xordæn, alternations can occur between two LVs involving clusters of islands. Second, as we saw with ke∫idæn, ændaxtæn, and other smaller alternation (usually single shared islands), systematic alternations can involve more nuanced parameters. Furthermore, in the first case, generalizations about the alternations are easier to extract in the form of grammatical features (e.g. agentivity in the majority of alternations between zædæn and xordæn) than if the islands are scattered, or are isolated. In other words, the more scattered the alternations, the more specific and nuanced the generalization. And since no LV alternates with all LVCs of another LV, but at most with a cluster of islands, we expect to get semantically complex generalizations attributed to the different alternations.

The existence of these alternations demonstrates that this is a generative system. Not only is there systematicity in the space of each LV, as seen in Chapter 3, but they can share spaces and interconnect. In Chapter 5, I will discuss semi-productivity and how opaque semantics does not necessarily hinder the productivity of a form. I will also return to the discussion of constructions and other basic linguistic ideas that are instantiated by this system.

149

5 CONSTRUCTIONS AND DISCUSSION

In the previous chapters we explored the semantic space exploited by the Persian verbal system. We investigated the organization of individual LVs as well as connections between certain LVs. We concluded that productivity and compositional semantics would result directly from the proposed model. In this chapter I address the transparency of LVCs, their uses as idioms, semi-productivity, and the status of constructions.

5.1 Transparency

One of the most interesting and intriguing properties we observe in Persian LVCs is their non-transparent semantics, or at best semi-transparent semantics. As we have seen in the last two chapters, it is often difficult to derive the meaning of an LVC directly through the meanings of its parts. In other words, the contribution to the LVC’s meaning by each of its two constituents is difficult to discern. It seems that the LVCs in Persian lie on a continuum, between idioms (opaque or frozen expressions) and fully transparent cases. A handful of examples can be found at the two extremes of transparency, however most LVCs fall somewhere in between.

5.1.1 Transparent LVCs

Transparent LVCs are constructions where the LV retains all its core meaning, and the resulting LVC has a compositional meaning. Examples of these forms include:

fekr kærdæn thought KÆRDÆN think (lit. do thought) kado dadæn gift DADÆN give a gift qævi ∫odæn strong ∫ODÆN become strong nun xordæn bread XORDÆN eat bread payin amædæn down AMÆDÆN come down, descend When the contribution of the two constituents of an LVC is clear, the island that the LVC belongs to tends to be highly productive. For example, any type of food occurring with xordæn will mean to eat that type of food. Here, xordæn explicitly expresses its core meaning to eat.

The LV needn’t retain its original meaning to occur in a transparent construction. Sometimes, it is sufficient to postulate a new verb meaning. For example, in zædæn, the island Musical Instrument produces a potentially infinite number of LVCs. The construction takes a musical instruments as PV and the verbal constituent, zædæn, is equivalent to the English use of the verb play or the French uses of the verb jouer to mean “to perform on an musical instrument.” In cases like these, positing constructions is equivalent to positing lexical rules

150

through which the constituents of LVCs combine. In other words, in these cases, the verb zædæn can be said to have a secondary meaning “to perform on a musical instrument.” 27

5.1.2 Idiomatic LVCs

At the other extreme of the transparency continuum, there are a handful of apparently opaque LVCs. Some examples include:

xo∫k zædæn dry ZÆDÆN freeze in fear left dadæn fat DADÆN drag out, procrastinate tæmiz dadæn clean DADÆN distinguish ab qure gereftæn verjuice GEREFTÆN cry gærdæn gereftæn neck GEREFTÆN declare oneself responsible ja xordæn place XORDÆN be very surprised ju∫ xordæn boil XORDÆN worry These types of expressions are referred to as idioms. The term idiom can be defined so that it covers only totally frozen expressions. The meaning of these expressions can’t be even remotely discerned from the meaning of its parts, and they don’t belong to productive constructions. They are stored individually in the lexicon.

There are many LVCs that are idiom-like, in that some aspects of the meaning are not predictable, but whose constituents do in fact contribute directly to the meaning of the whole. These constructions can also be productive. Most Persian LVCs fall in this in-between category, where the meaning of the construction is highly constrained by the contribution of its constituents (this contribution can be figurative or abstract), as well as the meaning of the whole. The meaning of these forms aren’t compositional, and can be best captured in the form of a construction.

5.2 Constructions

The most interesting cases of Persian LVCs are between the two extremes of transparency, presented above. Most often, the contribution of the two parts cannot be fully distinguished though they are not fully opaque either. The meaning of the forms is constrained by syntactic and sometimes idiosyncratic semantics, but the forms display more freedom (syntactic and semantic) than conventional idioms. As we saw in Chapter 1, Persian LVCs can undergo numerous grammatical processes signalling that the structure of the constructions are in fact transparent to these processes.

27 The reason why certain forms develop into the equivalent of secondary meanings is beyond the scope of this study, but one can surmise that long historical use of an LV with a large class of similar PVs will eventually produce something like a second meaning for the LV. Interestingly, these are the cases where the LV can be used along in a sentence without an explicit PV to express this secondary meaning. Consider:

Ali t∫i mizæne? “Ali what play?” What instrument does Ali play? Here, the verb zædæn refers to playing an instrument, and the actual construction is not explicitly needed to trigger this meaning. A similar case could be made regarding diseases and the LV gereftæn. Where the secondary meaning is sometimes used alone, such as in:

mæn æz mohsen gereftæm me from mohsen got I caught it from Mohsen This phenomenon is analogous to the English use of the verb catch or the French attraper used to express getting sick. It is important to note that in each of these cases, the context of the discussion plays a crucial role in determining the emergence of the correct verb meaning, as is typical in the case of polysemy.

151

5.2.1 Idiomatically Combining Expressions

Nunberg et al. (1994) propose that most of the expressions loosely categorized as idioms are in fact idiomatically combining expressions. In these expressions, conventional mapping from literal to idiomatic interpretation is homomorphic with respect to certain properties of interpretation of the components (Nunberg et al, 1994, p 503). Furthermore, they predict families of idioms where a verb occurs in different environments to form distinct but related idiomatically combining expressions. Unlike idioms, which they call idiom phrases, idiomatically combining expressions can undergo grammatical transformations.

The formation of families of related forms is reminiscent of the LVC islands formed in the Persian verbal semantic space. Certain properties of the core meaning of the LV emerge in LVCs, alongside a clear contribution by the PV. Persian LVCs seem to fall under this category of linguistic phenomena, as defined by Nunberg et al. (1994).

Karimi (1996) discusses the idiomatic status of Persian LVCs. She claims that only a handful of Persian LVCs can be considered idiomatic. The majority of LVCs are idiomatic combining expressions whose idiomatic meanings are composed on the basis of the meaning of their parts (Karimi, 1996, p 23). She discusses certain non-productive LVCs with opaque meanings (p 24):

Let us examine châne zadan (chin hitting=to negotiate) and xar kardan (donkey doing=to fool someone) with regard to their compositionality. Once we learn the meaning of these idiomatic CVs, we can see the relation between their parts. That is, châne zadan involves figuration indicating that negotiation requires moving the chin. Xar kardan, on the other hand, becomes transparent on the basis of the idiomatic meaning of xar ‘donkey’ which implies foolishness and stupidity: there is an apparent relation between the concrete and the abstract meaning of xar that provides the idiomatic meaning of the CV xar kardan.

Nunberg et al. (1994) mention that LVCs “often come in families, e.g. take a walk, take a hike, take a trip, take a flight. Listing each such expression results in considerable loss of generality and lack of prediction.” In other words, it is redundant to list each LVC as a separate entry, as if it was an idiom. Furthermore, it would not account for the production of new forms. Constructions capture generalizations. The properties of its constituents motivate the resulting meaning. In this sense, constructions are like lexical items, they must be learned and are idiosyncratic. Once learned, they can be used to produce new forms.

This notion of “family” clearly pertains to the islands of LVCs in Persian. Generalizations can be caught at this level to avoid redundant entries. Surely, these islands must be learned (though their motivation might be clear upon close examination), but they allow for the production of new forms.

In Constructions (1996, p154), Goldberg touches on this point:

If one knows that a construction has a particular form, then it is sometimes the case that one may reasonably infer that it has the particular interpretation that it has. However, it is fallacious to argue that because we may be able to pragmatically infer the meaning of a construction, its existence is therefore predictable rather than conventionalized. Such reasoning is based solely on a model of interpretation, yet we also must account for production.

152

In other words, even though motivations of a particular construction can be deduced post hoc, constructions are learned (we cannot predict their existence). However, the forms that result from the construction can be predicted. This approach avoids redundant entries in the lexicon, as well as avoiding multiple and idiosyncratic senses for certain verbs (p140):

the reason for postulating constructions is analogous to the reason why other researchers have wanted to postulate a lexical rule: in order to capture generalizations across instances. Moreover, it is claimed here that what is stored is the knowledge that a particular verb with its inherent meaning can be used in a particular construction. This is equivalent to saying that the composite fused structure involving both verb and construction is stored in memory. By recognizing the stored entity to be a composite structure, we gain the benefits [...] over a lexical rule account. For example, we avoid implausible senses such as “to cause to receive by kicking.” It is the composite structure of the verb and construction that has this meaning. We also allow other syntactic processes to refer to the inherent lexical semantics of the verb. Thus we do not lose the information conveyed by the verb, because the verb is not changed into a new verb with a different sense.

As we saw in previous chapters, Persian LVCs are built upon functional and sometimes perceptual traits of the PV. The semantics and the real-world knowledge attributed to the PV play a crucial role in determining the meaning of an LVC. To take an example from our analysis, we saw in Chapter 4 that one of the constructions with dadæn expresses the action of making an entity undergo a certain type of regimen.

værze∫ dadæn exercise DADÆN exercise pærvære∫ dadæn breed DADÆN raise ro∫t dadæn growth DADÆN grow ∫ost o ∫u dadæn cleansing DADÆN clean Instead of positing a highly idiosyncratic meaning of the sort “constrain an entity to adhere to the process of + PV,” the meaning of the construction results from the interaction of the PV and the LV. The speaker must be aware of the properties of the PV, namely that it must be a process of development or of physical conditioning. This knowledge and the knowledge that dadæn occurs in such a construction must be learned. However, once learned, new forms can be constructed based on this structure. Another example involves gereftæn when it occurs in a construction to express gathering to celebrate or commemorate something, specifically for socially important or festive occasions.

xætm gereftæn memorial GEREFTÆN have a memorial service ærusi gereftæn wedding GEREFTÆN throw a wedding tævælod gereftæn birthday GEREFTÆN have a birthday party If gereftæn simply meant to celebrate, then there wouldn’t be any restriction on the PVs it could combine with. However, only when combined with a specific type of PV does the meaning of the composite expression emerge. The meaning is allotted to the whole, and not to the LV itself.

Another important property of constructions is that they encode syntactic structure. Certain LVs can form both transitive and intransitive LVCs in their semantic space. This likens

153

constructions to grammatical processes of language, further blurring the distinction between grammar and lexicon.

5.2.2 Dual Nature of Constructions

The following set of constructions display this dual nature of constructions, residing somewhere between the grammar and lexicon of a language. While maintaining a status similar to that of lexical items (with explicit lexical information), LVCs also carry grammatical specifications. Certain specific syntactic structures can be observed with several LVs. This implies that there could be a hierarchy of constructions, where those higher in the hierarchy provide a general structure and meaning to those lower in the hierarchy. In other words, “the motivation for one construction having the form it does may come from the inheritance hierarchy itself, insofar as the form is inherited by a construction higher in the hierarchy” (Goldberg 2004, p3). For example, the LVCs presented below all share the same syntactic structure and very general semantics. These forms are particularly interesting because of their odd syntactic structure. Their sharing of this structure could imply that there is a construction higher in the hierarchy from which these constructions inherit their structure28.

The LVs gereftæn, amædæn, and sometimes ∫odæn combine with a nominal PV to form an LVC with a highly idiosyncratic syntactic structure. Many linguists exclude these forms from their analyses of LVCs due to their syntax, and because they form full sentences and aren’t considered verbal constructions. However, despite their unique syntactic characteristics, I propose that their study should be included in a comprehensive study of LVCs. This syntactic structure only occurs with LVs in the system, and not with any other verbal construction.

The meaning of these constructions cannot be deduced solely from the meaning of their parts, nonetheless, the constructions are productive. In general, in these constructions the LV always occurs in the third person singular, and the PV always takes a pronominal suffix. This suffix expresses the experiencer of the event. The LVCs with this structure express a seizure of some kind, whether it be emotional, physical, or psychological. Some examples are:

bæd-æm amæd bad-1SG come-3SG I disliked it. dærd-e∫ gereft pain-3SG obtain- 3SG He was seized with pain. hæsudi-etun ∫od jealousy-2PL become-3SG You were jealous. æz dust-e xahær-æm bæd-æm mi-ay-æd. from friend-GEN sister-1SG bad-1SG PROG-come.PRS-3SG. I disliked my sister’s friend. These impersonal LVCs cannot undergo any nominal derivation, and thus never occur as infinitives or participles. The PV cannot be modified, as is the case with most PVs implicated in other LVCs. In analogous forms from other languages, the experiencer or “subject” occurs

28 Goldberg (2004) discusses the general structure of Persian LVCs. In her article, she examines the dual nature of these constructions; namely, their phrasal and lexical behaviors. She concludes that these Persian verbs are indeed constructions stored in the lexicon. The general construction she proposes, higher in the hierarchy than those discussed in this study, captures generalizations in the system. These generalizations include “the formation of nominalizations, the resistance to separation, the stress facts and possible non-compositional semantics or argument structure” (Goldberg, 2004, p 17). The more specific LVCs presented here inherit these properties from the more general one.

154

in the dative or genitive case (Ghomeishi, 1996, p276). But, since Persian does not encode for case, the grammatical relation of the elements involved are unclear.

Vahedi-Langrudi (1996) likens these forms to what he calls defective, degenerate predication29. He claims that no lexical element can be in the subject position in these forms. However, the degenerate predicates he studies involve only ∫odæn and do not involve the subject pronominal enclitic attached to the PV as in the forms above. He explains that the impossibility of derivation in these forms is due to the lack of proper predication and this “prevents later application of word-formation” (Vahedi-Langrudi, 1996, p184).

Though syntactically identical, different LVs used in these constructions result in differences in meaning, while PVs combined with the same LV result in systematically related meanings (as in the set below). The construction formed with gereftæn seems to be most productive. This construction was presented in the analysis of gereftæn’s islands. Here, it is clear that this construction actually belongs to a bigger family of constructions that share the same syntactic structure.

With gereftæn, this construction expresses the sudden seizure of a psychological or biological state or pressure. The human subject of these forms experiences a state or condition that seizes or overwhelms him. The PV always expresses an involuntary state or urge that seizes the subject which usually needs to be manifested or released somehow. This may be a general state or action. The subject anticipates carrying out the action, he is seized with the urge to carry out the action.

oq-æm gereft gag-1SG GEREFT.3SG I felt like vomiting xænde-æm gereft laughter-1SG GEREFT.3SG I was seized with laughter bazi kærdæn-æm gereft play do-1SG GEREFT.3SG I became playful ∫a∫-æm gereft urine-1SG GEREFT.3SG I had the urge to urinate ræqs-æm gereft dance-1SG GEREFT.3SG I got the urge to dance gery-æm gereft cry-1SG GEREFT.3SG I wanted to cry æts-æm gereft sneeze-1SG GEREFT.3SG I had the urge to sneeze

In his article on gereftæn, Sharifi (1974, p475) says that “it is not entirely clear whether the other nominal unit should be considered a patient or as part of the compound verb.” He claims that the experiencer is indicated through the pronominal suffix. He makes an interesting conjecture on the semantics of these forms, positing that the meaning of the LVC “is similar to a kind of biological, emotional seizure, resulting in the action which is temporarily controlled” and in addition to the semantic feature “experiential” and “state,” these verbs might also be labeled as “delayed action.” In other words, the type of action expressed in these constructions with gereftæn is one of anticipated action. An internal emotional, psychological, or biological pressure seizes the experiencer who must temporarily restrain himself until he is able to release that pressure in an associated action.

With gereftæn, there is a similar construction that serves to express a cramping or clogging situation. In these forms, the PV itself is an experiencer or patient. The PV undergoes a clogging, choking, or cramping process. The 1SG marker here is the possessive, and not a clitic. See section 3.4.3 for more details.

29 cases where the subject position of the predicate is empty.

155

mast gereft yoghurt GEREFT.3SG the yoghurt set hæva gereft air GEREFT.3S the sky clouded up mah gereft moon GEREFT.3SG there was an eclipse of the moon mahit∫-æm gereft muscle-1SG GEREFT.3SG I got a cramp in my muscle næfæs-æm gereft breath-1SG GEREFT.3SG I had shortness of breath lule gereft pipe GEREFT.3SG the pipe clogged up The syntactic structure of the preceding LVC also occurs with the LV amædæn ‘to come.’ As in the forms with gereftæn, these constructions express the seizure of the subject by a state. However, with amædæn, the PV always expresses an emotional state that determines the attitude of the subject (expressed through the pronominal suffix) towards something.

xo∫-æm amæd good-1SG AMÆD.3SG I liked… bæd-æm amæd bad-1SG AMÆD.3SG I disliked… heyf-æm amæd pity-1SG AMÆD.3SG I felt unjust… xæsisi-æm amæd stingy-1SG AMÆD.3SG I became stingy... A small number of these forms occur with the LV ∫odæn. Again, the experiencer is seized by some force, here as a consequence of the environment.

hæsudiæm ∫od30 jealousy-1SG ∫OD.3SG I got jealous. dir-æm ∫od late-1SG ∫OD.3SG I’m running late31. særd-e∫an ∫od cold-3PL ∫OD.3SG They got cold. gærm-et ∫od hot-2SG ∫OD.3SG You got warm. Interestingly, the word xab ‘sleep’ is used in constructions of these kinds with several LVs. The following are the most common constructions this word appears in:

xab-æm gereft sleep-1SG got I got sleepy. xab-æm ræft sleep-1SG went I fell asleep. xab-æm bord sleep-1SG took I fell asleep. xab-æm amæd sleep-1SG came I was sleepy. In these LVCs, it is as if xab is a forceful, conscious entity acting as subject of the construction seizing the object. This is the only PV that combines with more than one LV in these constructions. Also, we must note that the LVs ræftæn and bordæn do not occur with this construction productively.

The existence of this construction suggests that a syntactic construction exists without a specific LV, and it higher in the hierarchy of constructions. The syntactic structure and a vague description of the semantics (experiencer being seized by the condition given by the PV) exist independently of the particular instantiations of it with explicit LVs. These constructions provide good criteria for the hypothesis that LVCs occur somewhere between the lexicon and the grammar of Persian.

30 This form can also be expressed with amædæn: hæsudi-æm amæd. 31 It is interesting to note that this form is acceptable, but *zud-æm ∫od (I am early) is not. This is probably due to the fact that being early doesn’t seize the subject as a feeling or as a necessity to act.

156

5.2.2 Motivations

One of the reasons the semantic space is divided into groups of branches lies in the similarities between the meaning of the LVCs. The LVCs grouped together tend to be similarly motivated with regard to the core meaning of the LV.

Goldberg posits polysemy links to illuminate the semantic relations between a particular sense of a construction and extensions of this sense (1995, p 76). Here, a similar relation or link holds, except that syntactic information is often not inherited from construction to construction. Since the LVCs presented in each semantic space occur at the same level, it would be difficult to find evidence that one construction occurs higher than another in some hierarchy (a prerequisite for inheritance). Furthermore, the semantics are only loosely related, and there is no dominating LVC, all constructions are motivated by the core meaning of the LV. Of course, motivations can result from different attributes of the core meaning. Spatial, temporal, dynamic, and aspectual attributes play a large role in their LVC occurrences. The contributing attributes can be quite abstract and in such cases the meaning of the LVC can deviate far from the core meaning of the LV.

There seem to be at least three non-exclusive types of meaning motivations: figuration, aspectualization, and instantiation of the core meaning. In figuration, there is a metaphorical image inspired by the physical or dynamic attributes of the LV. For example, the LV xordæn occurs in forms that seem to portray images of swallowing, or being penetrated or pierced by something. The same idea is present in ke∫idæn where there is an idea of pulling something with or towards the subject. Consider the following:

∫æm∫ir xordæn sword XORDÆN be wounded by a sword dæst xordæn hand XORDÆN be touched zæxm xordæn wound XORDÆN be wounded pip ke∫idæn pipe KE∫IDÆN smoke a pipe hæftir ke∫idæn gun KE∫IDÆN pull a gun on someone zæjr ke∫idæn torment KE∫IDÆN suffer In the cases with xordæn, there is an idea or image of being affected by something, being penetrated, or taking in. With ke∫idæn, the LV provides the idea of pulling towards the subject (as with pipe smoke), pulling out and across (as with gun), and pulling a psychological weight (as with torment).

Sweetser (1990) proposes that lexical content is structured by meaning schemata. In metaphorical constructions, the meaning schema of one item maps onto a different domain and is used to organize the new elements according to its structure. This seems to be the case in figurative motivations. Goldberg (1995, p81) captures this type of motivation through a metaphorical link. The attributes implicated in the mapping are specified in such a link.

In aspectual motivations, as with many LVCs formed with zædæn, only the aspectual traits of the action remain. Here, the idea is that of a quick action which is instantaneous and possibly repetitive. In the following forms, the idea of hitting, as a motion outward emerge in the figural sense. It is the temporal structure of the core meaning that remains. These actions are carried-out quickly and possibly repetitively. This type of link is similar to the figurative or metaphorical link above, but imagery is not the basis of the mapping.

157

t∫e∫mæk zædæn blink ZÆDÆN wink paru zædæn paddle ZÆDÆN paddle gaz zædæn bite ZÆDÆN bite And finally, there are cases where the core meaning of the LV explicitly motivates the meaning of the resulting LVC, though it becomes more nuanced. This is the case for some LVCs constructed with kærdæn and zædæn.

tæqaza kærdæn request KÆRDÆN request tærjome kærdæn translation KÆRDÆN translate eltemas kærdæn plea KÆRDÆN plead kotæk zædæn beating ZÆDÆN beat mo∫t zædæn fist ZÆDÆN punch sili zædæn slap ZÆDÆN slap This type of motivation can be captured by an instance link (Goldberg, 1995, p79) where the construction is a special case of another. In this case, the LVC is an instantiation of the LV’s core meaning. As can be seen above, zædæn occurs in LVCs with different types of motivations. This holds true for all LVs: they emerge in different LVCs expressing different attributes of their core meanings. The characterizations presented here show that there are several ways in which the LV can reemerge with respect to its core meaning. Variants must be learned on a case by case basis, but the core meaning determines and motivates the non-core meanings of the LVs.

In her discussion of Persian LVCs, Goldberg (2004, p3) states that

motivation can be provided by factors outside of the language-particular grammar, for example, by appeal to constraints on acquisition, principles of grammaticalization, discourse demands, iconic principles or general principles of processing or categorization. Alternatively, motivation may come from within the grammar. [...] Motivation is distinct from prediction: recognizing the motivation for a construction does not entail that the construction must exist in that language or in any language. It simply explains why the construction is “natural.”

One of the hypotheses of this thesis is that almost all LVCs are motivated by the core meaning of their respective LVs. As presented in Karimi (1996), cited above, even seemingly opaque LVCs display some sort of motivation from the core meaning of the LV. Most have clear semantic motivations (as with t∫ane zædæn ‘chin hit’ for bargaining, and xær kærdæn ‘donkey make’ for fooling someone (Karimi, 1996)). Furthermore, many isolated LVCs partake in alternations. Consider the following:

be doniya amædæn to world AMÆDÆN be born be doniya aværdæn to world AVÆRDÆN give birth to yad gereftæn memory GEREFTÆN learn yad dadæn memory DADÆN teach These forms are opaque, in that their meanings cannot be directly inferred from its constituents. Despite their opaque semantics, these LVCs partake in the productive

158

alternations presented in chapter 4. This fact shows that some part of the meaning of the LVs is present to allow for such an alternation.

5.2.4 Relations between Constructions

A different type of link exists between LVCs of different LVs. Some of the productive alternations presented in Chapter 4 display a subpart relation or link (Goldberg, 1995, p79). In these cases, where there are two alternating constructions, the structure of one construction is a subpart of another. This type of relation works for large sets of islands, as in the relation between zædæn and xordæn in the alternations where a whole group of related islands alternates with the same group in the other’s semantic space (see section 4.2.2). The LVCs formed with xordæn often share the same structure of the analogous LVCs formed with zædæn, except the former do not have explicit agents.

t∫aqu zædæn knife ZÆDÆN injure with a knife t∫aqu xordæn knife XORDÆN get wounded by a knife gul zædæn trick ZÆDÆN trick gul xordæn trick XORDÆN be tricked gereh zædæn knot ZÆDÆN tie a knot gereh xordæn knot XORDÆN get tied into a knot bæxiye zædæn stitch ZÆDÆN stitch bæxiye xordæn stitch XORDÆN get stitches However, more nuanced alternations cannot be captured by a general link. The links responsible for those alternations contain idiosyncratic information and must also be learned. But, it is important to consider that these links are not learned in isolation. Motivations for different constructions display patterns that may also be learned, which eases the acquisition of similar motivations. For example, when directed motion is known to be a crucial parameter for the alternations between gereftæn and dadæn and one alternation is learned, other shared islands will be easier to learn based on the established pattern.

5.3 The Persian Verbal System

5.3.1 What’s the Difference?

People are guided by what they know about linguistic components. This knowledge is a mixture of semantic and syntactic information, as well as real-world knowledge about the elements involved. The structure of Persian LVCs allows us to access at least two components of the verbal notion expressed. It seems almost as if we can observe the skeletal structure of the verbal notion, a clear image of the event or process. This transparency contrasts sharply with other systems that use mostly simple verbs.

Persian speakers do not necessarily have to learn thousands of relations between form and meaning, even if they must store all the LVCs in the lexicon. Even if we consider such massive storage, the patterns inherent to the organization of the system would guide the retrieval and production of grammatical forms more so than in a system with more arbitrary pairings between form and meaning. These patterns also provide a structure from which new forms can be evaluated and disambiguated. It is beyond the scope of this study to compare

159

and apply our findings from the large corpus of Persian LVCs to languages that utilize much fewer LVs. However, I propose some preliminary discussion about the differences between these two types of systems.

Surely, the meaning range of Persian LVCs is much narrower than the meaning range of simple verbs in, say, English or French. In these languages, simple verbs can have a large range of meaning. For example, the verb call can be used to mean talk on the phone, ask for the presence of, say in a loud voice, announce, request (e.g. she was called to the front of the room), dial, to name a few. In Persian, these verbal notions would be expressed through specific LVCs: telefon zædæn (telephone ZÆDÆN), seda zædæn (sound ZÆDÆN), dad zædæn (yell ZÆDÆN), dærxast kærdæn (request KÆRDÆN), ∫omare gereftæn (number GEREFTÆN), respectively. The LVCs which form verbal notions in Persian require a PV specified with an abundance of functional, perceptual, and concrete attributes. For this reason, an LVC can almost never be as general as its English counterpart.

Consider the fact that in English, you can say he buttered his bread to mean he spread butter on his bread. In the former case, a syntactic process transforms the noun into a verb. In Persian, an island exists with zædæn which expresses spreading a viscous material on a surface, of which butter or kære is an example. The verb zædæn does not mean spread. The verbal notion is expressed through the composition (zædæn, its particular PV, and the construction they appear in). New forms pertaining to spreading viscous materials are easily constructed in Persian (e.g. moræba (jam) zædæn, peanut butter zædæn). However, in English, it is impossible to say *he jamed his bread or *he peanut buttered his bread. One must use the verb spread which is more vague than the construction used in Persian since one can spread nearly anything, whereas the construction in Persian pertains to specific types of materials, and is thus more concrete. Furthermore, a word for the general notion of spread does not exist in Persian.

Interestingly, the French word tartiner expresses spreading (jam, butter, cheese, etc) on bread. This represents an arbitrary form to meaning pairing that must be learned which expresses the final result (une tartine). This verb is close to the Persian construction, but tartiner is more general, so one would have to say that il a tartiné son pain avec something (he tartiné his bread with something). Again, this is a general notion that must be combined with another element.32

Going back to the syntactic process that provides he buttered his bread in English, we observe that it is not productive, unlike the analogous Persian construction. This could be because syntactic processes are more difficult to master than semantic processes. Perhaps LVC formation, as in Persian, is easier than learning a syntactic process. Unfortunately, there is currently no acquisition data on Persian that can support or disprove this claim. It isn’t clear if Persian speaking children overgeneralize semantic processes (LVCs) as English speaking children overgeneralize syntactic processes in English. The steps and difficulties of acquisition are not known.

Since there are at least two lexical items in each construction in Persian, there is less polysemy, but more predictability. When two elements occur in a linguistic utterance, a context is made available for the speaker to partially disambiguate the constituents.

32 Notice that in French, one can say Ils sont en train de tartiner leur pain. Whereas, in English and in Persian, one would have to give a general word like something to express the same notion: They are spreading something on their bread. Or, in Persian: be nunha∫an ye t∫izi mizænæn. (to their bread something hitting).

160

In a system like the one in Persian, verbal forms are more predictable, but phonetically longer than those in other languages, such as English or French. English and French speakers, on the other hand deal with more verbal polysemy than Persian speakers, but they have phonetically shorter (often half the size) lexemes to memorize.

Persian speakers tend to gravitate toward LVCs versus simple verbs. Perhaps the decrease in ambiguity influences this tendency. For example, malidan means to knead, graze, spread, smear, and other related verbs. It’s nominal derivative, male∫ occurs with the LV dadæn to express to give a massage and nothing else. Furthermore, any type of physical manipulation occurring with dadæn (e.g. squeeze, tickle) specifically means to carry out that manipulation. The possible polysemy of lexemes like male∫ is resolved due to the existence of the construction with dadæn.

The fact that Persian LVCs are less polysemous than their English or French counterparts does not mean that LVCs cannot themselves serve as bases for metaphors. As we observed, Persian LVCs often have a certain imagery associated with them, usually stemming from the dynamic, spatial, and temporal characteristics of the LV. Certain instances of these constructions can be used abstractly, further extending the imagery. Consider the following forms:

zærbe xordæn shock XORDÆN be shocked => undergo a blow (psych.) gel gereftæn mud GEREFTÆN cover with mud => close indefinitely ju∫ aværdæn boil AVÆRDÆN boil => get fed up and angry atæ∫ zædæn fire ZÆDÆN set on fire => ruin, waste The occurrence of these forms in metaphorical sense is similar to the action of t∫une zædæn chin hit ‘to bargain,’ described in Karimi (1996, p 24, presented above). The image provoked by the LVC serves as a basis for a metaphorical mapping. This shows that Persian LVCs, though much narrower in meaning than their English or French counterparts, can indeed be used to express abstract verbal notions as do English and French verbs (e.g. break her heart, carry a burden, seal my lips). The transparency does not deter from this type of extension, though it might not be as commonplace as in English.

Finally, one might be tempted to say that Persian LVs represent verbal primitives. However, they cannot be considered as such. Each LV consists of several distinct features (e.g. temporal structure, dynamics, directed motion), which could each be described as a type of primitive. LVs provide a basis of features that can be exploited and merged with other elements in the language to produce new notions. As we saw in Chapter 3, different elements of the LV are exploited in different contexts, or constructions. If these LVs where in fact primitives expressing a single primitive feature, they would probably not allow such a rich variety of LVCs. We saw in Chapter 4, that even the most generic LV, kærdæn, doesn’t have a completely bleached meaning. If this LV simply expressed causation, then one would expect there not to be any unacceptable LVCs which involve causation. This proves not to be the case with internal states. As in English, one cannot say *he fiestied me to mean he made me fiesty (*atæ∫i-æm kærd). The unacceptability of the English form could be due to the morphological opacity of the expression, an issue we would expect to be resolved in a more transparent language like Persian. This fact supports the view that kærdæn, or any other LV, can’t be a primitive.

161

5.3.2 Why So Many LVs?

A crucial issue that arises when faced with a system of LVCs is the possible reasons why there are more than a dozen LVs. If the most important attribute to be expressed is transitivity or agentivity, there would be no reason to have more than two LVs in the system. Part of the answer to this question was presented in Chapter 4. We observed that though aspectual and grammatical attributes of an LV play a large role in contrasting LVCs with the same PV and a different LV, this explanation is not sufficient.

One could also argue that if everything is based on constructions, there needn’t be more than two or even just one LV in the system. All semantic information could be included in the construction and not depend on the contribution of the LV. Apart from the fact that a multitude of LVs allow for the expression of more nuanced meanings (as explained below), the knowledge of the core meaning of the LV provides an extra element to disambiguate the semantics. Once the LV is known, it drastically lowers the number of possible constructions the given LVC could belong to. Even if the LVC has never been heard before, when it occurs with a particular LV, the hearer can immediately call upon knowledge of the core meaning to begin to decipher the new form.

But, more importantly, and clearly visible in the data, there is an apparent need to express richer meanings, or more nuanced meanings. These nuances are encoded in the LV along with the construction it occurs in. The meaning of the LV stays in the background, though it isn’t compositionally present in its complete form with all its semantic dimensions. In other words, traces of the LV dynamics often emerge in LVCs. The image behind the LV remains, including transfer direction, emission, impact, and whether something is brought into possession or whether it is imposed. Thus, we cannot have only two LVs. If LVs where in fact bleached, we would expect there to be only two generic LVs, like kærdæn and ∫odæn described in Chapter 4. Consider the following forms:

dæst kærdæn hand KÆRDÆN put hand inside (e.g. glove) dæst dadæn hand DADÆN offer one’s hand (handshake) dæst zædæn hand ZÆDÆN touch otu kærdæn iron KÆRDÆN iron (usually carefully) otu zædæn iron ZÆDÆN iron (usually quickly) otu xordæn iron XORDÆN be ironed (focus: goal of action) otu ∫odæn iron ∫ODÆN be ironed (focus: general action) In the first set, the PV dæst ‘hand’ occurs with three agentive LVs (core meanings in English: do, give, hit). Each instance of the PV has a different meaning that depends on the core meaning of the LV it is working with, and specifically, a certain attribute of the LV. Different dynamic parameters are highlighted in each case above.

In the second set, the differences are more subtle. The two first forms are both agentive and express highly similar meanings. The contrast lies that in the first case, the LVC expresses a general action and in the second, there is a underlying implication that the action occurs quickly or is trivial (similar to the English give it a quick iron). These attributes emerge from the core meaning of zædæn.

162

In the second pair, the form with ∫odæn expresses a general action, for instance of a piece of clothing having been ironed. With xordæn, the LVC highlights having undergone the action, it can be used for a piece of clothing which hasn’t been fully ironed.

The stability of a multi-LV system lies in both the need to express more nuanced and unambiguous meanings and also the tendency to store less completely idiosyncratic meaning, in favor of having some predictability (from the core meaning of the LV). The meaning is not assigned by the whole expression or construction, but emerges as a result of the interaction between the constituents and the whole. As expressed by Victorri and Fuchs (1996), the meaning of an expression comes from the apport propre or contribution of its constituents and les régles or the rules that dictate how this contribution interacts with other elements in the context. In the case of Persian LVCs, the context is primarily its own structure. The PV and the LV contribute some meaning as does the construction they figure in. Hence, the meaning of an utterance can’t be reduced to the contribution of its parts nor can it be defined independantly (Fuchs and Victorri, 1996, p 38). As such, the meaning emerges from the two-way interaction between the part and the whole.

5.3.3 Productivity

It may be the case that Persian speakers actually store most LVCs in the lexicon. However, as we saw in the last subsection, even if this were the case, it would still not be as much of a burden to learn all form to meaning pairings. This is due to the non-zero contribution of the LV. Though the meaning is not compositionally present, knowledge of the core meaning alleviates the burden of learning a completely new meaning with each LVC. Also, a system like the one in Persian might make it possible to store more elements in long term memory by economizing the amount of totally new things. The LV isn’t a new form and neither is the PV. The only new thing is the meaning assigned to the whole (usually derived from a construction that generalizes over several LVCs). Many of the LVCs don’t occur in one of the islands presented in this study (labelled unclassified in the data available in the appendix). These must be learned on a case by case basis.

However, storing each LVC as a separate entity does not explain how new forms emerge and are understood. The productive nature of the system further supports the existence of constructions. One pressing issue is why certain LVCs are productive, and some aren’t. I am not ready to make any solid claims about this fundamental issue all linguists are faced with. However, I propose a few conjectures for the case of Persian LVCs.

One parameter that plays a role in productivity is frequency. The more frequent different instances of a construction are expressed (i.e. the more different tokens), the easier it is to make new forms based on existing ones. This could be partly because the more tokens a construction has, the more well defined the crucial attributes of the PV become. Conversely, if a construction includes a clear perceptual or functional definition of the PV required, resulting LVCs come about naturally. Some LVCs, like hærf zædæn (to talk) are very common, but do not come from a productive island. In these cases, it is more probable that each instance of the LVC is stored in the lexicon.

Another parameter that could play a role in productivity is imagibility. This parameter involves the link between the core meaning of the LV and PV and the image evoked by the action expressed by the construction. If the action is imagible (no matter if it is concrete or abstract), the resulting LVCs tend to be more productive.

163

Language does not work solely with generalizations. Speakers are sensitive to patterns, but natural language is full of exceptions. The framework adopted for this study provides the tools necessary to capture a large amount of the generalizations in the Persian verbal system, while allowing for exceptions to these generalizations to co-exist.

5.4 A Neurolinguistic Study

Neurolinguistic data on aphasic productions has revealed error patterns relevant the nature of the Persian verbal system. Nilipour & Raghibdoust (2001) report a patient who replaces the PV or the LV of certain LVCs with an empty pronoun (e.g. he substitutes the word ‘thing’ as PV, which might indicate a lack of access to the explicit PV). Another patient produces an LVC of similar meaning in place of a target simple verb (e.g. pare ∫od (rip become ‘ripped’) instead of ∫ekæst (‘broke’)). A third interesting case reported in this study is a patient who substitutes the LV with a wrong LV. A sample of his utterances appears below in (c) and (d). This pattern of substitution could reveal parts of the process of accessing the correct verb form.

(c) setta bæt∫e ra negah mi-dar-ænd. three kid ACC look PROG-have.PRS-3PL. They are looking at the three kids. (d) bæd ye dæfe ∫oru gereft-æm. then one time beginning get.PST-1SG. Then, I suddenly began. In (c), the aphasic has substituted the LV da∫tæn for the correct LV kærdæn. In (d), he substitutes the LV gereftæn for the correct LV kærdæn. This might be interesting because it might indicate productivity in the face of lexical loss. For the current study, it would be interesting to look at the features these LVs have in common to discern possible patterns. The above case is particularly interesting because the LV kærdæn is usually considered as a default LV, most neutral in meaning. Here, the patient substitutes LV with more richer semantics for the more generic one. Also, (d) makes sense in the light of another lexical item, æz sær gereftæn ‘from head obtain / to begin’ which might be triggered. Certain errors, like this one, could be due to the existence of the islands in the semantic space. In this case, the island serves as the basis for expressing taking something from a certain position, and the corresponding forms prevail (as with (d)) unless blocked by other knowledge (e.g. more appropriate lexical items, fixed expressions). The patient of this study might be resorting to more automatic responses according to the concept he wishes to express, exploiting his knowledge of islands when other information is not available due to the breakdown of his linguistic abilities.

Collecting a corpus of verbal errors will be the first step towards a neuropsychological analysis that might support or refute parts of the descriptive analysis provided in this thesis. It would therefore be necessary to manipulate current linguistic tests to concentrate on verbal utterances, and to provoke verbal errors from patients.

5.5 Chapter Review

In this chapter, I revisited the idea of constructions and discussed the consequences of constructions in the Persian verbal system. It is argued that most Persian LVCs are semi-

164

idiomatic, or idiomatically combining expressions. Though each construction has an idiosyncratic meaning assigned to it that must be learned, it provides a basis for productivity in the system. Furthermore, the idiosyncratic meaning is always motivated from the core meaning of the LV and semantic contribution of the PV. Though the meaning of the LV is not compositionally present in these constructions, the existence of the meaning in the background allows for partial predictability of the type of action expressed. The core meaning motivates the meaning of the LVCs it is included in through metaphorical, figural, and instance relations. Persian LVCs occur somewhere between the grammar and lexicon, and certain phenomena (e.g. multiple constructions sharing a peculiar syntactic structure) signal the existence of a hierarchy of constructions. LVCs occurring in the islands presented in this study inherit certain specifications from LVCs higher in the system.

The consequences of such a system are also discussed compared to other languages with less LVCs. It is argued that LVCs are less polysemous and thus less ambiguous than simple verbs in Persian and in other languages. Even if a Persian speaker must learn and store all the LVCs in the lexicon individually, the meaning contributed by the PV and the core of the LV allow for a some predictability of the meaning of the whole which is more narrow in range than most simple verbs. Also, Persian LVs cannot be considered verbal primitives, since each LV itself consists of a combination of spatio-temporal, dynamic, and aspectual attributes. The existence of many LVs, as opposed to a couple, is plausibly because of the need to express more nuanced meanings, and also because of the predictability provided by the core meaning of the LV. And finally, certain LVCs are more productive than others. I claim that frequency of different related LVCs and the imagibility of the action play a big role in determining the productivity of a construction.

165

6 THESIS SUMMARY

The analysis of LVCs provided in this thesis is based on the assumption that grammar consists of constructions and is not isolated from other cognitive processes. Like lexical entries, constructions are form to meaning pairs that must be learned. Unlike lexical entries, they specify for grammatical structure. Thus, the existence of constructions in grammar blurs the distinction between the grammar and the lexicon. Construction Grammar allows us to concisely capture linguistic generalizations and other fundamental properties of natural language, such as polysemy, compositionality, and productivity. Furthermore, it allows the use of real-world knowledge in language use, namely, functional, social, or perceptual attributes of the elements involved in an utterance.

Our goal has been to explore the semantic space of LVs in Persian. I provide a working definition of an LVC as any construction that includes a LV, a semantically weak verb, which combines with a non-verbal element to produce a novel verbal notion that is neither fully compositional nor completely idiomatic (a frozen expression). It was shown that LVs differ from full verbs and from auxiliaries. Each LV in Persian combines with a large variety of PVs to form different LVCs with a variety of meanings. It is counter-intuitive to propose that each LVC is stored as an entity in the lexicon. Such a proposition would fail to capture obvious patterns in the organization of the system.

The focus of this thesis has been to describe the mechanisms through which Persian speakers are able to disambiguate LVCs without having to posit redundant entries in the lexicon (i.e. storing hundreds of entries with the same LV and different PVs). These mechanisms must include adequate functions to account for productivity, polysemy, and non-compositional semantics.

I argue that the LVCs retain some elements of the core meaning of their respective LVs. The core meaning of an LV is the meaning uttered by a Persian speaker as the definition of the LV; this meaning is usually the meaning of the LV as used in non-LVC contexts (as a full verb). I then argue that the semantic space of the Persian verbal system consists of groups of islands. Islands are formed when a cluster of LVCs with a given LV express highly similar meanings. The LVCs that gather into islands result from a general construction which include an explicit LV, a type of PV defined by its functional, social, and perceptual attributes, and the meaning of the whole construction.

These constructions avoid positing highly idiosyncratic secondary meanings for LVs, such as “hit and pierce with a weapon like + PV” or “provide a lesson or advice in the form of + PV.” Constructions provide a blueprint for different verbal notions and include rich semantic and syntactic information.

In Chapter 3, I provide descriptions of the constructions allotted to each island of LVCs in four common LVs: zædæn, xordæn, gereftæn, and dadæn. I observed that, due to the variety of meanings expressed by each LV, it is impossible to extract overarching traits present for any of the LVs. Nevertheless, the organization of the space, and the existence of these constructions allows us to have a certain level of productivity and predictability of meaning. The non-compositional semantics is also taken into account since the meaning is inherent to the construction and does not result from the direct combination of the constituents. LVCs also capture polysemy in the sense that the islands are all linked to the core meaning of the LV, and constructions explicitly specify the context (i.e. the type of PV) in which a certain

166

meaning is expressed. Importantly, the meaning of the construction is not allotted to the LV itself, but to the construction in which it appears.

Having explored the mechanisms behind individual LVs, I show the existence of a different type of productivity in Chapter 4. An island, or groups of islands, in the space of an LV can alternate productively with islands in the space of other LVs. By alternation, I refer to types of PVs which can occur with different LVs to produce different, but related meanings. Since constructions provide a basis for productivity, the alternations in the system are also productive.

First, I explore alternations between zædæn and xordæn which involve groups of islands and which can be captured by general semantic attributes, such as transitivity or agentivity. Then, I show that there are single islands shared between these two and other LVs. The alternating constructions that result involve more nuanced parameters, often including dynamic and perceptual attributes of the actions. Since islands can be shared, it may host several constructions, depending on the number of LVs used.

The full set of LVCs of a given LV can never alternate with a full set of LVCs of another LV. As seen in Chapter 3, the variety of meanings expressed by an LV is too large to occur in parallel with the variety of meanings expressed by another LV. These two chapters show systematicity within the space of each LV, as well as systematic interactions between these spaces.

Finally, in Chapter 5, I return to the idea of constructions and discuss the system as a whole and the consequences of such a system. I highlight the fact that most Persian LVCs lie on a continuum between the two extremes of transparency, namely fully transparent and idiomatic. They can be considered idiomatically combining expressions and are most concisely captured within the framework of Construction Grammar.

Since the meaning of Persian LVCs are partially a combination of the LV and the PV in a non-compositional way, their constructions are semantically motivated. Motivations can be aspectual, and can include cognitive parameters, such as spatio-temporal configuration. LVCs are motivated by the core meaning of their respective LV. Alternations are also motivated, some by a simple generalization and others by more nuanced parameters.

I also describe the differences between a system like in Persian, and that of languages such as English and French. In Persian, there is less polysemy and more predictability, though the system renders it difficult to express very general actions. Persian LVs cannot be primitives since each LV is comprised of a set of more primitive features. The reason for a dozen LVs, and not just two in the system could be because of the need to express richer meanings with least amount of ambiguity. Certain LVCs are more productive than others. Factors effecting this could be frequency and imagibility of the action expressed.

This thesis illuminates an area of semantic investigation hereto overlooked for the Persian verbal system. It has been shown that the productive capacities of Persian speakers actually lie in the knowledge and development of a highly structured semantic space. This space contains many nodes that serve as attractors to certain types of verbal notions. My analysis shows that Persian LVCs contain semantic, lexical, and syntactic information, and that differentiation between the grammar and the lexicon is superfluous. This is the first study to investigate these properties of LVCs in Persian using and extending several cognitive linguistic theories, including Construction Grammar.

167

REFERENCES

Adams, V. 1973. An introduction to English Word-Formation. London: Longman.

Afshar, G.; Hakami, N; Hakami, N. 2002. Contemporary Persian Dictionary. Farhang Moaser Publishers : Tehran.

Bashiri, Iraj. 1981. Persian Syntax. Minneapolis: Burgess Publishing Company.

Barjesteh, Darab. 1998. Morphology, Syntax, and Semantics of Persian Compound Verbs: A Lexical Approach. PhD Thesis, University of Illinois.

Bauer, L. 1983. English Word-Formation. Cambridge: Cambidge University Press.

Beeman, William O. 1982. Culture, performance and communication in Iran Tokyo, Japan : Institute for the Study of Languages and Cultures of Asia & Africa.

Brugman, Claudia. 1981. The Story of Over. MA thesis, University of California at Berkeley.

Brugman, Claudia. 2001. Light verbs and polysemy. Language Sciences 23: 551-578

Butt, Miriam and Geuder, W. Light Verbs in Urdu and Grammaticalization. In R. Eckardt, K. von Heusinger and C. Schwarze (eds.) Words in Time: Diachronic Semantics from Different Points of View. Berlin: Mouton de Gruyter, 295–349.

Butt, Miriam. 1997. Complex Predicates in Urdu. In A. Alsina, J. Bresnan, and P. Sells (eds.), Complex Predicates, 107–149, Stanford, CA: CSLI Publications.

Cadiot, Pierre ; Lebas, Frank ; Visetti, Y.M. 2005. The semantics of the motion verbs: Action, space, and qualia. Space in Languages: Linguistic Systems and Cognitive Categories (ed. Hickmann, Maya and Robert, Stéphane). John Benjamins Publishing: Amsterdam, Netherlands.

Cadiot, Pierre and Visetti, Y.M. 2001. Pour une théorie des formes sémantiques - motifs, profils, thèmes. Paris, Presses universitaires de France.

Cadiot, Pierre. 1999. Les sens de jouer : esquisse d’une approche par le biais des attaches prépositionnelles. Recherches en linguistique et psychologie cognitive n°11. Reims: Presses Universitaires de Reims.

Cattell, Ray. 1984. Composite predicates in English. Orlando, FL: Academic Press Inc.

Chomsky, Noam. 1995. The Minimalist Program. Cambridge, MA: MIT Press.

Chomsky, Noam. 1965. Aspects of the Theory of Syntax. Cambridge, MA: MIT Press.

Chomsky, Noam. 1975. Reflections on Language. New York: Pantheon

Clark, Eve V. 1996. Early Verbs, Event Types and Inflections.Vol. 9: Children's Language. Mahwah, N. J.: Lawrence Erlbaum Associates.

168

Comrie, Bernard. 1976. Aspect. Cambridge, UK: Cambridge University Press

Dabir-Moghaddam, Mohammad. 1997. Compound Verbs in Persian. Studies in the Linguistic Sciences 27 (2): 25-59. University of Illinois at Urbana-Champaign.

Dehaene, Stanislas. 1997. La Bosse des Maths. Paris: Odile Jacob.

Dowty, David. 1991. Thematic Proto-Roles and Argument Selection. Language 67(3): 547-619.

Family, Neiloufar. 2001. The Cognitive Salience of Base Verbs in Persian Complex Verbs. DEA Thesis. Paris: Ecole des Hautes Etudes en Sciences Sociales.

Family, Neiloufar. 1998. Phonology Meets Morphology: An Interface Account of Persian Stress. Proceedings of Conference on Cognitive Science: Tehran, Iran.

Fauconnier, Gilles. 1997. Mappings in Thought and Language. New York and Cambridge: Cambridge University Press.

Fillmore, Charles J. & Paul Kay 1996. Construction Grammar. Manuscript, University of California at Berkeley Department of linguistics.

Fodor, Jerry. 2001. Language, Thought and Compositionality. Mind and Language 16(1): p 1-15.

Folli, Raffaella; Harley, Heidi; & Karimi, Simin. 2004. Determinants of Event Type in Persian Complex Predicates. Cambridge Occasional Papers in Linguistics 1:101-125. Cambridge, UK.

Frege, G. 1892. Uber Sinn und Bedeutung, Zeitschrift Philosophie und philosophische Kritik 100 (1892) 25-50. Translated as “On Sinn and Bedeutung” in Beaney 1997: 151-171 (The Frege Reader, Blackwell Publishers, Oxford, 1997).

Fuchs, Catherine and Victorri, Bernard. 1996. La polysémie : construction dynamique du sens. Paris : Hermès.

Ghadakpour, Laleh. 2002. Le système conceptuel, à l’interface entre le langage, le raisonnement, et l’espace qualitatif : vers un modèle de représentations éphémères. PhD thesis. Paris : Ecole Poytechnique.

Ghomeishi, Jila. 1996. Projection and Inflection: A Study of Persian Phrase Structure. PhD thesis, University of Toronto.

Goldberg, Adele. 1995. Constructions: A Construction Grammar Approach to Argument Structure. Chicago, IL: Chicago University Press.

Goldberg, Adele E.1996. Words by default: Optimizing Constraints and the Persian Complex Predicate. In Berkeley Linguistic Society 22.

Goldberg, Adele E. 2004. Words by Default: the Persian Complex Predicate Construction. Elaine Francis and Laura Michaelis (eds.) Linguistic Mismatches. CSLI Publications.

169

Goldberg & Jackendoff, 2004. The English Resultative as a Family of Constructions. Language 80 532-568.

Grimshaw, Jane and Mester, Armin. 1988. Light Verbs and Theta-Theory. Linguistic Inuiry 19: 205-232.

Haim, S. 1995. The Larger Persian English Dictionary. Tehran: Farhang Masser Publishers.

Haji-Abdolhosseini, Mohammad. 2000. Event Types in the Generative Lexicon: Implications for Persian Compound Verbs. Toronto Working Papers in Linguistics, Proceedings of NLS 2000.

Hale, Kenneth and Keyser, Samuel Jay. 1993. On Argument Structure and the Lexical Expression of Syntactic Relations. In K. Hale and S. J. Keyser, eds., The View from Building 20 : Essays in Linguistics in Honor of Sylvain Bromberger, p. 53-110. Cambridge, MA : MIT Press.

Hook, Peter Edwin. 1991. The Emergence of Perfective Aspect in Indo-Aryan Languages. In Approaches to Grammaticalization, ed. Elizabeth Traugott and B. Heine. 59-89. Amsterdam : Benjamins.

Hopper, Paul J. and Elizabeth C. Traugott. 1993. Grammaticalization. Cambridge: Cambridge University Press.

Ibrahim, Amr Helmy. 2002. Les verbes supports en arabe. Bulletin de la Société de linguistique de Paris, t. XCVII (2002), fasc. 1, p. 315-352.

Jespersen, Otto. 1965. A Modern English Grammar. Copenhagen : G. Allen and Unwin Ltd.

Karimi, Simin and Mohammad, Jan. 1992. Light Verbs are Taking Over: Complex Verbs in Persian. In Proceedings of WECOL 5, pages 195-212.

Karimi, Simin. 1997. Persian complex verbs: Idiomatic or Compositional. Lexicology 3(2): 273-318.

Karimi-Doostan, Mohammad-Reza. 1997. Light Verb Constructions in Persian. PhD Dissertation, University of Essex.

Keshani, Khosro. 1993. Farhange Farsiye Zansu. Tehran, Iran: Markaze Nashre Daneshgahi.

Khanlari, Parviz. 1986. Tarix-e zaban-e Farsi (A History of Persian). Tehran, Iran: Nashre No Publishing Company.

Lakoff, George. 1987. Women, Fire and Dangerous Things : What Categories Reveal About the Mind. Chicago: University of Chicago Press.

Lambton, A.K.S. 1961. Persian grammer. Cambridge, UK : University of Cambridge Press.

Langacker, Ronald W. 1987. Foundations of Cognitive Grammar. Volumes I & II. Stanford, California: Stanford University Press Larson, Robert. 1988. On the double object construction. Linguistic Inquiry 19:35-391.

Lazard, Gilbert. 1992. Grammaire du Persan Contemporain. California: Mazda Publishers

170

Lebas, Frank and Cadiot, Pierre. 2003. Monter et la constitution extrinseque du referent. Languages 130 : 9-30.

Levin, Beth. 1993. English Verb Classes and Alternations: A Preliminary Investigation. Chicago, IL: Chicago University Press.

Mahootian, Shahrzad, 1997. Persian. Routledge.

Megerdoomian, Karine. 2001. Event Structure and Complex Predicates in Persian. Canadian Journal of Linguistics 46 (1/2): 97-125.

Megerdoomian, Karine. 2002. Beyond Words and Phrases: A Unified Theory of Predicate Composition. PhD Dissertaion, University of Southern California.

Megerdoomian, Karine. 2004. A semantic template for light verb constructions. In Proceedings of First Workshop on Persian Language and Computers. Tehran University, Iran.

Megerdoomian, Karine. 2004. Finite-state morphological analysis of Persian In Proceedings of the Workshop on Computational Approaches to Arabic Script-based Languages. Coling 2004, University of Geneva.

Moin, M. 1997. (annotations and additions in: ) Borhan-e Qate (Dictionary of the Persian Language). Tehran : Amir Kabir.

Mohanan, T. 1997. Multidimensionality of Representation: NV Complex predicates in Hindi. In Alsina, A. et.al. (eds). Complex Predicates, 431-471. Stanford, CA: CSLI Publications.

Montague, R. 1970. Universal grammar. In R. Thomason (ed.) Formal Philosophy: Selected papers of Richard Montague. Yale University Press, New Haven and London.

Nilipour, R. & S. Raghibdoust. 2001. Manifestations of Aphasia in Persian. Journal of Neurolinguistics. 14 (2001) 209 - 230.

Nunberg, G., Sag, I., and Wasow, T. 1994. Idioms. Language 70: 3: 491-538.

Pinker, Steven. 1999. Words and Rules : The Ingredients of Language. Harper Perennial

Plag, Ingo. 2004. Productivity. Chapter 23 in Handbook of English Linguistics, ed. by Bas Aarts and April McMahon. Oxford, UK : Blackwell.

Prince, Alan and Paul Smolensky. 1993. Optimality Theory: Constraint interaction in generative grammar. Rutgers University Center for Cognitive Science Technical Report 2.

Rastorgueva, V. S. 1964. A Short Sketch of the Grammar of Persian. Indiana University, Indiana.

Rosch, E. (1988). Principles of Categorization. In A. Collins & E. Smith (eds), Readings in Cognitive Science: A perspective from psychology and artificial intelligence (p. 312-322), San Mateo, CA: Morgan Kaufmann, Inc.

171

Samvelian, Pollet. 2003. Le sujet, l’objet et l’inaccusativité dans les prédicats complexes nom-verbe en persan. Cahiers de Linguistique de l’INALCO: Paris, France.

Samvelian, Pollet. 2001. Les complèments d’objet “nus” en persan: Incorporation ou non? Bulletin de la Société de Linguistique de Paris.

de Saussure, F. (1915). Cours de Linguistique Générale. Paris : Payot.

Sharifi, Hassan. 1975. The Semantic Structure of the Persian Verb /gereftæn/ “to take.” Mid-America Linguistics Conference: 469-476.

Sharifi, Hassan. 1975. Persian Verbs: A Chafean Analysis. Mid-America Linguistics Conference: 459-461.

Stevenson; Fazly, A; and North R. 2004. Statistical Measures of the Semi-Productivity of Light Verb Constructions. In 2nd ACL Workshop on Multiword Expressions: Integrating Processing, pages 1-8.

Sweetser, Eve, 1990. From Etymology to Pragmatics. Cambridge: Cambridge University.

Tabaian, Hessam. 1979. Persian Compound Verbs. Lingua 47: 189-208.

Talmy, Leonard. 1988. Force Dynamics in Language and Cognition. In Cognitive Science, 12, 1, 49-100.

Talmy, Leonard. 2000 Towards a Cognitive Semantics: Volumes I & II. Cambridge, MA: MIT Press.

Vahedi-Langrudi, Mohammad-Mehdi. 1996. The Syntax, Semantics and Argument Structure of Complex Predicates in Modern Farsi. PhD Dissertation, University of Ottawa.

Viberg, Ake. 2002. The Polysemy of Swedish ge « give » from a Crosslinguistic Perspective. In Anna Braasch and Claus Povlsen (eds.). Proceedings of the Tenth Euralex International Congress. Vol. II. CST-Center for Progteknologi. Copenhagen University, Copenhagen. p. 669-682

Wierzbicka, A. 1982. Why can you have a drink when you can't *have an eat? Language 58:753-799.

Windfuhr, G. L. (1987). Persian. In Comrie, B., editor, The World’s Major Languages, pages 523–546. Oxford University Press, Oxford.

Wittgenstein, L. 1953. Philosophical investigations. (trans: Anscombe, G.E.M.). Oxford: Basil Blackwell.

173

APPENDIX

The LV analyses in this thesis are based on the following compiled data sets. This corpus includes forms compiled from many sources; namely, all the works pertaining to Persian presented in this thesis, as well as several dictionaries (Haim, 1995; Afshar, Hakami, Hakami, 2002; Keshani, 1993).

The diagram of the semantic space of the LV is provided followed by a list of example LVCs. The LVC data for each LV appear in groups, separated roughly into islands. The islands of the first six LVs are discussed in Chapters 3 and 4. For the eight other LVs shown here, I’ve divided the data into islands without further comments.

The reader will notice that some of the LVCs placed in certain islands don’t always correspond. The general semantics of the PV might differ slightly from the rest of the PVs in the groups, but the resulting meaning is usually very close to the others. This could signal that a particular LVC was inspired by the other members of the island. The goal of this classification is to get a global idea of the organization of the space.

Also, unlike the more rigorous analysis provided for the LVs in Chapters 3 and 4, the islands formed by the data here are not so much the result of similar PVs, but similar resulting meanings. Further analysis needs to be done to give a detailed and more precise description of the required PV.

Persian speakers will notice that the translations given for the LVCs are sometimes rough and unprecise, sometimes resembling notes rather than definitions. The purpose of this study is not to give a dictionary of Persian LVCs, but rather to provide the linguistic community with the general behavior displayed by these types of constructions.

han

d he

ld

coh

eren

t sp

eech

ZÆ

DÆ

N[to

hit]

rota

tion

rhyt

hm

ic

auto

m

otio

n

pro

tect

ing

prep

ara-

tion

pos

ition

emitt

ing

aura

l

visu

al

inst

rum

ent

hum

anso

unds

mus

ical

enh

anci

ng

pie

rcin

g an

d tr

ansf

erin

g

appe

ara

nce

text

ure

type

of

mus

ic

jaz

[jazz

]

se z

ærb

i [th

ree

beat

]muz

ik

[mus

ic]

tom

bæk

[tom

bak]

sænt

ur

[san

tur]

gita

r [g

uita

r]

bærq

[s

hine

]

jæræ

qe

[spa

rk]

su s

u [li

ght]

voca

l

exp

ress

ive

soun

ds

be!k

æn

[sna

p]

dæst

[c

lap]

kæf

[app

laus

e]

anno

ying

u

ttera

nce

hærf

[s

peec

h]

gæp

[wor

d/so

und]

laf

[vau

nt]

veng

[infa

n'ts

cry

]

qor

[mum

ble]

emot

iona

lno

ises

zar

[sob

]

qæhq

æh

[laug

hter

]

dad

[yel

l]

jiq [s

hout

]

ærb

ade

[yel

l]

nois

e

jez

[siz

zle]

liqui

d

tast

e!a

ne

[bru

sh]

mes

vak

[toot

hbru

sh]

lif [l

oufa

]

kære

[b

utte

r]

roqæ

n [o

il]

losi

on

[lotio

n]

næm

æk

[sal

t]

gola

b[r

ose-

wat

er]

ab

[wat

er]

ætr

[p

erfu

me]

dire

ct

stea

ling

trim

min

g

topo

logy

repa

iring

dam

agin

g

proj

ect

ile

we

apon

ri!

[bea

rd]

t!æm

æn

[gra

ss]

mu

[hai

r]

gere

h [k

not]

ta

[fold

]

fer

[cur

l]

bæxi

ye

[stit

ch]

peyv

ænd

[g

raft]

lætm

e [d

amag

e]

qat!

[cho

p]

sædæ

me

[dam

age]

tric

kery

bam

bul[c

heat

] gul

[fool

]

hoqe

[tric

k]

tir

[bul

let]

mu!

æk

[mis

sile

]

toop

[c

anon

]

kard

[k

nife

]

!æm

!ir [s

wor

d]

t!aqu

[kni

fe]

jib

[poc

ket]

bank

[b

ank]

gænj

[tr

easu

re]

pit!

[turn

]

qælt

[som

mer

saul

t]

varu

ne

[und

erw

ater

so

mm

ersa

ult]

bal [

win

g]

æsa

[c

ane]

tolo

mbe

[p

ump]

hæ!i!

[h

ash]

tæriy

ak

[opi

um]

ju!

[acn

e/bo

il]

kæpæ

k [m

old/

fu

ngus

]

tæbx

al

[col

d so

ar]

t!ado

r[te

nt]

tæxt

e [b

oard

]

may

e [s

tart

er]

qale

b [m

old]

pærd

e [c

urta

in]

sær

pa

[on

foot

]

zanu

[k

nee]

inje

ctin

g

dru

g

ampo

ol

[sho

t]va

ksæ

n [v

acci

natio

n]

vez

[fizz

le]

xam

e [c

ream

]

quic

k a

ctio

n

crea

tion

mea

sure

-m

ent

typ

e o

f hit

blun

t

zærb

e [p

unch

]

lægæ

d[k

ick]

kotæ

k[b

eatin

g]

t!æko

! [h

amm

er]

potk

[s

ledg

eham

mer

]

surf

ace

pa

[foot

]

dæst

[h

and]

suzæ

n [n

eedl

e]

ænd

aze

[siz

e]

!i!e

[gla

ss]

livan

[c

up]

mæ

vad

[sub

stan

ce]

fuel

ing

benz

in

[gas

olin

e]

gazo

il [d

eise

l]

agita

ted

indi

rect

ædv

ie'

[spi

ce]

affe

ctin

g

hurt

ing

bænd

[c

onne

ctio

n]

phys

ical

at

tack

mod

ifyin

g

AFFECTING: PROTECTINGordu zædæn camp set up camppærde zædæn drapery hang draperyt∫ador zædæn tent set up a tenttækiye zædæn tent setup tents/arrange for religious showtæxte zædæn wood planks cover by nailing wood planks

AFFECTING: APPEARANCE∫ane zædæn comb comb∫oxm zædæn plow plowqæltæk zædæn [steam] roller compress roof/road w/ a heavy rollerboros zædæn brush brushkise zædæn bag (rough fabric) exfoliatelif zædæn washcloth scrub with a loufasænge pa zædæn pomus scrub with a pomussombate zædæn sand paper sandmesvak zædæn toothbrush brush one's teethsohan zædæn file file smoothværdæne zædæn roller roll out dough

AFFECTING: TEXTURE∫ampo zædæn shampoo apply shampooahar zædæn starch apply starch (to clothes)dæva zædæn medicine apply topical medicinekære zædæn butter spread butterlak zædæn lacquer apply lacquer, nail polishlosion zædæn lotion apply lotionmoræba zædæn jam spread butterpomad zædæn gel apply gelræng zædæn color, paint coat s.t. with paintroqæn zædæn oil apply oil, lubricatesabun zædæn soap apply soapsiman zædæn ciment apply cimentsorme zædæn eye makeup apply eye makeupsorxab zædæn blush apply blushtæla zædæn gold cover in gold foilvær jæla zædæn polish polish by applying a compoundvær∫u zædæn white alloy cover with foils of this alloyvaks zædæn wax apply waxxaze zædæn mix of mud/plaster finish a surface with this mixzirlaye zædæn sealant apply a sealer coat of paint

AFFECTING: TASTEædviye zædæn spice add spice to foodnæmak zædæn salt add salt (us. by sprinkling)

ZÆDÆN 177

AFFECTING: LIQUID∫etæk zædæn splash dirty water splash polluted water when walkingab zædæn water sprinkle water, wet, splashætr zædæn perfume put on perfumegolab zædæn rose-water flavor or apply rose-waterqætre zædæn drop apply drops (to the eyes)sæm zædæn poison [insecticide] spray poison on plants, walls

AFFECTING: TRIMMINGbor zædæn cut cutting or shuffling cardschæmæn zædæn lawn mow the lawngærdæn zædæn neck behead, decapitatemu zædæn hair cut hairræg zædæn vein commit suicide, cut a vital veinri∫ zædæn beard shave beard

AFFECTING: SURFACEængo∫t zædæn finger to sign with fingerprintdændan zædæn tooth bite into to test texturedæst zædæn hand to touch (usually without permission)naxon zædæn nail to scratch s.o. or s.t.naxonæk zædæn little nail (pejor.) pick at a dish before servedqa∫oq zædæn spoon spoil an unserved dishsuzæn zædæn needle leave a mark with a needle

AFFECTING: TOPOLOGYfer zædæn curl curl (hair)gereh zædæn knot tie a knotta zædæn fold fold, crease

AFFECTING: DAMAGEasib zædæn injury, damage damagejer zædæn a rip renege, cheatlætme zædæn damage damage (e.g reputation, progress)qat∫ zædæn crack, split, slice split open a fruit, make slices of fruitsædæme zædæn injury hurt, damagezæxm zædæn wound damage, wound

AFFECTING: REPAIRINGbænd zædæn connection attach broken chainbæqælæk zædæn underarm pad sew on underarm padsbæxiye zædæn stiches stich wounds/make temporary stiches kuk zædæn stich stich a seampeyvænd zædæn graft graft plants, organspit∫ zædæn screw to attach/secure by a screwvæsle pine zædæn patch to patch over patch

ZÆDÆN 178

AFFECTING: INDIRECTabele zædæn pox develop pox blistersjavane zædæn sprout germinate, sproutju∫ zædæn boil develop pimples, boil (liquid)læk zædæn discoloration be discolored b/c of damage (fruit)pine zædæn callus develop a calluspoloq poloq zædæn bubbling sound to boil heavily (esp, a thick liquid)t∫ogoz zædæn infected boil develop a superficial infectiontæbxal zædæn cold sore develop a cold soretavæl zædæn blister develop a blistervaruk zædæn wart develop a wartxal zædæn mole, birth mark develop a molezæng zædæn oxidation oxidize (surface)

AFFECTING: TRICKERYbambul zædæn trick trickbelof zædæn bluff bluffgul zædæn trick trickhile zædæn trick trickhoqe zædæn trick cheat, slightkælæk zædæn trick tricknaro zædæn double cross double cross s.o.neyræng zædæn trick trick

AFFECTING: HAND HELD∫æm∫ir zædæn sword strike with a swordgu∫tkub zædæn masher to mash a stewkard zædæn knife stab with a knifemile zædæn rod to poke to open a clogged passageqæddare zædæn sword strike with a swordqæme zædæn dagger cut oneself in religious mouriningsix zædæn skewer prod with a pointed objectsok zædæn skewer prod with a pointed objecttæbær zædæn ax to hit with an axti∫e zædæn pickax to hit with a pickaxzærbæt zædæn stroke of sword strike with a sword

AFFECTING: PROJECTILEgule zædæn bullet shoot a bulletnavæk zædæn arrow shoot an arrowneyze zædæn spear spearni∫tar/nes∫tar zædæn lancet lance a boilsat∫me zædæn pellet shot pelletstir zædæn bullet, arrow hit with arrow or bullettup zædæn canon shot a canon

ZÆDÆN 179

AFFECTING: BLUNTpotk zædæn sledgehammer to hit with a sledgehammert∫æko∫ zædæn hammer hammert∫ub zædæn stick beat with a sticktæbær xun zædæn hardwood cane to use a cane as weapon in wartazyane zædæn whip whip

AFFECTING: TYPE OF HITkotæk zædæn hit beat somethinglægæd zædæn kick kickmo∫t zædæn fist punchordæng zædæn butt kick kick s.o. in the backpænjul zædæn claw scratch with the clawspo∫te pa zædæn back of foot trip s.o.qæfa zædæn nape dope slapsili zædæn slap slapsoqolme zædæn poke with elbow alert s.o. by poking w/ elbowt∫æk zædæn slap slaptæ∫ær zædæn angry words yell in anger and despisetipa zædæn toe kick overzærbe zædæn shock to shock

AFFECTING: UNCLASSIFIEDxærgah zædæn royal tent set up a royal camp∫abixun zædæn night attack surprise∫ax zædæn horn butt, buntbærge zædæn ticket write a ticket (sale, notice,…)dær zædæn door knock at the doordar zædæn tree, gallow execute by hanginggaz zædæn bite bitegol zædæn goal to score a goal in a ball gamemale zædæn trowel smooth out wet plaster, cement, mudnæm zædæn moisture wet, dabblenæzær zædæn view, opinion give the evil eyesine zædæn breast beat the breast in mourningtæmasxor zædæn look down make fun of s.o.tæ'n zædæn admonishment admonishtænz zædæn humor, allusion make fun of s.o.tormoz zædæn break stop quicklyzæban zædæn tongue lick, taste by lickingzæmin zædæn ground push, trip, throw down

ZÆDÆN 180

AUTO-MOTION: RHYTHMICæsa zædæn cane walk slowly w/a cane [with difficulty]bad zædæn wind, air fan, blow ontobad zædæn wind, air pump air into a tire, ballbal zædæn wing fly, flap wingsbil xædæn shovel shoveldæm zædæn breath breathejaru zædæn broom sweeplæng zædæn limp limpmik zædæn suck sucknæfæs næfæs zædæn breath pant for airpa zædæn foot pedal a bicyclepænbe zædæn cotton fluff contton by a vibrating chordpær zædæn feather to take flightpær o bal zædæn feather and wing thrash around in despairpær pær zædæn leaf, feather stuggle in pain, flutter in fearparu zædæn paddle paddlepelk zædæn blink blinkpok zædæn breath, inhale drag in smoke from a pipe/cigaretteqædæm zædæn step strollsæq zædæn roof of mouth masticate a hard candy or dried foodt∫æmt∫e zædæn ladel stir with a ladeltolombe zædæn pump pump water or airzoq zædæn throb throbing (from an injury)zur zædæn energy exert effort

AUTO-MOTION: ROTATIONmo’ællæq zædæn somersault turn a somersaultpit∫ zædæn turn to make a turnpit∫ o tab zædæn twist and curve roll in painpit∫ o xam zædæn twist and bend turn and twist, road, river, dancerpo∫tæk zædæn flip make a somersaultqælt zædæn roll roll in bed or on the floort∫ærx zædæn turn turn and whirlvaru zædæn upside down somersault in watervarune zædæn upside down somersault in water

AUTO-MOTION: POSITIONhælqe zædæn ring form a circle, encircle, knock with hookqamæt zædæn stature, body stand to start prayingsær pa zædæn squat to squat downt∫ombatme zædæn squat sit in a knee hugged positiont∫onbæk zædæn squat sit in a knee hugged positionzanu zædæn knee kneel

ZÆDÆN 181

AUTO-MOTION: UNCLASSIFIEDæqæb zædæn back, rear push back, drive backwardeynæk zædæn glasses put on glassesjelo zædæn front, ahead pass aheadneqab zædæn veil wear a veilpit∫e zædæn lacey face veil veil one's facequte zædæn float float aroundtotoq zædæn veil wear a veilvuj zædæn fidget fidgetvujvuj zædæn fidget fidgetvul zædæn slither move non-stop (e.g. a restless child)vulvul zædæn slither move and mix around quietly

EMITTING: TYPE OF MUSICgam zædæn octave play an octave in orderpænj nobæt zædæn five beat play a five beat rhythempærde zædæn a musical interlude to play a short piece of musicrædif zædæn a musical system play in a musical systemreng zædæn dance beat drum a dance beatse ∫e∫ zædæn 3/6th beat rhythm play a 3/6th rhythmse zærbe zædæn three beat rhythm play in three beatst∫æhar zædæn four play a four beat rhythemt∫æht∫æh zædæn nightingales song sing in a nightingale vibratotæfæ'ol zædæn music metric to play the tæfæ'ol rhythem

EMITTING: MUSICAL INSTRUMENT∫e∫ta zædæn guitar play the guitar∫eypur zædæn horn play a horndæf zædæn daf play the dafdæstæk zædæn small tonbak play the dastakdayere zædæn tambourine play the tambourinedohol zædæn drum play the drum (traditional)guitar zædæn guitar play the guitarkæmant∫e zædæn kamanche play the kamanche (traditional music)kus zædæn kettle drum beat a drum, pretendnaqare zædæn kettle drum beat the kettle drumney zædæn flute play the fluteorg zædæn organ play the organpiano zædæn piano play the pianosaz zædæn instrument play an instrumentt∫æqane zædæn spoon like instr. play the t∫æqanetæbl zædæn drum play the drumtænbur zædæn a string instrument play the tanburtar zædæn a string instrument play the tarxommæk zædæn small daf play a xommækxonbæk zædæn tombæk play the tombækzæng zædæn bell ringzæxme zædæn pick play instrument with a pick

ZÆDÆN 182

EMITTING: NOISEestart zædæn start sound of a starting enginejez zædæn crackling noise crackle

EMITTING: EMOTIONAL NOISESær zædæn donkeys bray yell loudly and out of placeærbade zædæn drunken yell yell in mindless angerbang zædæn bang yelldad zædæn yell yellhævar zædæn shout cry outjar zædæn loud cry cry out news or proclamations, hawkjiq zædæn cry scream, shrieknæ're zædæn clamour clamour, crynale zædæn moan moan from pain or miseryqæhqæhe zædæn laugh cackleqære zædæn roar to roar deeplyqæriv zædæn clamour to make yell out loadseyhe zædæn yell give a loud yelltæqæy'yor zædæn yell from throat yell out load and deepvile zædæn cry cryzar zædæn cry cry

EMITTING: ANNOYING UTTERANCEneq zædæn whine whine incessantlyqor zædæn mumble complaints mumble under ones breath in angert∫ar tækbir zædæn four part prayer recite the …. Prayertopoq zædæn error in speech speech faux pasvæng zædæn cat's desperate cry beg like a hungry catveng zædæn infant's cry beg helplesslyveq zædæn sound of a toy complain incessantlyver zædæn useless talk talk incessantly and uselesslyzænæx zædæn chin useless talk, drivelzer zædæn useless talk talk uselessly

EMITTING: COHERENT SPEECHgæp zædæn word/sound chit chathærf zædæn word/sound talklaf zædæn vaunt brag in speech

EMITTING: EXPRSESSIVE SOUNDSængo∫tæk zædæn snap snapbe∫kæn zædæn finger snap snap fingers with musicdæst zædæn hand applaudfændoq zædæn hazelnut snapkæf zædæn palm clap, applausemærhæba zædæn bravo clap hands in applause

ZÆDÆN 183

EMITTING: VISUAL∫o'le zædæn flame blazing flamesbærq zædæn shine to shine, lightingdud zædæn smoke leak smoke (as in a cracked stove pipe)kur su zædæn small light emit a low energy lightsu zædæn direction, sign, light shimmer of a weak lightsu su zædæn direction, sign, light shimmertænure zædæn chimney spin and rise like flames in a chimneyzæbane zædæn high flame rising flames

EMITTING: UNCLASSIFIEDbuq zædæn horn honk a car horn; blow a hornsæfir zædæn whistle whistlesut zædæn whistle whistlearogh zædæn burp burpemail zædæn email emailfaks zædæn fax send a fax ASAPkenaye zædæn allusion alluderip (colloqial) zædæn misfire an engine misfiringseda zædæn sound call someone (call their name)t∫e∫mæk zædæn wink winktelefon zædæn telephone call someone on the phonetelegraf zædæn telegraph send a telegraph

QUICK ACTION: PREPARATIONbar zædæn cargo load cargoduq zædæn yoghurt drink churn dug to separate butterqaleb zædæn mold to pour into a mold

QUICK ACTION: MEASUREMENTændaze zædæn size size up, quick estimate of sizepeymane zædæn measuring cup measure (lit.)qæd zædæn height measure heighttæxmin zædæn estimation estimatetæxte zædæn measure of fabric quickly measure some fabric

QUICK ACTION: CREATIONkorsi zædæn heated platform prepare a heated sleeping areapey zædæn foundation build a foundationpo∫tband zædæn buttress to brace with a buttresspol zædæn bridge to [create a] bridgetaq zædæn arch, ceiling build an arched ceiling, rooftiræk zædæn squat column provide support w/ a short column

ZÆDÆN 184

QUICK ACTION: UNCLASSIFIEDvær mal zædæn knead breadbuse zædæn kiss kiss lightly, as kissing handdid zædæn view take a suruptitious peekhæds zædæn guess guesslabxænd zædæn smile smilelæb zædæn lip tastemesal zædæn example give and examplepuz xænd zædæn smirk smirksær zædæn boss, superior check on, drop in to check on.t∫ort zædæn snooze snooze on and off

PIERCING: DRUGSbæng zædæn hashish smoke hashishbæst (t:) zædæn a dab of opium smoke a dab of opiummævad zædæn drugs inject drugs

PIERCING: INJECTINGampul zædæn shot to inject a shotni∫ zædæn sting stingvaksæn zædæn vaccin vaccinate by injectionvaksæn zædæn vaccination vaccinate

PIERCING: FUELINGbenzin zædæn gas fill the tank with gassuxt zædæn fuel put fuel into a vehicle

STEALINGbank zædæn bank hold up a bankbe jib zædæn into pocket steal; pocketjib zædæn pocket pick pocketskæf zædæn palm stealqænj zædæn treasure rob a treasureqap zædæn stealth catch to steal with stealth while in motion

UNCLASSIFIED∫elæng tæxte zædæn jump around rough play in mindless joyæbru zædæn eyebrow with a winkastin bala zædæn sleeve up begin purposefully, "roll up sleeves"bala zædæn up pull up (clothing)birun zædæn out rush out - person/ leak - gas or liquiddær ja zædæn in place march in place (soldiers)dæst o pa zædæn hand and foot working hard to stay afloat; flaildamæn zædæn skirt fan with a skirt, irritatedel [ra] zædæn the stomach get tired of; have enough ofgæmane zædæn guess sound outgærma zædæn heat heat strokedgoh zædæn shit mess up, make a big mistakegu∫e zædæn corner speak allusivelygu∫e o kenaye zædæn corner&euphemism innuendo, suggestionsja zædæn place, position substitute deceitfully; retreatkælle zædæn skull butt head, to rise uplas zædæn flirtation flirtloqme zædæn mouthful eat greedilymurt∫e pey zædæn ant foot not a close shavenofuse bad zædæn bad character forsee negative thingspa zædæn foot to cheat someone's account

ZÆDÆN 185

pænje zædæn claw strum a string instrumentpay zædæn foot ignore all rules, traditions, stricturespo∫t zædæn back to turn against s.t. or ideaqat zædæn mixed colloquial: be confusedsær o kæalle zædæn head and skull struggle or engage for a solutionsærbaz zædæn unrully to sherk ones dutysærma zædæn cold frost bittensaqær zædæn wine cup drink defiantlyshælvar bala zædæn pants up roll up the pants to cross the waterssur zædæn advantage hit with a better card in a card gamet∫æk o t∫ane zædæn slap and chin haggle, nickle and dimet∫e∫m zædæn eye give the evil eyetæfre zædæn running high jump to procrastinateto nobæt zædæn into turn cut in linetup zædæn yell (blast) to shout at someonetup o tæ∫ar zædæn yell and angry words to shout in anger and demeanva zædæn open, apart give up and stop, fall apartxo∫k zædæn solid, dry freeze with fearyæx zædæn ice freezeyekke zædæn shock shock s.o.zeh zædæn bow string break agreement, fail a taskaftab zædæn sun damaged by the sun (e.g. eye, fruits)qol zædæn gurgle boil (light liquid, such as water, milk)qolqol zædæn gurgle boil (thick liquid)atæ∫ zædæn fire set afirebæqæl zædæn bossom hug, lift up into the armsbe -[del] zædæn to heart get an idea, a crazy tought, an urgebe - zædæn pretend pretend to be [sleep, deaf, blind, …]be hæm zædæn together collide(intention); mix; break upbe kar zædæn to work utilizebe sær zædæn to head go crazydæstbord zædæn larceny robberyehtemal zadæn possibility guess the probabilityfal zædæn omen tell fortune by guessinggængæl zædæn jest jesthæm zædæn each other mix (e.g. food)hers zædæn glutteny, desire craven desirejæm zædæn total add up numberskart zædæn card (timecard) to stamp ones timecardlæb pær zædæn spillage spill from a overflowing containerlæh læh zædæn pant pantlis zædæn lick lickmix zædæn nail attach, reinforce with nailsmohr zædæn stamp stampnæfæs zædæn breath breathenæ'l zædæn horseshoe shoenæqb zædæn burrow burrow or tunnel throughout zædæn out of bounds kick the ball out of boundspæ∫æng zædæn small ramrod to poke a hole in a wallpæ∫m zædæn fur soften furballs pæhlu zædæn side claim equal standingpærse zædæn poor, vagabound move around aimlesslypæs zædæn back push back, rejectpey zædæn foundation destroy by destroying the founationpeymane zædæn wine cup drink wine (poet.)piyale zædæn cup to drink [wine, liquor]qælæm zædæn pen write, cross outqælæm zædæn type of item list items in a group

ZÆDÆN 186

qælæme zædæn sapling grafts cut saplings and plant themqælb zædæn heart to steal s.o.'s heart (poet.)qæng zædæn press mill stone exract oil by a press millqærbal zædæn sieve to pass through a sieveqeyb zædæn hidden to disappear, vanishqor’e zædæn lottery assignment cast lotsræqæm zædæn digits, writing paint, draw, writerah zædæn way, path holdup on the roadsæla zædæn invitation proclaimsekke zædæn coin stamp coint∫ænbær zædæn torous shape arrange in rings on top of each othertæxt zædæn shoe sole reinforce old shoe's soletaxt zædæn gallop galloptur zædæn net catch in a netxæt zædæn line cross outzærær zædæn loss cause a loss

ZÆDÆN 187

mod

ifyin

g

XO

RD

ÆN

[to

ea

t]

hur

t-in

g

suffering

mo

tion

cond

ition

emot

ion

wea

pon

type

of h

it

touc

hing

usurp

ing

fusi

ngrota

tion

gene

ral

fer

[curl]

gere

[k

not]

pit!

[tw

ist]

qose

[g

rief]

æsæ

f [r

egre

t]æzab

[torm

ent]

zæ

rbe

[shock]

goro

snegi

[hunger]

væ

sle

[connecti

on]

bæ

xiy

e

[sti

ch]

ju!

[weld

]

qæ

lt

[som

mers

ault

]

vul

[slit

her]

pit!

[tw

ist]

sekæ

ndari

[stu

mble

]

tæka

n

[motion]

ta

[fold

]

!æ

laq

[whip

]ti

r [b

ulle

t]

golu

le

[bulle

t]

pro

jec-

tile

hand

hel

d

kard

[k

nife]

sæt!

me

[pelle

t]

lægæ

d

[kic

k]

sili

[s

lap]

kotæ

k

[beat]

nozul

[inte

rest

]

re!ve

[bribe]

sæ

vari

[rid

ing]

tric

kgul

[trick]

færib

[trick]

qæ

bn

[cheat]

inst

rum

ent

t!æ

ko!

[ham

mer]

næ

zæ

r [o

pin

ion]

potk

[m

alle

t]æ

ngo!t

[fin

ger]

suzæ

n

[needle

]

naxon

[nail]

affe

cte

d

dam

agin

glæ

tme

[dam

age]

qat!

[c

hop]

sædæ

me

[dam

age]

gij

[diz

zin

ess

]

sæ

rma

[cold

]

resh

apin

g

AFFECTED: DAMAGINGzæxm xordæn wound be wounded∫ekæst xordæn defeat to be defeatedasib xordæn injury, damage get damagedt∫ak xordæn crack, cut sustain a crack, tear openjer xordæn a rip be rippedqat∫ xordæn splinter to splinter, to be splittæræk xordæn crack be crackedsædæme xordæn shock suffer shock, or blowzærbe xordæn shock be shocked

AFFECTED: FUSINGvasle xordæn patch rquire patching(as in badly torn clothes)peyvænd xordæn joining accepting of unitingbænd xordæn connection get connected, attachedbæxiye xordæn stiches requiring stichesdærz xordæn seam be pieced togetherju∫ xordæn weld, bond bond, as in two friends, befriend

AFFECTED: RESHAPINGfer xordæn curl become curledgereh xordæn knot knot up (reflexive), get tied uppele xordæn step appearance of stepspit∫ xordæn twist be twistedpit∫ o tab xordæn twist & curve have twists and turnspit∫ o xæm xordæn twist & turn [road] takes twists and turnst∫in xordæn pleat, crease be pleated, creasedta xordæn fold be folded, creasedtab xordæn warp, swing become wavy (hair), swing on a swing

AFFECTED: TOUCHINGængo∫t xordæn finger be toucheddæst xordæn hand touched; altered; molestednaxon xordæn nail to be scratchedsuzæn xordæn needle be punctured, be damaged by a needle

AFFECTED: TYPE OF HITbazu xordæn muscle receive a jostling blowkaj xordæn blow, slap receive a slap on the neckkotæk xordæn beating be beatenlægæd xordæn kick be kickedlætme xordæn slap be injured, slappedqæfa xordæn nape dope slappedsili xordæn slap be slappedgu∫mal xordæn punishment be chastised or punishedt∫æk xordæn slap to receive a slapzærb[e] xordæn impact receive a blowzærbæt xordæn sword strike be struck by a sword

XORDÆN 191

AFFECTED: HAND HELD∫æm∫ir xordæn sword be struck with a swordkard xordæn knife be wounded by knife∫ælaq xordæn whip be whipped

AFFECTED: PROJECTILEgule xordæn bullet be hit by a bulletsat∫me xordæn pellet get hit by pellets of a guntir xordæn bullet, arrow be hit by a bullet or arrow

AFFECTED: BLUNTpok (potk) xordæn mallet to get hit hard passivelyt∫æko∫ xordæn hammer be hammeredpotk xordæn sledgehammer to be hit by a sledge

AFFECTED: TRICKfærib xordæn trick be trickedgul xordæn trick tricknaro xordæn trick be trickedpa xordæn foot be stepped on, be cheated, be set asideqæbn xordæn cheating be cheated

AFFECTED: UNCLASSIFIEDnæzær xordæn eye be cursed by a bad eye (Mal occulae)t∫e∫m xordæn eye get effected by the evil eye (see nazær)aftab xordæn sun enjoy the sun, be cleansed by sunbad xordæn wind, air be exposed to air or windræng xordæn color, paint be painted, discoloredvaks xordæn wax be waxedbu xordæn smell get scorchedfoh∫ xordæn cuss to be sweared at (hessabi fohsh khord)otu xordæn iron get ironed

SUFFERING: EMOTIONæfsus xordæn regret regretænduh xordæn sorrow grieveæsæf xordæn regret regretæzab xordæn torment suffer alongdæriq xordæn regret regrethers xordæn annoyance annoynedamæt xordæn regret regretpæ∫imani xordæn regret regretqæm xordæn sorrow grieveqam xordæn sorrow to wallow in sorrowqose xordæn grief grievetæ'sof xordæn regret be sorrytimar xordæn nurse, attend console

XORDÆN 192

SUFFERING: CONDITIONdud xordæn smoke breathe or be exposed to smokegærma xordæn heat put up with the heatgorosnegi xordæn hunger starvesærma xordæn cold catch a coldgij xordæn absent minded reel, get dizzy

SUFFERING: UNCLASSIFIEDgol xordæn game goal be scored upon (e.g. soccer, basketball)dæstbord xordæn larceny become a robbery victim

USURPINGnozul xordæn interest living off of loan interest income sævari xordæn riding getting a (free) ride on somethingre∫veh xordæn bribe to live on bribes

MOTION: ROTATIONt∫ærx xordæn turn turned and whirled aroundqælt xordæn flip roll aroundqel xordæn roll to roll around (self or other)

MOTION: GENERALliz xordæn slippery slipsekændæri xordæn stumble stumblelæqze∫ xordæn trip tripsor xordæn slippery slidetækan xordæn motion to be shaken, jolter, movedtelo telo xordæn sway sway from left to rightvul xordæn slither moves none stop (passive, as an infant)qute xordæn float swim

UNCLASSIFIEDzæmin xordæn ground fallab xordæn water drink water, absorb waterbær xordæn on to run into s.o. or s.o..bazi xordæn game be trickedbe dærd xordæn to pain be useful, be a solutionbe gu∫ xordæn ear become aware of, learn of…be hæm xordæn together collide(accident); fall apart; dissolvebe kar xordæn to work be usefulforu xordæn down repress angergoh xordæn shit be in the wronghæm (be) xordæn churning be mixed upja xordæn place, position, roombe very surprisedju∫ xordæn boil, anxiety worry about somethingkafur xordæn campher to become impotentnan (kæsi) xordæn bread to be beholden by kindnesspele xordæn stairs a floor is higher and has stepssær xordæn head be discouraged, disappointedxun xordæn blood deeply afflicted, worry, sufferyekke xordæn shock be shockedxis xordæn wet soakpærse xordæn poor forced to hang around aimlessly / begzenhar xordæn agreement break an agreementruze xordæn fast not to fast when one should

XORDÆN 193

hæva xordæn air get some fresh airqol xordæn bubble rolling boilqæsæm xordæn oath take an oathsogænd xordæn oath to take an oath

XORDÆN 194

assu

me

GE

RE

FT

ÆN

[to o

btai

n]

gra

sp

ca

ptu

re

ou

tlaw

wild

an

ima

lfilm

[f

ilm]

vid

eo

[vid

eo

]

æks

[pho

to]

ma

hi

[fis

h]

!ekar

[hu

nt]

dozd

[t

hie

f]

æsir

[h

osta

ge

]

ze

nd

an

i [p

riso

nn

er]

se

rvic

e

rep

rese

nt

karm

æn

d[w

ork

er]

næ

qa!

[pain

ter]

pæ

ræsta

r [n

urs

e]

væ

kil

[la

wye

r]

gæ

va

h

[witne

ss]

va

se

t[m

ed

iato

r]

rem

oveg

row

th(e

xte

rna

l)

su

rface

æb

ru[e

ye b

row

]na

xo

n [n

ail]

gæ

rd

[dust]

læke

[s

tain

]

t!e

rk

[dir

t]

go

lab

[rose

-wate

r] !ire

[s

yru

p]

ab m

ive

[f

ruit juic

e]

pe

rfo

rmri

tual

so

cia

l

pers

on

al

æru

si

[wed

din

g]

jæ!n

[c

ele

bra

tion]

tævæ

lod

[b

irth

da

y]

du!

[sho

wer]

ruze

[fa

st]

rejim

[d

iet]

grip

co

nsid

ernæ

did

e

[unse

en]

gera

n

[de

ar]

ævæ

zi

[err

or]

ena

n

[rein

s]

torm

oz

[bre

aks]

ni!gu

n

[pin

ch

]

ob

tain

na

xo!i

[un

we

ll]

jan

[life

]

bu

[sm

ell]

se

ttle

men

t

loo

k

po

sitio

npo

se

sh

ap

e

positio

n/

pla

ce

rela

tive

cover

enro

be

se

ize

cra

mp

lesso

nre

spo

nse

mon

ey

he

lp

eja

re [re

nt]

no

zul

[in

tere

st]

va

m

[lo

an

]

yari

[h

elp

]

ko

mæ

k [he

lp]

re!v

e

[brib

e]

dæ

rs

[lesso

n]

sæ

bo

q

[pre

ce

de

nce

]

pæ

nd

[adv

ice

]æ

hv

al

[hea

lth]

jævab

[a

nsw

er]

xæ

bæ

r [n

ew

s]

ara

me!

[ca

lm]

næ

zm

[o

rde

r]

!oh

ræt

[fa

me]

qæ

rar

[po

sitio

n]

ja[p

lace

]

væ

qt

[tim

e]

tæræ

f [s

ide

]

don

bal

[aft

er]

æz s

ær

[fro

m h

ead

]

!ekl

[fo

rm]

su

ræt

[face

]

næ

x!

[desig

n]

jest

[ge

stu

re]

qolla

b

[hoo

k]

qiy

afe

[f

ace

]

næ

fæs

[bre

ath

]

xak

[du

st]

tæla

[gold

-lea

f]

qe

y

[ble

ar]

oq

[g

ag

]

vir

[cra

vin

g]

!a!

[urin

e]

beco

me

co

vere

d

co

ve

r

færm

an

[o

rde

r]

se

fare!

[ord

er]

dæ

stu

r [o

rde

r]o

rde

r

info

zæ

n

[wife

]

dou

st p

esa

r [b

oyfr

iend

]

mari

tal

extr

act

(inte

rnal)

em

plo

y

qæ

lb

[he

art

]

dæ

ma

q [n

ose

]

ga

t!

[pla

ste

r]

bæ

rf[s

now

]

a!q

al

[tra

sh

]

ca

tch

pæ

rvan

e

[butt

erf

ly]

ab

ele

[p

ox]s

æræ

tan

[c

an

ce

r]

dise

ase

ph

ysic

al

atæ!

[fire

]

tah

aræ

t[h

yg

en

e]

receiv

e

sa

fety

/cu

re

pe

rmis

sio

n

tra

it

sa

man

[s

ett

led]

æla

j [c

ure

]pæ

nah

[r

efu

ge

]!æ

fa

[cu

re]

roxsæ

t [p

erm

issio

n]

eja

ze

[pe

rmis

sio

n]

rese

rve

bin

d

qol [

pro

mis

e]

væ'e

de

[pro

mis

e]

ja [p

lace

]

væq

t [tim

e]

GRASP: CONSIDERævæzi gereftæn wrong mistake s.t. or s.o. for anothernadide gereftæn unseen pretend not to see, ignore∫ol gereftæn loose take something lightlyfærz gereftæn supposition supposegeran gereftæn dear consider dear

GRASP: CAPTUREfilm gereftæn film filmæks gereftæn picture take a picturevidio gereftæn video catch on video

GRASP: GRIPni∫gun gereftæn pinch pinch (part of body) between 2 fingerstormoz gereftæn break hit the breaksenan gereftæn rein take control

GRASP: OUTLAWæsir gereftæn prisoner capture war prisonerszendani gereftæn prisoner take a prisonerdozd gereftæn thief thief

GRASP: WILD ANIMAL∫ekar gereftæn hunt huntpærvane gereftæn butterfly catch a butterflymahi gereftæn fish fish

GRASP: UNCLASSIFIEDqænimæt gereftæn booty seize the moment, take bootyqafel gereftæn negligent catch/capture unawares

PERFORM RITUAL: MARITALfaseq gereftæn affair have a relation with a man other than husbandsiqe gereftæn temp wife marry a temporary wifezæn gereftæn wife (woman) marry (man)

PERFORM RITUAL: SOCIALarusi gereftæn wedding throw a wedding partyjæ∫n gereftæn party celebrateruze gereftæn fast fastt∫ele gereftæn fortieth day

of…hold 40th day of mourning ritual

mehmani gereftæn party to throw a party

PERFORM RITUAL: PERSONALtaharæt gereftæn hygiene practise personal hygienevozu gereftæn oblution carry out cleaning ceremony (islam) prior to prayingehram gereftæn pilgrimage complete the hajj rituralsdu∫ gereftæn shower take a showerhæmam gereftæn bath take a bath/showerrejim gereftæn diet regimen go on a diet

GEREFTÆN 197

COVER: COVERgæt∫ gereftæn plaster put in a cast, cover a wall in plastergel gereftæn mud cover with mud, close for goodqab gereftæn frame frametæla gereftæn gold wrap something in gold foil

COVER: BECOME COVEREDqey gereftæn rheum get bleary eyedsele gereftæn rough skin on

woundwound getting covered by a new rough skin

xak gereftæn dust get covered with dust

COVER: SEIZEoq gereftæn gag feel like vomiting, become disgustedvir gereftæn compulsion be compelled, be obsessed about somethingxab gereftæn sleep become sleepyxænde gereftæn laugh start laughing

COVER: CRAMPnæfæs gereftæn breath get breathtakenpa gereftæn foot cramp in the footqælb gereftæn heart have chest pains

COVER: UNCLASSIFIEDrah gereftæn way, path holdup on the roadru gereftæn face covering face in religious modestydærd gereftæn pain hurt (e.g. part of the body)bærq gereftæn electricity get electrocuted

OBTAIN: RESERVEotaq gereftæn room get a room (e.g. hotel)ja gereftæn place take up room, reserve a placenobæt gereftæn turn take a turnqærar gereftæn appointment to settlevæqt gereftæn time get an appointment

OBTAIN: PERMISSIONejaze gereftæn permission get permissionezn gereftæn permission ask for permissionroxsæt gereftæn permission to get permission

OBTAIN: MONEYnozul gereftæn interest charge interest on a loantavan gereftæn indemnity collect damagesre∫veh gereftæn bribe to accept a bribesjærime gereftæn fine collect a finesædæqe gereftæn alms collect alms to distributeæn'am gereftæn tip accept tip

GEREFTÆN 198

OBTAIN: LESSONyad gereftæn memory learnpænd gereftæn advice accept adviceændærz gereftæn advice accept adviceebræt gereftæn conclusion from

reviewlearn a lesson

tæ'lim gereftæn training to be traineddærs gereftæn lesson get a lessontæxasos gereftæn specificity to obtain a higher degree in a specific field

OBTAIN: RESPONSEæhval gereftæn health ask how someone's doingjævab gereftæn answer get the results of a test or requestxæbær gereftæn news to ask for news

OBTAIN: REPRESENTgovah gereftæn witness call as witness (I take god as my witness)zamen gereftæn guarantor find a gurantorgævahi gereftæn testimony to obtain testimonygæva gereftæn witness to take as witness, such as god or prophet

OBTAIN: EMPLOYbigari gereftæn chore employ without pay, slave, abusing, servicepæræstar gereftæn nurse hire a nursesævari gereftæn riding get a (free) ride on somethingkuli gereftæn rider on ones

shoulderto get a free ride on someone's back

OBTAIN: HELPariye gereftæn something

borrowedborrow

qærz gereftæn loan borrowvam gereftæn loan take out a loanyari gereftæn help accept aid, help

OBTAIN: SAFETY/CUREæman gereftæn safety taken into safetypænah gereftæn refuge take refuge∫æfa gereftæn cure get curedælaj gereftæn cure get cured

OBTAIN: BINDgero gereftæn pledge take in pledgeqol gereftæn promise make someone promisevæ'de gereftæn promise make someone promise

GEREFTÆN 199

OBTAIN: UNCLASSIFIEDtævan gereftæn power to become powerful, to bring back to lifebus gereftæn kiss get a kiss frombuse gereftæn kiss get a light kissbe kar gereftæn to work employ, exploitbeyæt gereftæn allegiance get votes for religious leadership (archaic)enteqam gereftæn vengeance take vengencelæb gereftæn lip kiss someone on the lipsmaye gereftæn principle kiss up?tælaq gereftæn divorce obtain a divorcedad gereftæn justice seek and receive justiceerteqa gereftæn promotion be promoted

RECEIVE: DISEASEabele gereftæn pox get the small poxgoh gije gereftæn confusion get dizzy, confusednaxo∫I gereftæn sick get sicksærætan gereftæn cancer get cancer

RECEIVE: ORDERdæstur gereftæn order accept order; get a precriptionfærman gereftæn order take ordersefare∫ gereftæn order take an order

RECEIVE: PHYSICALnæm gereftæn moisture to become wet, absorb humidityatæ∫ gereftæn fire catch on firebu gereftæn smell get smellyjan gereftæn life get animated, come to life, take life (kill)ræng gereftæn color, paint get tanned or color, fabric, leather, or human skinniru gereftæn power get energizedqovæt gereftæn power get power, energy

ASSUME: RELATIVEdombal gereftæn follow followæz sær gereftæn from head take it from the startfasele gereftæn distance distance onselfgu∫e gereftæn corner retire, to be shytæræf gereftæn side to take someone's side

ASSUME: POSITION/PLACEkenare gereftæn border keep aloofja gereftæn place, position,

roomtake up room, reserve a place

nobæt gereftæn turn take a turnqærar gereftæn appointment to settlevæqt gereftæn time get an appointment

ASSUME: SHAPEsuræt gereftæn face, order,

appearancea project being accomplished, performed, realized

tæ∫kil gereftæn take shape to come to order, as in a meeting, conferencenæx∫ gereftæn pattern to come together, as a solution appearing

GEREFTÆN 200

ASSUME:POSEæza gereftæn mourning set up a mourning wakematæm gereftæn mourning mourn or show much chagrinqiyafe gereftæn face make a disapproving faceqollab gereftæn hook interlock fingers with palms facing up for support of a

weightxæ∫m gereftæn rage display rage or anger (archaic)jest gereftæn pose to strike a pose with expression

ASSUME: TRAITæhæmiyæt gereftæn importance to become importantaram gereftæn calm calm downertefa' gereftæn altitude gain altitude∫etab gereftæn acceleration accelerate rapidlysebqæt gereftæn priority gain advance over something, speed past another carsor'æt gereftæn speed accelerate∫ohræt gereftæn fame become famousforuq gereftæn brightness become eminentasaye∫ gereftæn comfort relax

ASSUME: SETTLEMENTa∫iyane gereftæn nest settle down, nestnezam gereftæn order become orderedsaman gereftæn settled get settled

ASSUME: UNCLASSIFIEDqædæm gereftæn step keep pacezærb gereftæn beat play the beatreng gereftæn dance beat follow a dance beat togetherænjam gereftæn perform be finished, be performedons gereftæn intimacy become familiar and accustomedbærtari gereftæn superiority move up (socially, politically,…)

REMOVE: EXTRACTab gereftæn juice squeeze, juicegolab gereftæn rose-water extract rose-waterxun gereftæn blood draw blood (bleed a person)ab miveh gereftæn fruit juice juice fruitszæhr gereftæn venum collect venum∫ireh gereftæn sirop collect liquid from plants to create sirop (opium, maple,...)

REMOVE: GROWTHnaxon gereftæn nail clip nails, disposing of excessdamæg gereftæn nose blow ones noseæbru gereftæn eyebrow to trim ones eyebrows

REMOVE: SURFACEgærd gereftæn dust to get covered by dust (an object)læke gereftæn stain remove a staindamæg gereftæn nose nose to be blocked

GEREFTÆN 201

UNCLASSIFIEDemtiyaz gereftæn point gain a point or creditgor gereftæn flame burst into flamesbæqæl gereftæn bossom carry in the armsdu∫ gereftæn shoulder to hoist someone or something on ones shoulderbar gereftæn cargo take on cargotæsmim gereftæn decision make a decisionændaze gereftæn size measuresoraq gereftæn location inquire of something or someone's whereabout's, or

conditionkæm gereftæn few, small under estimatedæste kæm gereftæn on the small

sidehaving low expectations of s.o.

dud gereftæn smoke be filled with, engulfed in smokeeyb gereftæn defect critisizenokte gereftæn point find flaw in someone's speechqælæt gereftæn error correctxorde gereftæn minutia complain about, find small flawsdærz gereftæn seam sew two pieces of cloth together, shut up (zip it)kam gereftæn desire obtain gratificationnam gereftæn name be named, become well knownpærvaz gereftæn flight take flightvæ'de gereftæn promise,

appointmentobtain a promise, date

tæmas gereftæn contact contactemtehan gereftæn test give a testadæt gereftæn habit get into the habitfal gereftæn omen tell fortune by a method (cards,…)loqme gereftæn mouthful make a morsel, do something without approvalmæ'reke gereftæn show put on a magic showne∫ani gereftæn address ask or obtain an addressælaj gereftæn cure be curedbæla gereftæn misery befall with miserybe -[xæri, næfæhmi,

gereftæn assume take someone to be [stupid, dumb, weak,…]dar extiyar gereftæn in control take control ofeste'fa gereftæn resignation to force someone to resignesteglal gereftæn independence obtain independenceesteqrar gereftæn settlement to settleforsæt gereftæn occasion;

opportunityjump at an opportunity, catch

go∫n gereftæn semen, pollen to fecundated, impregnate by copulation, by spreading pollenkar (be) gereftæn work put to use

ne∫an gereftæn sign, target aim with a weapon: gun, bow and arrow, etc.pæs gereftæn behind, back take backpi∫i gereftæn ahead move aheadqaleb gereftæn mold make a mold from somethingrasmiyæt gereftæn official become officialreza gereftæn content to obtain forgivenesssur gereftæn advantage force a player to throw a card which one can trumpsuxt gereftæn fuel to get fueltæhvil gereftæn change, shift to accept for safekeeptæqlil gereftæn reduction be reducedaftab gereftæn sun sunbathegæriban gereftæn collar seize by the collar, accuse of somethingpiyale gereftæn cup to take up drinkingsær pa gereftæn squat hold up a child in a squat to pee

GEREFTÆN 202

bær gereftæn on choosedæst gereftæn hand latch onto something funnydamæn gereftæn skirt plead for mercyfæra gereftæn inside learn, take ingærdæn gereftæn neck declare oneself responsiblehal gereftæn health to make someone upsetko∫ti gereftæn wrestle wrestlemo∫t gereftæn wrist catch in the act, grabbing hand in mov'ttas gereftæn die cheat when rolling diegu∫ gereftæn ear accept advicegærm gereftæn warm be happyab qure gereftæn verjuice extract juice of unripe grapesbænd gereftæn connection stay putdæste payin gereftæn on the low side having expectations of low pricepær gereftæn feather set oneself freebær - gereftæn on- [du∫, po∫t, sar] hoist on the shoulders, back, headbala gereftæn up hold in high esteembazi gereftæn game to make a fool offæn gereftæn technique use a technical move in wrestlingfænd gereftæn trick to outsmart and trickfæra gereftæn surround cover (as a flood, army)ri∫ gereftæn beard to catch someone and demand an answersæboq gereftæn precedence to take a lessonpæhlu gereftæn side to bring to port (ship)pey gereftæn base follow up with persistancedæm gereftæn breath sing in harmony, control speech, attain harmony,

organisationsær gereftæn head take a good turnt∫e∫m gereftæn eye catch ones eye, attractnætije gereftæn result extract or deduce a conclusion, conclude

GEREFTÆN 203

regim

en

t

pæ

rvæ

re!

[ra

ise

]

up

on

oth

ers

a b

ind

pro

motio

n

DA

DÆ

N[to

giv

e]

pro

duce

pro

du

ct

miv

e

[fru

it]

gol [flo

we

r]

!ire

[sy

rup]

offe

r

impose

expre

ss

exp

lain

sa

y

kn

ow

led

ge

no

rishm

ent

pro

vid

e

he

lp

sa

fety

/cu

re

pe

rmit

result

em

an

ation

pre

se

nta

tion

em

be

lish

ren

der

tra

it

se

ttle

men

t

ple

asure

favo

r

lesso

n

respo

nse

rese

rve

so

cia

l

op

po

rtu

nity

rota

tio

n

ord

er

ran

k sh

ap

eexpo

su

red

isco

mfo

rt

topo

logy

ph

ysic

al

ch

an

ge

motio

n

ge

ne

ral

nu

r [lig

ht]

bu [

sm

ell]

næ

tije

[c

onclu

sio

n] s

ud

[pro

fit]

zæ

rær

[loss]

næ

zæ

r [o

pin

ion

]

pa

so

x

[re

sp

on

se

]

jæv

ab

[an

sw

er]

ne

da

[m

essa

ge]

xæ

bæ

r [

ne

ws]

e'e

lam

[an

no

un

ce

men

t]

tozih

[e

xp

lan

ation

]

!æh

r [d

elin

ea

tion

]

tæfs

il

[inte

rpre

tatio

n]

sæ

bo

q

[pre

ce

de

nce

]dæ

rs

[lesso

n]

ænd

ærz

[a

dvic

e] næ

maye!

[sho

w]

kon

fera

ns

[co

nfe

rence

]

so

xæ

nra

ni

[lectu

re]

eja

re

[re

nt]

pe

sta

n

[bre

ast]

qo

rs [p

ill]

na

n

[bre

ad]

mæ

dæ

d

[help

]ko

mæ

k

[aid

]

!æfa

[cure

]

pæ

na

h

[re

fug

e]

æla

j [c

ure

]

sa

ma

n

[se

ttle

men

t]næ

zm

[o

rga

niz

ation

]

tært

ib

[ord

er]

me

hm

an

i [g

et-

tog

eth

er]

!am

[d

inne

r]

!irin

i [s

wee

ts]

!oh

ræt

[fa

me]

so

r'æ

t [s

pee

d]

hal

[joy] kam

[s

atisfa

ctio

n]

mæ

ze

[ta

ste

]

eja

ze

[pe

rmis

sio

n]

roxsæ

t [p

erm

issio

n]

extiya

r [c

on

tro

l]

no

bæ

t [tu

rn]

ja [p

lace

]væ

qt

[tim

e]

!ekæ

nje

[to

rtu

re]

co

nn

ection

fors

æt

[cha

nce

]

em

kan

[p

ossib

ility

]

jai [p

lace

]

azar

[to

rme

nt]

go

rosn

eg

i [h

un

ge

r]

ræbt

[re

latio

n]

ju!

[weld

]

ete

sa

l [

join

]

pit!

[tw

ist]

fer

[curl

]

xæ

m

[be

nd

]

!ekl

[fo

rm]

læb

e[e

dge

]

næ

q!

[pa

tte

rn]

afta

b [su

n]

gæ

rma

[hea

t]

bu

[sm

ell]

qæ

lt [

flip

]

pit!

[tu

rn]

t!æ

rx

[ro

ll]

qe

ym

æt

[price

]

færm

an

[o

rde

r]se

fare!

[ord

er]

dæ

stu

r [o

rde

r]

tab

[sw

ing]

tæka

n

[jerk

]

gæ

rde!

[mo

ve

me

nt]

fase

le

[dis

tan

ce

]

na

m[n

am

e]

co

nd

itio

nlin

æt

[so

ftn

ess]

yo

bo

sæ

t [c

on

stip

ation

]

de

l d

ærd

[s

tom

ach

ach

e]

dis

pla

cem

en

t

ho

l[p

ush o

ff]

kut!

[mig

ratio

n]

ab

tæ

la

[gold

]

lo'a

b

[polis

h]ab

ro

qæ

n

[oil]

halæ

t[c

hara

cte

r]

ph

ysic

al

pers

on

al

linka

ge

vefq

[co

ord

ination

]ræ

bt

[re

latio

n]

a!ti

[re

united]

ara

mesh

[ca

lmn

ess]

relin

quis

h

volu

nta

rily

in-

vo

lunta

ry

reza

ya

t[a

gre

e]

de

l[h

ea

rt]

jan

[life

]

ve

kalæ

t[a

tto

rne

y]

conduct

pro

gre

ss

eve

nt

co

mp

letion

ænja

m[f

inal]

ænja

m[p

erf

orm

]

eda

me

[co

ntinu

ation

]

suræ

t[fa

ce

]

feysæ

le[c

losu

re]

xa

tem

e[e

nd

ing]

mon

ey

pro

duct/

se

rvic

e

ke

rye

[le

t]

nozul

[usu

ry]

malia

t [ta

xes]

mo

ne

tary

tæxfif

[dis

co

unt]

ærz

an

[ch

ea

p]

justice

en

sa

f[h

on

esty

]

hæ

q[r

ight]

up

on

se

lfso

gæ

nd

[pro

mis

e]

væ

'ed

e[p

rom

ise

]

tæla

q[d

ivorc

e]

se

da

[no

ise

]

qo

l[a

ssu

ran

ce

]

qæ

sæ

m[s

wea

r]

sa

ze!

[re

united

]

væ

rze!

[tra

inin

g]

xa

k [d

ust]

rejim

[die

t]

ab

use

mis

use

zæ

hm

æt

[hard

ship

]

jaye

zeh

[prize]

dæ

reje

[ra

nk]

ert

eq

ah

[pro

motio

n]

oka

zyo

n[p

ossib

ility

]

han

dle

fe!a

r [s

que

eze]

male!

[massa

ge]

qe

lqelæ

k

[tic

kle

]

po

sitio

n

ja

[pla

ce

]

qæ

rar

[positio

n]

OFFER: EXPLANATION∫æhr dadæn narration narratefærz dadæn supposition supposemæsal dadæn example give examplemesal dadæn example give an exampletæfsil dadæn details tell in detailtozih dadæn explanation explain

OFFER: SAYbe∫aræt dadæn good news bring good newselam dadæn announcement advertiseente∫ar dadæn propagation propagatefoh∫ dadæn cuss cussmesaj dadæn message leave a messagemojde dadæn good news bring good newsmonada dadæn messages bring inspired messagesneda dadæn message announcenehib dadæn loud sound call out loudpæyam dadæn message leave a message, send a messagepeyqam dadæn message leave a messagesæla dær dadæn proclamation proclaimsælam dadæn peace salute

OFFER: RESPONSEæqide dadæn opinion give opinionjævab dadæn answer answernæzær dadæn opinion express an opinionpasox dadæn answer to answerxæbær dadæn news give sign of life, give news

OFFER: LESSONændærz dadæn advice advisedærs dadæn lesson teachebræt dadæn warning to teach a lessonmæ∫q dadæn homework assign homeworknæsihæt dadæn advice advisepænd dadæn advice advisepi∫næhad dadæn suggestion suggestsæboq dadæn precedence teachsoq dadæn guidance guide toward a goaltæ'lim dadæn knowledge teachyad dadæn memory teach

OFFER: PRESENTATIONæra'e dadæn presentation presentgozare∫ dadæn report reportkonferans dadæn conference deliver a conferencenæmaye∫ dadæn show give a showne∫an dadæn sign, target show, point out

DADÆN 207

OFFER: MONEYæn'am dadæn tip tipbaj dadæn toll pay tribute or tollfoqolade dadæn allowance give allowancehævale dadæn money order transfer of money through an orderhoquq dadæn salary pay salaryhoquq dadæn salary pay sallaryjire dadæn allowance pay subsistence salarykabin dadæn dowry pay dowrykæfare dadæn penance to pay penancenozul dadæn interest pay interestpada∫ dadæn award awardpi∫æki dadæn beforehand pay in advancere∫veh dadæn bribe bribesædæqe dadæn alms give almssæhm dadæn share pay a sharetænzil dadæn interest loan on interesttæsædoq dadæn alms give alms for a wishtavan dadæn indemnity pay damagesvæzife dadæn pension pay pension

OFFER: NOURISHMENT∫ir dadæn milk milk (breastfeed)ab dadæn water water (e.g. flowers)nan dadæn bread provide forpestan dadæn breast breastfeedqæza dadæn food to feedsaqær dadæn cup offer wine, cupzæhr dadæn poison poison

OFFER: PRODUCT/SERVICE∫æhadæt dadæn witness attestejare dadæn rent rent outkeraye dadæn let letkuli dadæn ride on shoulder to carry one (figurative or actually)

OFFER: HELPæriye dadæn rent pay rentariyæt dadæn borrowed lend somethingbæradæri dadæn brotherhood give supportbazu dadæn upper arm help, assist (give a hand)dad dadæn justice do justicedeldari dadæn compassion show compassiongævahi dadæn testimony testify; bear witness to someones charactermædæd dadæn help helppæhlu dadæn side provide supportqærz dadæn load lendvam dadæn loan give a loanyari dadæn help aid, help

OFFER: SAFETY/CURE∫æfa dadæn healing healælaj dadæn cure cureæman dadæn safety provide safeteltiyam dadæn conciliation conciliation, cause to healmænzel dadæn home provide a homenejat dadæn salvation savepænah dadæn refuge give refuge

DADÆN 208

OFFER: SOCIAL∫am dadæn dinner invite for dinner∫irini dadæn sweets provide sweets for some successmehmani dadæn get together have a get togethersur dadæn celebration offer a celebration of a success

OFFER: SETTLEMENTesteqrar dadæn settlement cause to become settlednezam dadæn order put in ordersaman dadæn settlement settle someone downsazman dadæn organisation organizetærtib dadæn order take care of

OFFER: TRAIT∫etab dadæn hurry speed upenerzhi dadæn energy energizemæsuniyæt dadæn immunity make immuneniru dadæn power energizetævan dadæn power to bestow power, bring back to life againqodræt dadæn power give powerronæq dadæn prosperous make prosperous∫æxsiyæt dadæn dignity give dignity

OFFER: PLEASUREhal dadæn health to give pleasurekam dadæn desire gratify someone’s wisheskeyf dadæn intoxication give pleasurelezæt dadæn pleasure pleasemæze dadæn taste taste goodsæfa dadæn purity make pleasant

OFFER: OPPORTUNITYemkan dadæn possibility give possibilityforsæt dadæn occasion allow, give a chancekar dadæn work employ/ Put to usemeidan dadæn opportunity encouragemohlæt dadæn remaining time give a chancevaqeyæt dadæn reality make happenehtemal dadæn possibility give possibility

OFFER: MONETARYærzan dadæn cheap make cheapertæxfif dadæn discount give a discount

OFFER: JUSTICEensaf dadæn honesty be honesthæq dadæn right give righthoqm dadæn judgement give a judgement

OFFER: PERMISSIONæzn dadæn permission permitejaze dadæn permission grant permissionextesas dadæn allocation allocate, specify for somethingrah dadæn path let someone throughreza dadæn content agreeroxsæt dadæn permission to permit

DADÆN 209

OFFER: PROMOTIONdæræje dadæn rank give higher rankerteqa dadæn promotion promotekeyfær dadæn reward rewardpost dadæn position hire for a bureacratic positionejr dadæn reward reward

OFFER: RESERVEmæjal dadæn time, possibility give a chancenobæt dadæn turn give a chance or an appointmentvæqt dadæn time give an appointment

OFFER: UNCLASSIFIEDera'e dadæn offering offerfætva dadæn arbitration announce an arbitrationho∫dar dadæn warning warnjan dadæn life give life, energizejayeze dadæn prize grant a prizemo∫arekæt dadæn partnership sharemoræxæsi dadæn vacation allow a vacation (employees)ruh dadæn soul animatetæqviyæt dadæn strength strengthentæsæla dadæn consolation consoletæsliyæt dadæn condolences give condolencestæxasos dadæn specificity to allocated to a specific use (automobil ra be o tæxasos dadand)tæzækor dadæn warning warnviza dadæn visa grant a visa

RELINQUISH: VOLUNTARILYesnad dadæn attribute attribute to someoneextiyar dadæn control give controltæxsis dadæn allocation allotvekalæt dadæn attorney give power of attorneytævækol dadæn representation put in hands of lawyer

RELINQUISH: INVOLUNTARILYdel dadæn heart become attached (emotion)jan dadæn life to die

IMPOSE: ABUSE∫ekænje dadæn torture torture∫ekæst dadæn destroyed destroyæzab dadæn torture tortureazar dadæn bother bother someonedeq dadæn grief cause grieffærib dadæn lie trickgorosnegi dadæn hunger starve someonegu∫mali dadæn punishment chastisehers dadæn annoyance annoyqose dadæn grief cause someone grieftæsdi' dadæn worrying give pain or troublezæjr dadæn torment make someone suffer, torment s.o.zærær dadæn loss incur a loss

IMPOSE: MISUSEdærdesær dadæn headache cause an inconviniencezæhmat dadæn labour to trouble someone

DADÆN 210

IMPOSE: PHYSICALettesal dadæn connection to connect electrical wiresju∫ dadæn soder soder two thingspeyvæst dadæn connection connecttæmas dadæn contact to make a tangent (in geometry)

IMPOSE: PERSONALa∫ti dadæn forgiveness arrange forgiveness between two partiessaze∫ dadæn mediation to medaite, to harmonize

IMPOSE: LINKAGEertebat dadæn connection relate facts or issuesnesbæt dadæn relation relateræbt dadæn relation relatevefq dadæn coordination coordinate

IMPOSE: REGIMENT∫osto∫u dadæn cleansing clean∫osto∫uyemæqzidadæn brainwashing brainwashadæt dadæn habit condition someone/somethingpærvære∫ dadæn raise, breed raiseqosl dadæn ablution absolve (generally self) by dipping in waterro∫t dadæn growth bring up, growtæmrin dadæn exercise exercise somethingværze∫ dadæn exercise train, exert, exercise a part of the body

IMPOSE: TOPOLOGY∫ekaf dadæn split splitbore∫ dadæn cut make a cutettesa' dadæn expansion expandfer dadæn curl to curl hairgereh dadæn knot to tie, to knotjer dadæn rip ripkez dadæn singe shrink something by applying heatpit∫ dadæn twist twistqat∫ dadæn chop, crack make a crack in somethingt∫ak dadæn crack crack something aparttab dadæn warp or swing warp or swing, warp a rope, hair, etc. or push a swingtæra∫ dadæn shave shavetiz dadæn sharp distinguishtowse'e dadæn expansion expandxæm dadæn curve create a curve or bend (in something)xæra∫ dadæn scratch irritate, scratch

IMPOSE: SHAPE∫ib dadæn cline slopeform dadæn form give shapelabe dadæn edge, ledge give an edge to a formnæx∫ dadæn pattern design a pattern on somethingsuræt dadæn face produce, accomplishtæ∫kil dadæn take shape give shape, bring togther (conference, class,…)tærh dadæn design make a design, blueprint

DADÆN 211

IMPOSE: HANDLE∫ok dadæn shock to apply a shock (jolt someone)fe∫ar dadæn pressure squeezegir dadæn hold hang temporarily, catch, constrainhol dadæn push pushke∫ dadæn elastic stretch out (an elastic)male∫ dadæn massage massagemasaj dadæn massage massagemo∫t o mal dadæn massage massagenaværd dadæn knead knead, massageqelqelæk dadæn tickle tickle

IMPOSE: EMBELLISHab noghre dadæn plated silver to silver plateab roqæn dadæn water oil pour oil on cooked riceab tæla dadæn plated gold to gold platelo'ab dadæn enamel to enamelpærdaxt dadæn finish polish, shine, finishzær o bærq dadæn shine give lustrezinæt dadæn beauty embellish

IMPOSE: UPON SELFqol dadæn promise promisevæ'de dadæn promise promise

IMPOSE: UPON OTHERSqæsæm dadæn oath to put under oathsogænd dadæn oath to put under oath

IMPOSE: EXPOSUREaftab dadæn sunshine expose something to the sunbad dadæn wind expose to wind, or airboxur dadæn steam aromatherapeuticdud dadæn smoke fumigate, smoke (food)gærma dadæn heat expose to heathæva dadæn air expose to airkafur dadæn campher alleviate painsærma dadæn cold expose to cold

IMPOSE: ROTATIONqælt dadæn summersault roll somethingqel dadæn roll roll somethingqute dadæn swim make a thing or person floatt∫ærx dadæn turn rotate

IMPOSE: DISPLACEMENTazadi dadæn freedom give freedomfærar dadæn escape make someone escapegozær dadæn passage allow to passkut∫ dadæn migration force animal migrationobur dadæn passage pass or move someone/somethingvel dadæn loose let loose, let go free

IMPOSE: GENERALgærde∫ dadæn voyage make something move abouthærekæt dadæn movement set in motionpærvaz dadæn flight initiate flight, bring to flighttækan dadæn motion shake, jolt, move

DADÆN 212

IMPOSE: RANK∫erafæt dadæn nobility give nobilitylæqæb dadæn title bestow title or nicknamenam dadæn name to nameonvan dadæn member bestow a titleqædr dadæn value value somethingrasmiyæt dadæn official to make official by sanctiontæ'mid dadæn baptism baptise, make official

IMPOSE: POSITIONfasele dadæn distance separateja dadæn place find a place, make something fitjai dadæn place give a placeqærar dadæn appointment allot a spot for something, put in place

IMPOSE: ORDERdæstur dadæn order orderfærman dadæn order give an ordersefare∫ dadæn order order

IMPOSE: UNCLASSIFIEDbæst dadæn expansion extend, or expandfærq dadæn difference make differencejesmiyæt dadæn solidity solidify, give bodyju∫ dadæn boil boillængær dadæn weight causing unblanacesuze∫ dadæn burn cause burningtæfrih dadæn leisure allow leisuretæft dadæn heat stir frytæhrir dadæn modulation modulate a voice to sing, print in caligraphytæqir dadæn change change somethingtæqlil dadæn reduction diminish, reducetaxt dadæn gallop whip into a gallop

CONDUCT: EVENTons dadæn initmacy make intimate, accustomtælaq dadæn divorce divorcetæm∫iyæt dadæn promoting organize, get off the ground

CONDUCT: PROGRESSænjam dadæn perform perform somethingedame dadæn continuation continue

CONDUCT: COMPLETIONænjam dadæn completion accomplish somethingextetam dadæn end bring to an endpayan dadæn end endxateme dadæn end bring to an end

PRODUCE: PRODUCTbærg dadæn leaf blossomgæ∫n dadæn male fecundategol dadæn flower blossomhasel dadæn crop yieldmive dadæn fruit bear fruitsæmær dadæn yield yieldsele dadæn rough skin when body produces the thick skin on wound while healing

DADÆN 213

PRODUCE: EMANATION∫ekæm dadæn belly swell out (as in a building)avaze dadæn voice call outbirun dadæn out emitboxar dadæn steam expose to steambu dadæn odor emanate a smell, scorch (e.g. fish)foruq dadæn brightness give off lightnæm dadæn moisture emit moistureseda dadæn sound make a noise

PRODUCE: RESULTfayde dadæn profit be advantageousnatije dadæn conclusion provide a conclusionsud dadæn benefit be beneficial (financial)zærær dadæn loss cause a loss (financial)

PRODUCE: CONDITIONarame∫ dadæn calmness calm someoneasaye∫ dadæn comfort comfortdeldærd dadæn stomach ache causea stomach achego∫aye∫ dadæn relief give reliefhæsasiyæt dadæn allergy trigger an allergylinæt dadæn smoothness causing loosening of boweltæskin dadæn soothing sootheyobosæt dadæn constipation cause constipation

UNCLASSIFIEDæz dæst dadæn from hand loseatæ∫ (be) dadæn fire to waste (be atæ∫ dadæn)bær bad dadæn on wind lose by wastingbærfab dadæn snow water dissapointbar dadæn baggage give audience, or fructify (botany)bazi dadæn game let in on a gameboruz dadæn appearance divulgedæm dadæn breath allow speech, give a momentdær dadæn in submit, acceptdæst dadæn hand shake handsenan dadæn leash become subservientfæra dadæn above pay attention to, focus in a certain dir.foru dadæn down swallowgærdæn dadæn neck accept burdengo∫ad dadæn loose give way, in backgammon: leave empty spacegu∫ dadæn ear listenlæm dadæn lean lay around lazilyleft dadæn fat in stew drag out, kneedleft o lo'ab dadæn fat and jelli go through senseless ritualslegam dadæn rein give inlo dadæn trap tell on, snitchmeydan dadæn range (city) provide freedompa dadæn foot opportunity knockingpær dadæn feather set freepoz dadæn appearance show offqælæt dadæn mistake lead to errorqurt dadæn swallow swallowræm dadæn stamppede to cause frenzy by fearru dadæn face allow someone to get too comfortablesær dadæn head release, liberate

DADÆN 214

sær o samandadæn head and settle someone downtæmiz dadæn clean distinguishtas dadæn die coming up lucky in dicetul dadæn length take a long time∫ohær dadæn husband marry off (a girl)æhæmiyæt dadæn importance care, consider importantælamæt dadæn signal signalærzehal dadæn plea pleadbærekæt dadæn bless blessbus dadæn kiss kissdærxast dadæn application applyeftexar dadæn pride give pleasureemtehan dadæn test take a testenteqal dadæn transfer transfere'tebar dadæn validity authenticateetminan dadæn assurance assurefæraz dadæn again give backhædær dadæn waste wastehedye dadæn gift give a giftjælal dadæn glory glorifykahe∫ dadæn slowdown lessenkud dadæn manueur spread manueurlæb dadæn lip make your lip available for a kisslæm dadæn lean lean onmæmuriæt dadæn commission give a missionmæ'ni dadæn meaning give meaning tomanur dadæn manueur spead manueurmosabeghe dadæn competition competemosaf dadæn opposition opposenæmayændegidadæn delegation delegaten'emæt dadæn abundance bestowed in abundance (by god)nosxe dadæn copy / order prescribepæs dadæn behind return a merchandisepas dadæn guard guard, protect, shift responsibility to otherspas dadæn pass (as in ball) sneak out of responsibilitypo∫ti dadæn back rest supportræhn dadæn pawn pawnræm dadæn rear back make a horse rearrævaj dadæn propogation propegaterox dadæn face occur, happensærayæt dadæn contagion pass on a diseasesoq dadæn move on implementtæhvil dadæn change, shift deliver for safekeeptækiye dadæn support lean ontæmærkoz dadæn concentration bring togethertæ'mim dadæn generalisation applying a rule to everyone equallytæmiz dadæn clean distinguishtænæzol dadæn degredation reduce

DADÆN 215

KÆ

RD

ÆN

[to

do]

state

undergo

xo!k

[dry

]kæ

j [be

nt]

pærd

axt

[pay

men

t]

action

qæ! [

fain

t]qo

lqol

[b

ubbl

e]

mon

tæqe

l [tr

ansf

er]

xæbæ

r [n

ews]

tæb

[feve

r]

ræha

[fr

ee]

geriy

e [c

ry]

unergative

zend

egi

[life

]

!uxi

[jok

e]

ACTION∫ane kærdæn comb combbolænd kærdæn tall liftdævæt kærdæn invitation invitegero kærdæn pledge to mortgagehæbs kærdæn imprisonment imprisonjæfa kærdæn oppression oppressjæng kærdæn war have warjaru kærdæn broom sweepmæqlub kærdæn defeat defeatmærju’ kærdæn returned return somethingmat∫ kærdæn kiss give a kissmotaele kærdæn investigation investigateotu kærdæn iron ironpay kærdæn purpose trace, suivrepayin kærdæn down put downpeykar kærdæn war go to warpeymane kærdæn measure measureqæbn kærdæn cheating cheatqærq kærdæn drowned drown (trans)tæhædi kærdæn disrespect disrespecttæhækom kærdæn command dominate, commandtæhrir kærdæn writing writetæqaza kærdæn request applytærjome kærdæn translation translatetayp kærdæn type typevared kærdæn arriving put something in something elsexun kærdæn blood commit murderazmaye∫ kærdæn exam test somethingnehzæt kærdæn movement set off on a marchqætar kærdæn train set up in line, one after the otherruy kærdæn face revealzærær kærdæn loss to sell at a loss

STATE∫ærmsar kærdæn shamed give shameæziyæt kærdæn bother bother, annoybænd kærdæn lace fastenbaz kærdæn open openbidar kærdæn awake wake someone updæmær kærdæn flat set flatdelxor kærdæn dismayed dismay someonegæriban kærdæn collar tear, rendgærm kærdæn warm make warmgerd kærdæn round make round, assemblegereftar kærdæn stuck capture, put into difficultyhæraj kærdæn sale put on salekæj kærdæn crooked bend, curvekæm kærdæn little substractkenar kærdæn side put asidelule kærdæn tube roll up (trans)pæhn kærdæn spread spreadpor kærdæn full fill upqæbul kærdæn assent acceptræha kærdæn liberated set freeræng kærdæn color paint, colorro∫æn kærdæn light turn on

KÆRDÆN 219

sæbok kærdæn light humiliate, make lighhtsetæm kærdæn oppression oppresssir kærdæn satisfied fill, satisfytæmiz kærdæn clean cleanxærab kærdæn broken destroyxub kærdæn good healing someone, making betterzæjr kærdæn torment persecute, drive outmoqer kærdæn confessor compel to confessbirun kærdæn out firedær kærdæn in shoot, draw outfæraz kærdæn closed closeforu kærdæn into thrustnæm kærdæn moisture dampen, moisten

UNDERGObad kærdæn wind swell, inflate (also to show off)deldel kærdæn wavering be ambivalentgerieh kærdæn weeping crygir kærdæn hold get stuckhæras kærdæn fear be alarmedkæf kærdæn palm foamkeyf kærdæn intoxication to pleasureqæ∫ kærdæn faint faintqolqol kærdæn gurgle boiltæb kærdæn fever get a fevervæhm kærdæn fear fear

UNERGATIVE∫ærm kærdæn modesty be modest∫uxi kærdæn joke jokeærz kærdæn presentation make a humble remarkætse kærdæn sneeze sneezeaqaz kærdæn beginning begindærhæm kærdæn confused confuse, mix upderiq kærdæn refusal hold backebraz kærdæn declaration declareelan kærdæn announcement announceemtehan kærdæn test test, tryextelat kærdæn mingle mingleextiyar kærdæn choice choosefærahæm kærdæn collected gather, assemblefekr kærdæn thought thinkgæman kærdæn opinion suspectgærde∫ kærdæn voyage travelgonah kærdæn sin singozær kærdæn passage pass bygu∫ kærdæn ear listenhærekæt kærdæn movement movehamam kærdæn bath take a bathlæj kærdæn grudge hold a grudgenæf’ kærdæn benefit get a benefitnale kærdæn moan moannaz kærdæn coquetry feign disdainpæsænd kærdæn admiration choose, selectpæsændaz kærdæn save savepo∫t kærdæn back turn aroundqælæt kærdæn mistake make a mistake

KÆRDÆN 220

sæbr kærdæn patience waitsælam kærdæn peace say hellosorfe kærdæn cough coughsoxænrani kærdæn lecture give a lecturetæjob kærdæn surprise be surprisedtæma∫a kærdæn watch watchtæmæ’ kærdæn covetousness covettæn kærdæn body put on (e.g. clothes)tæqir kærdæn change changevæqt kærdæn time find a leisure opportunityvelvele kærdæn clamor uproarxahe∫ kærdæn request ask politelyxodahafezi kærdæn goodbye say goodbyezendegi kærdæn life livedæst kærdæn hand put your hand in somethingdav kærdæn move increase the stake, make a movefærz kærdæn supposition suppose

KÆRDÆN 221

!OD

ÆN

[to b

eco

me

]

sta

te

inte

rnal

xo!k

[d

ry]

kæ

j [b

ent]

pæ

rdaxt

[p

aym

ent]

underg

o

montæ

qel

[tr

ansfe

r]

ræha

[fre

e]

so

rx

[blu

sh]

zeræ

ng

[c

leve

r]

atæ!i

[en

raged]

INTERNALhævar ∫odæn collapse collapseate∫I ∫odæn enraged be angereddælæme ∫odæn gelatine coagulateaftabi ∫odæn sunny become sunny

STATEmæqbul ∫odæn accepted be accepteddelxor ∫odæn annoyed fill up with indignationvared ∫odæn arriving enterjæm ∫odæn assembly gather in a group (reflexive)qæbul ∫odæn assent be acceptedænjomæn ∫odæn association gathergom rah ∫odæn astray be misleadbæstæri ∫odæn bed-ridden get sickbozorg ∫odæn big growæziyæt ∫odæn bother be bothered∫ekæste ∫odæn broken break downdeq mærg ∫odæn bruised death die of sorrowaram ∫odæn calm calm downbatel ∫odæn cancelled be cancelled, expirebæste ∫odæn closed close upfærahæm ∫odæn collected be gathered, assembledmoqer ∫odæn confessor to confess, aknowledgedærhæm ∫odæn confused get confused, displeasedkæj ∫odæn crooked bend, curvea∫ofte ∫odæn distracted be dishevelledforu ∫odæn down descendqærq ∫odæn drowned drownasan ∫odæn easy be facilitatedtæmam ∫odæn end endagah ∫odæn enlightened opening of consciousnessængixte ∫odæn excited get excitedgeran ∫odæn expensive rise in price, become (cost) dearælave ∫odæn extra be addedbænd ∫odæn fasten staykæm ∫odæn few be diminished, become lessdæmær ∫odæn flat lie flatrævane ∫odæn flowing to get goingpa ∫odæn foot get uppa berehne ∫odæn foot bare become barefootæz xod bi xod ∫odæn from self without self lose normal statepor mive ∫odæn full fruit be full of fruit (of tree)dæst pat∫e ∫odæn hand heel rushdæst be yæxe ∫odæn hand to collar get into a fight, conflictsær ∫odæn head come to an end, expirebolænd ∫odæn high get upavare ∫odæn homeless become a wandererræha ∫odæn liberated get liberated∫ol ∫odæn loose lose energy, get listlessgom ∫odæn lost get lostfæravan ∫odæn many multiply, become moreaqe∫te ∫odæn moist become moistnæzdik ∫odæn near get nearer, draw nearmane’ ∫odæn obstacle to prevent, hindera∫kare ∫odæn obvious be revealed∫aqel ∫odæn occupying get employed, have a job

∫ODÆN 225

baz ∫odæn open get openedabadan ∫odæn populated get populatedabestæn ∫odæn pregnant get pregrantamade ∫odæn ready get readyaraste ∫odæn refined be refinedrefuze ∫odæn refused failmærju’ ∫odæn returned get returnedgerd ∫odæn round become roundfærari ∫odæn runaway be a fugitive, runawayhol ∫odæn rushed rushsir ∫odæn satisfied be satisfiedmariz ∫odæn sick get sickzæde ∫odæn sick get sick of somethingnærm ∫odæn soft get softvelo ∫odæn spread get spread outgereftar ∫odæn stuck be captured, get into difficultymo’ællæq ∫odæn suspended be suspended (fig. as well)amixte ∫odæn taught be taughtte∫ne ∫odæn thirsty become thirstyqætar ∫odæn train get in line, one behind the otherlule ∫odæn tube roll upgærm ∫odæn warm get warmab ∫odæn water meltgeryan ∫odæn weeping weepfetile ∫odæn wick roll up or offhæm ræng ∫odæn same color become one withkælafe ∫odæn exhausted get exhaustedkæm ∫odæn few diminishlazem ∫odæn necessary become necessaryleh ∫odæn crushed be crushedmæ∫qul ∫odæn busy get busymo∫tæri ∫odæn client become a faithful clientmot∫ale ∫odæn crumpled get crumplednæ∫'e ∫odæn high get highnegæran ∫odæn worried get worriedpæ∫iman ∫odæn regretful regretpæri∫an ∫odæn chaotic get chaoticpenhan ∫odæn hidden be hiddenpu∫ide ∫odæn covered get coveredru be rah ∫odæn face to road get ready to gosaken ∫odæn resident become a residentvækil ∫odæn lawyer get a job as a lawyervel ∫odæn loose be let loose

UNDERGOelam ∫odæn announce be announcedbidar ∫odæn awake wake upjaru ∫odæn broom get sweptænjam ∫odæn conclusion be accomplishedtaxliye ∫odæn evacuated to get evacuatednemune ∫odæn example be an example, idolhæbs ∫odæn imprisonment be imprisonedxæbær ∫odæn news get warned, get newse∫are ∫odæn point be pointed outsævar ∫odæn ride mount to ride (e.g. horse, car)telesm ∫odæn spell spellstruckne∫ane ∫odæn target be a targetændær ∫odæn within get in

∫ODÆN 226

motmæ'en ∫odæn certain become surepærdaxte ∫odæn transfered be transferredqafel ∫odæn prisoner become a prisoner of

∫ODÆN 227

em

it

KE

ıIDÆ

N[to

pu

ll]

anim

al

pull

su

ffe

r

in

out

alo

ng

dra

w

so

und

em

otio

nal

yell

ex-

tra

ct

instr

um

en

t

tou

ch

ca

rry

hois

t

cre

ate

insta

ll

dra

w

sm

oke

wea

pon

me

nta

l

se

rve

phy

sic

al

su

ffe

r

!ire

[s

yru

p]

go

lab

[ro

se

-w

ate

r]

roqæ

n [o

il]qæ

za

[fo

od

]

!æm!ir

[s

wo

rd]

sig

ar

[cig

are

tte]

hæ!i!

[h

ash]

pip

[p

ipe

]

qæ

lya

n

[hoo

ka] in

ha

le

næ

fæs

[bre

ath

]

xæ

miy

aze

[y

aw

n]

hæ

ftir

[sho

tgun

]

kæ

ma

n

[bow

]

otu

[ir

on]

lif

[loufa

]

jaru

[b

roo

m]

æng

o!t

[fin

ge

r]d

æst

[han

d]

bar

[ca

rgo]

mæ

hm

el

[ca

rria

ge

]

pæ

rde

[curt

ain

]b

ad

ba

n

[sa

il] ælæ

m

[ban

ne

r] div

ar

[wall]

jade

[ro

ad]

sim

[w

ire

]

lule

[p

ipe

]ka

bl

[ca

ble

]

næ

q!i

[desig

n]

xæ

t [lin

e]

æbru

[e

ye

bro

w]

!eih

e

[ne

igh]fæ

ryad

[y

ell]

dad

[s

ho

ut]

hæ

var

[sho

ut]

her

[la

ug

h]

hura

[h

orr

ay]

nale

[m

oan

]

zæ

jr

[su

ffra

nce

]

goro

sn

eg

i [h

un

ge

r]

sæ

dæ

me

[inju

ry]

te!n

eg

i [th

irst]

!ærm!æ

ri

[sha

me]

færa

qæ

t [s

epa

ratio

n]

hæ

sræ

t [jea

lousy

]

sæ

xti

[ha

rdsh

ip]

DRAW: HOISTælæm ke∫idæn banner hoist a flagbadban ke∫idæn sail prepare the sail of a boatpærde ke∫idæn curtain pull the curtains

DRAW: CREATEjædvæl ke∫idæn curb build curbs on sidewalks and aquaductsdivar ke∫idæn wall build a wall (usually to enclose or separate)jade ke∫idæn road build a roadsorahi ke∫idæn baluster build a balustrade or guard rail (protect)

DRAW: INSTALLkabl ke∫idæn cable lay down cableslule ke∫idæn tube install pipessim ke∫idæn wire stretch a wire, installbænd ke∫idæn masonry joint grout masonry jointszeh ke∫idæn gut string string a bow, string a tennis racket

DRAW: DRAWæbru ke∫idæn eyebrow draw in eyebrowsnæq∫e ke∫idæn plan plan, scheme (fig. & lit.)næqa∫i ke∫idæn drawing drawnil ke∫idæn blue paint draw tracery on mosaic, pots, etc.rimel ke∫idæn mascara apply mascarasorme ke∫idæn collyrium anoint the eyes with collyrium, eyelinervæsme ke∫idæn indigo leaves dye eyebrows with woad or indigo leavesxæt ke∫idæn line draw a line, cross outqælæm ke∫idæn pen cross out, elimiate

SUFFER: SUFFERarezu ke∫idæn wish wish futilelydærd ke∫idæn hurt suffer paindærde sær ke∫idæn hurt of head go through difficultyentezar ke∫idæn waiting long forjæfa ke∫idæn oppression be oppressedmælamæt ke∫idæn blame be blamed, suffer blamemehnæt ke∫idæn misfortune suffer a hardshipmennæt ke∫idæn favor implore, beg for a favorrænj ke∫idæn suffering endure a sufferingsæxti ke∫idæn hardship endure hardshipsetæm ke∫idæn oppression endure oppressionxærhamali ke∫idæn drudge do drudgeryzæhmæt ke∫idæn effort work hard with endurancezæjr ke∫idæn torment sufferzærær ke∫idæn loss bear a loss

SUFFER: PHYSICALbigari ke∫idæn chore suffer from not being paid for laborgorosnegi ke∫idæn hunger suffer hungerqolle ke∫idæn strong thirst endure thirstsædæme ke∫idæn injury suffer, experience hardshipte∫negi ke∫idæn thirst suffer from thirst

KE∫IDÆN 231

SUFFER: MENTAL∫ærmsari ke∫idæn shame be ashamedælæm ke∫idæn pain sufferæzab ke∫idæn torture suffer pain or punishmentfæraqæt ke∫idæn separation endure separationhæsræt ke∫idæn regret long for, wish nedamæt ke∫idæn regret suffer regretsqæm ke∫idæn sorrow grieveqorbæt ke∫idæn distant place be homesickxejalæt ke∫idæn shame be embarrassed

EMIT: EMOTIONALah ke∫idæn sigh sighher ke∫idæn laugh laugh mockinglyhura ke∫idæn hooray cheer onmerno ke∫idæn cat's call meow under sexual heatnale ke∫idæn groan groanqeyhe ke∫idæn excited yell to yell excitedly in joy or anger

EMIT: ANIMAL∫eihe ke∫idæn neighing neighzuze ke∫idæn howl howl, whimper

EMIT: YELLdad ke∫idæn yell yellfaryad ke∫idæn yell yellhærdud ke∫idæn attack call attack while yellinghævar ke∫idæn shout cry outjar ke∫idæn proclamation proclaim unimportant factsjiq ke∫idæn scream scream (see zædæn)næ’re ke∫idæn roar roar threateninglyqiye ke∫idæn cry cryya hu ke∫idæn name of the lord asking for lord's attention

EMIT: UNCLASSIFIEDmut∫ ke∫idæn kissing noise attract by making kissing noise∫o'le ke∫idæn flame flame, blazehort ke∫idæn sipping sound slurpnotoq ke∫idæn speech talk, speak (same as notoq zædæn)sæfir ke∫idæn whistle make a whistling sound (arrow, bullet, cannon ball…)sut ke∫idæn whistle be surprised by a large billzæbane ke∫idæn tongue, wedge shoot up (flames)

PULL: INHALEbu ke∫idæn smell smell, sniffdæm ke∫idæn moisture be infused, steeped, or preparednafas ke∫idæn breath breathexamiyaze ke∫idæn yawn yawn

KE∫IDÆN 232

PULL: SMOKEhæ∫i∫ ke∫idæn hash smoke hashpip ke∫idæn pipe smoke a pipeqælyan ke∫idæn hookah smoke a hookahsigar ke∫idæn cigarette smoke cigarettest∫opoq ke∫idæn pipe smoke a pipevafur ke∫idæn opium pipe smoke an opium pipe

PULL: WEAPON∫æm∫ir ke∫idæn sword draw a sword∫ax∫an'e ke∫idæn horn/rib bone take a threatening posture (fig. & lit.)hæftir ke∫idæn pistol pull out a gunkæman ke∫idæn bow draw a bowkard ke∫idæn knife pull a knife on someoneneyze ke∫idæn javelin ready to hurl a javelinnimt∫e ke∫idæn musket draw a musketqæddare ke∫idæn dagger threaten with a daggert∫aqu ke∫idæn knife pull a knife on someonetiq ke∫idæn blade cut something with a blade, draw a sword

PULL: EXTRACThærf ke∫idæn letter coerce to speakgolab ke∫idæn rose-water extract rose-waterroghæn ke∫idæn oil drip oil out of somethingkar ke∫idæn work overwork someonekeyfær ke∫idæn reward be rewarded

PULL: SERVEmey ke∫idæn wine pour winepeymane ke∫idæn wine drink winepiyale ke∫idæn glass (wine) drink offqæza ke∫idæn food serve food onto a plate from a pottæ’am ke∫idæn food serve food∫ilan ke∫idæn royal feast throw a royal feast

PULL: INSTRUMENTkise ke∫idæn exfoliating glove exfoliate with a cleaning gloveqæltæk ke∫idæn roller flatten with a roller (dough, asphalt, roofing…)jaru ke∫idæn broom sweep with a broomlif ke∫idæn sponge clean with a spongeotu ke∫idæn iron iron

PULL: TOUCHængo∫t ke∫idæn finger effacedæst ke∫idæn hand pull hand across, extricate from a taskpa ke∫idæn foot pull foot across

PULL: CARRYbar ke∫idæn cargo, baggage carry a load, lit. & fig.mæhmel ke∫idæn carriage carry a carriage on ones shoulder

KE∫IDÆN 233

PULL: UNCLASSIFIEDfærahæm ke∫idæn collected gather, assemble, wrinklenæm ke∫idæn moisture become dampbirun ke∫idæn out pull outdændan ke∫idæn tooth pull out a toothdær ke∫idæn in draw in, holddær bær ke∫idæn in side embrace, take someone into one’s bosomforu ke∫idæn into pull down, sink in, inhalepæs ke∫idæn behind pull backpayin ke∫idæn down pull downpi∫ ke∫idæn ahead bring up (a subject)qor’e ke∫idæn lottery decide a lotteryt∫ek ke∫idæn check, draft write a check on an accountnure ke∫idæn depilatory apply a depilatory creammil ke∫idæn applicator needle anoint with collyrium

UNCLASSIFIEDænan ke∫idæn rein stop a process, call a haltbe gænd ke∫idæn to rottenness pollutingdamæn ke∫idæn skirt move off, withdrawdamæn dær ke∫idæn skirt in pull up skirtderaz ke∫idæn long lie downgærdæn ke∫idæn neck rebel, revoltleqam ke∫idæn rein stop a processmot∫ ke∫idæn wrist drag someone alongnaz ke∫idæn demand adoration satisfy all of s.o.'s whimspa ke∫idæn foot avoid a placeqædæm ke∫idæn step cease to visit a placesæræk ke∫idæn head (dim) check surreptitiouslysix ke∫idæn skewer skewer (kabob)sine ke∫idæn chest land rising in a slope facing the sun kenar ke∫idæn side pull asidesemat ke∫idæn table cloth spread a table clothab ke∫idæn water rinseægæb ke∫idæn back retreatbala ke∫idæn up embezzlebesær ke∫idæn onto head drape over ones head (blanket, veil, scarf)dærhæm ke∫idæn confused pucker, frown, contractenteqam ke∫idæn vengeance take vengeancehæm ke∫idæn each other draw togetherkæbadeh ke∫idæn chained exercise bow to dance with a kæbadehkin ke∫idæn revenge take vengeancekoduræt ke∫idæn indignation be indignantlæ∫kær ke∫idæn army mobilize and drive the armypær ke∫idæn feather take flightqæba ke∫idæn cloak wear a cloak (religiousqæd ke∫idæn size, length grow tallersær ke∫idæn head gulp downtænure ke∫idæn chimney rise in a columns (fire, smoke, cloud, fog…)tir ke∫idæn shot hurt with a shooting paintul ke∫idæn length take a long timevær ke∫idæn side pull up (shoes), to get serious about s.t.be dændan ke∫idæn to tooth taste, eatsæf ke∫idæn row line up

KE∫IDÆN 234

happenin

g

ÆN

DA

XT

ÆN

[to th

row

]

insta

ll

appeara

nce

pre

positio

n

ju! [b

oil]

fro

m a

sta

te

ca

st

t!e!m

[e

ye

]tæ

k o

tab

[fe

ren

ic

action

]

tæm

æ'

[co

ve

tousn

ess]

kar

[wo

rk]

geri

ye

[cry

]

go

d

[dee

p]

topo

logy

t!u

ruk

[wrin

kle

]

pit!

[tw

ist]

sp

rea

da

ble

ph

ysic

al

ab

[wate

r]

t!o

[ru

mo

r]læ

k [sta

in] x

æt

[scra

tch

]

situ

ation

fort

une

gir [

tra

p]

exte

laf

[diffe

rence

]

hæ

t!æ

l [t

rap

]

xu

b

[go

od

]

bad

[b

ad

]

qor'e

[lot]

ja [p

lace

]

!i!e

[w

indo

w]

po

sitio

ndu

r [f

ar]

æqæ

b

[beh

ind]

shin

e

move

me

nt

nu

r [lig

ht]t!e

raq

[lig

ht]

ne

ga

h [lo

ok]

t!e!m

[e

ye

]

to a

sta

te

sim

[w

ire]

insta

llspre

ad

cro

wd

ve

lvele

[c

lam

our]

qolq

ole

[c

row

d]

thro

w

fær!

[ru

g]

rum

izi

[tab

leclo

th]

bæ

stæ

r [b

ed]

hit

wea

po

n

jelo

[fr

on

t] tof [s

pit]

tup [

ca

non

]

navæ

k [

arr

ow

]

læg

æd

[kic

k]

mo!t

[fist]

t!æ

ng

[scra

tch]

shednaxon

[n

ail]

obje

ct

pust [s

kin

]

tas [d

ie]

tæsbih

[r

osa

ry]

toxm

[se

ed

s]

bæ

t!e [

ch

ild]

CAST: MOVEMENTt∫e∫m ændaxtæn eye cast eyes on, looknegah ændaxtæn look glance at

CAST: SHINEnur ændaxtæn light shine light on somethingsaye ændaxtæn shadow cast a shadowt∫eraq ændaxtæn light shine light

THROW: POSITIONrah ændaxtæn way start (e.g. a car)æqæb ændaxtæn behind delay, push backhæva ændaxtæn sky throw upwardjelo ændaxtæn front change (e.g. an appointment) to a sooner onepi∫ ændaxtæn front advancezæmin ændaxtæn ground throw down, dropdur ændaxtæn far throw away

THROW: HITmo∫t ændaxtæn fist punchlægæd ændaxtæn kick kickt∫æng ændaxtæn scratch scratch, snatch

THROW: WEAPONkæman ændaxtæn arrow shoot an arrownavæk ændaxtæn arrow shoot an arrowtir ændaxtæn bullet shoot a guntof ændaxtæn saliva spit on somethingtup ændaxtæn cannon fire cannons atxænjær ændaxtæn dagger stab with a dagger

THROW: OBJECTkæmænd ændaxtæn lasso trow a lassolængær ændaxtæn anchor drop anchortas ændaxtæn die roll the dicetasbih ændaxtæn rosary slide beads on a rosaryab ændaxtæn water to irrigate (a field)

PREPOSITION: TO A STATEbe dam ændaxtæn entrap trapbe tæmæ' ændaxtæn covetousness make covetbe gerieh ændaxtæn weeping drive to tears

PREPOSITION: FROM A STATEæz ju∫ ændaxtæn from boil stop from boilingæz pa ændaxtæn of feet wear outæz t∫e∫m ændaxtæn from eye worsen the opinion towards something

ÆNDAXTÆN 237

APPEARANCE: PHYSICALxæt ændaxtæn line scratchlæk ændaxtæn stain cause a stain, develop a stainxæra∫ ændaxtæn scratch scratch

APPEARANCE: SPREADINGab ændaxtæn water make water (e.g. mouth)xun ændaxtæn blood cause to bleedt∫o ændaxtæn rumor spread rumors

APPEARANCE: TOPOLOGYpit∫ ændaxtæn twist cause a twist to appeart∫uruk ændaxtæn wrinkle cause to become wrinkledgod ændaxtæn deep dish cause a shallow indentation

HAPPENING: CROWDvelvele ændaxtæn clamor cause a tumultqolqole ændaxtæn commotion cause a commotion

HAPPENING: SITUATIONgerate ændaxtæn obstacle make problemsgir ændaxtæn hold to betrayhæt∫æl ændaxtæn ditch put into hard to escape situationbæd ændaxtæn bad make something go wrongtæfræqe ændaxtæn discord create discord

HAPPENING: FORTUNEpe∫k ændaxtæn lots throw lots, lotteryxub ændaxtæn good cause good luckbæd ændaxtæn bad cause bad luckqor’e ændaxtæn lottery cast lots

INSTALL: INSTALL∫i∫e ændaxtæn glass install a windowja ændaxtæn place fit something in its placejam ændaxtæn glass glaze a windowpine ændaxtæn patch sew on a patchsim ændaxtæn wire string

INSTALL: SPREADfær∫ ændaxtæn rug spread a rugbæstar ændaxtæn bed prepare a bedsofre ændaxtæn table-cloth spread out food on a clothlo ændaxtæn rumor spread rumors

SHEDbæt∫e ændaxtæn child have an abortionpust ændaxtæn skin shed skin (snakes, or sun burned humans)naxon ændaxtæn finger nail lose a nail

ÆNDAXTÆN 238

UNCLASSIFIEDpit∫ ændaxtæn twist, curve cause difficulties∫elæng ændaxtæn long strides walk rapidlyælæm ændaxtæn flag declare defeatæng ændaxtæn essence be badly in needbær ændaxtæn on abolish, overthrowbæsat ændaxtæn items display ones ware for saledær ændaxtæn inside get someone entangled in a skirmishdæst ændaxtæn hand make fun ofgærdan ændaxtæn neck decapitategol ændaxtæn flower redden (e.g cheeks)jer ændaxtæn tear cause discordkola ændaxtæn hat give inlegam ændaxtæn rein put on the controlsmot∫ ændaxtæn wrist arm wrestlepænje ændaxtæn paw attempt to conquerpæs ændaxtæn behind give birth (derogatory)pelos ændaxtæn rough mat live vagrantlypo∫t hæm ændaxtæn back to back weave liespær ændaxtæn feather admit one’s weakness, grow weakru ændaxtæn face show boldnessvær ændaxtæn side abolish, overthrowabru ændaxtæn eyebrow lift an eyebrowtoxm ændaxtæn seeds implant, inseminategerye ændaxtæn tears bring to tearsbænd ændaxtæn string remove hair using a stringbar ændaxtæn load unloadfetne ændaxtæn sedition sew seditionlerd ændaxtæn impurity give up impurities as sedimentslong ændaxtæn towel throw in the towelqælæt ændaxtæn mistake lead to errorsepær ændaxtæn shield throw in the gauntlettowq ændaxtæn necklace develop a stain ringvaj ændaxtæn prayer pray for someone

ÆNDAXTÆN 239

happenin

g

OF

TA

DÆ

N[to

fall]

attitude

appeara

nce

pre

positio

n

ju!

[bo

il]fr

om

a s

tate

cast

t!e!m

[eye]

tæk o

ta

b

[fe

renic

a

ctio

n]

tæm

æ'

[co

ve

tou

sne

ss]

kar

[work

]

ge

riy

e [

cry

]

god

[d

eep

]to

polo

gy

t!uru

k

[wrin

kle

]

pit!

[tw

ist]

spre

ad

ab

le

phy

sic

al

ab

[w

ate

r]

t!o [

rum

or]

t!o [

rum

or]

læk [

sta

in] x

æt

[scra

tch]

situa

tion

fort

un

e

pe

yka

r [b

attle

]

exte

laf

[diffe

ren

ce

]

hæ

t!æ

l [tra

p]

xub

[g

oo

d]

ba

d [

ba

d]

qo

r'e

[lo

t]

læj

[gru

dg

e]

ne

da

mæ

t [r

em

ors

e]

positio

n

dur

[fa

r]

æq

æb

[b

ehin

d]

jod

a

[se

pa

rate

]

sh

ine

hap

pe

n

upo

n

nur

[lig

ht]t!era

q [lig

ht]

aft

ab

[s

un

sh

ine]

neg

ah [

look]

gozæ

r [p

assa

ge

]

t!e!m

[e

ye

]

to a

sta

te

cro

wd

velv

ele

[c

lam

ou

r]

qo

lqole

[c

row

d]

ATTITUDElæj oftadæn grudge develop a grudge, develop a disliket∫æp oftadæn left develop a grudgenedamæt oftadæn remorse attack of remorse

CAST: HAPPEN UPONgozær oftadæn passage happen to pass byt∫e∫m oftadæn eye look uponnegah oftadæn look glance

CAST: SHINEt∫eraq oftadæn light be shone uponaftab oftadæn sun covered by sunshinenur oftadæn light shine onto

CAST: POSITIONdæmær oftadæn flat fall flatzæmin oftadæn earth fallæqæb oftadæn back fall back (schedule)birun oftadæn out isolateddær pey oftadæn in pursuit follow, get afterdur oftadæn far left behindfæraq oftadæn far become separatedjelo oftadæn front fall aheadjoda oftadæn separate become separatedpæs oftadæn behind fall behindpi∫ oftadæn in front move aheadqælæt oftadæn mistake happen, fall in wrong placedonbal oftadæn pursuit pursue, followja oftadæn place settle in, fit, properly cookedrah oftadæn path start working, leaving, start walking

PREPOSITION: TO A STATEbe gerieh oftadæn weeping be driven to tearsbe tæmæ’ oftadæn covetousness become glutonouskar oftadæn work start operating (organization, factory,…)kam (be) oftadæn palette to fall into prey's (mouth) traptæk o tab (be) oftadæn ferenetic action become disheartenedtæmæ' (be) oftadæn covetousness become covetous

PREPOSITION: FROM A STATEæz ju∫ oftadæn from boil stop boiling, calm downæz kar oftadæn from work break downæz pa oftadæn from foot run out of energyæz pa dær oftadæn from foot in run down, to grow extremely feebleæz pærde be dær oftadæn out of the veil a secret revealedæz qælæm oftadæn from pen forget to account for an itemæz t∫e∫m oftadæn from eye fall from high esteemtæk o tab (æz) oftadæn ferenetic action act out of desperationtæmæ' (æz) oftadæn covetousness loose one's desire

OFTADÆN 243

APPEARANCE: PHYSICALxæra∫ oftadæn scratch get scratchedlæk oftadæn stain develop a stain, get stainedxæt oftadæn line get scratched

APPEARANCE: SPREADINGt∫o oftadæn rumor spreading of rumorsxun oftadæn blood start to bleedlit∫ oftadæn slime become covered with slimeab oftadæn water drool (person), over ripened (fruits)

APPEARANCE: TOPOLOGICALgod oftadæn deep sink, (usually, sunken eyes)pit∫ oftadæn twist develop difficulties, curl up (hair)t∫uruk oftadæn wrinkle become wrinkled

HAPPENING: SITUATIONgir oftadæn hold be captured, get stuckhæt∫æl oftadæn well, trap run into severe difficultiespeykær oftadæn battle a (unexpected) battle brew upetefaq oftadæn event happenehtiyaj oftadæn need become necessaryextelaf oftadæn difference develop disagreementspit∫ oftadæn twist develop difficulties, curl up (hair)

HAPPENING: FORTUNEgerate oftadæn obstacle run into problemsqor’e oftadæn lottery win a lottery, get luckytelesm oftadæn magic spell face mysterious hurdles

HAPPENING: CROWDvelvele oftadæn clamor uproarhæras oftadæn fear engulfed in fear (society, a group)qolqole oftadæn commotion become restless (crowd)

UNCLASSIFIEDmæqbul oftadæn accepted be acceptedmoæser oftadæn effective to have a positive effectqæbool oftadæn accept become acceptablegere oftadæn knot encounter a hurdlekæj oftadæn crooked take a bad turn, become obstructivepa oftadæn foot, step lucky trunqætæk (bær) oftadæn roller a process becoming facilitatedtuye dæhæn oftadæn in mouth become rumoredvær (see bær) oftadæn side become unfashionabledær oftadæn in get entangledbær oftadæn side become unfashionable, dissolved (gov, org)naf oftadæn navel loosing the umbilical cord

OFTADÆN 244

DA

!TÆ

N[h

ave]

attr

ibut

e

cons

ider

inh

eren

t

æzi

z [d

ear]

mo'

æza

m

[hig

hest

est

eem

]

qæni

mæ

t [w

indf

all]

pose

ssextiy

ar [c

hoic

e]

væqt

[tim

e]

peyq

am

[mes

sage

]

decl

arat

ion

tran

sien

t

orze

[abi

lity] sæ

br [p

atie

nce]

qeyr

æt

[hon

or]

tærd

id

[hes

itanc

e]

arez

ou [w

ish]

mey

l [ap

etite

]

hærf

[wor

d]

ebra

z [d

ecla

ratio

n]

doxæ

n [s

peec

h]

tra

it

stat

e of

m

ind

tab

[sw

ing]

!ohr

æt [

fam

e]zu

r [fo

rce]

CONSIDERæjab da∫tæn strange consider strangeaziz da∫tæn dear hold deargerami da∫tæn dear to hold dearhormæt da∫tæn reverence regard with reverencejayez da∫tæn appropriate deem appropriatemahsub da∫tæn counted consider consideredmo'æzam da∫tæn highest esteem hold in high esteemmo'af da∫tæn exempt hold exemptmohkæm da∫tæn strong hold stronglyna∫tiz da∫tæn worthless deem worthlessnang da∫tæn shame hold an act shamefulnaxo∫ da∫tæn ill take it as a bad signnik da∫tæn good take as good omenqæbul da∫tæn assent maintain as trueqænimæt da∫tæn windfall to consider s.t. a windfallræva da∫tæn accepted deem acceptablevajeb da∫tæn necessary hold requisitexo∫ da∫tæn happy take it as a good sign

POSESSIONbaqi da∫tæn remainder have remainder (a story, a math division)bimar da∫tæn patient to have (care of) a sick relativeedrar da∫tæn urine having to urinateejaze da∫tæn permission have permissionetela da∫tæn information know, have awarenessextiyar da∫tæn choice have the choiceezn da∫tæn permission have been permittedforsæt da∫tæn occasion have a chance govah da∫tæn witness have witnessja da∫tæn place have a place to stay, or in s.o. heartkar da∫tæn work be busylæqæb da∫tæn title be known as…mæjal da∫tæn chance get a chancemæ'moriyæt da∫tæn mission given a missionmærqum da∫tæn written have it writtenmæskæn da∫tæn residence residemul da∫tæn mistress have a mistress on the sidenæsib da∫tæn share own a shareniaz da∫tæn need needpænah da∫tæn harbor be harboredpayam da∫tæn message have a messagepeyqam da∫tæn message have a messageqærar da∫tæn appointment stand, sit (the cup sits on the table)tæsælot da∫tæn control be in control oftimar da∫tæn mental illness have a mental patientvæqt da∫tæn time have a chance, have time

DECLARATIONærze da∫tæn display offer for sale, displaydoxæn da∫tæn speech have something to sayebraz da∫tæn declaration declaree'elam da∫tæn announcement announce, proclaimesrar da∫tæn insistence be insistenthoj'æt da∫tæn plea argue for a pleahoj'b da∫tæn forbidden declare s.t. as forbiddensohbæt da∫tæn speech have things to say

DA∫TÆN 247

TRAIT∫ekæm da∫tæn belly be convex∫ohræt da∫tæn fame be famousæhami'æt da∫tæn importance be importantænjomæn da∫tæn association be organizede∫aræt da∫tæn reference be well knowne∫tehar da∫tæn notoriety be well knownedame da∫tæn continuation have continuationertebat da∫tæn connection be connectedesalæt da∫tæn nobility be of noble descentesteb'dad da∫tæn dictatorship act dictatoriallyet'tesa da∫tæn expansiveness have multi-meaninget'tesal da∫tæn connection be connected (cross-connected wires)extesas da∫tæn specialty ear markedfæ'aliæt da∫tæn activity be activegonah da∫tæn sin vulnerable, sinful to mistreatgonjaye∫ da∫tæn capacity capability for a task or knowledgehozur da∫tæn presence be presentkefa'æt da∫tæn equality be equalkefayæt da∫tæn sufficiency be sufficientkeyfiy'æt da∫tæn quality be well qualifiedliaqæt da∫tæn deserving be deserving oflozum da∫tæn necessarries be neededmæharæt da∫tæn expertise be an expertmoqayeræt da∫tæn contradiction be in contradiction to (law, logic, sense,…)næm da∫tæn moisture be moistnætije da∫tæn result be fruitfulnævæson da∫tæn oscillation oscillatenemud da∫tæn appearance stand outniyabæt da∫tæn deputized decreed to act in place ofpayab da∫tæn stamina strongpey da∫tæn foundation have a basispeykar da∫tæn follow-up be assiduouspo∫te kar da∫tæn assiduousness be assiduousqabeli'æt da∫tæn fitting be fit forqobn da∫tæn defraud be defraudedræmæq da∫tæn energy have the required energysæfa da∫tæn cleanliness be refreshingsæfavab da∫tæn good deed count as a pious actsal da∫tæn age be oldsaman da∫tæn home be setteledsar o kar da∫tæn dealing have a dealing withsepas da∫tæn gratitude be thankfulsud da∫tæn profit be profitable, be good fort∫am da∫tæn order be orderly, prosperoustab da∫tæn curl have curlstæ∫æxos da∫tæn notoriety be well respectedtæ'adol da∫tæn balance be well balancedtæb da∫tæn fever run a fevertæfavot da∫tæn difference be differenttæhædob da∫tæn convexity be convextæhæmol da∫tæn forbearance be forbearingtæ'm da∫tæn taste have good tastetæmæ'vol da∫tæn wealth be wealthytæqædom da∫tæn priority have prioritytæzad da∫tæn contrast have contrast, self contradictingvæj'he da∫tæn high position trustworthy

DA∫TÆN 248

vagir da∫tæn contagion be contagiousvir da∫tæn desire have a strong desirevojud da∫tæn existence existzarfiyæt da∫tæn capacity be capable enoughzur da∫tæn power be powerful

TRANSIENT ∫ærm da∫tæn shame be shameful of∫egoft da∫tæn surprise be in awe ofa'dæt da∫tæn habit be habituated toæza da∫tæn mourning be in mourningarezou da∫tæn wish wish, be wishfulbak da∫tæn fear be fearfulbavær da∫tæn belief believedelalæt da∫tæn guidance suggest (one fact suggests another), point toderiq da∫tæn attachment unwilling to part withe∫q da∫tæn love admiree∫teha da∫tæn appetite have appetitee∫tiaq da∫tæn eagerness be eagereba'e da∫tæn refusal refrain with disdainehteram da∫tæn respect be respectfulehtiyaj da∫tæn need be in need oferadæt da∫tæn admiration be a frienderade da∫tæn will be determinedeta'æt da∫tæn obedience be obedientetminan da∫tæn confidence be confident ofgæman da∫tæn opinion suspect, have an ideahosele da∫tæn patience to have patience forjædal da∫tæn fight be feistyjorræt da∫tæn courage be bravematæm da∫tæn mourning be in mourningmeyl da∫tæn appetite have appetite formosælæm da∫tæn certain be certain of a factmul da∫tæn hesitation be hesitantomid da∫tæn hope be hopefulozr da∫tæn excuse to show and excuseqæm da∫tæn sorrow be sorrowfulsorfe da∫tæn cough have a coughtæ'æsob da∫tæn prejudice be prejudicedtæmæ' da∫tæn gluttony have excessive desiretænæfor da∫tæn hatred hatetæqaza da∫tæn request requesttærdid da∫tæn hesitancy be hesitanttæsmim da∫tæn decision have decidedtævæqo' da∫tæn expectation expecttæzælzol da∫tæn weakness be weakenedtahayor da∫tæn astonishment view with astonishmenttakædor da∫tæn offense be offendedvoquf da∫tæn knowledge knowxæbær da∫tæn new be informed of the new, know a factxiz da∫tæn uprising, swelling ready to pounceyæqin da∫tæn certainty be certain

DA∫TÆN 249

INHERENT∫ærm da∫tæn modesty be modestædæb da∫tæn manners be well manneredærzani da∫tæn generosity give generouslyazærm da∫tæn piety be kind, pious, respectful, generousbolhævæsi da∫tæn capriciousness be capriciousgozæ∫t da∫tæn forgiveness be generous, be forgivinghæng da∫tæn personality have many positive characteristicshæras da∫tæn fear be fearfuljæriyan da∫tæn flow have movement (e.g. a lake, body of water)jedi'yæt da∫tæn seriousness work seriouslykeyf da∫tæn intoxication to give pleasurenæxvæt da∫tæn narcissism be narcissisticorze da∫tæn ability be capableostovar da∫tæn solid keep with principalsqæræz da∫tæn bad intention have bad intentsqebahæt da∫tæn shame be shameful (an act, a deed)qeflæt da∫tæn negligence be inattentiveqeyræt da∫tæn honor be protective of ones honorqorur da∫tæn pride have excessive pridesæb'r da∫tæn patience be patienttakæbor da∫tæn arrogance to act arrogantlytaqæy da∫tæn patience have endurancezærær da∫tæn defect, loss be hazardous, be loss producing

UNCLASSIFIEDærz da∫tæn scroll have a long requestæyadi da∫tæn hands have an involvmentamæd da∫tæn appearance of good omenar da∫tæn disdain disdain (az in kar ar dasht)bær da∫tæn above pick upbær jay da∫tæn in place maintainbær pa da∫tæn on feet, standing to organize and hold a celebaration or meetingbar da∫tæn load be pregnantbaz da∫tæn open preventbe yad da∫tæn to memory rememberbozorg da∫tæn large hold in high esteemdæst da∫tæn hand be involved, influencedel da∫tæn heart be sensitivedur da∫tæn far keep apart, advise againstdust da∫tæn friend like, loveehtemal da∫tæn probability be probableejri da∫tæn ration be in military (ration eater)ersal da∫tæn send sendfæda da∫tæn sacrifice give in sacrificefæra da∫tæn over hold in high position, esteemfæraqæt da∫tæn separation have free time (from work)fareq da∫tæn separation hold apartfayede da∫tæn payback usefulforu da∫tæn into hold downgæ∫ da∫tæn body essence imbalance of body essencesguv bændi da∫tæn nefarious deal partake in a nefarious dealhajæt da∫tæn need have a request from the lordhal da∫tæn liveliness feel welliman da∫tæn belief be religious, be a believerkæmin da∫tæn ambush wait in ambushke∫ik da∫tæn guard duty be on guard duty

DA∫TÆN 250

kermæk da∫tæn wiggles itching to cause troublelazem da∫tæn needed to have a need formæbzul da∫tæn bestowed to bestowmæfuf da∫tæn concatenate drop the last letter of some Arabic wordsmæhfuz da∫tæn safe keep safemæhrum da∫tæn deprived deprive s.o.mæhzun da∫tæn sorrowful make sadmærdom da∫tæn people be sociablemætva da∫tæn tanglees curls caught in a tanglemæxfi da∫tæn hidden keep hiddenmæzbut da∫tæn sealed hold under sealmæ'zur da∫tæn one with excuse hold s.o. without guilt due to an excusemanzur da∫tæn notion mean something by inferencemaye da∫tæn starter be profitable (leave some capital behind)miyane da∫tæn relation to have a relationmoa∫eræt da∫tæn social interaction be sociablemonafat da∫tæn discord not get alongmoqædam da∫tæn front runner give deferencemoqarar da∫tæn required to impose requirementsnæhan da∫tæn hidden keep secretnæzær da∫tæn look to have a plan for s.t.nan da∫tæn bread be self sufficientnefræt da∫tæn disgust hold in disgustnegæran da∫tæn worried cause worrinessnegah da∫tæn look keepnesbæt da∫tæn relation be relatedpær o bal da∫tæn feather and wing be agilepærva da∫tæn worry, concern not act due to worrypenhan da∫tæn hidden keep hiddenpeyda da∫tæn apparent believepu∫ide da∫tæn covered keep covered, keep secretqædæm da∫tæn step bring good luckqærabæt da∫tæn separation avoid intimate relationsræbt da∫tæn connector be connectedræft o amæd da∫tæn go and come have frequent contact or visits withreqbæt da∫tæn relish relishru da∫tæn face be cheekysabok da∫tæn light weight assume something less worthy than it ist∫æ∫m da∫tæn eye covett∫am da∫tæn coquettish play hard to gettælatom da∫tæn rough waves be unsettledtæmayol da∫tæn bent be predisposedtæqdim da∫tæn offering offer as gifttætæmeh da∫tæn leftover to have leftoverva da∫tæn open preventvær da∫tæn side pick upyad da∫tæn memory knowzæhre da∫tæn bile fearless

DA∫TÆN 251

bring

fort

h

AV

ÆR

DÆ

N[to

brin

g]

exp

ressio

n

bring

abo

ut

ph

ysyca

l

tra

it

ha

l [h

ealth

]

dæ

rd

[pain

]

xo! [g

ood

]

ba

d [

sw

ell]

xa

b

[sle

ep]

ju!

[boi]

bæ

nd

[b

locke

d]

appeara

nce

phy

sic

al

sta

te

of

min

d

taqæ

t [to

lera

nce

]

ete

qad

[b

elie

f]qe

ira

t [p

ride

]

fort

un

e

xu

b [

go

od

]

næ

hs [

ba

d]

nik

[g

oo

d]

positio

nen

tity

em

otion

to

positio

n

gir

[g

rasp

]

næ

zd

ik

[clo

se

]

payin

[d

ow

n] b

e d

ony

a

[to

worl

d] be d

æst

[to

ha

nd

]

be

hu!

[to

co

nscio

usn

ess]

sæ

fa [jo

y]

læk

[sta

in]

t!æ

rbi

[fa

t]pu

st

[skin

]æ

ba

[a

bh

ora

nce

]

bæ

t!e

[ch

ild]

em

tiya

z [p

oin

t]

miv

e [

fru

it]

elæ

t [e

xcu

se

]

ho

jæt

[re

aso

n]m

esa

l [e

xa

mple

]

POSITION: POSITIONpayin aværdæn down lowerbala aværdæn up vomit, throw upfæraz aværdæn again bring backforud aværdæn down cause to descendgerd aværdæn round accumulate, bring togethergir aværdæn hold bring into one’s grasp, findhæm aværdæn together bring to closure, close a woundjæm'e aværdæn assembly gather together, assemble togethergozar aværdæn passage pass bypæs aværdæn behind return a purchased articlepi∫ aværdæn front bring up (for discussion), table an issuefærahæm aværdæn collected bring togetherhojum aværdæn mass rush (of the masses), attack for pillage

POSITION: TObe dæst aværdæn to hand obtainbe donya aværdæn to world give birthbe ja aværdæn to place recognize (I finally placed him… (English))be meidan aværdæn to square prepare something for work, proposebe næzær aværdæn view imagine, (o dærehaye himaliya ra be nazar avard)be næzm aværdæn to order rewrite a story if form of a poembe sær aværdæn to head bring to a head or an endbe tæng aværdæn press press s.o.'s patiencebe xater aværdæn to mind rememberbe yad aværdæn to memory recallbe zæban aværdæn to tongue saytaxt (be taxt) aværdæn gallop bring to a gallop

APPEARANCE: FORTUNEtas aværdæn dice having good like with dicedæst aværdæn hand have a good hand in a card gamebæd aværdæn bad have bad luck

APPEARANCE: STATE OF MINDtaqæt aværdæn staying powerendure difficultieseteqad aværdæn belief come to believe innam aværdæn name become famousnaz aværdæn coy act coy or coquetish (less common than naz kærdæn)qeiræt aværdæn honor become honor bound

APPEARANCE: PHYSICALgu∫t aværdæn meat gain weightab aværdæn water excrete serum from an infection, tear from an infectionlæk aværdæn stain develop a dark spont or patch on the skinpih aværdæn cellulite develop cellulite lumpspust aværdæn skin develop a scab on a woundt∫ærbi aværdæn fat develop fatty lump

BRING ABOUT: EMOTIONæba aværdæn abhor abhorsæfa aværdæn joyousness bring joyxo∫ aværdæn pleasure please

AVÆRDÆN 255

BRING ABOUT: TRAITke∫ aværdæn elastic stretch out (an elastic)bænd aværdæn blockage blockbad aværdæn swell be affected with swellingvær aværdæn separation ferment doughfeshar aværdæn pressure put pressure onzur aværdæn force put under pressure, push hardju∫ aværdæn boil bring to a boil (water, soup)

BRING ABOUT: PHYSICALxun aværdæn blood cause bleeding (scratch until you bleed, be-xaroon ta xoon biyari)dærd aværdæn pain to bring pain, cause painhal aværdæn moment bring joy, make the momenttæb aværdæn fever cause a fever, develop a fever

BRING ABOUT: UNCLASSIFIEDhæmle aværdæn attack wild attack (said when the target is easily within grasp)yad aværdæn memory remind, jog s.o. memoryræhm aværdæn mercy act mercifulræhmæt aværdæn mercy act with compassion

BRING FORTH: ENTITYtohfe aværdæn unique gift give a cheap gift (said in derision)emtiaz aværdæn point score a point, win a point, gain a pointbæt∫e aværdæn child have a childmive aværdæn fruit bear fruit

BRING FORTH: EXPRESSbæhane aværdæn pretext make excuses to evade (evade reposibility)elæt aværdæn excuse give excuseshojæt aværdæn reason give logical explanationmesal aværdæn example give an examplepasox aværdæn answer convey an answerozr aværdæn apology give an excuse to avoid (avoid an invitation)govah aværdæn witness cite as evidenceqæriv bær aværdæn cry on cry out

UNCLASSIFIEDpænah aværdæn refuge take refuge, reach for protectionpædid aværdæn visible bring to life, create- bær aværdæn on bring up, cause (qoqa, færyad, poost, sæfir, qæriv, pineh, abele)- bær aværdæn casue (commotion) or be forced into a different state- dær aværdæn in sar-discover a secret, dom-go wild, dæbe-renege, bazi-fool around- dær aværdæn get onself into a different state, surprise - dær aværdæn seda, næva, hærf, (make up sound, melody, rumor)- dær aværdæn shæx (horn) be so utterly amazed- dær aværdæn quz (hunch) develop a hunched back, gol (to flower)∫abixun aværdæn night attack surpriseæz pa dær aværdæn from foot in defeat by wearing downæz xod dær aværdæn from self in make up (stories)boz aværdæn goat have a cow (have bad luck)dær aværdæn inside bring outja aværdæn place cook properlykæm aværdæn few run short ofqobar aværdæn dust become dim (e.g. eyes)ruy aværdæn face ask for support, to appeal for help, be masjed rooy avardan

AVÆRDÆN 256

sær dær aværdæn head in learn the secret of, understandsepele∫k aværdæn snake eye have a bad turn (luck)tab aværdæn patience put up with, remain patienttæng aværdæn tight need to go to the bathroomtæ∫rif aværdæn pomp grace with one's presence, arrive, visitbar aværdæn fruit raise, nurture, cultivatepayin aværdæn foot become active and inquisitivesær aværdæn head be in a hurried rush sær bær aværdæn head on come into being, generally out of place (become an upstart)pey aværdæn foundation learn a secret (vs. pey bordan, happen onto a secret)vared aværdæn arriving impose something (usually neg.)lam aværdæn dervish belt pretence of piety

AVÆRDÆN 257

tra

it

AM

ÆD

ÆN

[to c

ome]

pro

pert

y

act

ove

rcom

e w

ith

phy

syca

l

he

sita

tion

qelq

elæ

k

[tic

kle

]

dæ

rd

[pa

in]

zu

r [e

ffort

]

xo! [g

oo

d]

heyf

[cha

rity

]

xab

[s

leep

]

ju! [b

oi]

e!v

e

[co

que

try

]

t!e!m

o æ

bru

[e

ye a

nd

eybro

w]

poz [

sh

ow

off

]

bæ

nd

[blo

cke

d]

kæ

m

[little

]

appeara

nce

xun

[blo

od

]

bu [sm

ell]

em

an

atio

n

na

ture

bad

[w

ind]

bara

n [ra

in]

ze

lzele

[e

art

hqu

ake]

fort

un

e

xub

[g

oo

d]

næ

hs [b

ad

]

nik

[go

od

]

positio

n app

ea

r

pæ

did

ar

[app

are

nt]

pe

yda

[fo

und

]

be

co

me

piru

z

[vic

tori

ou

s]

fa'e

q

[exce

llen

t]

naye

l [a

chie

ve

r]

turn

o

ut

vare

d

[kno

win

g]

laze

m

[ne

ce

ssa

ry]

ka

rgæ

r [w

ork

er]

em

otio

n

hæ

sud

i [jea

lousy

]

to

fro

m

po

sitio

n

ke

na

r [s

ide

]

næ

zdik

[c

lose

]

pa

yin

[d

ow

n]

be

don

ya

[to

wo

rld

] be

dæ

st

[to

han

d]

be h

u!

[to c

on

scio

usne

ss]

æz p

æs

[fro

m b

ehin

d]

æz p

ey

[fro

m p

urp

ose

]

æz p

a d

ær

[fro

m f

oo

t in

]

meyl

[app

etite

]a

ar

[dis

da

in]

ACTe∫ve amædæn coquetry to act coquettishlypoz amædæn attitude to show offqæpi amædæn exaggeration self aggrandizeqer-o-qæmze amædæn dance and wink act coyly and evasivelyqopi amædæn brag to bragt∫e∫m-o-æbru amædæn eye and eyebrow to act coquettishlyt∫osi amædæn fart self aggrandize, make empty threatslap amædæn cheat, renege renege on an agreement, cheat

POSITION: POSITIONgir amædæn hold be obtainable (nune xub gir miayæd?)næzdik amædæn near draw nearpayin amædæn low move lower (temp., prices, person...)kenar amædæn side step aside, come to an agreementsær amædæn head end

POSITION: TObe dæst amædæn to hand to be obtainedbe donya amædæn to world be bornbe hoosh amædæn to aware regain consciousnessbe jan amædæn to life be pushed to extremitiesbe kar amædæn to work become usebe meidan amædæn to square be prepared to workbe qeyræt amædæn to zealousness become zealousbesær amædæn to head reach the limit (endurance, patience, life, forebearance)be tæmæ amædæn desire become voracious, gluttonousbe tæng amædæn to tight become impatientbe xod amædæn to self come to, recoverbe zir amædæn to below come down, fallbehæm amædæn to each other go well with each otherbesær dær amædæn to head in to fall

POSITION: FROMæz pa dær amædæn from foot in break downæz pæs amædæn from behind follow behindæz pæs bær amædæn from behind able to manage s.o. or s.t.æz pey amædæn from purpose pursuit

POSITION: UNCLASSIFIEDændær amædæn within exitgerd amædæn round convene, assemblejæm amædæn assembly gather in a group (reflexive)færahæm amædæn collected become availableba bad amædæn with wind dropping into ones lap (free)bær amædæn above be capable of, come offbær sær amædæn on head to happen, befallbahæm amædæn together come together, unitebaz amædæn again returndær amædæn in come out to be+adj.,to exit, to be solvedfæra amædæn over, upper transcend, overcomeforu amædæn into come downgerde hæm amædæn around other assemble for a meetinghæm amædæn each other close up, bind togetherpæs amædæn behind come back, returnpi∫ amædæn ahead occur

AMÆDÆN 261

APPEARANCE: APPEARpædid amædæn visible appear, originatepædidar amædæn visible appear, originatepeyda amædæn visible be foundbahu∫ amædæn intelligent seem intelligentnapæsænd amædæn unacceptable appear unacceptablepæsænd amædæn admiration become acceptablegozær amædæn passage come to passæjæb amædæn strange appear strangeqærib amædæn strange appear weird

APPEARANCE: FORTUNEnik amædæn good turn out goodvæqt amædæn time be a blessed time for a deednæhs amædæn bad have bad fortune

APPEARANCE: NATUREbad amædæn wind blowing of windbaran amædæn rain to rainzelzele amædæn earthquake earthquake occur

APPEARANCE: EMANATIONbu amædæn smell wafting odorxun amædæn blood bleed

APPEARANCE: UNCLASSIFIEDneda amædæn call receive spiritual messagegovah amædæn witness be affirmed, as by God's callinggereftar amædæn stuck run into difficultymane' amædæn impediment put up an impediment

OVERCOME WITH: EMOTION∫ærm amædæn shame beholden to shamexo∫ amædæn good turn out wellbæd amædæn bad dislike/ bad turn of eventsmeyl amædæn inclination become favorably inclined

OVERCOME WITH: HESITATIONaar amædæn disdain disdain to actdæriq amædæn thrift unconvinced to act or shareekrah amædæn aversion resist out of aversionheyf amædæn waste be reluctant to let gokerahæt amædæn squeamish become squeamishzæhmæt amædæn hardship feel disinclinedgonah amædæn sin feel morally unacceptable

OVERCOME WITH: PHYSICALdærd amædæn pain hurtqelqelak amædæn tickle be tickledxab amædæn sleep feel drowsy

AMÆDÆN 262

OVERCOME WITH: UNCLASSIFIEDyad amædæn memory recall∫egeft amædæn wonder wonderajez amædæn desperate run out of patience

TRAIT: TURN OUTdo∫var amædæn difficult end up difficult to [do, accept, workout…]dærxor amædæn worthy be worthy of s.o.govar amædæn digestible turn out healty or deliciouskarbær amædæn applicable end up being usefulkargær amædæn worker end up being functionallazem amædæn necessary turn out to be key, essentialnafe'e amædæn useful become usefulpor amædæn full turn out knowledgeablevajeb amædæn imperative become imperative

TRAIT: PROPERTYbænd amædæn blockage cease to flowbæs amædæn sufficient sufficehal amædæn spirit be rejuvenatedja amædæn place be restored (hale∫ jan amæd)ju∫ amædæn boil reach boiling point - compared to zædænkæm amædæn few run shortke∫ amædæn elastic stretchqaser amædæn failing fall short, fail expectationsbæsandeh amædæn fitting fit the purpose

TRAIT: BECOMEfa’eq amædæn excellent excelnayel amædæn achiever overcome, achievepiruz amædæn victorious become victoriousmoqer amædæn confessor be compelled to confessionvaqef amædæn aware become aware of

UNCLASSIFIEDkutah amædæn short accept a compromise (accept less than initial position)maye amædæn starter seed rat on, insinuate against, (plant seed of resentment)rah amædæn path cooperate, go alongsir amædæn satiated become fed upvær amædæn side come off, rise (for dough fermentation)færaz amædæn above, height arriveforud amædæn descend come down, descendgæ∫n amædæn male have a rut, to rutqaleb amædæn conqueror prevailqeyræt amædæn honor become defensive about ones honoroxt amædæn intimacy get used tovared amædæn arriving befall a bad occurancegeran amædæn dear tolerate reluctantlybar amædæn yield to be raisedgæzænd amædæn damage sustain damage or injury

AMÆDÆN 263

BO

RD

ÆN

[to c

arry

]

from

si

tuat

ion

wea

r of

f

ræng

[c

olor

]

taqæ

t [pa

tienc

e]

mæ

qz

[bra

in]

dire

ctio

n

prep

ositi

on

emot

ion

to

bene

fit

from

pæs

[ret

urn]

pi! [

ahea

d]

foru

[dow

n]

sud

[pro

fit]

be s

ær

[to h

ead]

be k

ar

[to w

ork]

be q

æni

mæ

t[to

loot

] æz

bein

[fr

om b

etw

een]

æz

pi!

[from

ahe

ad]

æz

miy

an

[from

bet

wee

n]

susp

ect

zen

[judg

emen

t]gæ

man

[s

uspi

cion

]

!æk

[dou

bt]

ers

[inhe

ritan

ce]

næf'

[ben

efit]

lezz

æt

[ple

asur

e]

rænj

[s

uffe

ranc

e]

hæsr

æt

[reg

ret]

capt

ure

exte

rnal

self

gero

gan

[hos

tage

]

gero

[col

late

ral]

del [

hear

t]

xab

[sle

ep] mat

[aw

e]

requ

ire

eteq

ad

[faith

]væ

qt [t

ime]

ener

ji [e

nerg

y]

DIRECTIONforu bordæn down insert, stab, immersejelo bordæn ahead carry forwardpæs bordæn behind return for refundpi∫ bordæn ahead make progress, take ahead

REQUIREenerji bordæn energy require energyeteqad bordæn belief require dedication or beliefvæqt bordæn time take time

PREPOSITION: TObe kar bordæn to work usebe qanimæt bordæn as loot lootbe sær bordæn to head pass, spend time

PREPOSITION: FROMæz miyan bordæn from between destroyæz pi∫ bordæn from forth accomplishæz rah be dær bordæn to door corrupt s.o. (idiom)æz bein bordæn from between destroy

FROM SITUATION: EMOTIONhæsæd bordæn envy suffer in envyhæsræt bordæn regret long for, wish lezzæt bordæn pleasure enjoy, take pleasure inrænj bordæn sufferance suffertæmæ'e bordæn greed have greedxejlæt bordæn shame be shamed

FROM SITUATION: BENEFITers bordæn inheritance inheritnæf’ bordæn benefit get a benefitnatije bordæn result learn a lesson, get what's deservedsud bordæn profit make a profit, take a windfalltæmæto' bordæn profit benefit, take advantage

SUSPECT∫æk bordæn suspect to be suspicious of somethinggæman bordæn superficial judgment suspectzen bordæn subjective judgment become suspicious

CAPTURE: SELFmat bordæn puzzled become bewilderedxab bordæn sleep fall asleep

CAPTURE: EXTERNALdel bordæn heart steal someone’s heartgero bordæn collateral confiscate as collateralgerogan bordæn hostage take hostages

BORDÆN 267

WEAR OFFgu∫ bordæn ear of talk someone's ear offmæqz bordæn brain talk s.o.'s brains outræng +DO bordæn color remove the tint, make pale, (fade as in the Sun)sær + DO bordæn head of talk someone's head off, bore someonetaqæt +DO bordæn patience of wear s.o. patience out

UNCLASSIFIEDbu bordæn smell become suspicious, catch wind ofdær bordæn in carry, gaindæst bordæn hand modify surreptitiously in order to cheatfærman bordæn order take order, be obedientgero bordæn pledge win the stakehæmle bordæn attack attackhesab bordæn calculation be afraid ofhojum bordæn swarm rush onto to horde, mass attacknam bordæn name namepa-be-pa bordæn in lock-steps accompany closelypænah bordæn refuge seek refugepæyam bordæn message carry a messagepey bordæn foundation discover the reason for, become awarerah bordæn path guide, drivesær bordæn head drive fast, be in a hurrysær dær bordæn head in understand, make outsojud bordæn prostration kiss s.o. feet, prostrate in prayertæ∫rif bordæn pomp go away with pomp taxt bordæn gallop attack in a hordetohfe bordæn unique gift give a cheap gift (derisive)

BORDÆN 268

RÆ

FT

ÆN

[to g

o]

go

disa

ppea

r

ræng

[c

olo

r]

bu [s

mel

l]

yad

[mem

ory]

sta

te

prep

ositi

on

out

to

dire

ctio

n

from

gij

[diz

zine

ss]

hædæ

r [w

aste

]zæ

f [w

eakn

ess]

payi

n [d

own]

æqæ

b [b

ack]

næzd

ik

[nea

r]bi

run

[out

]gæ

rde!

[e

xcur

sion

]

nævæ

rd

[trav

ersa

l]

be x

ab

[to s

leep

]

be k

ar

[to w

ork]

be ta

b[to

cur

l] æz

yad

[from

mem

ory]

æz

xod

[from

sel

f]æ

z dæ

st

[from

han

d]

do

pi!v

az

[gre

etin

g]

!irje

[div

e]

nævæ

rde

neza

mi

[mili

tary

mot

ion]

DO∫irje ræftæn dive divenæværd-e-nezami ræftæn military motion exercise military manouverspi∫vaz ræftæn greet at arrival go out to welcome a guest

GO: OUTdædær ræftæn stroll (kid talk) go out for a strollgærde∫ ræftæn excursion go out for a stroll or excursionnæværd ræftæn traverse traverse, cross

GO: DIRECTIONæqæb ræftæn back retreatæqæbi ræftæn backward move backwardbala ræftæn above ascendbirun ræftæn out go outforu ræftæn into sink,kæj ræftæn crooked go the wrong waynæzdik ræftæn near go nearpæs ræftæn behind retreatpayin ræftæn down descendpi∫ ræftæn ahead make progressfæra ræftæn above transcend

GO: UNCLASSIFIEDrah ræftæn path walkdær ræftæn door escape

DISAPPEARbu ræftæn smell lose smellræng ræftæn color lose color, fadeyad ræftæn memory forget

STATEsær gij ræftæn confused become dizzydær ræftæn in, within get dislocated (a joint)hædær ræftæn waste go to wasteqæ∫ ræftæn faint laugh to the point of faintingzæf ræftæn weakness feel faint, feel famishedva ræftæn apart become weak, give upbaz ræftæn open fall apart

STATE: UNCLASSIFIEDja ræftæn place fall into placesær ræftæn head overflow

PREPOSITION: TObe kar ræftæn to work be usedbe tab ræftæn to curl curl up, slither away

PREPOSITION: FROMæz dæst ræftæn from hand lost (usually an opportunity)æz xod ræftæn from self faintæz yad ræftæn from memory be forgotten

RÆFTÆN 271

PREPOSITION: UNCLASSIFIEDbær bad ræftæn on wind be ruined, perished

UNCLASSIFIEDtond ræftæn fast act impertinentlyab ræftæn water shrinkke∫ ræftæn elastic swipe, smuggleke∫ o qows ræftæn twist and turn beat around the bushpit∫ o tab ræftæn twist and turn act evasivelyvær ræftæn side meddle with, play withxab ræftæn sleep fall asleepdel male∫ ræftæn stomach churning have hunger pangsdel zæf ræftæn stomach weakness feel faint heartedhærz ræftæn misguided be lost out of negligencemale∫ ræftæn kneed, massage spasm of an internal organ

RÆFTÆN 272

The Case of Light Verb Constructions in Persian - Relais d

Documents

Transcript of The Case of Light Verb Constructions in Persian - Relais d