Picard - Issues in the Glottalic Theory of Indo-European (1995)

download Picard - Issues in the Glottalic Theory of Indo-European (1995)

of 12

Transcript of Picard - Issues in the Glottalic Theory of Indo-European (1995)

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    1/12

    Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=rwrd20

    Download by: [173.93.163.225] Date: 03 June 2016, At: 06:16

    WORD

    ISSN: 0043-7956 (Print) 2373-5112 (Online) Journal homepage: http://www.tandfonline.com/loi/rwrd20

    Issues in the glottalic theory of Indo-European: Thecomparative method, typology and naturalness

    Marc Picard

    To cite this article: Marc Picard (1995) Issues in the glottalic theory of Indo-European:

    The comparative method, typology and naturalness, WORD, 46:2, 225-235, DOI:10.1080/00437956.1995.11435943

    To link to this article: http://dx.doi.org/10.1080/00437956.1995.11435943

    Published online: 16 Jun 2015.

    Submit your article to this journal

    Article views: 20

    View related articles

    http://www.tandfonline.com/doi/mlt/10.1080/00437956.1995.11435943http://www.tandfonline.com/doi/mlt/10.1080/00437956.1995.11435943http://www.tandfonline.com/action/authorSubmission?journalCode=rwrd20&page=instructionshttp://www.tandfonline.com/action/authorSubmission?journalCode=rwrd20&page=instructionshttp://dx.doi.org/10.1080/00437956.1995.11435943http://www.tandfonline.com/action/showCitFormats?doi=10.1080/00437956.1995.11435943http://www.tandfonline.com/loi/rwrd20http://www.tandfonline.com/action/journalInformation?journalCode=rwrd20

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    2/12

    M RC PIC R D

    Issues in the glottalic theory

    of

    Indo-European: The comparative

    method, typology and naturalness

    Abstract . The Glottalic Theory of Proto-Indo-European, whereby the

    traditional voiced series

    of

    stops *b, *d, g has been replaced by the

    ejectives *p',

    *t',

    *k', has been argued for mainly on typological

    grounds. What I will try to show here

    is

    that since any proposed sound

    change must satisfy the condition

    of

    naturalness every reconstructed

    segment or class

    of

    segments, no matter how typologically plausible,

    must be derivable in each daughter language in conformity with this

    fundamental principle. To illustrate this, I will present a case from

    Proto-Algonkian showing how both the comparative method and typo

    logical considerations are insufficient to help us choose whether 6 or *i

    should be reconstructed since these two factors are not necessarily de

    terminative

    of

    naturalness.

    1.

    Introduction. Although there have been different reconstructions

    of the Proto-Indo-European (PIE) stop system since the Neogrammar

    ians posited two series

    of

    plain stops and two series

    of

    aspirated stops,

    the accepted standard reconstruction is generally recognized to be the

    following:

    *p

    (*b)

    *bh

    *t

    *d

    *dh

    *k

    g

    gh

    *T

    *D

    *Dh

    As noted by Salmons,

    the

    reconstruction of basic PIE phonology is

    treated with great and justifiable reverence by historical linguists as

    one

    of

    the most monumental achievements

    in

    the history

    of

    the study

    of human language. As a result, the reconstruction is sometimes re

    garded as more solid than it actually is because even though

    the

    triumphs

    of

    the reconstruction

    of

    PIE obstruents were many, the sys

    tem was never accepted as set in stone by Specialists (1993:6).

    After having looked at some of the things that have made a num

    ber of indo-europeanists realize that the comparative method alone is

    225

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    3/12

    226 WORD

    VOLUME 46, NUMBER 2 (AUGUST, 1995)

    insufficient to insure phonetically real reconstructions, and that as a

    consequence it must be supplemented by typology, I will try to show

    that there is another crucial and indispensable factor that must be taken

    into account, one that has been, notwithstanding a few scattered allu

    sions to its desirability, almost totally neglected by historical linguists.

    This factor is naturalness.

    2 Problems with the reconstructed stop system

    o

    PIE

    Among

    the numerous criticisms that were leveled at this reconstruction, the

    two that are most often

    quoted-and

    that are remarkably similar in

    nature-come

    from Martinet

    a nd

    Jakobson. In the mid-fifties, the

    former expressed the opinion that ' 'a series

    of

    the type bh, dh, gh only

    appears to be attested in languages in which there also exist a series

    of

    voiceless aspirates, ph, th, kh (1955: 114-5). A little later, Jakobson

    observed that ' 'no language adds to the pair / t /-/d/ a voiced aspirate

    /dh/ without having its voiceless counterpart /th/ ; therefore the

    ories operating with the three phonemes / t /- /d/- /dh/ in Proto-Indo

    European must reconsider the question of their phonemic

    essence''

    (1962:532). A few years later, Emonds remarked that any quick

    inventory

    of

    stop systems in existing natural languages throws doubt

    on whether the T[raditional] I[ndo] E[uropean] stop system is even

    possible in natural language

    (1972: 110). In a nutshell, then, it

    seemed that ' 'the comparative method was at odds with known facts

    of

    language structure because the reconstructed PIE stop system

    con

    tained a typologically unbalanced structure which is not characteristic

    of natural languages (Baldi & Johnston-Staver 1989:88).

    Another problem was the apparently arbitrary absence of *bin the

    voiced stop series, something which is apparently without parallel in

    any known language which has otherwise regular stop series. In Mad

    dieson' s Patterns o Sounds (1984) where he surveys 317 phonemic

    inventories, only two show a bilabial gap but, unlike PIE, they are also

    missing another stop somewhere else. As summed up by Salmons, a

    voiced bilabial gap represents a notable oddity in the world's

    languages, even if we could identify one clear case (1993:53). A

    further anomaly was that the standard PIE reconstruction created some

    very unusual restrictions on root structures, the most notable being the

    so-called *deg constraint, i.e. the fact there were no roots containing

    two voiced stops (*DVD). Nobody seemed to be able to find any

    parallels to this kind of restriction in any known languages, so some

    thing was definitely amiss. Finally, there was the peculiar

    ''exclusion

    of the plain voiced stops from inflectional affixes and pronouns''

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    4/12

    PICARD: IE GLOTTALIC ISSUES

    7

    (Barnhard 1986:67-8), which is unusual for unmarked consonants like

    *b, *d, *g.

    Following is a sample of what phonologists have had to say about

    these types

    of

    reconstructional incongruities:

    ''Although it is necessary to make inferences about the phonetic

    systems which we reconstruct, these inferences must be solidly

    grounded in a theory

    of

    universal processes and phonetics'' (Her

    bert 1986:269).

    ''Reconstructed languages should be thought of as real languages

    in every sense

    of

    the term. This means that we should be very

    careful not to reconstruct anything that

    is

    not characteristic

    of

    language

    in

    general: our goal should be to strive for reality in our

    reconstructions (Barnhard 1986:281).

    Nothing (no event, sequence

    of

    events, constellation of prop

    erties, general law) that cannot for some good reason be the case

    in the present was ever true in the past (Lass 1986:22). Lass

    calls this the principle

    o

    pan-temporal uniformity which in turn

    yields a simple rule for reconstructive procedure, namely that

    the

    impossible never happened (1986:26)

    What all this shows, as a growing number

    of

    Indo-Europeanists

    came to realize, is that the venerable, reliable, tried-and-true method of

    comparative reconstruction was not sufficient to insure plausible pro

    tosystems. One also had to take typology into consideration if some

    kind

    of

    phonetic reality was to be achieved. This

    is

    essentially what led

    to the development

    of

    what has come to be known as the Glottalic

    Theory of Indo-European (also sometimes called the Ejective Model).

    In 1973, it certainly seemed like an idea whose time had c'ome since it

    was proposed completely independently by Hopper in the United

    States and Gamkrelidze Ivanov in the Soviet Union. There were

    minor differences, of course, but the essence of the proposal was the

    same.

    3 Typology and the Glottalic Theory

    This proposal centered on

    the hypothesis that PIE did not have plain voiced stops but rather

    voiceless ejectives

    which, phonetically, are produced with a

    glottalic

    airstream mechanism which involves the movement

    of

    pharynx air by

    the action of the glottis (as opposed to most sounds which are produced

    with a pulmonic airstream mechanism). This is why it has become

    known as the Glottalic Theory.

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    5/12

    228 WORD VOLUME 46, NUMBER 2 (AUGUST, 1995)

    The replacement of

    *0

    by *T' provides a very plausible explana

    tion for the distributional anomalies that were noted above. First

    of

    all,

    the case of the missing *b becomes much less aberrant

    if

    we substitute

    *p' since it turns out that a gap in the labial series of a language with

    glottalized stops is a common situation (Hopper 1973: 155). There is

    a very simple phonetic reason for this, as explained by Salmons:

    Ejectives involve the closure and raising

    of

    the glottis while the

    primary occlusion of the airstream is occurring . . . This has the effect

    of

    compressing the air in the chamber, which when released, provides

    the popping sound characteristic of

    ejectives. This requires enough

    compression so that the pop is

    a1 1dible

    and is thus easier to produce at

    more posterior points of articulation. The chamber created by a bilabial

    ejective is large enough-hence weak enough-that it is entirely lack

    ing in some systems (1993:51-2).

    Secondly, the singular constraint against *DVD roots gains a very

    straightforward explanation

    if

    we posit *T'VT' instead since

    ' ' in

    lan

    guages having ejectives, there is a widespread constraint against the

    cooccurrence of two ejectives in a

    root

    (Bombard 1986:68). Finally,

    concerning the fact that the traditional voiced stops are inexplicably

    missing from certain morphological categories, this

    is

    accounted for by

    the observation that cross-linguistically glottalics as a class occur

    only very rarely in inflectional affixes and pronouns'' (Bombard 1986:

    68).

    The Glottalic Theory did not gain immediate universal acceptance

    despite its apparent advantages, however. There were a number

    of

    criticisms from various quarters, and three basic types of counterar

    guments were put forth, namely that (1) *b was not absent but simply

    rare, or that /b/ gaps exist in attested languages; (2) a voiced aspirate

    series is possible without a voiceless aspirate series; and (3) typology

    has no business in linguistic reconstruction. A review of the first two

    issues would lead us too far afield, and I will not deal with them here,

    especially since they have been discussed at some length by Salmons

    (1993). I will turn directly to the third issue.

    4 Naturalness in phonological change As we have seen, the Glot

    talic Theory of PIE has been argued for mainly on typological grounds.

    In his overview

    of

    the debate that has raged around this model since its

    inception some twenty years ago, Salmons (1993) surveys some of the

    critical responses to its usefulness and viability. One

    of

    these stems

    from the view expressed by some Indo-Europeanists that typological

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    6/12

    PICARD: IE GLOTT ALIC ISSUES

    229

    considerations are useless in general and especially in the case at hand,

    and that the comparative method alone is suitable for reconstruction.

    However, it seems obvious to anyone who wishes protosegments

    to be more than simple abstractions with no phonetic reality that ty

    pology can play a supporting role to the comparative method in lin

    guistic reconstruction. As stated by Hopper, few comparative lin

    guists these days would be content with a purely algebraic account of

    sound changes. In recent years, linguists have insisted with increasing

    emphasis that language change and hypothetical proto-languages

    should not only have internal consistency, but should also conform to

    some standard

    of

    'naturalness', i.e. should have generalizable analogs

    in

    documented systems and diasystems (1973: 143).

    Having recognized that typology can serve as a complement to the

    comparative method, can we then go on to claim that it is the only such

    adjunct? This would certainly seem the view of Baldi and Johnston

    Slaver if we are to judge by their statement that 'the introduction of

    synchronic typological analysis into diachronic linguistics results from

    . . . the fact that there exists no mechanism within comparative re

    construction methodology to establish the correctness of a recon

    structed system, or the relative superiority of one proposed system over

    another' ( 1989:85). Similarly, Gamkrelidze has claimed that all recon

    structional decisions

    ' 'must

    rest wholly with typological consider

    ations (1988:6, 1989: 118). What I will try to show here is that other

    crucial factors, which seem to have been almost completely neglected

    in

    the disputation over reconstructional accuracy, must be taken into

    account.

    As I have argued elsewhere (cf. Picard 1984, 1990, 1994), any

    proposed sound change must satisfy the condition of

    naturalness

    Very

    briefly, a natural sound change is one that obeys the canons of mini-

    mality

    and

    directionality

    Thus, a diachronic correspondence between,

    say, u and Iii would automatically be deemed to have taken place in

    at least two stages, i.e. with fronting to Iii (perhaps via /tt/) and

    unrounding to Iii occurring separately, since the requirement that any

    change be minimal would normally dictate that a vowel cannot alter

    more than one feature at a time (though further conditions may be

    necessary for consonants, as explained in Picard 1994: 16-20).

    On the other hand, restrictions on the direction of sound changes

    would preclude

    us

    from positing that a correspondence between i and

    lui could have had

    Iii

    as an intermediate stage given that such a shift,

    although minimal, does not appear to be attested (for reasons that have

    yet to be determined as far as I know). Every reconstructed segment or

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    7/12

    230

    WORD

    VOLUME 46, NUMBER 2 (AUGUST, 1995)

    class o segments, then, no matter how typologically plausible, must

    be derivable in each daughter language in conformity with this funda

    mental principle. To illustrate this, I will present a case from Proto

    Algonkian which evidences the fact that typological considerations

    alone are insufficient to insure phonetically realistic reconstructions

    since they are not necessarily determinative o naturalness.

    5. Typology and naturalness

    n

    Proto-Algonkian. Bloomfield s

    reconstruction

    o

    Proto-Algonkian (P A) assumes the following set

    o

    consonants:

    *p *t

    *v

    c

    *k

    *6 *s *s *h

    1

    *m *n

    *y *w

    To my knowledge, no one has ever questioned the typological plausi

    bility

    o

    this system. All

    o

    these segments are found in Modern

    English, for example. Yet there exists a strong possibility that one

    o

    the obstruents, namely *6, could be wrong because it yields changes

    that violate the minimality constraint.

    Bloomfield s (1946) reconstruction was mainly achieved on the

    basis o four languages o the Central group, viz. Fox, Cree, Meno

    mini and Ojibwa (FCMO). One

    o

    the problems he faced in this en

    deavor was to determine the proper proto-segments for the various

    correspondences shown below:

    FOX CREE MENOMINI

    OJIBWA

    (1)

    askoteewi iskoteew eskootEEW iskotee fire

    2)

    ineniwa iyiniw enEEniw

    10101

    m n

    3)

    ayaapeewa ayaapeew ayaapEEW

    ayapee

    buck

    4) ineniwa

    iyiniw

    enEEniw inini

    m n

    5) anemwa a im aDEEm

    anim

    dog

    Sets

    1,

    2 and 3 obviously reflect PA *t, *n and *y respectively.

    For set 4, he selected 1 on the basis o the reflex ll found in languages

    other than FCMO, e.g. Abenaki /aln-/, Miami-Illinois /alenia/, Mic

    mac /lnu/ etc. The most problematic set was 5 since most languages

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    8/12

    PICARD: IE GLOTTALIC ISSUES

    23

    were found to have either /n/,

    It/

    or

    Ill

    (as in Delaware /all urn/ or Miami

    /alemwa/). The sole exception was Arapaho with /6/, i.e. /he6/

    dog ,

    and it

    is

    on this basis that 6 was reconstructed for correspondence sets

    like 5.

    In spite

    of

    the fact that, within PA, 6 fits in perfectly, there are

    good reasons for rejecting it

    as

    the direct antecedent of all the modern-

    day variants. First of all, since /n/ in this series of correspondences (5)

    is in

    all probability the reflex

    of

    an earlier

    Ill,

    the shift

    1

    In being

    widespread throughout Algonkian as evidenced by series 4, we can

    infer the following scenario:

    6

    I\

    I

    I \

    6 t I

    >n)

    Of

    the two diachronic changes this entails, 6

    It/

    poses no

    problem since it involves a single phonetic property, namely manner of

    articulation or,

    in

    other terms, the feature [continuant]. The shift *6

    Ill,

    on the other hand, must be viewed with the utmost suspicion since

    it

    is

    far from being minimal. As far as I know, no such development

    has ever been observed

    in

    any language, which should come as no

    surprise given that it would imply the simultaneous modification

    of

    no

    less than four features, viz. [sonorant], [voiced], [lateral] and [distrib-

    uted]. In other words, point

    of

    articulation, manner

    of

    articulation and

    voicing would all have to change at the same time, which is hardly

    conceivable. Moreover, it

    is

    difficult to come up with any reasonable

    intermediate stage(s) between

    6 and

    Ill.

    At this point, it

    is

    instructive to go back to Bloomfield's original

    statement concerning the origin of the /6/:/t/:/l/:/n/ synchronic corre-

    spondence. With characteristic terseness, he wrote: 'PA 6 (unvoiced

    interdental or

    lateral?)

    (1946:87) (my emphasis). Because he actually

    used the symbol [6] throughout, however, most Algonkianists seem to

    have either taken it at face value, i.e. as a voiceless interdental frica-

    tive, or to have been content to regard it as phonetically indetermi-

    nate (Goddard 1990: 102). One notable exception was Siebert who,

    while continuing to represent the protosegment with a theta, opined

    that 'Bloomfield's second surmise

    is

    the preferred solution'

    in

    that

    1*61

    was not an interdental but a voiceless lateral and the unmarked

    member of the contrast 1*11:1*61 (1975:451). Adding phonetic speci-

    ficity to these constructs, he stated that

    /*II

    [was] a voiced lateral,

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    9/12

    232

    WORD VOLUME 46, NUMBER 2 (AUGUST, 1995)

    probably apico-alveolar, and /*9/ a voiceless lateral or lateral fricative,

    probably apico-alveolar also' (1975:300).

    Without being aware

    of

    Siebert's conclusions, I set out to show

    in

    Picard ( 1984) that, on the basis

    of

    evidence both internal and external

    to

    PA proper, the most likely protosegment was *i because its dual

    lateral/fricative nature, combined with its [+anterior], [+coronal],

    [+continuant] features, could easily have led to a shift

    to Ill

    on the

    one hand, and /9/, on the other, with the latter further evolving to It/

    in certain languages,

    as

    summarized below:

    Arapaho e

    Cree, Atsina, Cheyenne, etc. t

    i

    Miami, Delaware, Abenaki, etc.

    In sum, naturalness can be seen to be an indispensable factor in

    reconstruction.

    s

    Salmons has observed,

    it

    is

    a kind of diachronic

    typology'' in the sense that when one tries to establish naturalness, one

    tries to show that a given proposed change occurs across various lan

    guages of the world in similar fashion (1993:68). That is not its only

    role, however, for it also constitutes the most accurate way of consis

    tently establishing the all-important distinction between sound corre-

    spondences and sound changes. Historical phonologists are starting to

    pay much more attention to the phonetic content

    of

    proto-segments and

    intermediate stages ofdevelopment, and such a perspective requires that

    maximal use be made

    of

    all available tools and mechanisms.

    6 Naturalness and the Glottalic Theory

    f we

    now take a look

    at

    the Glottalic Theory in terms of naturalness,

    we

    find that there are two

    important issues which closely parallel the PA situation we have just

    examined, and which have led to skepticism

    of

    this model in various

    quarters. First

    of

    all,

    as

    noted by Salmons, critics

    of

    the Glottalic

    Theory have often charged that positing glottalization means positing

    a feature for the proto-language that has disappeared from all of the

    daughter languages except Armenian (1993:43). The reason voiced

    stops were posited in the first place, of course,

    is

    that this

    is

    what most

    of

    the daughter languages have (cf. Meillet 1964:88, Baldi 1983:6-7,

    131, Watkins 1985:111).

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    10/12

    PICARD: IE GLOTIALIC ISSUES

    233

    However, there

    is

    no a priori motive for excluding the possibility

    of

    reconstructing protosegments which have changed across the board

    if the criterion of naturalness warrants it. This was certainly the case

    for PA

    *i,

    a segment for which one has to go outside the Algonkian

    family proper and into Ritwan, the other branch of the Algie family, in

    order to obtain direct evidence of synchronic manifestation from Yurok

    (cf. Picard 1984:435 -6). Another example of this type is from Proto

    Muskogean where, given Choctaw, Koasati, Hitchiti lbl and Creek

    lkl

    Haas (1947) reconstructs

    kw

    which yields the well-attested shifts of

    labial fronting and voicing in one case, and delabialization in the other.

    Better yet, there are cases

    of

    this type in well-recorded language

    families like Romance where the original Latin segment is obviously

    not reconstructed. As noted by Hall, for example, the developments

    of lui w] show either the labio-dental fricative Ivi or the bilabial

    plosive lbl in the various Romance languages (1976:59), e.g. Latin

    vocem

    >French

    voix but Spanish voz Such a development necessarily

    implies an inter-mediate bilabial fricative IBI as Hall himself suggests,

    which is the product

    of

    fricativization, and which either becomes stri

    dent or non-continuant, and it shows conclusively that

    it

    is

    thoroughly possible that a particular feature

    of

    a proto-language can

    fail to survive in any (attested) daughter language (Salmons 1993:

    44).

    The second major issue regarding naturalness and the Glottalic

    Theory involves the development of the voiceless ejectives into voiced

    stops. This seems like a pretty radical change and one that can hardly

    be conceived to have occurred in one fell swoop. As pointed out by

    Baldi and Johnston-Staver, there is such articulatory distance be

    tween explosive glottalized stops and voiced stops that a direct change

    seems phonologically impossible (1986:84).

    If

    we assume, in accor

    dance with the criterion of minimality, that any change which affects

    obstruents can never simultaneously involve more than one

    of

    its major

    phonetic properties, namely voicing, point of articulation and manner

    of articulation, then a shift of *T' to D is in clear violation

    of

    this

    constraint.

    In order to entertain the possibility of such a development, one

    must postulate one or more intermediate stages. The most plausible

    scenario would appear to be by way of implosives These are stop

    consonants which differ minimally from both

    T'

    and D in that they are

    produced with an ingressive glottalic airstream like the former, and

    voiced like the latter. Not surprisingly, according to Salmons,

    the

    change from implosive to plain

    voiced

    stops is very common and

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    11/12

    234 WORD VOLUME 46, NUMBER 2 (AUGUST, 1995)

    can be straightforwardly motivated and described, phonologically and

    also articulatorily (1993:53). Finally, those few languages such

    as

    Hittite and Germanic that have T would simply have undergone de

    glottalization.

    7 Conclusion In order to achieve the greatest possible degree of

    phonetic realism and accuracy in our reconstructions, which is some

    thing a growing number of historical linguists feel can and should be

    done, the time-honored comparative method must be supplemented by

    typological considerations and phonological naturalness. As stated

    most appropriately by Salmons, ':the

    comparative method

    and internal

    reconstruction often admit multiple reconstructions, i.e., they are often

    inconclusive.

    Typology

    like

    naturalness

    and so on, may help

    us

    elim

    inate some proposed reconstructions (1993:63) (my emphasis). Not

    only that but they will also undoubtedly help us to postulate in their

    stead reconstructions that have real phonetic substance and factuality.

    Department

    o

    Modern Languages and Linguistics

    Concordia University

    Montreal Quebec

    REFERENCES

    Baldi, Philip. 1983. An introduction to the Indo-European languages. Carbondale: Southern

    Illinois University Press.

    Baldi, Philip, and Ruth Johnston-Staver. 1989. His toric Italic phonology in typological per

    spective.

    The new sound

    o

    Indo-European: Essays

    n

    phonological reconstruction.

    Ed.

    Theo Vennemann. Berlin: Mouton de Gruyter. Pp.

    85-101.

    Bloomfield, Leonard. 1946. Algonquian. Linguistic structures o Native America. Eds. Harry

    Hoijer et al. New York: Viking Fund Publications in Anthropology 6. Pp.

    85-129.

    Bombard, Allan. 1986.

    The

    aspirated stops of Proto-Indo-European. Diachronica. 3:67-79.

    Emonds, Joseph. 1972. A reformulation of Grimm's law. Contributions to generative pho

    nology. Ed. Michael Brame. Austin: University

    of

    Texas Press. Pp. 108-22.

    Gamkrelidze, Thomas

    V

    1988.

    The

    Indo-European glottalic theory in the light

    of

    recent

    critique. Folia Linguistica Historica 9:3-12.

    Gamkrelidze, Thomas

    V

    1989. Language typology and Indo-European reconstruction. The

    new sound o

    Indo-European: Essays in phonological reconstruction.

    Ed. Theo Vennemann.

    Berlin: Mouton de Gruyter. Pp. 117-21.

    Garnkrelidze, Thomas V., and Vyacheslav V Ivanov. 1973. Sprachtypologie und die Rekon

    strucktion der gemeinindo-germanischen Verschli isse. Phonetica

    27:150-56.

    Goddard, Ives. 1990. ' 'Algonquian linguistic change and reconstruction.'' Linguistic change and

    reconstruction methodology. Ed. Philip Baldi. Berlin: Mouton de Gruyter. Pp. 99-114.

    Haas, Mary. 1947. Development of Proto-Muskogean *kw. International journal o American

    linguistics 13:135-37.

    Hall, Robert A., Jr. 1976.

    Proto-Romance phonology.

    New York: Elsevier.

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6

  • 8/16/2019 Picard - Issues in the Glottalic Theory of Indo-European (1995)

    12/12

    PICARD: IE GLOTT ALIC ISSUES

    235

    Herbert, Robert. 1986. Language universals, markedness theory, and natural phonetic pro

    cesses. Amsterdam: Mouton.

    Hopper, Paul. 1973. "Giottalized and murmured occlusives in Indo-European."

    Glossa

    7:141-

    66.

    Jakobson, Roman. 1962. "Typological studies and their contribution to historical comparative

    linguistics." Selected writings, val.

    I

    The Hague: Mouton.

    Lass, Roger. 1986. "Conventionalism, invention and 'historical reality': some reflections on

    method.

    Diachronica

    3:15-41.

    Maddieson, Ian. 1984. Patterns o sounds. Cambridge University Press.

    Martinet, Andre. 1955. Economie des changements phonetiques. Berne: Francke.

    Meillet, Antoine. 1964 [1937]. Introduction a / etude comparative des langues indo-europe

    ennes. University

    of

    Alabama Press.

    Picard, Marc. 1984. On the naturalness of Algonkian

    *i.

    International journal

    o

    American

    linguistics

    50:424-37.

    Picard, Marc. 1986. On the structure of the lower numbers in pre-P

    A.

    International journal

    o American linguistics 52:72-76.

    Picard, Marc. 1990.

    On

    the evaluation of competing analyses in historical phonology: natu

    ralness, minimality and the case

    of

    Armenian /erk ." Language Sciences 12:85-99.

    Picard, Marc. 1994. Principles and methods in historical phonology: From Proto-Algonkian to

    Arapaho. Montreal: MeGill-Queen's University Press.

    Salmons, Joseph. 1993. The glottalic theory: Survey and synthesis. (=Journal

    o

    Indo-European

    studies, Monograph series, Number 10). McLean, VA Institute for the Study of Man.

    Siebert, Frank

    T.,

    Jr. 1975. "Resurrecting Virginia Algonquian from the dead: the reconstituted

    and historical phonology of Powhatan." Studies in Southeastern Indian languages. Ed.

    James M. Crawford. Athens: University

    of

    Georgia Press. Pp. 285-453.

    Watkins, Calvert, ed. 1985. The American heritage dictionary o Indo-European roots. Boston:

    Houghton Mifflin.

       D  o  w  n   l  o  a   d  e   d   b  y   [   1   7   3 .   9

       3 .   1

       6   3 .

       2   2   5   ]  a   t   0   6  :   1   6   0   3   J  u  n  e   2   0   1   6