The Stoic Verbal System

21
The Stoic Verbal System Author(s): C. H. M. Versteegh Source: Hermes, 108. Bd., H. 3 (1980), pp. 338-357 Published by: Franz Steiner Verlag Stable URL: http://www.jstor.org/stable/4476172 . Accessed: 01/05/2013 19:41 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. . Franz Steiner Verlag is collaborating with JSTOR to digitize, preserve and extend access to Hermes. http://www.jstor.org This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PM All use subject to JSTOR Terms and Conditions

Transcript of The Stoic Verbal System

Page 1: The Stoic Verbal System

The Stoic Verbal SystemAuthor(s): C. H. M. VersteeghSource: Hermes, 108. Bd., H. 3 (1980), pp. 338-357Published by: Franz Steiner VerlagStable URL: http://www.jstor.org/stable/4476172 .

Accessed: 01/05/2013 19:41

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Franz Steiner Verlag is collaborating with JSTOR to digitize, preserve and extend access to Hermes.

http://www.jstor.org

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 2: The Stoic Verbal System

THE STOIC VERBAL SYSTEM*

The difficulties one meets in studying Stoic grammatical theory may well be illustrated by the verbal system, whose reconstruction has been undertaken in various ways. It is our aim in this paper to study first the data provided by the Greek grammarians, and to determine the influence of Stoic theories on this corpus with the help of the direct quotation in the scholia (scholia D. T. 250, 26 - 251, 25). Then we will analyse the data from Varro (De L. L. VIII - X) and from the Latin Corpus Grammaticorum, in connection with the direct quotation by Priscianus (Inst. 414, 21 sqq.). Finally, we will compare the various reconstructions which have been proposed, and give our own proposal.

The Stoic definition of the verb (Aiga), or rather the definition given by Diogenes of Babylon in his Peri Phones, runs O5iga' toTt gpog X6you oalrntvov &C'i v9sTov xaTwnyo6pnja . An isolated verbal form - including the infinitives2 - is nothing more than a xaTly6prtla without construction (6&rnV9vTOV, cDnv'TUxTov), i.e., not consisting of a noun and a copula3. Other Stoics, among them Apollodoros, introduced a morphological element into the definition: (Ofi5i' toTt) ototX?tov X6you dnT-oTov aigtatv6v Tt oCuvTaxT6v nspit Ttv6; i TtV6Vv4. The Stoic functional definition of the verb

* I wish to thank prof. dr. J. C. F. NUCHELMANS, Nijmegen, for having read an earlier version of this paper, and for his very useful comments concerning the presentation of the material. All responsibility rests, of course, mine.

I SVF 3, 213, 32; cf. SCHMIDT, 1839, 61 -64. Cf. also the definition attributed to Dionysios Thrax by scholia D. T. 106, 6-8: piigd tart M,ig -xatily6prga rnilgivouoa (DI BENEDETTO,

1958, 207-208). 2 Cf. Apollonios Dyskolos, synt. 43, 15 - 44, 1: cIyE xai oil dltn Tf4 Zrod4 au,r6 gv

xacXouorn Opa, T6 8t niptinarI yp6ip&?1 xarly6pgra ila 4 o6aga, xai Urt Tr; dit6 tOrTov

kyxXietc,; xrX; cf. also scholia D. T. 356, 10- 11, where it is stated that the Stoics called the verb xatiy6pilga. We believe that what Apollonios intends to say is that the Stoa called finite verbs xa-r,nyopiAgata, if and in so far as they are used in a sentence; forms such as lrFpilnctCi, ypci9&, one could say, always fulfil this condition. Just as one can discuss the 6voLx >>ExoxpdTril; and the nTT()o0r >>Eo)xpadTnl(, there is also a OfAia >>ypct(PCv<( (>>ypdpsi<), as well as the xarisy6- pilga >>ypW(pct<(. The OMija may be given in the form ypdt(ptv or in the form ypdpst, since our concern is not with the morphological root, such as Varro, De L. L. 9, 102, nor with the YcvIXo- tatov jfiga, mentioned by Apollonios, e. g. adv. 129, 17, but with the semantic group rPAP -,

which may be represented by each of its members. 3 Cf. EGLI, 1967, 16- 17 >>Verbum ist ein AuBerungsteil, welcher ein unzusammengesetztes

Pradikat bezeichnetx<. Cf. SCHMIDT, 1839, 63. According to LONG, 1971, 104- 106, the complete sentence consists of a n7roat and a xatily6pnga; a verbal form taken in isolation is a odeficient lektonx< (?XT6v tXXiirEc), contrary to the common opinion that a deficient XExr6v may also be an isolated noun. Another interpretation of the term dto6vGErov: TELEGDI, 1976, 286, n. 3, who explains this term as >without an accusative complement<.

4 SVF 3, 213, 33-34; cf. SVF 2, 183 for Apollodoros' definition of the xzrialy6pnga; cf. EGLI, 1967, 17: >>ein Element einer AuBlerung, das keine Falle annimmt und etwas bedeutet, das mit einem oder mehreren (Subjekten) zusammengesetzt werden kann(x.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 3: The Stoic Verbal System

The Stoic Verbal System 339

may be contrasted with the Aristotelian one, which has an essential character: verbs signify time5. In the Techne we find a morphological description, as well as the essence of the verb: it signifies an action (tvtfpysta) or a passion (nitaog)6. This type of definition was to remain the standard one: verbs signify actions (cp6yiara); they have certain morphological properties; and they signify time7.

The definition of the verb is followed by the chapter of the accidents (netpFsi6Fva, oui4L43rix6ta). BARWICK has shown that the theory of the accidents of the noun in the Techne is, apart from a few details, based on Stoic doctrine, so that we may expect the same influence to be present in the theory of the accidents of the verb8. There is indeed a remarkable similarity between the tapsL6,Fetva of the Techne and Varro's species declinationum, which is explained by the fact that Varro's theories are strongly influenced by the Stoa9. The Techne mentions eight nrapFs6giva (tyxXiost4, 6tatoE14, st6n, ax1 tatx, &ptggoi, np6owna, %p6vot, ouvuyiat) 10: of these, only the oX1iaTa and the ouCu'yiat have no precedent in Stoic doctrine'. The most troublesome accident is that of the xp6vot; in order to understand the Stoic conception of verbal tenses, we must first look into the Stoic ideas about time.

According to Chrysippos real time - physical time - is the interval of the movement of the kosmos; time is one of the four tocof gara 12. With the void (xcv6v) it shares the property of being infinite into two directions, stretching

s E.g. Aristot. poet. 1457a 14-15: fiULa 8t (powt ouvG9kPl oTljavrtxi gteta Xp6vou oW6v glEpog, atlaivFt xaG'a6n.r6

6 Techne 46, 4 -5: ,i6td tatt Xt4tq dlnroro4 tIr6sxTtxi xp6vcov 'e xat npooailwv xcLi

diptii6v, lvipyesav j n6Go4 iaptordca. 7 To quote only two examples: verbum est pars orationis administrationem rei significans

cum tempore et persona numerisque carens casu (Charisius, 209, 24-27); verbum est pars orationis cum temporibus et modis, sine casu, agendi vel patiendi significativum (Priscianus, 8, 1; 369, 2 - 3; cf. a shorter version, 466, 11 - 12).

8 Cf. BARWICK, 1957, 36 -39. 9 Cf. BARWICK, 1957, 44sqq.

10 Techne 46, 5 - 47, 2. 1 1 As for the aXgarca: they indicate the difference between simple and compound words; the

Stoic grammarians did not include this difference within their concept of XXOL4 (declinatio). Varro's paragraphs about the compound words (De L. L. 8, 61 - 62) are ascribed by BARWICK (1957, 38- 39) to another source than the adjacent paragraphs, which have a distinctly Stoic, c. q. anomalist flavour. The division of the compound words with Charisius (194, 24; cf. Consentius, 350, 13) together with its precedent in Quintilianus' remarks on grammar (Inst. orat. 1, 5, 68) is apparently related to the Techne (30, 1 -4) and perhaps the result of Palaemon's influence in Latin artigraphy. As for the oubuyias: it is not known whether the Stoa divided the verbs into various conjugations. The system as it is found in the Techne found its way into Latin grammar via Varro's coniugationes and Palaemon's ordines (BARWICK, 1922, 236 - 237).

12 SVF 2, 331. Chrysippos' definition: SVF 2, 510; cf. CHRISTENSEN, 1962, 25-26; RIST, 1969, 273-288; GOLDSCHMIDT, 1953; SAMBURSKY, 1971, 98- 108.

22

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 4: The Stoic Verbal System

340 C. H. M. VERSTEEGH

towards the past and towards the future; the present (tv?aMoiX6;) is only an intersection of two infinities, itself without dimensions and without existence from a geometrical point of view 3. On the other hand, one has to concede that the past and the future are without real existence, and only possess an abstract, derived being (6(p artxtvat) whereas the present is really >>present<( (VD ptEl) 14. Another opposition is that between nFnFpaagvog and dn&Fpog: only the present represents a definite point in time, which makes it Es,npaG- gvo;,; the past and the future are infinite, and thus ds?tpot 15. These two terms are used for physical times, not for grammatical tenses, so that they cannot be identical with the grammatical opposition between n)ptoltivo; and

6ptaotog; otherwise, past tenses of the verb could never be Ci)ptovot 16. The precarious existence of the present time from a physical point of view

- already commented on by Aristotle 17 - entails the impossibility of present actions. Still, in everyday speech we do use present verbal forms, but these denote a grammatical tense, which differs from the physical time. This >>im- proper(< time or tense is the contraction of the last part of the past and the first part of the future, or, in the words of Priscianus: Praesens tempus hoc sole- mus dicere, quod contineat et coniungat quasi puncto aliquo iuncturam prae- teriti temporis etfuturi nulla intercisione interveniente. . ., ut si in medio versu dicam oscribo versumo priore eius parte scripta, cui adhuc deest extrema pars, praesenti utor verbo dicendo >scribo versum(<, sed imperfectum est, quod deest adhuc versui quod scribatur 18. We may conclude from this that the present verb is necessarily imperfectum (dtcXF?kt), because part of it is still in the future.

Greek grammarians recognized the necessity of a present tense in gram- mar. They argued that the present tense may be nonexistent from a physical point of view, but on the other hand it is possible to speak of the present hour, day, week, month, or year, so that in actual speech the present time is treated

13 SVF 2, 509; 2, 520. 14 SVF 2, 164, 26-27. 15 SVF 2, 520. 16 Contra COLLART, 1954, 187-188. 17 Aristot. phys. 219b - 220a; 233b 33sqq.; cf. STEINTHAL, 18902, 1, 312sqq. For Aristote-

les' theories concerning time: CONEN, 1964. 18 Prisc. 8, 52; 415; cf. Plutarchos' quotation from Chrysippos, de comm. not. 41 = SVF 2,

517. This concept of present time also found its way into Arabic philosophy, cf. Firabi's definition in his commentary on Aristoteles' Peri Hermeneias, ed. W. KUTZSCH / S. MARROW,

Bairut, 19712, 41, 2-4: >>If we take the time with a definite distance in the past from the present moment, which is the end and the beginning, and we join it with the same time in the future, and if their distance from the present moment, which is the end and the beginning, forms one single distance in past and future, and if they are taken together, then this time is the present timex" (my translation); cf. the remarks, VERSTEEGH, 1977, 75-77.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 5: The Stoic Verbal System

The Stoic Verbal System 341

as a reality 9. This grammatical present tense, which is called by the philo- sophers &xaptato; gp6vog and by the grammarians nkatvx6; Xp6vo20, is also needed when we wish to account for the endings of the present verb, which differ from the endings in the other tenses21.

In the Techne the major division of the verbal tenses is into tvCorTd,

napsXvicukg, iilk)ov. The past tense is subdivided into four categories, the present and future tenses cannot be subjected to further division, because of the uncertainty we have concerning the character of the actions which take place during these times. The four categories of the past tense are: tnepouvts-

IXt6;, 7rapwrnv.X6;, 7tapaX?LtEVo;, d6pt.ot 22. The criterium in the division of the tenses is the chronological relationship between them in terms of their distance from the present, which is used as the point of departure for the chronological frame: in the past, we first have the pluperfect, then the imperfect, and finally, nearest to the present, the perfect; the aorist does not indicate the temporal distance between the past action and the present moment, although it can indicate this distance in combination with a temporal adverb, such as dtptt, iniXai23. The six tenses form three pairs according to both morphological and semantic criteria. Both present and imperfect are durative (tpapactxic6;); they differ in that the duration of the present points towards the future, whereas the duration of the imperfect stretches towards the present24. Both perfect and pluperfect are covtsX15tC6; (Tt?1to;), the difference between them consisting in their relative distance from the present: the perfect is used for those actions which have taken place recently, the pluperfect for those actions which took place a long time ago (Xctt)25. Both aorist and future are i6ptoto;, i. e., they do not indicate any temporal

19 Scholia D. T. 559, 4 - 8; 248, 21 - 22, in keeping with a very old, popular tradition about time, which may be dated back as far as Homeros, cf. Plato, Soph. 262D; Parm. 1 51 E - 152A, cf. COLLART, 1954, 182, n. 6.

20 Choirob. 2, 12, 2sqq.; cf. also Joh. Philop. comment. in Aristot. phys. 703, 16. The grammatical present tense is discussed in the scholia D. T. 248, 13 sqq.; 249, 3 - 8; 404, 26 - 28; 559, 4sqq.; cf. also Plutarchos' critical remarks against Archedemos, de comm. not. 41 = SVF 3, 263, 14.

21 Scholia D. T. 248, 13 sqq. 22 Techne, 53, 1 - 3. 23 Scholia D. T. 249, 19 - 23. 24 Scholia D. T. 250, 5 - 14. 25 Scholia D. T. 249, 14- 16; 22-25. The relevance of the perfect for the present time is

evidently restricted to that of temporal closeness, i. e., what has also been called )>perfect of recent past<< (COMRIE, 1976, 60- 61), or >hot news perfect< (MCCAWLEY, 1971, 348 - 350); in the Grammaire Generale et Raisonnee of Port-Royal, we find the so-called ))rule of twenty-four hours<<: we say e. g. j'ai ecrit ce matin, but j'ecrivis hier, )oce qui ne se dit proprement que d'un temps qui soit au moins eloigne d'un jour de celuy auquel nous parlons(< (ed. H. BREKLE,

Stuttgart-Bad Cannstatt, 1966, 108- 109; cf. DONZA, 1967, 121 sqq.).

22

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 6: The Stoic Verbal System

342 C. H. M. VERSTEEGH

distance from the present; in the case of the future the designation &6pltcvo; is regarded as unnecessary, because it is well known that actions in the future cannot be fixed temporally26. This is the division as it is used for the indicative mood. In the case of the other moods, it seems that the dichotomy was nrapaTaT,Xrx / aUVrXtixt627.

From the terminology of this division alone one can conclude that its origin lies in Stoic grammatical theory. What is more, there is a direct quota- tion in the scholia D. T. to prove the connection. In the discussions concern- ing the treatment of the Greek verb in the Techne there is one particularly important passage, where the Stoics are mentioned explicitly, and where the following description of their doctrine is given: T6v vcaoT5ota oi Ertoxoi tvcatbTa tapaTatix6v 6pi4ovTat, 6tt napacT?;VFtTal i ci; icapsXnxiu-

o`ta xal ci; agkAovta 6 ydp Xtyov >>?7LoC<< xcai OTIt ?not'l? tl (patvct xai 6-tt otofftj cV 6v &t 1apaaTrtx6v nap(pXTIijvov iuaparantx6v. 6 ydzp Xycov >iotoitouv<< &uTt -6 ltlov ?itno;locv ?t(paivct, oti5cu 8t & iiXiipcoxctV, dXXa? inot?asi gv, v 06XLyo & Xpovcv ?i dyip u6 napcpXnt?vov nltov, u6 tInov 6XUyov 6 xati ipoaXrq(p9v 7counoct TX?1tov appq)Xrx6ra, O6v ytypapa, 6; xaXsituact napaxcijicvo; ut ̀r6 nktXroiov tXv T'v ouvI'uX?tav Tfi tv;vppysia;' 6 roivuv tvoTorb xcat tapa-tacrtx6; b;o& dt&-sXst; djq(p ouyy?-vst;, 6t6 xat

'ro; aTCoi; auwp(p6wvol; XpbCV'rat, olov rtiu'( tTutr'tov. '0 6t nlapaxciiisvo; XaXs1Tat EVcOT-r0; ouV'sItx6O;, roVTrou 6t ntap(px1n1Fvo; 6 i)iUcpouVcusXt6;

,tsi: OUV tXa-rspo0 'rl?Ft;(W 7apCi%llyral, ouy'ycVE1; xal TOt; Xapaxrilptort-

xoIt ocotyxcioti Xp6i)tsvoI tot; awoIt; (paivovTit, OlOV TtTUs(pa tT-s'rpPtv ...

'O ? di6picvro; xara' 'rfV 40topICrTiaV T(i) p 0VXXOV'rt oyUYvYiF c;) y&p TOV

nrotcow r6 iroa6v 'roID gtVTovor ltt6Optocov, oGTCo roT tno'iTjaa r6 TOI

napqxiluvou. Tov dpit ToIvuv Tr( a&opio'arq St&6ogFuvou yivTrat iapaXFit[i- vo;, olov FAoirnGa tp-ut - icco0il]xa, TOO5 &? aXQat iupoovegogFi1vou 6 67usp-

auv'usXtx6; yivc-rai, olov tnOoiria naiXat - ?t?rsotlluxctv ..; 'A6ptoro; 8t

?xlGitl nCp6; a'v-rt6ao-coXiv 'rob irapaxctgtvou xai i5iuspOuvr?Xtxoi3

6ptlO6v-rwv TroV XpvOU TTIWa, oT0 gdv Tr6 6pt auvvoot)gsvov ?XovIo;, otu

Xcy60tsvov, TroV3& tus pouv'sXtxo3 Tr6 n1arX28. We may conclude that both the Fv?sacvuo4 inepiT-rcaTx;O and the iapaxsijicvo; naparartx6; are ducXii the former in the present, the latter in the past. The tvcvcd)& csuv-rstx6o; and the napax4tcpvog auv-Xtx;Og are both -?X?cto;, again with the opposition between present and past. Moreover, they must be w' pioal?vo; , since they are said to >>determine the distance of time (sc. between the action and the present time)(< (6ptlOVT'rV To Xp6vou 'rjiga). The 60pia'To; is 'rt?to;, past, and

26 Scholia D. T. 249, 16-22. 27 Cf. STEINTHAL, 18912, 2, 303-304. 28 Scholia D. T. 250, 26 - 251, 20.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 7: The Stoic Verbal System

The Stoic Verbal System 343

indeterminate as to the distance of time, which is indicated by its name. The gllXov is also t6ptoro;, although this term does not figure in its name, since there is no determinate future, and, consequently, no dvt6StatoXi' between the F?XXov and another future tense29. Moreover, the p'XXwv is non-past, and, of course, e1txi.

According tO Dl BENEDETTO, the technical-grammatical part of the Techne does not stem from Dionysios Thrax (1st cent. B. C.), but is the result of a compilation of a later period (3rd. - 5th cent. A. D.)30. The modification of the Stoic system in the Techne, which we find also in the writings of Apollo- nios Dyskolos31, is, therefore, not an innovation of the Techne. It is also rather improbable that the innovation stems from Dionysios Thrax, since those of his opinions which are cited elsewhere follow the Stoic doctrine rather strictly32. The definition of the verb in the Techne is more or less Aristotelian, and in the division of the verbal tenses we see that the Aristotelian concept ?time< replaces the Stoic aspectual categorisation. This modification must have taken place sometime between Dionysios Thrax and Apollonios Dysko- los, i. e., probably in the first century B. C. 33.

Within the newly introduced chronological categorisation the most salient difference between the Stoic system and the system as it is found in the Techne and with Apollonios Dyskolos concerns the position of the perfect, which is no longer regarded as a perfect of the present tense (~vsotd)g ouvOtcItxO6;), but as a past tense - because lcapaxstltt aiTCt i irXpol; 3. This different approach may have been the result of a different conception of linguistic ana- lysis: the Stoic analysis took as its point of departure the shape of the words (the perfect is the present of the pluperfect, just as the present is the present of the imperfect), whereas later analysis was more concerned about the meaning

29 Scholia D. T. 251, 21-25. 30 Cf. Di BENEDETTO, 1958 - 59; cf. for a later evaluation SIEBENBORN, 1976, 69; we agree

with SIEBENBORN'S conclusion that the arguments advanced by DI BENEDETTO against the authenticity of the Techne are rather conclusive.

31 Apoll. Dysk. fragm. 78 - 84; Apollonios Dyskolos still uses the Stoic term ntctpo.Xtnvog instead of the term in the Techne, irapcXi?XkuM, which is also used by Aristoteles (Tryphon has n7appXllrtvog, fragm. 80 DE VELSEN).

32 Cf. Di BENEDETTO, 1958, 207 -208; e.g. Dionysios' opinion on the distinction between 6vogea and 1npoonyopia, his combination of pronoun and article into one part of speech, and his definition of the verb, cf. above note 1.

33 Maybe by Tryphon? The scholia D. T. report a discussion between Tryphon and an unknown Stoic (356, 21 sqq.; cf. VERSTEEGH, 1977, 48, n. 77), which shows that Tryphon was opposed to the Stoic method of categorizing the parts of speech. Tryphon's contribution to the doctrine of the criteria of correct speech is stressed by SIEBENBORN, 1976, 161 - 162. On Tryphon: cf. RE VIl A, 1, 726-744 (WENDEL).

34 Apoll. Dysk. fragm. 82, 10; the term nrnpctxFigvo4 is not used by the Stoa.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 8: The Stoic Verbal System

344 C. H. M. VERSTEEGH

of the words (concerning the past vs. concerning the present time)35. As re- gards the position of the perfect, we may also point out that in Hellenistic times the perfect gradually lost its present relevance, so that it finally replaced the aorist as the normal narrative tense (e. g. in New Testament Greek). This reinforced, of course, its past character, and caused the analysis of the perfect verbal forms as past tenses 36.

Stoic influence reached Latin grammar in two ways. Leaving aside the question of the authenticity of the discussions between analogists and anoma- lists in Varro's De Lingua Latina37 we may safely state that a good deal of primary Stoic influence is to be found in his work. The Latin grammarians, on the other hand, were dependent on the Greek school tradition, which was probably introduced in Rome by Palaemon 38. The data concerning the verbal system to be found in Varro's writings are not very extensive, but they do have a certain relevance for the reconstruction of the Stoic system. In book X of De Lingua Latina we find a complete picture of Varro's conception of the verbal system in chapters 47 -48, where he says that the three tenses of the verb legebam - lego - legam constitute an analogia coniuncta triplex: legebam: lego = lego: legam. He adds in hoc fere omnes homines peccant, quod perperam in tribus temporibus haec verba dicunt, cum proportione volunt pronuntiare39. A similar analogy is found in the tenses of the perfect: lego:legebam = legi:legeram, and so on. The point at issue is the trium temporum verba pronuntiare, in which the anomalist argument in the ninth book follows the series lego - legi - legam, whereas the analogists argue that in comparing verbal forms we should stay within one and the same category: neque ex divisione si unius modi ponas verba, discrepant inter se40. This implies that in establishing a grammatical analogy we should distinguish between infect and perfect forms.

From the foregoing it appears that Varro's principle in setting up a divi- sion of the Latin verbal tenses was a primary division into infect and perfect,

3s Cf. scholia D. T. 214, 29- 31, where it is asserted that not the morphological properties

(1taprn6liEtva), but the semantic ones (rt tftca) should form the basis for our division of the parts

of speech, cf. the discussion, scholia D. T. 356 sqq., where Tryphon is mentioned. 36 This expansion of the functions of the perfect was followed later by the final victory of the

aorist as the main historical tense, and the gradual disappearance of the perfect, which was to be

replaced by periphrastic forms in its function of perfect aspect, cf. DEBRUNNER, 1954, 121 - 122.

Concerning the differences between Stoic and later division, cf. TRAGLIA, 1956, 72-75. 37 Cf. SIEBENBORN, 1976, 6- 13. 38 On Palaemon: RE I A, 1, 596-597 s. v. Remmius, 4 (WESSNER); cf. BARWICK, 1922,

145 sqq. According to BARWICK, Latin artigraphy had been under Stoic influence until the time of

Palaemon, when it received an influx of Alexandrian doctrines. 39 Varro, De L. L. 10, 47; cf. also 9, 96. 40 The anomalists' objection is directed against the fact that the perfectum should have only

one form and the infectum only one (contra STEINTHAL, 18902, 1, 316, n.)

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 9: The Stoic Verbal System

The Stoic Verbal System 345

each divided into past, present, and future tenses. It is understandable, though, that he sometimes forgot his own principle: the character of the Latin perfect as the primary historical and narrative tense practically leads to a series lego - legi - legam. And so we find with Varro legi cum de lego decli- natum est, duo simul apparent, quodam modo eadem dici et non eodem temporefactum and ... ut est lego legi legaM41. The same kind of contradic- tion is found in his incidental use of praeteritum instead of praesens perfec- tum 42.

The verbal system followed by the Latin grammarians is more or less an adaptation of the system of the Techne to the needs of the Latin language. According to the theory of DI BENEDETTO the similarities between the two systems could be explained, if we assume that Apollonios Dyskolos was the most influential Greek grammarian in the Latin tradition43. There are three tenses according to the consensus of the Latin grammarians, praesens (instead of this term Palaemon used a literal translation of the Greek tvsato;, instans44), praeteritum, futurumr45. These are the so-called tempora naturalia, i. e., the physical times. In speech, however, the praeteritum shows a division into three differentiae, (distinctiones, species), which are said to be the tempora in declinatione verborum46. These differentiae (the 6ta(popai of the Techne) are: imperfectum, perfectum, plusquamperfectumr47, also used in combination with the term praeteritum. These terms indicate tenses, not real

41 Varro, De L. L. 8, 3; 8, 9; cf. also 9, 34. 42 Varro, De L. L. 8, 20; 8, 58-59; 9, 104. 43 Cf. Di BENEDETrO, 1959, 99. 44 Cf. BARWICK, 1922, 113; 130. 45 Char. 214, 9- 13 (168 K.); Diom. 335, 20 - 336, 7; Probus, 155, 36- 37; Don. 360, 9- 10;

(Serg.), 507, 37; Pomp. 234, 36-37; Cons. 377, 17-18; Mar. Vict., 199, 24-25; Sac. 432, 10- 13; Dos. 406, 9 - 10; Asp. 51, 8 - 9; Comm. Eins. 209, 6 - 33 (cf. 256, 17 - 29). There is only one division - connected with the name of Cominianus - with two tenses, tempus praesens and tempus praeteritum, the latter divided into three species: perfecta (absoluta), inchoativa, recordativa. According to this division the future tense does not exist as such, but it is only a mood of the present tense, the so-called modus promissivus. This doctrine is mentioned and mostly rejected by later grammarians (cf. [Serg.], 503, 29 - 504, 3; Pomp. 214, 10 - 215, 17; Cons. 374, 21 -31; Sac. 432, 20-24). We do not believe that this division with its highly idiosyncratic terminology reflects Stoic doctrine; rather, it is an individual development, apart from the main trend of Latin artigraphy.

46 Don. 360, 11- 13; Asp. 51, 9-12. 41 Diom. 335, 10 - 336, 7; Prob. 159, 39 - 160, 1; Don. 360, 11-13; (Serg.), 507, 38;

Pomp. 234, 37-38; Cons. 377, 8-12; Mar. Vict. 199, 25-26; Sac. 432, 10-13; Dos. 406, 10- 12; Asp. 51, 9-12; Comm. Eins. 209, 6-33. Another division is given by Charisius (214, 13 - 17; 168 K.): he distinguishes between four differentiae: incohativasive imperfecta (legebam), praeterita (legi), oblitterata (legeram), recordativa (legerim). The formula which introduces this passage (aliis ita disserere placuit) indicates that it does not stem from Palaemon (cf. BARWICK,

1922, 65).

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 10: The Stoic Verbal System

346 C. H. M. VERSTEEGH

times, in other words, they are a construction of the grammarians, they are artificales48.

The division into five grammatical tenses is a chronological one; the plu- perfect is repeatedly stated to be the >>oldest(< tense, quasi vetustissima in Consentius' words49. The pluperfect is used when we wish to indicate an action which took place a long time ago, just as the perfect is used when we talk about actions which have taken place only a short time ago5O. Diomedes compares the pluperfect with the Greek ~ntspauvtrX1tx64 in spite of its quite different character51. The chronological principle of the Latin division is also apparent in the alternative name for the imperfect, minus quam perfecta 12. As for the praeteritum and its three differentiae: one gets the impression that gradually the denominations imperfectum, perfectum, plusquamperfectum lost their function of differentiae of the past tense, and became names of tenses in their own right. The addition praeteritum is often left out. Diomedes explains the meaning of these terms as follows: Praeteritum enim imper- fectum est quidem praeteritum, non tamen perfectum, cum quasi praeterisse tempus adfirmamus; alioquin coepimus nec perfecimus, quasi legebam et scribebam et similia .. . perfectum etenim tempus, cum tempus quo egimus eo quod egimus finitum est. Item praeteritum plusquamperfectum, cum tempus iam pridem exactum demonstramus quo quid legimus. Hoc enim distat a per- fecto, quod superioris temporis recens potest videri actus, sequentis longa intercapedo. Id enim Graeci viiepavvreA6ixv appellant, quasi 6InSp -rv UvvTe- Aoiv-ra Xp6vov, quod nos praeteritum plusquamperfectum dicimus. At vero tempus perfectum apud nos dopiaqco xai irapaxeiguvco valet53.

The major difference between this system and that of Varro concerns the position of the futurum exactum: in modern grammars it is customary to differentiate between laudavero (indicativus futuri exacti) and laudaverim (coniunctivus perfecti), even though the rest of the inflection of these two forms is identical. As we have seen, Varro had incorporated the futurum exactum in his system of infecta / perfecta 54. But later grammarians almost unanimously agreed in making laudavero a subiunctivusfuturi55. The reason for rejecting the indicative and future character of the form laudavero was that most grammarians considered such a form as a contradiction in terms: an

48 Cons. 377, 18. 49 Cons. 377, 19. 50 Pomp. 235, 27 - 29: perfectum est quod completum est, sed paulo ante; plusquamperfec-

tum quod completum est olim; Cf. Diom. 336, 35: cum quod egimus tempore inveteraverit; Comm. Eins. 209, 32: quando actio nostra iam inveterascit.

51 Diom. 335 - 336. 52 Prob. 159, 39 - 40. 53 Diom. 335, 30-336, 17. 54 Cf. Varro, De L. L. 9, 97. 5S Asper, 551, 27-28; cf. SCAGLIONE, 1970, 63 and n. 24.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 11: The Stoic Verbal System

The Stoic Verbal System 347

action cannot be at the same time perfect and future, or at least that is what Priscianus told the Greeks who were unfortunate enough to have just such a form in their language (tstt5u'xogat): melius tamen Romani consideratafuturi natura ... simplici in eo voce utuntur nec finiunt spatium futuri56. The resulting grammatical system of the verb did not show the same symmetry as the Varronian system, which appears to have been forgotten.

The position of Priscianus differs from that of the rest of the Latin gram- marians: he studied and lectured in Constantinople (6th century), and must have had more direct contact with Greek materials. His professed intention was to adapt the writings of the two great Greek grammarians, Apollonios Dyskolos and Herodianos, to the needs of the Latin language57. His remarks in the 8th book of his Institutiones concerning the verbal tenses are very rele- vant for our aims, since he mentions the Stoic verbal theory explicitly. Priscia- nus' own division is more or less the same as that of the other grammarians. He divides the principal tenses into three: praesens, praeteritum, futurum58 The praeteritum possesses three differentiae: praeteritum imperfectum, prae- teritum perfectum, praeteritum plusquamperfectum, whose function is de- scribed as follows: Itaque quod accidit ipsis rebus quas agimus, nomen tem- pori ipsi imponimus, praeteritum imperfectum tempus nominantes, in quo res aliqua coepit geri necdum tamen est perfecta, praeteritum vero perfectum, in quo res perfecta monstratur, praeteritum plusquamperfectum, in quo iam pri- dem res perfecta ostenditur59. But Priscianus shows his interest in comparing Greek and Latin by warning us that the Latin perfect not only indicates a per- fect action, ... sed etiam pro dopikaov accipitur, quod tempus tam modo per- fectam rem quam multo ante significare potest60. Both Latin praeteritum per- fectum and Greek di6ptaoo; are indefinite tenses, because they need an addi- tional adverb of time in order to indicate the length of time that passed be- tween the action and the present time.

As for the tempus futurum, we hear from Priscianus that the future tense is related to the present tense, since they are both used for actions whose outcome is as yet uncertain: in praesenti enim etfuturo pleraque incerta nobis sunt angustissimaque est eorum cognitio nobis et dubia plerumque61. On the other hand, the future tense is also related to the praeteritum perfectum, i. e., the Latin perfect, when this has the function of the Greek ei6pioTo;C: ... nam futuro quoque tempori cognatio est cum praeterito perfecto quantum ad infinitatem temporis, hoc est ad d6picrzov. Sicut enim praeteritum perfectum

56 Prisc. 8, 38; 405, 17-19; Cf. SCAGLIONE, 1970, 65 and n. 25. 57 Cf. GLOCK, 1967, 57-58 and n. 6-7. 58 Prisc. 8, 38; 405, 8. 59 Prisc. 8, 39; 406, 1-6. 60 Prisc. 8, 54; 415, 25-27. 61 Prisc. 8, 38; 405, 13- 14.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 12: The Stoic Verbal System

348 C. H. M. VERSTEEGH

vel modo vel multo ante possum intellegere, sic etiam futurum vel mox vel multo post intellegere possumus futurum62. We may deduce from these remarks that the future tense is both ?>imperfect<< and >>indefinite<<. The present tense is conceived of in the same way as in the Stoic system: the Stoics, according to Priscianus, called this tense instansperfectum, since it represents a gradual transition from future to past tense63. The present tense indicates an action which has begun, without having been finished yet, as for instance the action of writing a verse64. This means that the present tense is >>imperfect<(; when an action undertaken in the present time has been finished, we may use the perfect tense, which in that case has the same function as the Greek napax4t[Evo0 - in Stoic terms tvForC'T auvtsFtx64 -, namely to indicate a res modo-perfecta65.

Our problem is now how to incorporate these data within a consistent whole. Various attempts at a reconstruction have been made, by METTE66, POHLENZ67, BARWICK68 (and STEINTHAL69); they tried to represent the hier- archical and symmetric ordering of the Stoic >>tenses<< by means of a tree-dia- gram. We will now discuss the arguments for and against their reconstruc- tions, referring to diagrams I to III.

In METTE'S system we find an opposition past/present, symmetrically divided into three aspects, napaa-rtx6g, ouvtXtx6;, Miptotog. The future (gtllov) is replaced by a hypothetical Avsotr4 &6ptoTo;. That the Stoa, however, did know a future tense, is shown by the remarks of the scholiast concerning the reason why the gtllcov is not called gllXcov &OptoTo;. These remarks are corroborated by Priscianus' testimony and by the Varronian division, in which the future is an essential part70. METTE'S solution would imply that the tv6a;Co5 is 6ptoeIivo;, which is not true: only the past tenses

62 Prisc. 8, 55; 416, 26 - 417, 3; cf. also 8, 40; 406, 8 - 10:futurum quoque cum incertum sit ... et infinitum utrum paulo post an multo erit, non potuit discretis quibusdamfinibus declinatio- nibus uti. We may add here that according to Priscianus (8, 55; 417, 3-4), the form scripsero should be regarded as a future, just as scribam. This means that he does not follow the opinion of the other grammarians who regard this form as a subjunctivus perfecti; cf. SCAGLIONE, 1970, 69-81.

63 Cf. above, note 17, and also the remark, Comm. Eins. 413, 8- 10: praesens imperfectum esse dicimus sicut et futurum eo quod non manet sed quodammodo transit. Unde a philosophis instans dicitur. The addition a graecis d6ptarov eo quod non stat is obviously a mistake!

64 Prisc. 8, 52; 415, 3-6; and cf. above, note 17. 65 Prisc. 8, 53; 415, 11-12; 8, 54; 415, 23-25. 66 METTE, 1952, 4, n. 4. 67 POHLENZ, 1939, 177; cf. TRAGLIA, 1956, 72. 68 BARWICK, 1957, 51-52. 69 STEINTHAL, 18902, 1, 311 -315. 70 We regard Cominianus' modus promissivus as an independent development in Latin

grammar, cf. above, note 45.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 13: The Stoic Verbal System

The Stoic Verbal System 349

Diagram I (Mette):

XPONOI

AVLOUb; ~~~~ncl(pxrpq.lvo;

sLapaxctrlX64 ouvrcAtx6; &6pto;oq napatcattx6; ouvt6Xtx6r &6picTo;

are divided according to the opposition bptopLvo;/a16ptoco;, whereas the present and the future are neutral for this aspect.

According to POHLENZ71, the Stoic tenses are divided into bptaovot and dt6ptorot. The former group is subdivided into nopaTartxoi and ouvmiktxoi, both with a present and a past variant, whereas the latter group is represented by the (reapwxyrivo;) t6picro; and the gtXXov (dt6pioro4). The Xp6vot na- pwractxot are, however, neutral for the opposition 63ptogvog/di6picro;,

71 An interesting detail in POHLENZ' expose is his reference to the Semitic aspectual system. POHLENZ points out that most of the early Stoics came from a Semitic milieu, and hence he concludes that in setting up a Greek verbal system they may have been influenced by their Semitic mother tongue: >>Die Stoa nimmt also in dieser Frage eine merkwiUrdige Ausnahmestellung ein. Wie sie dazu gekommen ist, ergibt sich . . . sofort, wenn wir an das Semitische denken. Denn der semitische Tempusgebrauch ist vom indogermanischen eben dadurch unterschieden, daB die Grundformen qtl und yqtl nur den durativen oder abgeschlossenen Charakter der Handlung bezeichnen, wahrend die Zeitstufe, die Zugehorigkeit zu Gegenwart, Vergangenheit oder Zukunft, sich erst aus dem Zusammenhang ergibt. Ohne Zweifel hat hier die Stoa semitisches Sprachempfinden in die Auffassung des griechischen Sprachbaus hineingetrageno (POHLENZ,

1939, 177- 178). The comparison is rather tempting, but nevertheless premature. In the first place, there is no Semitic aspectual system as such: one should'compare the Greek verb with some Aramaic dialect, in this case. In the second place, the grammatical analysis of a language is not influenced by the categories of one's native speech, by an instinctive )>SprachbewuBtseinxx, but rather by the categories of grammatical education in the schools, and there can be no doubt that the language which the early Stoics learned at school was Greek. Finaly, there is no consensus at all among Semitists, on how we must analyse the categories of the verbal system in a Semitic language. It is interesting to note that for a later period it has been surmised that the first Arabic analyses of Arabic verbal forms were influenced by the Greek concept of verbal tenses, cf. VERSTEEGH, 1977, 75 sqq. On the Semitic background of the early Stoics: POHLENZ, 1926; on the aspects of the Hebrew verb: RUNDGREN, 1961; BROCKELMANN, 1956, 37-39 (relevance of Greek data for the analysis of the Hebrew verb).

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 14: The Stoic Verbal System

350 C. H. M. VERSTEEGH

Diagram 11 (Pohlenz):

XPONOI

dptagtvo4 d&6p1too,

napaotzntx64 ouvVaIX64 7rapwpxnTivo 4 Ixx)v

tvsoTW5 napqX1plVog tVrT(b@4 napcX11pvoc

since they to not determine the distance of time between the action and the present. We must also point to the fact that the dOplaToCo must be categorized as a Xp6vo; OuvTs?.Xtxo6, as is proved by the remark of the scholiast concern- ing the name of this tense: tOpiotoC, (xakCTiat) tr6 Toov gt 6pi4?lv 76tET

TET?sagvov Moy? T'6 irpayj.Ca 72. We may conclude that the opposition 60 pta-

gtvo;/dt6picTog is always a subdivision of the past-perfect tenses. BARWICK divides the Stoic tenses into present, past, and future; in each

group there is a nopaTaTix6;/avsvTXltx6 opposition, as well as one repre- sentant of the Wbpltogtvo/&t6picTo4 opposition (in the present bptogtvo;, in the past &6piato;, in the future (d&6ptaTo;): >>Da alles Geschehen sich in der Zeit vollzieht, und damit in einer der drei Zeitstufen, so sind naturgemaB die beiden Aktionsarten den Zeitstufen untergeordnet<< 73. However, unless there is evidence to the contrary, there is no reason to assume a priori that the Stoics, because of their natural awareness of three physical times, used these as the basis for their verbal system. There are other arguments against BAR- WICK's reconstruction. In the first place, there is the explicit testimony of the scholiast that the g?X,o)v does not tolerate a further subdivision74, which means that there can be no opposition between a gtlXlov napacnaix6; and a gtlXcov ouvT?Xtx6 75: not only must it be said that BARWICK's hypothetical

72 Scholia D. T. 404, 7. 73 BARWICK, 1957, 51. 74 Cf. above, note 29. 75 In Greek there is nofuturum exactum, as in Varro's system; the so-calledfuturum atticum,

mentioned in the scholia under the name of sT'61iyov gtAXXcov, seems to be regarded by the scholiast as a later development, mentioned solely for the sake of curiosity. Moreover, according

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 15: The Stoic Verbal System

The Stoic Verbal System 351

Diagram III (Barwick):

XPONOI v

Avea-rcb^^ :::rxlgvo g" v

xrapatat16q - etpcr'taTtx6; &6ptoror (apaTcvrIl6;) (6&6piTorom)

ouvTltx6r; ((bpiogvo;) ouvtsXtX6u - (auvtcXXt64 ) -

two future tenses >>nicht ausdrucklich bezeugt werden<, but on the basis of our evidence we must say that the existence of two different future tenses is downright impossible. Concerning the opposition 4ptopgvo;/&6pio;o5 BAR- WICK then remarks: >>Der Beginn eines Geschehens in der Gegenwart ist nach stoischer Auffassung immer bestimmt, da die Gegenwart selber bestimmt, d. h. durch Vergangenheit und Zukunft begrenzt ist (iFnFpacgtvor,, Diog. La. 7, 141). Dagegen sind Vergangenheit und Zukunft ihrer Natur nach, von der Gegenwart aus gesehen, unbegrenzt (dnF,tpog, Diog. La. 7, 141) ... Daher ist der Beginn eines Geschehens in Vergangenheit und Zukunft an sich, ohne eine nahere Bestimmung ... immer unbestimmt, dt6ptoro,< 76. But, as we have seen above77, the opposition 6ielpor,/nts7npaogtvog is not identical with the opposition dt6ptoror/boptogtvor, so that the fact that the present time is ntcitspaagtvo; does not warrant the conclusion that the present tense is bptopitvo;. Actually, we have seen that the present tense is neutral for the opposition bptogtvog/&0ptoTo4. Moreover, it is not correct to assume that, since the past time is infinite (6itctpog), the past tense is necessarily indefinite (&0ptcnO;). This is shown by the remark of the scholiast that the aorist toi'-

to the scholiast, this future tense determines the distance of time between the future action and the present, so that it would be a >>tlkXwv

' ptogtvo4(x, cf. scholia D. T. 215, 16- 18, and also

Priscianus' remarks concerning the form re-r6Wvoett, above, note 56. 76 BARWICK, 1957, 51 -52. 77 Cf. above, note 16.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 16: The Stoic Verbal System

352 C. H. M. VERSTEEGH

aa acquires the same meaning as a pluperfect when the adverb na'iat is 78 added78

According to BARWICK the division of the Stoic verbal tenses by STEIN- THAL >>hat in der Oberlieferung keinen Anhalt und ist auch sachlich unmog- lich< 79. On the whole, STEINTHAL'S division seems to be identical with that of POHLENZ, and consequently, the same arguments can be brought against his reconstruction as against that of POHLENZ. STEINTHAL'S position is not al- together clear, though, since he states first, correctly in our view: >>Die Xp6vot sind teils 4vrFX?Ct,, teils ovTXtoI entsprechend den infectis und perfectis

Varrons s80. But he adds: >>Diese beiden Arten aber stehen gar nicht parallel, sondern sie liegen alle in der einen Linie der Zeit, und die beiden genera oder modi temporum (nicht etwa actionum oder gestorum) bei Varro, namlich die infecta und perfecta entstehen durch eine Teilung dieser Linie (ex divisione). Auf der einen Halfte lagen die xpovot 6T?rcCt, oder narpaxartxoio, auf der andren die tt)?tot oder auvreXtxot<<81. It is not clear whether STEINTHAL prefers a division according to tense or to aspect (naparctrtxoi/a/uvt.Xtxoi). However this may be, his final solution shows a division according to the opposition cbptogvog/d6pttoS;82.

In our view, the main division of the Stoic verbal tenses is that into tempo- ra perfecta and tempora infecta, i. e., into covrXtxotI and itaparartxoi. The

first group includes the aorist, the perfect, and the pluperfect; the second group includes the present, the imperfect, and the future. Within the group of the tempora perfecta there is a further division into those tenses which express and determine the relative distance of time between action and present (perfect, pluperfect), and the tense which is indeterminate for this aspect (aorist). The future shares with the aorist the indeterminateness, but for the future this feature is redundant, since there is no determinate future tense. The resulting scheme is not a symmetric one, but there is no reason at all, why it should be. To demand a symmetric result would be aprioristic. Our propo- sal seems to be the only possibility to solve the puzzle. The opposition cbptapti-

vo,/dt6platoo as we have seen above, can only be applied to past tenses, whereas a basically temporal division is out of the question, since it would make the perfect (vFvso'ri; OuvTEXu5S6g) a present tense: this is contradicted by

the discussion in the scholia, from which it is obvious that the perfect is used

78 Scholia D. T. 251, 12-14. 79 Cf. BARWICK, 1957, 52, n. 1. 80 STEINTHAL, 18902, 1, 31 1. 81 STEINTHAL, 18902, 1, 311. 82 STEINTHAL, 18902, 1, 314: >>Hiernach ist anzunehmen, daB3 man die vier oben genannten

Tempora, die beiden 6lTeXEtct und die beiden t?XEIot zusammengefal3t habe als bptogvoi, finita;

und daB man der Bestimmtheit (8tacdwpllet4) der Zeit die doptotia gegenilbergestellt habe,

eigentlich in doppelter Form als nip rjXpvoq &6picto; und als glXov d6pto4o((.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 17: The Stoic Verbal System

The Stoic Verbal System 353

Diagram IV (Versteegh):

XPONOI

lnapataTI,rx64, ovrFX,t6q

Ovaot6i4 ntCLPo'XllgtVO9 gtlWAO)V

dbpIogtvo4 6tcr

tvcar(b Wrnpfv)X11gL0vo

for past actions. In our reconstruction the perfect is a perfective tense, which is marked as a present vis-A-vis the pluperfect. There is no formal or semantic relationship between the present and the perfect, in spite of the apparent rela- tionship of the names tvsoac'b capatattx6; and tvsot6cio ooVTs.Xtx6g. This was already observed by STEINTHAL,WhO points out the absence of a pair gflXov napa-tatxo6, and >lfXZv ouvtsXtx6;. He concludes from this >>dal3 der Parallelismus der Namen tvFoTrCi, napa'artx6;, und tvFotiC9 trXlo4, usw. ein rein zufalliger, aus der Empirie absichtslos entsprungener ist, der auch eben darum nicht bemerkt ward und auch nicht einmal hinterher eine Construction veranlaB3te, weil die Tatsachen einen weiter fortgesetzten Paral- lelismus nicht begtinstigten(< 83. There is, in fact, a wide gap between the actual theories of the Stoa and our attempts to reconstruct them, which often tend to oversystematisize and to produce symmetric schemes, where nothing of the kind had been intended in the first place. The subdivision of the future tense in BARWICK's reconstruction is a good example of this phenomenon. The data from Varro support our thesis that the opposition napaatrtx6;/auvtsXlx64 formed the basis of the Stoic verbal system. The later development, represented by Apollonios Dyskolos, the Techne and the Latin Corpus Grammaticorum, temporalizes, as it were, this division.

83 STEINTHAL, 18902, 1, 315.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 18: The Stoic Verbal System

354 C. H. M. VERSTEEGH

In his short note on the Stoic tenses COLACLIDES compares his version of the Stoic system with the Indoeuropean verbal system, and he concludes that they are almost identical84. According to him the Stoic system may be recon- structed as follows:

imperfective def. perfective indef. perfective

present perfect -

imperfect pluperfect aorist

future

It is typical of most Indoeuropean languages with an aspectual system that they lack a present perfective85, and in so far COLACLIDES' identification is certainly to the point. But we must be careful not to suppose a priori that the Stoic categories rnapacntxo64 / ouv-rsXIxo6 were identical with the modern terms perfective / imperfective. This can be clearly seen with regard to the position of the future tense: as we have stated repeatedly above, this tense is neutral for the opposition definite / indefinite, and it must accordingly be categorized as an imperfective tense. Interestingly enough, as COLACLIDES points out, Modern Greek does possess a distinction between an imperfective and a perfective future tense (the latter being in this case indefinite) 86. This modern distinction is, however, irrelevant, for the Stoic verbal system, where the opposition nraparnTtx6;, / auvtEXtxo; leaves no place for a )>gXtxxv ouvtsXtx6;oq :Euv'r,Xtx6g is applied to actions in the past only, whereas nLapactattxo; can be used for actions in all three tenses87

After the completion of the manuscript of this article I received a copy of KLAUS SCHOPSDAU'S

publication >Zur Tempuslehre des Apollonios Dyskoloso (Glotta, 56, 1978, 273 -294). The reconstruction of Apollonios' theory of tenses, and particularly the fact that Apollonios did not use the opposition imperfect/aorist lead the author to the conclusion 1) that Apollonios followed a non-Alexandrian tradition, probably the Stoic system (p. 289), and 2) that the Stoic system was

84 COLACLIDES, 1966, 215, referring to KURYLOWICZ, 1964, 92- 93. 85 Cf. COMRIE, 1976, 53 about the confusion regarding the terms >>perfect(< and operfective<x

(for the perfective: ib., 16 - 21): the perfect is defined as expressing ))a relation between two time- points, on the one hand the time of the state resulting from a prior situation, and on the other the time of that prior situationo<, whereas the perfective >Andicates the view of a situation as a whole, without distinction of the various separate phases that make up that situation(<. A present

perfective does not exist in most lndoeuropean languages, although it does e. g. in Bulgarian, cf. COMRIE, 1976, 67- 70.

86 Cf. COMRIE, 1976, 95-97. 87 In other words, the meaning of the term ouvtrX1x6r, is exactly the one given by many

people incorrectly to the term >>perfective<(, namely that of )>completed(<, cf. COMRIE, 1976, 18- 19. Note that a perfective future is perfectly well possible, whereas a hypothetical >>gtxwv ouvt?c)tx6r,< would be a contradiction in terms.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 19: The Stoic Verbal System

The Stoic Verbal System 355

based on the opposition paratatik6s/syntelik6s. The only difference between SCHOPSDAU'S

reconstruction of the Stoic system (p. 293) and mine concerns the position of the future tense, which he regards as syntelik6s and aoristos (but cf. ib. p. 294, n. 37). ROLF HIERSCHE in his article >'Aspekt' in der stoischen Tempuslehre?<< (Zeitschrift fur vergleichende Sprachforschung, 91, 1977, 275 -287) is not primarily concerned with the reconstruction of the Stoic verbal system, but tries to show that the Stoics did not attach aspectual values to their division. According to HIERSCHE the main opposition is that of h6rismtnos/a6ristos, the former indicating tenses which possess a >>Quantum der Zeitx< (282 - 283), the latter the aorist and future tenses which do not possess such a )>Quantum<<. The chr6noi h6rismtnoi are enhest6s paratatikos, parOichEmenos paratatik6s, and enhest6s syntelikos. The terms paratatik6s and syntelik6s are not aspectual ones: paratatikos does not mean )>durativ<(, but is used for >>(der) Erstreckung einer Handlung in der Zeit, die unterschiedlich fur die jeweiligen Tempora bemessen wird(< (p. 285).

REFERENCES

BARWICK, K. (1922), Remmius Palaemon und die romische Ars Grammatica Leipzig, 1922 (Hildesheim, 19672).

BARWICK, K. (1957), Probleme der stoischen Sprachlehre und Rhetorik. Berlin, 1957. Di BENEDETTO, V. (1958 - 59), Dionisio Trace e la Techne a lui attribuita. Annali della Scuola

Normale Superiore di Pisa, Classe di Lettere, Storia e Filologia, 27, 1958, 169- 210; 28, 1959, 87 - 118.

BROCKELMANN, C. (1956), Hebraische Syntax. Neukirchen, 1956. CHRISTENSEN, J. (1962), An essay on the unity of Stoic philosophy. K0benhavn, 1962. COLACLIDES, P. (1966), On the Stoic theory of tenses. Quarterly Progress Report of the MIT

Research Laboratory of Electronics, 80, 1966, 214 - 216. COLLART, J. (1954), Varron, grammairien latin. Paris, 1954. COMRIE, B. (1976), Aspect. An introduction to the study of verbal aspect and related problems.

Cambridge, 1976. CONEN, P. F. (1964), Die Zeittheorie des Aristoteles. Munchen, 1964. DEBRUNNER, A. (1954), Geschichte der griechischen Sprache. 11. Grundfrage und Grundzuge des

nachklassischen Griechisch. Berlin, 1954. DONZI, R. (1967), La Grammaire Gentrale et Raisonnte de Port-Royal. Contributiona l'histoire

des id&es grammaticales en France. Berne, 1967. EGLI, U. (1967), Zur stoischen Dialektik. (diss. Bern), Basel, 1967. GLOCK, M. (1967), Priscians Partitiones und ihre Stellung in der spatantiken Schule.

Hildesheim, 1967. GOLDSCHMIDT, V. (1953), Le systeme stoicien et l'ide de temps. Paris, 1953. KURYtOWICZ, J. (1964), The inflectional categories of Indo-European. Heidelberg, 1964. LONG, A. A. (1971), Language and thought in Stoicism. In: LONG, A. A. (ed.), Problems in

Stoicism, London, 1971, 74- 113. MCCAWLEY, J. D. (1971), Tense and time reference in English. In: Fillmore, C. J. /

Langendoen, D. T. (edd.), Studies in linguistic semantics. New York, 1971, 96- 113. METTE, H. J. (1952), Parateresis. Untersuchungen zur Sprachtheorie des Krates von Pergamon.

Halle, 1952. POHLENZ, M. (1926), Stoa und Semitismus. Neue Jahrbilcher fur Wissenschaft und Jugend-

bildung, 1926, 257-270. POHLENZ, M. (1939), Die Begrundung der abendlandischen Sprachlehre durch die Stoa. Nachr. d.

Gesellsch. d. Wiss. zu Gottingen, phil.-hist. Kt., N. F., 3, 1939. RIST, J. M. (1969), Stoic philosophy. Cambridge, 1969. RUNDGREN, F. (1961), Das althebraische Verbum. Abriss der Aspektlehre. Uppsala, 1961.

23

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 20: The Stoic Verbal System

356 C. H. M. VERSTEEGH

SAMBURSKY, S. (19712), Physics of the Stoics. London, 19712. SCAGLIONE, A. (1970), Ars grammatica. The Hague / Paris, 1970. SCHMIDT, R. (1839), Stoicorum grammatica. Halle, 1839 (Amsterdam, 1967). SIEBENBORN, E. (1976), Die Lehre von der Sprachrichtigkeit und ihren Kriterien. Studien zur anti-

ken normativen Grammatik. Amsterdam, 1976. STEINTHAL, H. (18902 / 18912), Geschichte der Sprachwissenschaft bei den Griechen und

R6mern. Berlin, 18902 / 18912. TELEGDI, Zs. (1976), Zur Herausbildung des Begriffs 'sprachliches Zeichen' und zur stoischen

Sprachlehre. Acta Linguistica Academiae Scientiarum Hungaricae, 26, 1976, 267 - 305. TRAGLIA, A. (1956), La sistemazione grammaticale di Dionisio Trace. Studi Classici ed Orientali,

5, 1956, 37 - 78. VERSTEEGH, C. H. M. (1977), Greek elements in Arabic linguistic thinking. (diss. Nijmegen),

Leiden, 1977.

Apoll. Dysk. adv. - Apollonios Dyskolos, De Adverbiis. Ed. R. Schneider. Grammatici Graeci, 11, 1, 117-210.

Apoll. Dysk. synt. = Apollonios Dyskolos, De Constructione. Ed. G. Uhlig. Grammatici Graeci, 11, 2, 1-497.

Apoll. Dysk. fragm. = Librorum Apollonii deperditorum fragmenta. Ed. R. Schneider. Gram- matici Graeci, 11, 3, 1 - 288.

Asp. = Aemilius Asper, Ars. Ed. H. Hagen. Grammatici Latini, Suppl. 39-61. Asper = Asper junior (?), Ars. Ed. H. Keil. Grammatici Latini, V, 547 - 554. Char. = Charisius, Artis grammaticae libri V. Ed. C. Barwick. Leipzig, 1964 (K. = ed. H. Keil,

Grammatici Latini, 1, 1 - 296). Choir. = Choiroboskos, Prolegomena et scholia in Theodosii Alexandrini Canones Isagogicas

de flexione verborum. Ed. A. Hilgard. Grammatici Graeci, IV, 2, 1 - 371. Comm. Eins. = Commentum Einsiedliense in Donati Artem minorem et maiorem. Ed.

H. Hagen. Grammatici Latini, Suppl. 202-266. Cons. = Consentius, Ars de nomine et verbo. Ed. H. Keil. Grammatici Latini, V, 338 - 385. Diom. = Diomedes, Artis grammaticae libri III. Ed. H. Keil. Grammatici Latini, 1, 297 -529. Don. - Donatus, Ars grammatica. Ed. H. Keil. Grammatici Latini, IV, 353 -402. Dos. - Dositheos, Ars grammatica. Ed. H. Keil. Grammatici Latini, VII, 376-436. Gramm. Graeci = Grammatici Graeci. Leipzig, 1883- 1901 (Hildesheim, 1965). Gramm. Lat. = Grammatici Latini. Leipzig, 1855 - 1880 (Hildesheim, 1962). Joh. Philop. = Johannes Philoponos, Commentaria in Aristotelis Physica. Ed. H. Vitelli.

Commentaria in Aristotelem Graeca, XVI/XVII. Berlin, 1888. Mar. Vict. = Marius Victorinus, Artis grammaticae libri 111. Ed. H. Keil. Grammatici Latini, VI,

1 - 184. Plut. de comm. not. = Plutarchos, De Communibus Notitiis. Ed. M. Pohlenz/R. Westmann.

Leipzig, 19592. Pomp. = Pompeius, Commentum Artis Donati. Ed. H. Keil. Grammatici Latini, V, 95 -312. Prisc. = Priscianus, Institutionum grammaticarum libri XVIII. Ed. M. Hertz. Grammatici

Latini, I1/111. Prob. = Probus, Instituta Artium. Ed. H. Keil. Grammatici Latini, IV, 45- 192. Quint. inst. orat. = Quintilianus, Institutiones oratoriae libri primi capita de grammatica

(1, 4-8). Ed. M. Niedermann. Neuchatel, 1947. Sac. = Sacerdos, Artium grammaticarum libri tres. Ed. H. Keil. Grammatici Latini, VI,

427 - 546. Scholia D. T. = Scholia in Dionysii Thracis Artem Grammaticam. Ed. A. Hilgard. Grammatici

Graeci, I, 3, 1 - 586.

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions

Page 21: The Stoic Verbal System

Notes on Cicero 'Epistulae ad familiares' 357

(Serg.) = (Sergius), Explanationum in Artem Donati libri Il. Ed. H. Keil. Grammatici Latini, IV, 486- 565.

Serv. = Servius, Commentarius in Artem Donati. Ed. H. Keil. Grammatici Latini, IV, 403 - 448. Sext. Emp. adv. math. = Sextus Empiricus, Adversus Mathematicos. Ed. transi. R. G. Bury.

Cambridge (Mass.) / London, 1961. SVF = Stoicorum Veterum Fragmenta. Ed. J. von Arnim *. Stuttgart, 1903 (Stuttgart, 1968). Techne = Dionysios Thrax, Ars Grammatica. Ed. G. Uhlig. Grammatici Graeci, I, 1, 1 -101. Varro, De L. L. = Varro, De Lingua Latina. Ed. transi. R. G. Kent. Cambridge (Mass.)/

London, 1958.

Nijmegen C. H. M. VERSTEEGH

* Two numbers: volume, fragment; three numbers: volume, page, line.

NOTES ON CICERO 'EPISTULAE AD FAMILIARES'

1, 4, 2. Haec tamen opinio est populi Romani, a tuis invidis atque obtrectato- ribus nomen inductum fictae religionis non tam ut te impediret quam ut ne quis propter exercitus cupiditatem Alexandriam vellet ire.

After religionis S. B.* inserts idque susceptum ab aiiis on the ground that the motive of Lentulus' ill-wishers was surely hostility to Lentulus. I doubt whether this insertion is justified. In 1, 1, 3 (written a few days before our letter) Cicero says to Lentulus cui (sc. Pompeio) qui nolunt, idem tibi, quod eum ornasti, non sunt amici. There were therefore some senators who were against both Pompey and Lentulus. These people had two motives: first, to prevent Pompey from getting an army; second, to obstruct Lentulus. Cicero here assures Lentulus that their first motive was stronger than their second; he naturally would do so in writing to Lentulus.

1, 5 b, 1. Hic quae agantur quaeque acta sint, ea te et litteris multorum et nun- tiis cognosse arbitror; quae autem posita sunt in coniectura quaeque videntur fore, ea puto tibi a me scribi oportere.

agantur M: aguntur GR ea te M: ex te GR

The manuscript evidence supports an indirect question in the first sentence (and a relative clause in the second); therefore ea must either (with WESEN- BERG) be deleted as an erroneous anticipation of the following ea or be

' S. B. = Cicero, Epistulae ad Familiares, ed. D. R. SHACKLETON BAILEY, 2 vols., Cambridge, 1977.

23*

This content downloaded from 128.228.173.41 on Wed, 1 May 2013 19:41:46 PMAll use subject to JSTOR Terms and Conditions