1

Having your cake and being it too: the syntactic, semantic, and morphological history of possession and existence in Armenian1 Bert Vaux, UWM Workshop on Armenian Syntax, Pithiviers May 24, 2005 êßå æågeraj ƒödnor udä, kålöƒå …avi gu: eßå æågeraj ƒödnor udä, kålöƒå …avi gu. The ass who eats inedible grass gets a headache. åmån miruk ónno@ äj æä: åmån miruk onno@ äj æä. Not all bearded creatures are goats. 1. Introduction (1) Benveniste 1952: “transitive perfect” : possession nora ê gorceal nora ê handerj noRA e goRtseAl noRA e hAndERdz 3sg.GEN be.3sg do-ppl. 3sg.GEN be.3sg garment ‘(s)he has done/accomplished’ ‘(s)he has a garment’ • In Benveniste’s analysis, the transitive perfect:

o is a possessive construction, parallel to the have-perfect of European languages such as French, Italian, German, Dutch, and (older stages of) English.

o is part of a split ergative system (with transitives behaving differently than intransitives), parallel to (but independent of) what we find in other languages in the region.

• Benveniste’s generalization remains influential in the theoretical syntactic literature up to the present day (cf. Kayne 1993, Thibault and Sankoff 1997, etc.).

(2) TODAY’S POINTS: 1. The possession analysis of the transitive perfect was already known by the Armenian monks

in Vienna by 1866. 2. More concrete evidence for the possession analysis occurs in Koriwn, Middle Armenian, and

at least two modern dialects, which use lexical ‘have’ to form the perfect with “transitive” verbs.

3. These perfect constructions actually distinguish unaccusatives and passives (which select ‘be’) from transitives and unergatives (which select ‘have’). [Unergatives may have the option of selecting either genitive or nominative subject; cf. Sigler on SWA.]

4. The distinction between two classes of intransitive verbs—unaccusative and unergative—shows up in several other situations in Armenian: i. compound verb formation in many modern dialects, e.g. thəmaχ aniel ‘be greedy’ vs. mayil elniel ‘wonder,

be amazed’ (Van dialect, Ačařyan 1952:189) ii. progressive aspect formation with guni in Hamshen (Ačařyan 1947:140-1) iii. formation of perfects of witnessed vs. unwitnessed events in some dialects (Grigoryan 1957:171-2)

1 Thanks to Jos Weitenberg for his assistance with the Dutch forms and the grabar corpus.

The wheel is come full circle: I am here.

King Lear V.iii.176

2

iv. formation of subject vs. object participles in nominal relative clauses in Standard Western Armenian (Sigler 1997)

v. selection of active vs. passive meaning with the -man participial suffix in the New Julfa dialect (Ačařyan 1940)

vi. analytic vs. synthetic causative formation in Modern Armenian (Comrie 1981:182) 2. Indo-European Background (3) unim ‘have’ • Has an Indo-European history (*h3ôp-ne- ‘obtain’; cf. Skt. ¶pnoti ‘obtains’ (Meillet 1929),

Latin opes ‘wealth, possessions’, Hittite happinant- ‘rich’) • Odd infinitive—unenal (first appears in Middle Armenian, apparently) • Takes accusative object, yet belongs to -i- conjugation (normally associated with

intransitives) • Suppletive: the aorist stem is kal- (cf. kal-uac ‘possession, property’; cf. also Karabagh

(Vankh) inÏ-kyal- ‘own’, literally ‘self-have’; cf. English my own book…) • Has a related derivative ånd-unim ‘receive’ (ånd- ‘to’) • Its descendant in one of the modern Karabagh dialects selects a dative subject and first-

person verb agreement: inj onem ‘I have’ (literally 1sg.DAT have-1sg), etc. (Grigorjan 1957:262)

3. The “Transitive Perfect” 3.1. Classical Armenian (4) Standard description: split ergative perfect construction, with transitives taking genitive subject and intransitives taking nominative subject.

• Čhaləχean 1886:159: the -eal participle takes a nominative subject especially when the verb is čhezokh ‘neutral’2 or kraworakan ‘passive’.

• Aytånean 1866.2:96-97: “The modern European languages seem to parallel our 12th-century language, using have with transitive verbs and be with intransitives and middles.”

(5) Some examples of Classical Armenian perfects a. transitive with auxiliary ‘be’ oæ i…ê ån†er…eal jer zor arar Dawi† otSH itsHe əntHERtsHeAl dzER zoR ARAR dAwitH neg be.sj read.ppl 2pl.gen acc.rel did David ‘have you not read what David did?’ (Matthew 12:3) b. transitive without auxiliary ew no…a †o@eal va@va@aki zgorcisn gna…in zhet nora Ew notsHA tHo“eAl vA“vA“Aki zgoRtsisn gnAtsHin zhEt noRA and 3pl.gen leave.ppl immediately acc.net.accpl.def went after 3sg.gen ‘and they immediately left their nets and went after him’ (Matthew 4:20)

2 I think he means it in the sense of ‘intransitive’; cf. on p. 150 he states that a verb is čhezokh if it can’t take an accusative case object.

3

c. intransitive with auxiliary zi oæ e†ê me®eal ê a@Àikd ayl nnÀê zi otSH EtHe mereAl e A“dZikd Ajl nndZe for neg that die.ppl be.3sg girl.2 but sleep.3sg ‘for the girl is not dead, but is sleeping’ (Matthew 9:24) d. intransitive without auxiliary ew matu…eal dpir mi asê …na Ew mAtutsHeal dpiR mi Ase tsHnA and approach.ppl scribe one say.3sg to.3sg.acc ‘and a scribe came up and said to him…’ (Matthew 8:19) • N.B. the auxiliary verb does not agree with the subject in person or number in transitive

constructions of types (a-b); in these cases, the verb is 3rd singular. • Benveniste’s analysis of the transitive perfect has been influential, but was already made by

Aytånean 1866 and Čhalåƒean 1886: (6) Čhalåƒean 1886:159 “We saw above that compound tenses are formed with the forms of ‘be’ ê ‘is’, êr ‘was’, i…ê ‘will be’ and a genitive [subject]: gorceal ê im, i…ê Ïo teseal, a®eal i…ê nora, etc…The correct translation of these is gorcac æunim / ißlediyim yoq ‘I haven’t done’, tesac unis / gördüyün var m¥ ‘have you seen?’, because ê im, ê Ïo mean unim ‘I have’, unis ‘you have’ [respectively].” 3.2. Support for the possessive theory (7) formation of the present progressive in Christian Hamshen dialect (Aça®yan 1947:140-1): 1. with ‘have’: •erim guni ‘I am carrying’, etc. (carry-1sg imperfv-have-3sg) 2. (old Janiktsis) Nom. subject. and 3sg verb: yes eguß ä, etc. 3. (young Janiktsis) possessive pronoun + 3sg verb: ims eguß ä, etc. 4. cf. Trabzon forms present progressive with -er, but imperfect progressive with uni

(Petrosyan Armenological Dict., 587) (8) soyn órinak ew amenayn girÏ hogepatumÏ` nßanakeal unin zÏaÀu†iwns amenayn zóra… soyn ôrinak ew amenayn girÏ hogepatumÏ` nßanakeal unin zÏaÀu†iwns amenayn zôra… ‘this exemplar and all divinely-narrated books have noted the brave deeds of all the

soldiers’ (Koriwn, page 9 of Venice 1894 edition) (9) Aytånean 1866.2:96-97: a. “[In some varieties of Modern Armenian] the verb unim [‘have’] also builds witnessed

and non-witnessed perfects with the -ac participle: tesac æunim [‘I haven’t seen’], ditac unis [‘have you noticed?’], aynæa§ †ßnamea… dêmn elac æunêin [‘they hadn’t faced such enemies’]. These perfects indicate that it never happened/was, and generally are used interrogatively or negatively.”

b. “[Middle Armenian also formed perfects with the -aj participle + unim:] zÏa@aÏn a®ac unêin ‘they had taken the city’, i nerÏsê pahac uni ‘he has kept [it] inside’.”

4

c. “This also existed in the Classical language: a®eal unim [‘I have taken’], and especially in the very frequent forms ê im teseal [‘I have seen’], æiÏ im teseal [‘I haven’t seen’], i…ê Ïo teseal [‘you will have seen’], which is the same as in Modern Armenian and Turkish.”

(10) Karst 1901:373—“the -aj participle also appears with unim, e.g. Smpad’s Chronik

zÏa@aÏn a®ac unêin ‘they had taken the city’ (Dulaurier 650). Aytånean (II 97) linked this to the western European possessive + preterite participle construction. This must not have been a normal tense in Middle Armenian, because it is foreign in Modern Armenian.”

(11) Muslim and Christian Hamshen dialects form (some) perfects with ‘have’, e.g. khiadz

unim ‘I’ve written’ (Dumézil 1963:15; cf. also Aça®yan 1947:144-45, œaribjan 1953:402, Grigorjan 1957:422) o v. infra for more precise discussion of when they use ‘have’ vs. ‘be’ here

3.3. Transitive/intransitive, or unaccusative/everything else?

o After Benveniste’s time, it was realized (Burzio 1986?) that the division in the W. European perfect is not transitive:intransitive, but unaccusative:non-unacc.

o Can/should this discovery apply to Armenian as well? 3.3.1. Two kinds of intransitive verbs (12) Unergative (SV): often have agentive argument, shows protagonist control, internal

causation; take ‘have’ in perfect in Basque, Italian, French, Dutch, etc. dine, golf, work, telephone, sleep, shine, laugh, bleed, smile, quarrel, travel, joke, chat,

bark, dance, sing, emission verbs (ring, flare, groan, creak, gurgle, shine, sparkle, stink, boil, bubble), cough, shiver, snore, tremble, yawn, lie (down), sit, kneel, stand, shout, yell…

(13) Unaccusative (VO): movement, the argument undergoes a change of state, temporally

bounded eventualities, telic, argument is a patient; take ‘be’ in perfect in Basque, Italian, French, Dutch, etc.

passives, raising verbs (believe, seem, etc.), arrive, come, go, return, leave, die, fall, enter, grow, arise, emerge, ensure, begin, exist, occur, follow; (sink, open, close, increase, break, drop)…

And I hear, moreover, his highness is fallen into this same whoreson apoplexy.

King Henry IV, I.2.34

5

(14) useful diagnostics and mnemonics • Pesetsky: unaccs never take cognate objects, unergs sometimes do • Partee: unergs can’t show causative alternation (e.g. as is possible with ‘break’) • Levin and Rapoport: only unaccs can form resultative construction (the river froze solid

vs. *we yelled hoarse) • only unergs can make false reflexive resultative: we yelled ourselves hoarse • Horn: unergs can’t form resultatives with -ee, -able, un-, re- • Burzio: only unergs can form agentives in -er (jumper) • Levin and Rapoport: unergs can’t make -ed adjective (frozen lake vs. *jumped child)

3.3.2. Dialectal Armenian perfects (15) Aça®yan 1947:144-45—[Christian] Hamshen normally forms the perfect, etc. with unim ‘have’: …onaj unim ‘I have strewn’ etc. [All of his examples are transitive verbs.] (16) œaribjan 1953:402, Grigorjan 1957:422—In Hamshen the perfect tenses are formed with the perfect participle and [im or] unim (they state that both auxiliaries are possible for each verb) (17) Dumezil 1963:15—Muslim Hamshen also uses the perfect tense (passé composé) with ‘have’, but only with transitive verbs: kiaj unim ‘I’ve written’, etc. (18) Homshetsma (Muslim Hamshen, spoken in Köprücü) a. jija@-aj-a he laughed (unergative) (-a = ‘is’) mer-aj-a he died (unaccusative) eg-aj-a he came (unaccusative) b. jija@-aj-ui I (have) laughed *mer-aj-ui I (have) died *eg-aj-ui I (have) come ‘do’: aajui, aajues, aajua…

What is the -ui, -ues, -ua, etc.? < unim ‘have’ !

6

(19) In Xotorjur, all “transitive” verbs take unim ‘have’ instead of em ‘be’ to form periphrastic tenses (Hulunean and Haçean 1964:408; they add (p. 409) that many dialects use unim in this way):

a. PERFECT kerac unim ‘I have eaten’ PLUPERFECT kerac unêi ‘I would have eaten’ kerac piti unenam kerac piti unenayi kerac unenam kerac unenayi b. mays t@ay berac uni ‘my mother has born (lit. ‘carried’) a boy’ go@eyn erku mard spanac unêin ‘the thieves had killed two men’ duÏ oæinæ ima…ac æuniÏ ‘you haven’t learned anything’ menÏ lsac unenanÏ piti ‘we should have listened’ c. All “intransitive” and “middle” verbs select ‘be’ instead (acc. to H and H 1964): t@ên Ïun e®ac a ‘the boy was sleeping/asleep’ (unergative!!) kovn korac êr ‘the cow was lost’ oæƒayn gtnuac êr ‘the sheep was found’ (passive) 3.3.3. Classical and Middle Armenian perfects (20) Aytånean 1866.2:96-97 o “The modern European languages seem to parallel our 12th-century language, using have

with transitive verbs and be with intransitives and neutral/middles.” (21) Avetisyan and Ghazarjan 1986:178-9: • “often the nouns and pronouns used as subjects in the 3rd singular of the present, past, and

future tenses formed with the verb em ‘be’ and the past ppl [i.e. -eal] are put in the genitive case, by which are expressed the number and person of the verb. For example, instead of es teseal em, dukh teseal ēkh, arkhaj teseal ē we have teseal ē im/dzer/arkhaji. Only nergortsakan [transitive] and čhezokh [neutral/?middle] meanings are expressed by these forms...e.g. im ethē čhēr ekeal ew χōsetsheal ənd nosa (John 15:22).”

o NB this example suggests that subject choice is controlled by the last verb o NB nergortsakan = energia ‘having internal working’ (i.e. transitive), which

parallels Levin and Rappaport’s observation that unergatives have internal agent

7

(22) What exactly is čhezokh? • Petrosjan, Linguistic Dictionary 250: verbs that don’t require an object to carry out an

operation: sleep, smile, be afraid; also includes reciprocal and reflexive verbs (kiss/hug each other, wash oneself)

Problem: Ajtənean groups neutrals with intransitives, but Avetisjan and Ghazarjan

group them with transitives. Which is correct? (23) unergatives with genitive subject erewil erewil ‘appear’ [takes avoir in French, though listed elsewhere as unacc]

Agathangelos 786, 2 Եւ անդ երեւեալ դիւացն ի մեծ եւ ի բուն մեհենացն Հայոց թագաւորացն, ի տեղիս պաշտամանցն, յԱնահտական մեհենին, յԵրէզն աւանի.

açaparel ačapareal ‘hurry’ (cf. Dutch heeft zich gehaast)

Eghishe Ew e†ê æêr mer açapareal ew i §axust darjeal, mium i mênÀ oæ tayin aprel Ew e†ê æêr mer açapareal ew i §axust darjeal, mium i mênÀ oæ tayin aprel:

Movsēs Xorenatshi, Patmuthiwn Hajotsh 3, 45, 2 Քանզի աճապարեալ Սամուելի Մամիկոնենի մտերմի Արշակայ, յոլովաձեռն գնդաւ զհետ մտեալ փախստեայս արկանէ յամուր այրս ի գաւառին Մանանաղւոյ. ‘Samuel Mamikonean, a friend of Arshak’s, quickly gave pursuit with a powerful force and sent them fleeing to an inaccessible cave in the province of Mananaghi’

nayil najil ‘look’ (cf. Dutch heeft gekeken)

Agathangelos 733.5 եւ իմ նայեցեալ տեսի զկերպարանս նորա Ew im nAjEtsHeAl tEsi zkERpARAns noRA

a¬aw†el A“ōtHel ‘pray’

Patmuthiwn vratsh, L.Mr. Vit. Reg., 100, 16 Եւ նորա աղաւթեալ յերկար եդ զխաչն ի վերայ նորա չորեքկուսի, եւ նոյնժամայն նստաւ ողջացեալ:

Corpus textorum hagiographicorum armeniacorum: Mart. Suk., 6 (AH, 35, 4, 26) Եւ աղօթեալ նոցա, ամենեցուն առ տէր Աստուած ասեն.

nstil nstil ‘sit’ [cf. Dutch heeft gezeten, is gaan zitten]

Corpus textorum hagiographicorum armeniacorum: Mart. Theod., 4a (AH, 68, 10) Եւ դատաւորին նստեալ յաթոռ ատենի, հրամայեաց ածել զԹէոդորոս.

xawsil χawsil ‘speak’ [cf. Dutch heeft gepraat; also unergative in Basque]

Čhaləχean 159 æiÏ im amenewin erdueal ew oæ sut xóse…eal

8

tSHikH im AmEnEwin ERdueal Ew otSH sut XosEtsHeal Matthew ??.?? [here in the sense of ‘crow’, which should be unergative]

զի յայսմ գիշերի մինչչեւ հաւու խօսեալ իցէ, երիցս ուրասցիս զիս ‘for this evening before the cock crows, you will deny me thrice’

Agathangelos 71.6 ոչ խօսեալ նոցա երբեք եւ ոչ իմացեալ եւ ոչ զմտաւ ինչ ածեալ, ոչ վասն քո եւ ոչ վասն իմ

erdnul erdnul ‘swear’ (cf. Dutch heeft gevloekt)

Chaləχean 159 æiÏ im amenewin erdueal ew oæ sut xóse…eal tSHikH im AmEnEwin ERdueal Ew otSH sut XosEtsHeal [both of these verbs are unergative!]

Eghishe M®næelov ba®ba® arjakeal ew asê. Erdueal im yaregakn, i mecn astuac M®næelov ba®ba® arjakeal ew asê. Erdueal im yaregakn, i mecn astuac

Movsēs Xorenatshi 2.79.14 [cite # may be wrong] վասն զի երդուեալ էր նմա հօրն իմոյ ի լոյսն արեգական.

cxel tsχel ‘smoke’

Exodus 20:18 ew amenayn ±o@ovurdn tesanêr zjaynn...ew zlea®nn cxeal ‘When the people saw the thunder and lightning and heard the trumpet and saw the mountain in smoke’

(24) Problem 1: unergatives with nominative subject iĵanel ‘descend’ [takes avoir in French, but appears to be unacc in Armenian]

Mt 14.29 Եւ իջեալ ի նաւէն Պետրոս. ‘Peter got out of the boat’ čanaparhordel ‘travel’

Luke 10:33 (travel should be unergative; cf. Dutch heeft gereisd) Samara…i omn çanaparhordeal` ekn noyn a® novaw ‘a certain Samaritan traveled and came to where he was’

Proposal: unergatives optionally take either, as in SWA relative clauses

9

(25) optionality with ‘hurry’ Eghishe

Ew e†ê æêr mer açapareal ew i §axust darjeal, mium i mênÀ oæ tayin aprel: Ew EtHe tSHER mER AtSApAReal Ew i pHAXust dARdzeAl, mium i mendZ otSH tAjin ApREl

Movsēs Xorenatshi, Patmuthiwn Hajotsh 2, 55, 4 Աստ աճապարեալ Արտաշէս մեծապէս ընծայիւք իջանէ ընդ առաջ նորա, զմեղանս յանձն առեալ, հանդերձ հարկօք անցելոց ամացն լինի առաջի նորա. ‘in haste Artashes came down to meet him there with magnificent presents, and accepting the responsibility for defaulting he appeared before him with the tribute of the previous years’

(26) Problem 2: unaccusatives with genitive subject mtanel mtanel ‘enter’

Exodus 5:15 ew mteal dpratsh ordwotshn Israjeli ‘then the scribes of the sons of Israel came’

*one should expect mteal to be unaccusative… Ωo¬ovel Zo“ovEl ‘gather, meet’ [NB normally takes nominative subject in grabar]

John 18.2 զի բազում անգամ ժողովեալ էր անդր Յ[isus]Ի աշակերտաւքն հանդերձ ‘And Judas also, who betrayed him, knew the place: for Jesus many times had gathered there with his disciples.’

4. Other Manifestations of the Unaccusative/Unergative Distinction 4.1. Personal Relative Clauses in Western Armenian (Sigler 1997) (27) (yes) (maro-i-n) *(namag-må) @årge-…-i 1sg.NOM Maro-DAT-def letter-a send-aor-1sg ‘I sent a letter (to Maro)’ (28) a. (im) (maro-i-n) @årg-aj namag-ås 1sg.GEN Maro-DAT-def send-ppl letter-1sg ‘the letter that I sent (to Maro)’ b. (im) *(namag) @årg-aj zarmig-ås 1sg.GEN letter send-ppl cousin-1sg ‘the cousin that I sent a letter to’ (29) ing-aj ßiß-å (internal argument of unaccusative) fall-ppl bottle-def ‘the fallen bottle’

10

(30) kan†-åv-aj gamuræ-å (internal argument of passive) destroy-pass-def bridge-def ‘the destroyed bridge’ If the verb has an external argument, that argument cannot head the PRC; it can only be the

possessor: (31) ÏirÏ-i-n Ïån-aj gin-å (subject of transitive) book-GEN-def buy-ppl woman-def cannot mean: ‘the book that the woman bought’ must mean: ‘the woman that the book bought’ (32) *bora-c-aj tåraci-n (subject of unergative) shout-aor-ppl neighbor-def ‘the neighbor that shouted’ 4.2. Compound verbs with do/be • Levin and Rappaport Hovav 1995:140: Basque expresses agentive unergatives with light

verb construction headed by ‘do, make’ + noun • In Turkish, unaccusatives take olmak ‘be’, unergatives take etmek ‘do’ (Özkaragöz 1986) (33) Verbs formed from Turkish and Persian roots in the Armenian dialect of Van (Aça®yan 1952:189): a. active/transitive → aniel ‘do’ †åmaƒ aniel be greedy, hoard ƒäbär aniel inform sus aniel be quiet (??) §aralamiß aniel tear, devour §özmiß aniel annihilate Ïüfür aniel curse b. neutral or passive/intransitive → elniel ‘be’ ƒavaslamiß elniel desire mayil elniel marvel, wonder, be amazed §a®lamiß elniel shine (should be unergative!)

11

(34) Homshetsma (cf. Dumézil 1963:18-19) a. beda enuß ‘wait’ emane† enuß ‘entrust’ (Tk emanet etmek) haz enuß ‘love’ k¥zd¥rm¥ß enuß ‘anger’ b. Ïayilmiß elluß ‘slip’ (Tk kay¥lmak) aßu@ elluß ‘be in love’ (Tk aߥk olmak) inanmiß elluß ‘believe’ anlaßmiß elluß ‘understand’ dußunmiß elluß ‘think’ †aßinmiß elluß ‘move’ (35) Mush (Ba@dasaryan-ÿa§al…yan 1958:237ff.) †ambi eniel order †aslim eniel trust, depend on, deliver, give (Tk. has both teslim etmek and teslim olmak, both meaning ‘surrender’) †arif eniel tell, narrate ƒåsa eniel tell, speak b. hibålmiß e@nal to be very surprised (36) Tigranakert (Hanejan 1978:106, 199ff) a. ängäÀ enil listen bazar enil come to an agreement bäÏlämiß enil wait äÀälä enil hurry aßi enil vaccinate bäæ enil kiss deynåm enil swear (oath) zan enil think, suppose b. †iÏsåmiß e@näl be disgusted, loathe ®azi e@näl agree ayle-vayle e@näl be altered dänåßmåß e@näl become informed about something 4.3. Witnessed and Non-Witnessed Perfects Grigoryan 1957

• Some dialects (Ararat, Tiflis, Shamaxi, Astraxan, Julfa, Urmia, Maragha, Xoy) form resultative perfects (harakatar) primarily with passives and middles; transitives lose their transitivity in this construction. (171)

• In some dialects (Agulis, Karabagh, Ardvin, Shaghax, Karchevan, Meghri, Hadrut, kå-dialects), transitive verbs with the harakatar participle form the va@akatar, whereas

12

passives and intransitives can have harakataror va@akatar sense (e.g. Agulis tåvec åm = SEA tvel em) (172)

• problem: it is unclear whether harakatar/va@akatar is witnessed/non-witnessed, perfect, pluperfect, or something else

4.4. Causatives?

• standard wisdom: Modern Armenian uses -tsh- for causatives • tal ‘give’ can also be used (SWA (Sigler 2003), Tigranakert (Haneyan 1978:106), etc.)

o ‘give’ = ‘cause to have’? • Comrie 1981:182: -tsh- used with intransitives; transitives prefer tal

(37) What about tal as a marker of intransitives? a. Karabagh vaz tal ‘run’ (NB ‘run’ can be telic or atelic, hence unergative or unaccusative) b. Ararat (Markosyan 1989) bol†i tal ‘wander aimlessly’ (323) ©ålƒapindå tal ‘strike the final heavy blow’ (324) zuza tal ‘howl’ (said of wolves and dogs) (327) †ev tal ‘encourage, animate’ (328) la§in tal ‘be stormy; when waves strike the shore’ (331) loÏ tal ‘jump with big long strides’ (332) cikal tal ‘let out a sharp sound’ (335) kådi tal ‘ƒtut tal’ (336) (kådi ‘goat’) c. SEA lujs tal ‘shine (intr)’ (but §ayl tal ‘shine (tr)’) kul tal ‘swallow’ 4.5. Number agreement in SWA? NPs that are logically plural but morphologically singular take plural agreement with transitive verbs, but singular agreement (normally) with intransitive and passive verbs (Bardakjian and Thomson 1977:30, 38) (38) jERgu hokhi Ajth AXtSHigə gə siREn two person that girl-definite imperfective love.3pl ‘two people love that girl’ (39) jERgu dARin Sud gAntsHni two year.definite quickly imperfective.pass.3sg ‘the two years will pass quickly’

13

5. Conclusions • Ample evidence for unaccusative/unergative distinction in Armenian • Classical and dialectal perfect systems behave like W. European systems in

distinguishing unaccusative vs. unergative, not active:passive • Perhaps one should not translate nergortsakan as “transitive”, so as to avoid excluding

unergatives? • Meillet was wrong to say that one cannot fruitfully study the syntax of Classical

Armenian because it is calqued on Greek and/or Syriac. References Ačařyan, Hračhya. 1947. Khnnuthyun Hamšeni Barbaři [Study of the Hamshen Dialect]. Erevan:

Haykakan SSH Gituthyunneri Akademiayi Hratarakčhuthyun. Ačařyan, Hračhya. 1952. Khnnuthyun Vani Barbaři [Study of the Van Dialect]. Erevan: Haykakan SSH

Gituthyunneri Akademiayi Hratarakčhuthyun. Ajtənean, Arsen. 1866. Khnnakan kherakanuthiwn ašχarhabar kam ardi hayerēn lezui [Critical grammar of

the modern Armenian language], 2 volumes. Vienna: Mχithareantsh Tparan. Reprint 1987, Erevan, Erevan University Press.

Avetisyan and Ghazaryan. 1986. Grabari je®nark. Benveniste, Émile. 1952. The passive construction of the transitive perfect. BSL 48:52-62. (= chapter 15

of his Problems in General Linguistics) Benveniste, Émile. 1960. The linguistic functions of ‘to be’ and ‘to have’. BSL 113-134. (= chapter 16

of his Problems in General Linguistics) Burzio, Luigi. 1986. Italian Syntax. Dordrecht: Reidel. Čhalåƒean, Vrthanês. 1886. —erakanu†iwn Haykaznean Lezui [Grammar of the Armenian Language].

Vienna: Mχitharean Tparan. Dulaurier, Ed. 1869-1906. Recueil des historiens des croisades: documents armeniens. Paris: Imprimerie

Nationale. Dumézil, Georges. 1963. Notes sur le parler d’un Arménien musulman de Hemßin. Mémoires de

l’Académie Royale de Belgique 57.4. Brussels: Palais des Académies. Hulunean, Jaruthiwn and Matthêos Haçean. 1964. Yußamatean XotorÀuri [Memorial Book of

Xotorjur]. Vienna: Mekhitarist Press. Karst, Josef. 1901. Historische Grammatik des Kilikisch-Armenischen. Strassburg: Verlag von Karl J.

Trübner. (reprinted 1970, de Gruyter and Co., Berlin) Kayne, Richard. 1993. Toward a Modular Theory of Auxiliary Selection. Studia Linguistica 3-31. Levin, Beth and Malka Rappaport Hovav. 1995. Unaccusativity. Cambridge: MIT Press. Lyonnet, Stanislas. 1933. Le parfait en arménien classique. Paris: Librairie Ancienne Édouard

Champion. Markosyan 1989. Araratyan barbaŕ [Ararat dialect]. Erevan: Luys. Meillet, Antoine. 1929. Sur l’étymologie d’arménien unim. MSL 23:276. œaribjan, Ararat. 1953. Hay barba®agitu†yun [Armenian Dialectology]. Erevan: Haykakan SSÂ

Petakan He®aka Mankavar±akan Institut. Özkaragöz, I. 1986. The Relational Structure of Turkish Syntax. Doctoral dissertation, UCSD. Schmalstieg, William. 1993. More on the Old Armenian transitive perfect. In Proceedings of the Second

International Symposium on Armenian Linguistics (1987) ed. by L. Hovsepian, N. Parnasian and S. Simonian, pp. 102-113. Eerevan: Armenian Academy Press.

Sigler, Michele. 1997. Number Agreement and Specificity in Standard Western Armenian. Doctoral dissertation, MIT.

Thibault, Pierette and Gillian Sankoff. 1997. A sociolinguistic study of verb morphology in Montreal French: reanalyzing the alternation between avoir and être. 21st Annual Penn Linguistics Colloquium.