CC BY
27
ARMENIAN ONOMATOPOEIC VERBS1
Резюме: Статья посвящена вопросу о морфологическом статусе и происхождении основообразующих элементов -(a)c-, -(a)nc-, -(a)r-(a)c- и -(a)r-(a)nc- в древнеармянских звукоподражательных словах. Рассматривается вопрос о возможности сопоставления этих элементов с древнеармянскими презентными суффиксами -(a)c- и -(a)nc- незвукоподражательных глаголов. При этом в рамках гипотезы о происхождении презентного суффикса -с- от стативного/инхоативного *-ehj-ske/o- ставится вопрос о возможной связи протоармянских звукоподражательных глаголов со стативом; рассмотрены свидетельства других индоевропейских языков, в частности, звукоподражательные глаголы, засвидетельствованные у Гомера в форме перфекта со значением настоящего времени типа веврще ‘ревет’.
Ключевые слова: индоевропейский перфект, звукоподражательные глаголы, древнеармянский язык.
In 1989, A. Meillet published an article on Armenian etymology where he pointed out the resemblance of hunc ‘noise’, hncem ‘to resound’ and sunc ‘breath’ and expressed skepticism about the attempts to establish the origin of that type: “les deux mots sont trop pareils pour n’avoir pas ete modifies l’un par l’autre, mais il est inutile d’essayer de faire des hypotheses sur l’histoire et la nature de ces influences reciproques trop compliquees pour se laisser deviner.” Further, he compared sunc to svasiti ‘to hiss’, Skr. svanas ‘noise’, Lat. sonus ‘sound’ with the anlaut sound change as in sun, Skr. sva ‘dog’, and claimed that the final -c- originated in *-sky-. Despite the cited disclaimer, Hr. Acaryan presented a systematic study of many more onomatopoeic verbs and their nominal derivatives (Acaryan 1899: 202-205) integrated later into Acayan’s dictionary. The topic
1 The paper presents results obtained within research programs supported by the Ministry of education and science of the Russian Federation (contract nr. 16.740.11.0292, research leader Petr Kocharov) and the Research Council of the President of the Russian Federation (grant nr. MK-389.2011.6, research leader Alexander Nikolaev). I am grateful to Prof. N. N. Kazansky and Dr. S. D. Kleiner for their valuable comments.
was revised again in 1995 by J.C.A. Greppin who referred to HAB but not to Acaryan’s earlier article.
Hr. Acaryan noticed the high productivity of onomatopoeic verbs in the classical language and the relevance of this verbal class for comparative studies. He then proposed a classification of 29 derivational models attested for the Armenian onomatopoeics. Apart from the reduplication typical of this lexical class, the following stem extensions have been attested since the early classical Armenian texts: -c-, -ac-, -nc-, -anc-, -re-, -rnc-, -arac-, -aranc-. These extended stems followed -e- or -a- conjugations and served as the basis for the regular aorist forms in -ec- and -ac- in a way that their suffix status cannot be justified from the synchronic point of view (hncem, hnceci).
As Meillet justly noted, the interference of these derivational models should have led to a developed pattern according to which some of extensions could derive secondary synonymous variants from the same roots. Thus, gocel ‘to cry (of man)’, probably from *wokw-ye/o- of PIE *wekw- ‘speak’ (LIV: 673-674), modified into gorocel ‘id.’ (Sapuh Bagratuni, 9 c.)2, goroc ‘cry’ (Nerses Palienc', 14 c.) by analogy to -ac- : -arac- variation attested elsewhere. This interference, together with the onomatopoeic nature of the roots themselves, hinders any etymological attempts. However, the morphological pattern itself should have been motivated from the Proto-Armenian perspective and requires an explanation. The main question that arises in this respect is whether or not the onomatopoeic root extensions were motivated by the verbal present suffixes -c- and -nc-. If the answer is positive, further suggestions can be made about the grammatical background of this morphological phenomenon.
While Meillet and Acaryan seem to favor the comparison of -c-and -nc- in onomatopoeics and the respective suffixes of non-onomatopoeic verbs, Greppin, on the contrary, remains skeptical in view of the lack of secure etymologies; G. Klingenschmitt noticeably does not approach this issue in his monograph on the Armenian verb.
2 This and the other references to the Classical Armenian sources are based on the Leiden Armenian Lexical Textbase: http://sd-editions.com, ed. by Jos Weitenberg, and the on-line concordance of the Classical Armenian Bible (Constantinople 1895): http://212.34.228.170/bible_27E/about.htm, ed. by Thomas J. Samuelian.
2. Following Meillet, Acaryan directly compared -c- in seel ‘to hiss (of astonished or angry person; of snake)’ (Bible), beel ‘to bellow (of cow)’ (John Chrysostom, 5 c.), eeel ‘to scream (of travailing or mourning woman; of swallow)’ (Bible), etc. to the imperfective suffix -ei- (e.g. in pax-e-el ‘to flee’, aor. pax-e-ay). Likewise, he compared the element -nc- of sncel ‘to blow (of wind, God’s breath)’ (Bible), bncel ‘to bellow (of calf)’ (Anania Sanahnec'i, 11 c.), mncel ‘to mourn (of dove)’ (Bible), etc. to the suffix of erk-nc-el ‘to fear’ (aor. erk-e-ay), kor-nc-el ‘to disappear’ (aor. kor-e-ay), mart-nc-el ‘to fight’ (aor. mart-e-ay), and yar-nc-el ‘to belong to’ (aor. yar-e-ay). It should be noted that all of the abovecited verbs had *-i- aorist stems that, most probably, continued the PIE stative *-e- stems.
However, the origin of the present suffix -c- remains problematic. Unlike Meillet, who reconstructed *-sk-ye/o-, some scholars preferred to reconstruct *-sk- in front of the front vowels (Olsen 1988), others - *-k-, or *-kw- in front of the front vowels or *-ye/o-. Thus, R. Godel compared the suffix -c- to Gk. -aa-, cf. Arm. t'ak'cim, Gk. mtfaaw (Godel 1982: 22). Finally, H. Pedersen assumed that the suffix -c- should be compared to *-k- in Gk. ёбц-к-а ‘I have put’, Lat. fe-c-i ‘I did’ (Pedersen 1906: 348).
It should be noted, that the latter sound change finds support in some verbs of sound performance: gocem ‘cry’ from a denominative *wokw-ye/o- or an iterative *wokw-eye/o- (Ved. vacayati), PIE *wekw- ‘speak’3, and pcem ‘blow, breath out’ < PArm. *phukh-ye/o-in view of puk ‘breath, wind, fart’ (Meillet 1909: 357; Godel 1982 (=1965): 22). The assumption of PIE *-ky- > Arm. -c- offers a new explanation of alacem ‘to prey, lament’ < *plh2k-ye/o- ~ OCS плакать ‘cry, lament’ < *pleh2k-, see (Derksen 2008) for Slavic comparanda. To this may be added a tentative comparison of the Armenian onomatopoeic root b(i)- (bcem, bibem, see below) to PIE *bhi-k- of OCS. бьчела, OIr. bech ‘bee’ (ESSJa, 3: 104-105; Derksen 2008: 72) or PIE *bhowk- of PSlav. *bucati, Lith. baUkti ‘to roar’ (ESSJa, 3: 74; Derksen 2008: 67).
3 B. Olsen (1988: 8) explains kocem ‘call, invite’ as an irregular reflex of
*gwet-ye/o-, PIE *gwet- ‘speak’ (IEW: 480f.) influenced by semantically close gocem, cp. (Belardi 2009: 106). G. Djahukyan seems to accept *-ty- > -c- for this and other lexemes, cf. cic ‘cry’ < *gi-tyo- or *gi-kyo- (fl^ayK^H 1982: 59, 171).
G. Djahukyan accepted seem < PArm. *sicem < *kwikid < PIE *kwey-k-, cf. Lith. svyksti ‘to whistle’ and sunc ‘breath’ < PIE *kwonkyo-, cf. Lith. svanksti ‘breath, snuffle’ (^^ayK^H 1982: 58, 171). W. Belardi suggested that both Armenian verbs may be ultimately related to each other and to sulem ‘to whistle’ < PIE *kow-lo-, cf. Skt. sut-kara- ‘whistling’, svasati ‘to hiss’, thus, positing a Proto-Indo-European onomatopoeic root *kew- ‘to hiss’ (Belardi 2009: 150, 154-155; Martirosyan 2010: 522). The comparison of Arm. sc el to Gk. oifa ‘to hiss’ suggests the root *kw-. However, further comparison of the Armenian -c- to the Greek -Z-, probably, from *-g-ye/o- is difficult to account for on formal grounds (Greppin 1995: 62; Sihler 1995: 516-517)).
3. The origin of -nc'- is even more controversial; its -n- has been interpreted as an infix, as part of the root, or as a separate derivational element. A detailed discussion of the present suffix origin is provided in (Klingenschmitt 1982: 78-84), see also (Olsen 1999: 522). The existence of erkcot ‘frightened; timid’ suggests that the stem of erk-ncim (aor. erk-eay) ‘to fear’ is secondarily derived from Proto-Arm. *erkc- -Gk. SsSioKo^ai ‘to frighten’. However, this hypothesis does not explain why an infix was not added to the majority of -c- stems even when the root ended in a voiceless obstruent, cf. xrtcim ‘be(come) confused’.
Concerning the onomatopoeic extension, G. Djahukyan derived Arm. drncem ‘to make ring’ (Philo, 6 c.) and drncim ‘tingle (of ears)’ (Basil of Caesarea, 5 c.) from PIE *dhren-k-ye-, cf. Toch. B trenk- ‘to speak’ (for semantics cf. Rus. trendet' ‘to jingle’ > colloquial ‘to chat’), OIrl. drecht ‘song’, and further PIE *dhren-, OInd. dhranati ‘to sound’, Gk. dpfjvog ‘weep’, Goth. drunjus ‘sound’, etc. (^^ayK^H 1982: 182-183; Belardi 2009: 54). Moreover, this etymology, if correct, provides a Proto-Armenian model for the reanalysis of -rn- on which see below. However, the forms dandacem ‘to ring (a bell)’ (Yeznik, 5 c.) and dandacecuceal ‘having caused (a bell) to ring’ (Anania Narekac'i, 10 c.) show that drncem could be part of a derivational pattern for the onomatopoeic root d-, as assumed in (Greppin 1995: 56-57), and not belong to the cited Indo-European comparanda.
In some cases, a phonetic explanation of the rise of the auslaut -nc- is possible. The Classical Armenian reduplicated stems could contain expressive alternation in the final consonant of the second root, notably from -r- to -j-, cf. bar-baj-ank' ‘aberration’; dr-dj-iwn ‘nonsense’. In his thorough analysis of such and other Classical Armenian reduplicated stems, M. Leroy assumes that -n- could be
inserted in front of the alternant -j- in onomatopoeic verbs, cf. mr-mr-am (Gk. nop/ivpw, Lat. murmur) ^ mr-munj, mr-mumn (Leroy 1986: 69). Alternatively, mrmnjem has been explained from *mur-mun-ye- with a dissimilation of sonants and hardening of a glide, see (Martirosyan 2010: 481-482). Given the interchange of -c-/-j- (e.g. papacem, papajem ‘to bleat (of goat)’) and -nc-/-nj- (sncem, ssnjem ‘to whisper’) attested for onomatopoeic verbs, it is not impossible that -j- is original in these forms.
4. Furthermore, Acaryan identified the element -ac- in sacel ‘to rumple (of cart)’ (Bible), bacel ‘to cry (of animal)’ (Vahram Rabuni, 13 c.), t'acel ‘to mumble (of men)’ (Jacob of Nisibis, 5 c.), hacel ‘to groan (of man)’, etc. with the suffix in can-ac-el (aor. can-e-ay) ‘to understand; know’. However, the segmentation of can-ac-el is probably no longer tenable, at least from the diachronic point of view; the present stem here looks as an Indo-European archaism, from *gnh3-sk(-y)e/o-, cf. Gk. yiyvwoKw, Lat. nascor, Toch. B knastar (LIV: 163-164). The same is true of alacel ‘to prey, supplicate’, where the internal -a- should be a reflex of a laryngeal, if from *slh2-sk(-y)e/o- (cf. Gk. iXdoKo^ai ‘to appease’ < *si-slh2-ske/o-(Klingenschmitt 1982: 68)) or *plh3-sk(-y)e/o- (cf. Lat. ploro ‘to weep’), see (Martirosyan 2010: 27-28) for a detailed etymological analysis. The dialect forms with -n- (cf. T'iflis alancem) may be given as an evidence that the verb was considered a member of the onomatopoeic stock for which such stem variation is typical. Both verbs, canacel and alacel, have *-H-sk-(y)e/o- as the source of -ac-.
5. Acaryan further compared -anc- in gancel ‘to cry out (of human)’ (Dawit' Anyalt', 6 c.), t'ancel ‘to mutter (of animals)’ (John Chrysostom, 5 c.; Ephrem, 5 c.), xancel ‘to grunt (of ass, boar)’ (Bible; Agat'angelos, 5 c.), and kancel ‘to cry (of human, lion)’ (Agat'angelos, 5 c.; Grigoris Arsaruni, 7 c.; T'ovma Arcruni, 10 c.) to mel-anc-el ‘to sin’ where the suffix can be synchronically segmented in view of mel(k') ‘sin’.
Apart from (Klingenschmitt 1982: 84), the issue of melanc el ‘to sin’ has been recently addressed by B. Olsen (Olsen 2011) who interprets it as a -ci- derivative from a nominal stem, the -n- of which continues Hoffmann’s suffix *-h3n- - *mel-h3n-ehi-ske- ‘make enter into a state of defilement’. The grammatical value of Hoffmann’s suffix, a marker of quality adjectives, contributes to the functional relation of the present suffixes -c- and -nc - (connected to *-ehi- perfective stems) to the stative, or the perfect. One may wonder if the adjective suffix -unc attested in anterunck' (Bible, Jer. 2.31) ‘without a master’ belongs here as well, see (Greppin 1975:
139; Olsen 1999: 612-613). In any case, the relation between the present suffix -anc'- and the onomatopoeic stems with -anc- remains dubious.
6. No explanation has been proposed for the -(a)f- component of the remaining extensions: -fc- (kfcel ‘to caw (of crows)’ (Philo, 6 c.)); -fnc- (kfncel ‘to scream (of man), to creak (of dog), to caw (of crow) (Bible; Philo; T'ovma Arcruni)’, pfncem ‘to make a noise’ (Bible), xfncel ‘to grunt (of ass)’ (Philo)); -afac- (t'afacel ‘to lament’ (Nerses Lambronac'i, 12 c.; Barsel Maskeronc'i, 14 c.), kafacel ‘to cry (of man)’ (Aristakes Lastivertc'i, 11 c.; Nerses Lambronac'i, 12 c.), hafacem ‘to groan’ (Bible)); -afnc- (xafncel ‘to grunt (of ass)’ (Vardan Aygekc'i, 13 c.)), and -afanc- (t'afancel ‘to lament’ (John Chrysostom, 5 c.; Severian of Gabala, 6 c.), kafancel ‘to cry’ (Yovhannes Mandakuni, 5 c.)).
Due to the productivity of the stem variation pattern, the morphological borders within these root extensions became blurred. In some cases, the appurtenance of varying stems to the same onomatopoeic root is unmistakable, e.g. t'acem, t'ancem, t'afacem, t'afancem ‘to sigh, lament’ or xancel, xafncel ‘to grunt (of ass)’. However, this is not always so. An example of the ambiguities involved here may be illustrated by the onomatopoeic root b- ‘to bellow’ (NBH, I: 500; HAB, I: 458) that is attested in b-nc -em ‘to bellow (of cows and calfs)’ (Anania Sanahnec'i, 11 c.); b-c-em ‘to bellow (of cow)’, b-c-am ‘to bellow (of bull)’ (John Chrysostom, 5 c.); b-j-em ‘to bellow (of cow)’ (Vardan Arewelc'i, 13 c.); b-b-c-am ‘to growl (of lion)’ (Grigor Tat'ewac'i 14 c.); b-b-c-em ‘to growl (of beast)’ (Karapet Sasnec'i, 12 c.), bi-b-em ‘to cry (of calf)’ (John Chrysostom, 5 c.). Now, the question is whether we shall, following Hr. Acaryan, attribute bafacem (Ephrem, 5 c.; Basil of Caesarea, 5 c.; Kirakos Drazarkec'i, 11 c.) and bafajem ‘to bleat (of a ram)’ (Eusebius of Caesarea, 5 c.) to the same root b- ‘to bellow; growl’, or to a distinct root baf- ‘to bleat’. A similar case of morphological ambiguity is presented by munc- (NBH, II: 287) whence mncem ‘to mourn (of dove)’ (Nahum 2:7; Ezek 7:16), and mrunc- (NBH, II: 302; I: 301; Hubschmann 476), whence mrncem ‘to roar (of lion, desperate man)’ (Bible) and mfncam ‘id.’ (Kirakos Drazarkc'i,
11 c.).
In a number of contexts, forms without -(a)r- and with it appear next to each other in a way that allows for the interpretation of -(a)r-forms as markedly expressive. Cf. John Crysostom: Ew mek' yorzam zart'nucumk, bazum zams 3nd koreln ew 3nd koskopeln, ew 3nd t'anceln ew 3nd tafanceln hanemk ‘And when we would awake,
there would be much time for stretching and yawning, and to utter sighs and laments’ (trans. (Greppin 1995: 57)); John Crysostom: Yorzam i sud mtanices, hacesces, haraces darnapes “When you would enter into mourning, you will cry out and sigh bitterly” (trans. (Greppin 1995: 60)); Bible (Jer. 47, 3): I jayne aspataki nora, i zinuc otic nora ew i saceloy karac nora ew i saraceloy anuocn nora “At the noise of the stamping of the hoofs of his strong (horses), at the rushing of his chariots, (and at) the rumbling of his wheels”; T'ovma Arcruni: Ibrew zxoz vayreni kncer krncer “he grunted like a wild pig” (NBH, I: 1127). Notably, hacel appears along with haracel in all the contexts (John Chrysostom, 5 c.; Yovhannes Mandakuni, 5 c.; Yovhannes Drasxanakertc'i, 10 c.; T'ovma Arcruni, 10 c.) provided in (NBH, II: 48).
7. Although no decisive solution have been proposed concerning the relation between the present suffixes and onomatopoeic stem extensions, some very tentative considerations are presented below that attempt to find functional links between the alleged present suffix prototype *-ehrsk- (matched by aorists in *-ehr) and the onomatopoeic verbs in other Indo-European languages.
The overview of the Indo-European words for speech and sounds provided in (Mallory, Adams 352-367) leaves an impression that the lexical stock in question didn’t have any prominent morphological features except reduplication. Some onomatopoeic verbs with reduplication can be safely reconstructed for Proto-Indo-European, e.g. *khakha-, *lala-, *murmur-. Therefore, onomatopoeic verbs, in principle, could have been partly integrated into the perfect in some of the daughter languages due to the superficial morphological similarity of reduplicated stems rather than grammatical properties. Even though the question of relation between the Late Indo-European perfect and the Anatolian hi-conjugation remains largely unsolved, it may be noted that a number of Old Hittite verbs of speech or sound performance, including the reduplicated ones, follow the hi-conjugation, cf. wai-/wi- ‘to cry (out)’, halzai- ‘to cry out, shout’, ishamai- ‘to sing’, mema-/memi- ‘to speak’, wewakk- ‘to demand, ask’, wiwa-/wiwi- ‘to cry (out)’, see (Kloekhorst 2008, s.vv.).
A more specific evidence comes from the Greek language that was probably the closest relative of Armenian and, thus, may reveal shared Late Indo-European innovations of the lexical and grammatical properties of the perfect. Apart from the stative and resultative (subject- and object-oriented) usage of the perfect amply attested in Homer (Perelmuter 1983), a group of verbs point to other archaic
functions of the Proto-Greek perfect. Thus, one finds perfect forms with purely present value, cf. ysynOa ‘to rejoice’, SsSopKa ‘to observe’, zszpiya ‘to chirm’, etc., see (Wackernagel 1926: 167; Chantraine 1927: 157-164). Even when a present by-form is available it is difficult to detect the difference between the grammatical nuances expressed by the two stems4.
The verbs attested in such perfect forms noticeably belong to several distinct lexical groups: 1) stable qualities: soma ‘to be similar’, srnOa ‘to be accustomed to’, etc.; 2) physical states: SsSna ‘to burn; to be burnt’, asanna ‘to be rotten’; 2) states of mind, emotions, and perceptions: ysynOa ‘to rejoice’, SsiSia ‘to fear’; 3) auditory signals: fisfipvxa ‘to roar’ (of waves or warriors), ysywva ‘to cry’, *KSKknya, *XsXnKa ‘to scream’ (attested in the form of active participles), /ismKa ‘to blead’, [ASfivKa ‘to moo’, zszpiya ‘to utter a shrill cry’ (along with pres. zpiZw) (Perelmuter 1977: 5-6).
The first two groups belong to prototypical statives, the third one suggests object-oriented resultative interpretation, however, the last one remains problematic. It is difficult to treat it as a result of the secondary spread of the perfect morphology from stative or resultative verbs, and yet its general imperfective meaning is not coherent with the known lexical and grammatical features of the Indo-European perfect. One can hardly agree that the logical relation between the verbs of sense perception and those of sound performance could be a sufficient ground for their grammatical unification following the examples provided in (Lommel 1922), cf. Av. goas ‘to hear’, Ved. ghosati ‘to sound’.
Yet, another interpretation is possible. In some contexts, where perfect forms are used to express the present tense, they have habitual aspect and seem to be in contrast to present forms with the grammatical value of the actual present, cf. II. 8, 559: ...navza 8s siSszai dazpa, vsvnOs 8s zs ppsva noi/i^v “...and all stars are seen,
4 The resultative evaluation of the perfect was proposed already in (Brugmann 1913: 83), similar reconstruction was accepted by (Hoffmann 1970), (Szemerenyi 1970), etc. W. Cowgill insists on the stative value rather than resultative (Cowgill 1974: 563). J. Kurylowicz qualified the perfect-present verbs as having a double function of durative (primary) and resultative (secondary) (Kurylowicz 1964: 98, n. 4). See (Perelmuter 1977: 5-31) for detailed discussion of the approaches to the Homeric perfecta tantum with bibliographic references and textual evidence.
and the shepherd joyeth in his heart”5 (in the fragment giving an idyllic image of a clear night); II. 14, 139-140: ‘AzpsiSn vvv Ы\ пои Axillfoq oXoov Kfp vnOsi... “Son of Atreus, now in sooth, methinks, doth the baneful heart of Achilles rejoice within his breast” (with an explicit reference to the actual present). To this may be added the reduplicated Indo-Iranian perfect stem derived from PIE *wekw- ‘to speak’ (LIV: 673-674) and reconstructed on the basis of Ved. (pra) vavaca (RV 1.67.8b) and OAv. (pard.)vaoxdma (Y. 1.34.5c). In both contexts, the perfect forms denote habitual actions, cf.: RV 1.67.8: a
r r
vi ye crtanti rta sapantah b at it vasuni pra vavaca asmai “They who release him, paying sacred rites, — truly to such doth he announce great wealth” (trans. R.T.H. Griffith); Y. 1.34.5c: par§ va vispais (pard.)vaoxdma daёuuaisca xrafstrais masiiaisca “We have (always) declared you (to be) over and above all the creepy old gods as well as (their?) men” (Skjaervo 2006: 45). This grammatical value of a nonactual present, pointed out already in (Renou 1924), could have been part of the basic functions of the Indo-European perfect. Since the verbs of sound performance are typical of different species of animals (only sheep bleat) or emotional states of humans, they may be interpreted as predicates of quality rather than predicates of intentional action. Thus, ftsfipvxa may be considered not as a result of an action performed by a sheep, but as a result of its being a sheep.
References
Джаукян 1982 - Джаукян Г. Б. Сравнительная грамматика армянского языка. Ереван: Изд. АН СССР.
Перельмутер 1977 - Перельмутер И. А. Общеиндоевропейский и греческий глагол. Видо-временные и залоговые категории. Л.: Наука. Перельмутер 1983 - Перельмутер И. А. Статив, результатив, пассив и перфект в древнегреческом языке (язык Гомера) // Типология результативных конструкцией (результатив, статив, пассив, перфект). Под. ред. В. П. Недялкова. Л.: Наука. С. 142-148.
Acaryan 1899 - Acaryan Hr. Hayereni bnajayn ew krknawor barers (630 bari knnut'iwn) // Handes amsoreay 13: 202-207, 232-236.
Belardi 2009 - Belardi W. Elementi di armeno aureo, III. Roma: Calamo.
5 The text and translation, here and below, are cited after: Homer. Homeri Opera in five volumes. Oxford, Oxford University Press. 1920. Homer. The Iliad with an English Translation by A.T. Murray, Ph.D. in two volumes. Cambridge, MA., Harvard University Press; London, William Heinemann, Ltd. 1924.
Brugmann 1913 - Brugmann K. Grundriss der vergleichenden Grammatik der indogermanischen Sprachen. 2 Bearb. Bd. 2. T. 3. Strassburg.
Chantraine 1927 - Chantraine P. Histoire du parfait grec. Paris: Champion.
Cowgill 1974 - Cowgill W. More evidence for Indo-Hittite: The tense-aspect systems // L. Heilmann, ed., Proceedings of the Eleventh International Congress of Linguistics. Bologna: Sicieta editrice di Mulino. Vol. 2. P. 557-570.
Derksen 2008 - Derksen R. Etymological dictionary of the Slavic inherited lexicon. Leiden: Brill.
Drinka 2003 - Drinka B. The development of the perfect in Indo-European: stratigraphic evidence of prehistoric areal influence // Henning Andersen (ed.). Language Contacts in Prehistory: Studies in Stratigraphy. P. 77-105. Amsterdam: Benjamins.
ESSJa - Трубачев О. Н. (отв. ред.). Этимологический словарь славянских языков. М.: Наука. 1974-.
Godel 1965 - Godel R. Les origins de la conjugaison armenienne // Revue des etudes armeniennes, n.s. 2. P. 21-41 (= R. Godel. Linguistique armenienne. Etudes diachroniques. Paris: Freres Ghoukassiantz, 1982. P. 19-38).
Greppin 1975 - Greppin J. A. C. Classical Armenian nominal suffixes. Wien: Mechitharisten-Buchdruckerei.
Greppin 1995 - Greppin J. A. C. The Armenian onomatopoetic verbal infix in -(a)(n)c'- // Annual of Armenian Linguistics 16. P. 53-66.
HAB - Acharian, Hrachia. 1971-1979. Hayeren armatakan bararan. 4 vols. Erevan.
Hoffmann K. 1970. Das Kategoriensystem des indogermanischen Verbums // Mmchener Studien zur Sprachwissenschaft, 28. S. 19-41.
Kloekhorst 2008 - Kloekhorst A. Etymological dictionary of the Hittite inherited lexicon. Leiden: Brill.
Kurylowicz 1964 - Kurylowicz J. The inflectional categories of Indo-European. Heidelberg: Winter.
Leroy 1986 - Leroy M. Le redoublement comme procede de formation nominale en armenien classique // Maurice Leroy and Francine Mawet (eds.), La place de l’armenien dans les langues indo-europeennes. Louvain: Peeters. P. 62-75.
LIV - Lexicon der indogermanischen Verben. 2. ed. by H. Rix. Wiesbaden: Dr. Ludwig Reichert Verlag, 2001.
Mallory, Adams 2006 - Mallory J. P., Adams D. Q. The Oxford introduction to Proto-Indo-European and the Proto-Indo-European world. Oxford University Press.
Meillet 1898 - Meillet A. Etymologies armeniennes // Memoires de la societe de linguistique de Paris, 10. P. 274-82.
— 1909. Armeniaca // Memoires de la societe de linguistique de Paris, 15. P. 353-357.
NBH - 1979-1981. Nor bargirk' haykazean lezui (Новый словарь древнеармянского языка). 1-2. Ереван: Изд. Ереванского университета.
Neu 1968 - Neu E. Das hethitische Mediopassiv und seine indo-germansichen Grundlagen (= Studien zu den Bogazkoy-Texten, Heft 6). Wiesbaden: Harrassowitz.
Olsen 1988 - Olsen B. The formation of the subjunctive paradigm in Classical Armenian // Annual of Armenian Linguistics 9. P. 7-14.
Olsen 1999 - Olsen B. The noun in Biblical Armenian. Berlin, New York: Mouton de Gruyter.
Olsen 2011 - Olsen B. Contributions to the Armenian etymology. Handout of a talk presented at the 23rd Annual UCLA Indo-European conference on October 28th.
Renou 1925 - Renou L. La valeur du parfait dans les hymnes vediques. Paris: Champion.
Sihler 1995 - Sihler A. New comparative grammar of Greek and Latin. Oxford University Press.
Skjaervo 2006 - Skjaervo P.O. An Introduction to Old Avestan http://www.fas.harvard.edu/~iranian/ OldAvestan/oldavestanprimer.pdf.
Stang 1932 - Stang Chr. S. Perfektum und Medium // Norsk Tidsskrift for Sprogvidenskap, 6. S. 29-39.
Wackernagel J. 1926. Vorlesungen uber Syntax mit besonderer Beruck-sichtigung von griechisch, lateinisch und deutsch. 2 Aufl. Basel : Birkhauser.
Watkins 1973 - Watkins C. Hittite and Indo-European Studies: The Denominative Statives in -e- // Transactions of the Philological Society 1971. P. 51-93.
