CC BY
13
Acta Linguistica Petropolitana. 2022. Vol. 18.1. P. 267-281 DOI 10.30842/alp23065737181267281
Greek ^arc- and its congeners
Alexander Nikolaev
Boston University, (Boston MA, USA); alexander.s.nikolaev@gmail.com
Abstract. Ancient Greek (aor.) ¡ans/o- 'to seize' ([Hes.] Sc. 231, 252, 304) has not yet received a satisfactory etymology which this paper aims to provide. Ever since antiquity it has been argued that aor. ¡an8/6- is a secondary replacement of *^apns/o-made from nearly synonymous and much better attested verb ^apnxro 'seize, overtake, strike' (for which only sigmatic aorist is attested). This view does not carry conviction; the present paper argues that aor. ¡iaTC8/o- is real and its root ¡¡an- is distinct from that of ¡apnxro for which a new etymological analysis is offered (PIE *merkw- 'snatch, take away', Toch. A mark- 'to take away'). My proposal is to view ¡an- as a zero-grade root allomorph *mmp- formed from a full-grade root *memp- (< *menp-) cf. na08/6-(< *bhiidh-) : nsv0- (< *bhendh-), SpaKs/o- (< *drk-) : SspK- (< *derk-), etc. The hypothetical full-grade root *memp- (< *menp-) can be analyzed as a verbal governing compound of PIE *men- 'hand' (Lat. manus 'hand', Olr. muin 'protection', OE mund 'hand', Hitt. maniyahh- 'to hand over') and the root *hiep- 'take' (Lat. coepi 'I began', apio 'I tie', Hitt. epp-z 'grabs', Ved. apa 'has reached'): *men-hip- 'take with one's hand; seize'. Similar compounds with PIE *men- 'hand' used as the first member are reflected in Lat. mancipium 'laying hold of a thing; ownership' (quasi *men- + *keh2p-) and mandare 'to hand over' (quasi *men- + *dhehi-). This nominal compound underlies the neo-root *menhip-; compare Ved. gup- 'to protect' (based on the compound go-pa- 'cow-herd'), Lat. cred- 'believe' (< *kreds-dhehi-) or PIE *uelhibh- 'speak lies' (< *uelhi-bheh2-). In aor. *(hi)e-menhip-t 'seized' the laryngeal would be lost in a heavy CHCC cluster by Schmidt-Hackstein's rule, hence 3 sg. *(hi)e-menp-t (> * (hi)e-memp-t with place assimilation), 3 pl. *(hi)e-mmp-ent, remade as thematic aorist ¡ans/6-. A neat parallel to the proposed development may be found in another verbal governing compound of PIE date, namely, *men(e)sdhi- 'to implement thinking' > ¡sv0- (remade as ¡av0-) which in turn engendered a new zero-grade ¡a0-.
Keywords: Greek epic poetry, Greek etymology, compounds, formulaic collocations, laryngeal theory, Hittite, Tocharian.
Acknowledgements. It is a pleasant duty to thank Roberto Batisti, Hannes Fellner, Stefan Hofler, Sergio Neri, Norbert Oettinger, and Michael Weiss for their help and advice. Responsibility for any errors is, of course, mine alone.
© Alexander Nikolaev, 2022
К этимологии др.-гр. цая-
А. С. Николаев
Бостонский университет (Бостон, США); alexander.s.nikolaev@gmail.com
Аннотация. Древнегреческий глагол ¡апе/6- (aor.) 'хватать' ([Hes.] Sc. 231, 252, 304) до сих пор не получил удовлетворительного этимологического толкования, предложить которое—цель настоящей статьи. Начиная с античности, тематическая аористная основа ¡шпе/6- объяснялась как вариант основы *^арпе/6-, образованной от хорошо засвидетельствованного глагола ¡арпхю сходного значения 'хватать, нападать'. Однако глагол ¡шрпхю образует только сигматический аорист, а возможность выпадения р в гипотетической основе *^арпе/6- не подтверждается параллелями; столь же маловероятно, что в эпической поэме VI века до н. э. мог сохраниться такой просодический архаизм, как /mrpe/o-/ со слоговым сонантом. В силу этих причин традиционное объяснение не выдерживает критики. В настоящей статье утверждается, что корень ¡ап- этимологически не связан с корнем глагола ¡арпхю, для которого предлагается новое сопоставление с тох. A märk- 'забирать' (и.-е. *merkw-). Аористная основа ¡ап- рассматривается как отражение алломорфа *mrnp- с нулевой ступенью аблаута в корне, образованного от корня *memp- по модели па0е/6- (< *bhtidh-) : nsv0- (< *bhendh-), 5раке/6- (< *drk-) : SspK- (< *derk-) и т. д. Корень *memp- (< *menp-) восходит к древнему композиту, состоящему из и.-е. *men- 'рука' (лат. manus, др.-ирл. muin2, др.-англ. mund) и и.-е. *hiep- 'брать' (лат. coepi 'я начал', хетт. epp-z 'хватает', др.-инд. äpa 'достиг'). В хеттском языке глагол ep(p)-/ app- употребляется в сочетании со словом для руки (kissarta ep(p)-), а в латинском языке мы находим manu capere, равно как и следы сходной синтаксической конструкции с корнем *men- в сложных словах mancipium 'покупка, право собственности' и mandäre 'вручать'. Реконструкция композита *men-hip- 'брать рукой, хватать', тем самым, может быть подтверждена на синтаксическом уровне. Этот композит лег в основу корня *menhip-; в качестве параллели к появлению новых глагольных корней из сложных слов можно привести др.-инд. gup- 'защищать' (из go-pä-), лат. cred- 'верить' (из *kreds-dhehi-) или и.-е. *uelhibh- 'говорить неправду' (из *uelhi-bheh2-). В аористе *(hi)e-menhip-t 'схватил' ларингал был закономерно утрачен согласно т. н. правилу Шмидта—Хакштайна, следствием чего явилась парадигма 3 sg. * (hi)e-memp-t (с ассимиляцией по месту артикуляции), 3 pl. *(hi)e-mmp-ent, в древнегреческом языке отразившаяся как тематический аорист ¡шпе/6-. Предлагаемый в статье сценарий находит полную параллель в еще одном праязыковом сложном слове, а именно, *men(e)sdhi- 'вкладывать
ум' (др.-инд. medha- 'мудрость') > др.-гр. / ¡av0- и аорист с нулевой ступенью аблаута ¡а0е/о-.
Ключевые слова: греческая эпическая поэзия, греческая этимология, композиты, ларингалы, хеттский, тохарский.
The root ¡an- 'to seize' is scarcely attested in Ancient Greek and its forms have not yet received a satisfactory explanation. Thematic aorist ¡arce/o- is attested in the epic poem Shield of Heracles traditionally (and wrongly) ascribed to Hesiod, but now usually dated to the first third of the sixth century BCE: 1
(1) ropyovs^ arcXntoi ts Kai ou ^axai epproovro ie^evai ¡arcesw
'The Gorgons, dreadful and unspeakable, were rushing after him (scil. Perseus),
eager to catch him.' (Sc. 230-1, ed. and trans. Most)
(2) Toi S' ¿KurcoSa^ Xayo^ fipsuv avSps^ 0npsuTai, Kai KapxapoSovTs Kuvs npo, ie^evoi ¡anesiv 2, o'i S' ie^svoi urcaXu^ai.
'Others, huntsmen, were overtaking swift-footed hares,
and there was a brace of jagged-toothed dogs in front,
eager to catch them — while the others (scil. hares) were eager
to escape.' (Sc. 302-4).
Two more forms of this aorist stem have been tentatively identified in fragments of archaic Greek lyric poetry, but in both cases the text is
1 The Shield of Heracles was probably composed between 570 BCE (according to the scholia, Stesichorus knew the poem, and the name of centaur Melanchaetes on François vase is based on Sc. 186 ¡sXay%avrnv ts Mi^avxa) and ca. 590 BCE (if Cycnus' demise has anything to do with the fall of Crisa in the Sacred War in 591 BCE); see [Bing 2012: 179] for further discussion and bibliography.
2 ¡anssiv fl : ¡apnxssiv Leid. Voss. Q 59, Ambros. E 39; ¡apnssiv Vatic. 1332, acc. to Russo's app. crit.
uncertain in the extreme. 3 A reduplicated aorist is found in Sc. 252 ov Se npraxov ^e^dnmev* 'whomever they (= Keres) caught first'.
Next to the verbal root ^arc- we find the adverb e^^anera^ (- w w -) 'quickly, readily, hastily, promptly, unverzüglich reagierend',4 attested from Homer on, e.g. Il. 5, 836 e^arcero^ änopouoev 'he speedily leapt down'; Od. 14, 485 e^arcero^ ündKouoe 'at once he gave ear'.5 The connection with ^arce/o- 'to seize' is beyond doubt: for semantics cf. Latin rapidus 'quick' from rapid 'to seize'.6 Finally, a form of ^arc- may be found in Laconian personal name XIAOIOMATOX if Wachter [2001: 162] is justified in taking it as 'the one who seizes silphion'.7 Let us take a closer look at all these forms.
3 Archil. 48, 3 West (via P.Oxy. 2311) has ¡¡an ( ) srap and a verb meaning 'to seize' vel sim. appears appropriate in view of the clearly sexual narrative of the poem (for which see [Swift 2019: 270-1]). Alcman fr. 120 [PMGF] (= 177 Calame) is a hopelessly corrupt quotation transmitted by Herodian as xrô Ss ■fano^wBia Kax' av râppav ¡âproçf entais where Martin West (apud [PMGF] and [1990: 216]) replaced the unclear fxâproç with ¡¡anoîa' (but there is no shortage of alternative solutions for which see the app. crit. in Calame's edition and [Sitzler 1883: col. 932]).
4 Paraphrased in the scholia as svspyrôç, Ta%sroç, sioi^raç, anouSairoç, ¡sxà anouSfç.
5 Another adverb often compared with ¿¡¡anéroç is ¡ay 'in vain, without result, aimlessly, recklessly' and its derivatives: ¡aytSioç (î) 'vain, empty, false', ¡ayiStroç 'randomly' (Od. 3, 372), ¡ayiXôyoç 'vainly speaking' h. Merc. 546, ¡ayuXâKaç 'barking for no reason' Sapph. 58, 21+, PN *Mayi%oç (cf. nomen gent. Mayi%iSai, Delos, 3rd cent. BCE [IG 112: 205 ff.]). While in theory the meaning of ¡ay (on which see [Spatafora 1997] and [Radif 1998]) can be reconciled with that of ¿¡¡anéroç under the assumption that ¡ay qualifies an action done in haste and therefore not achieving the desired result, the semantic distance remains considerable; I therefore prefer the interpretation advanced by le Feuvre [2015: 535-553] who argued that the original meaning of ¡ay was 'condemnable' and derived it from the root of ¡¿¡^o^ai: *mrnbh-s > 'de façon blâmable'.
6 Compare further SCr. hitar 'hastily', hitnja 'haste' from the same root as OCS xitati 'seize, snatch'; OE scëot 'quick' from the same root as Ved. cud- 'to press forth'; OHG behende 'fast' < bi hendi 'at hand'; Gk. ¿¡nX^Kiraç 'rashly, madly' from ¿¡nX^aaro 'attack'; or Italian immantinente 'immediately' < in manu tenente 'while holding in hand'.
7 Cypriot PN ma-pi-so-ni-yo (Kafizin) is probably non-existent: Neumann [1989: 168-169] suggests ]-ma pi-so-ni-yo, see also [Egetmeyer 2010: 353].
Adv. ¿¡¡arcero^ 'quickly' looks like a derivative made from an s-stem adjective (cf. aoKsXero^ 'stubbornly' from aoKsXf^ 'withered').
However, does not guarantee a simplex nominal s-stem
(cf. yevo^ : suysvf^) since the second members of verbal governing compounds of this type were often formed directly from thematic aorist stems in Greek, cf. 0u^oSaKf^ 'biting the heart' (aor. SaKe/o-), Sio0avf^ 'twice dead' (aor. 0ave/o-) or aivorca0f^ 'suffering dire ills' (aor. na0e/o-) (see [Meissner 2006]). ^ ¿¡¡arcero^ is therefore best taken as based
on thematic aor. *^ane/o-. 8
For ¡s^arcoisv at Sc. 252 several manuscripts 9 offer a different reading, namely, ¡s^aprcoisv made from a nearly synonymous and much better attested verb ¡apmro 'seize, overtake, strike'.10 Even though the form would not scan properly (¡e^apn- would fill the biceps of the fifth foot), this reading has been accepted by some authorities, notably by Jacob Wackernagel who viewed it as another instance of "avSpornTa-scansion" (viz. memrp- ^ ^ ).11 But a sixth-century poem is an unlikely place to look for a precious archaism of prosody, and the irregular scansion is better explained with Hackstein [2002a: 4, fn. 7] as "ad-hoc-Flexion" of versefinal ¡¿¡aprcsv used a few lines earlier. This said, the editors of the Shield usually print the reading ¡e^anoisv of the paradosis at Sc. 252, and with
8 See [Blanc 2018: 424]: "la dérivation directe à partir du verbe est probable, bien que la maigreur des données ne permette pas de la démontrer." In theory, ¿¡¡ansroç could also be kunstsprachlich in which case the derivational basis of the adverb would not have to be an s-stem (cf. ànxspéroç or npoqipovsroç that are not based on *ràTspfç or *npoq>povfç).
9 See F. A. Wolff apud [Ranke 1840: 225]. For the reading ¡apnssw offered by a single codex at Sc. 303 see fn. 2 above.
10 ¡флтю is well attested in early epic, lyric, and tragic poetry; the reliable уХюа-aai ката nôXsiç attribute the verb to Cypriot dialect (Kunpirov. s^apysv sXaPsv, see [Bowra 1970 (1934): 43]), and Ruijgh [1957: 166] assigned this word to the "Achaean" lexical stock. Egetmeyer [2010: 488] points out that Xa^Potvro is not attested in Cypriot inscriptions possibly because it was replaced by ¡шрлтю in everyday use as the default verb 'to take'.
11 See [Wackernagel 1914: 113 (= 1955: 1170), fn. 1], following [Ehrlich 1910: 31]; on the problem of avSpoTfra see [Barnes 2011; Maslov 2011].
good reason: verse-final ¡e^arce/o- can be unproblematically taken as a metri causa formation created from ¡arce/o- on the model of Xaxe/o-'to obtain (by lot)' : XeXaxe/o- 'to grant'. Given its ambiguity, the reduplicated form is best left aside in the discussion that follows and we can focus on the plain thematic aorist stem.
Ever since antiquity, 12 ¡ane/o- has been taken as a secondary replacement of *^aprce/6- made from verb ¡apmro, mentioned in the previous paragraph, even though only sigmatic aorist stem is attested for the latter verb in Greek. In support of this view Friedrich Schwarz [1932: 81-82] marshalled the parallel context in a Hesiodic fragment describing sons of Boreas yearning to catch the Harpies: ie^evoi] ¡apyai (fr. 150, line 29 Merkelbach-West = fr. 98 Most = fr. 63 Hirschberger).13 But even if the poet of the Shield used the verse from the Catalogue as his model and adopted the phrase with verse-initial ie^evoi (> -ai) from it, why did he replace ¡apyai with a different yet synonymous aorist infinitive? An additional problem with Schwarz's view is that an infinitive in -eeiv can only be made from a light monosyllabic root, cf. iSeeiv, ^ayeeiv, PaXeeiv [Niko-laev 2013], and while the poet of the Shield is ostensibly quite fond of the distended ending (cf. Sc. 240 rcpaQeeiv, 332 Xmeeiv, 337 eXeeiv), in order to use it here he would have to create a root ¡an- from ¡apn- by dropping the medial -p-, which defies belief as it would be an unparalleled poetic license. We do not expect the poet of the Shield—who is deeply steeped in traditional epic language—to engage in bold linguistic experiments or create completely artificial forms.
Since a linguistically compelling scenario for either a change of (unattested!) aor. *^aprce/o- to ¡ane/o- or a creation of root allomorph ¡an- from ¡apn- does not present itself,14 and the adverb ¿¡¡arcero
12 For Ioannes Pediasimus (scholiaparaphrastica on Sc. 231) ¡¡anseiv comes "ano xoC ¡¡¿pro»" with "anoPoXfl xofi p"; this formulation is mirrored by Curtius [1873: 26] who speaks of an "Ausstossung eines p".
13 Mason [2015: 30] similarly argues that the poet of the Shield adapted the line from the Catalogue; he even suggests that it may have been the same poet [2015: 310].
14 To cite one example of an unconvincing solution, Mahlow [1926: 404] assumed the following analogical proportion: Xa^P(dvro) : XaPs/o- = ¡apn(xro) : X, where X is
independently confirms the existence of a root ¡an-, it appears prudent to accept the linguistic reality of aor. ¡ane/o- and view it as etymologi-cally distinct from ¡dpmro.15 It remains entirely possible that these two very similar-sounding roots were confused in early Greek and the meaning of ¡an- was secondarily adjusted to match the semantics of ¡apn-, but due to the scarce attestation of the former root this hypothesis cannot be verified.16
What is the etymology of the root ¡an- 'seize'? 17 Since zero grade of the root is the canonical ablaut marker of thematic aorist in Greek, the default assumption is that aorist stem ¡ane/o- goes back to either *mNp-e/o- or *mNkw-e/o- made from the root *meNp- or *meNkw-,1S cf. aor. SpaKe/o- 'to see' (< *drk-) : prs. SspKe/o- (< *derk-), aor. Xine/o-'to leave' (< *likw-) : prs. Xsins/o- (< *lejkw-), aor. eX^e/o- 'to come' (< *hiluch-) : fut. ¿Xsuos/o- (< *hileudh-), etc. The hypothetical root
resolved as ¡ane/o-, the relationship between Xa^avro and Xape/o- having been reinterpreted by the speakers as showing a loss of root-internal consonant in the aorist stem.
15 Cf. [Beekes 2010: 903]: "[i]n spite of its semantic agreement with ¡apnTro, ¡a-nesiv can hardly be connected with it in formal terms."
16 The verb ¡apnTro has no etymology: ancient connection with (non-existent) ¡apn 'hand' can be safely discarded, while Egetmeyer's hesitant comparison with ¡apva^i < *merh2- is a counsel of despair [2010: 488]. I would like to draw attention to Toch. A mark- 'to take away' (which is etymologically distinct from B mark- 'to besmirch', see [Malzahn 2010: 755-756]; perhaps from *h2nmerg- 'to touch', Ved. mrj-): a privative compound sne-marklune (A 359.15) translates Skt. aharyo 'not to be taken away' and pret. III markas is found in A 120 b 5: tam prastpenu sundari nandespapsune markas "Zu dieser Zeit hatte Sundari das sittliche Verhalten des Nanda weggenommen ..." (trans. by Carling [2000: 285]). PIE *merkw- 'to snatch, take away' would be expected to give PToch. *markw- > A mark- with elimination of labial coartic-ulation before another consonant: 3 sg. aor. *merkw-s(t) > Toch. A pret. III markas, cf. *hsdkw-s 'eye' > Common Tocharian *ak > Toch. A ak, B ek (for the phonology see [Pinault 2008: 456-457]). The reconstruction *merkw- is found in [Kosling 1998: 247], but it is not supported by any comparative material.
17 Cf. [Beekes 2010: 903]: "etymology unknown"; [Blanc 2018: 424]: "pas d'etymologie".
18 Since the root is not attested in Mycenaean, the choice between root-final *kw and *p cannot be made on Greek-internal grounds.
*memp- (< *menp- with place assimilation) can be given an Indo-European etymology. I propose to analyze it as a verbal governing compound with PIE *men- 'hand' 19 as its first member (Lat. manus, -us 'hand', Olr. muim 'protection', OE mund 'id.' (< *mn-to-), Hitt. maniyahh- 'to hand over' and perhaps maninkuwa- 'near, close' 20) and PIE *hiep- 'take' as its second member (Hitt. epp-z 'grabs, seizes', Lat. apiscor 'I seize', coepi 'I began', Ved. apa 'has reached', OAlb. ep 'gives', [LIV2: 237]).
Even though the meaning of the postulated verbal governing compound *men-hip- 'take with one's hand', 'seize' 21 is fairly trivial, this reconstruction would still look more plausible if it were supported with evidence for a syntagma parallel to this compound. Such evidence may be available in Latin, Hittite, and Old English although, as usual with formulaic collocations, both constituent parts of the collocation have been subject to lexical renewal.22 In Latin where the root cap- encroached
19 For this root see the detailed discussion by Rikov [2003] not all of whose conclusions are endorsed here, as well as [Neri 2013]. The reconstruction *meh2-r, *meh2-n-'hand' ([Kroonen 2013: 375-376], with many predecessors) seems hardly viable to me: si)¡apf<; 'easy' can be explained without positing any etymological connection with the word for 'hand' (see [Blanc 1992]), while alleged Pindaric ¡ápr| 'hand' is a vox nihili (see [Forssman 1966: 135-140]), despite its resuscitation by Puhvel [2004] who compared it with Hitt. (GIS)mari- 'weapon' (which may in reality be Luw. marit-, see [Starke 1986: 162]); [Opfermann et al. 2022] take Hitt. mari- from PIE *(h2)mohi-ri- 'cutter').
20 Cop [1964: 64] was the first to compare the root of Hitt. maninkuwa- with PIE *men-, Lat. manus. The suffix of maninkuwa- has been compared to Lat. -inquus (e.g. propinquus 'near'), Gk. -ano^ (e.g. áXXoSanó^ 'belonging to another people or land; foreign' < *aliod-nkwo-) and Ved. -añc (e.g. pratyáñc- 'turned towards') ever since Benveniste [1954: 41]. An alternative analysis of maninkuwa- has recently been proposed by Frotscher et al. [in press] who have posited a nasal-infixed stem *maninku-'to shorten' made from PIE *menkw- 'to lack'.
21 According to Jochem Schindler's theory, this type of verbal governing compounds developed from original possessive compounds ('Erfassen mit Hand habend' > 'mitHand erfassen'), see [Lindner 2018: 52-54].
22 For a methodological discussion see [Gercenberg, Kazansky 2005: 1085-1087] and [García Ramón 2010].
on ep-/ap- in the meaning 'to take',23 we find manu capere (e.g. Verg. G. 3, 420; Liv. 33, 7; Plin. Nat. 13, 19, 8, etc.); in Latin legal language we also find mancipium 'laying hold of a thing; ownership' (cf. Var. L. 6, 85 mancipium, quod manu capitur) and manceps 'contractor, renter' (quasi *men- + *keh2p-). In Hittite the default word for 'hand' is ke/issar,24 and the verb ep(p)- / app- is frequently construed with it to form phrases meaning 'take with one's hand'. 25 Finally, the root *hiep- 'take' was lost in Germanic, but the second member of the Old English compound mund-gripe (m.) 'grasp, grip of the hand' may be viewed as its "Ersatzkontinuant" 26
On the basis of this data we can theorize that beside the well-established PIE collocation 'to place in one's hand' (*ghes- + *dhehi-: Ved. haste / hastayoh dha-, YAv. zastaiio da-, Gk. ev xsipi / xspoi 0n-, Hitt. kissari da-; 27 *men- + *dhehi-: Lat. mandare, Osc. manafum 'to hand over' 28) Proto-Indo-European had a collocation with the opposite meaning, namely, *men- + hip- 'take with one's hand'. It served as the basis of an (originally possessive) compound *m(e)n-hip- which, in turn, evolved into a neo-root *menhip- 'take hold of, seize'. This process would be fully parallel to the development of PIE collocation *kred(s)- + *dhehi- 29 that served to express the trust between strangers in a hospitality
23 In accordance with Kurylowicz's Fourth Law of Analogy, Lat. ep-/ap- was relegated to a secondary semantic function reflected in apio 'fasten', coepi 'began', etc.
24 Even though Anatolian preserved a reflex of PIE *men- in Hitt. maniyahh-, etc., see above.
25 E.g. KBo 25.1 a 2 LU-as assu kissarta epzi 'the man takes good by hand'; KUB 12.63 Vs. 26 UR.BAR.RA kissarta epten 'catch a wolf by hand'; KUB 45.3 i 10-11 nu LUAZU GUB 1 MUSEN epzi ZAG-it=ma=z kissarit 'the magician grabs one bird with his left hand'; KBo 3.13 rev. 14 SU-mit eppun 'I seized with my hand'.
26 E.g. Beowulf380: Dxt he prittiges manna mxgencrxft on his mundgripe hxbbe 'he possesses the might of thirty men in the grip of his hand'.
27 For the reconstruction of this collocation see [Gercenberg 1972: 113; Schutzeichel 2014: 211-214].
28 On Lat. mandare see [Neri 2013: 198 with fn. 105].
29 On which see [Weiss 2020: 269-280].
relationship: 30 in Indo-Iranian we find both the collocation (e.g. Ved. srad ... dhatta 'trust') and the compound (Ved. sraddha-, OAv. zrazda-), while in Italic and Celtic *kreds-dhehi- developed into new roots (Lat. credere 'to believe', OIr. creitid 'id.') which spawned further morphological derivatives 31.
We have thus arrived at a neo-root *menhip- 'take hold of, seize'.32 For a root with a punctual meaning like this we expect a root aorist: 33 in 3 sg. *(hi)e-menhip-t the laryngeal would be lost in a heavy CHCC cluster by Schmidt-Hackstein's rule, 34 hence *(hi)e-menp-t 'seized' (> *(hi)e-memp-t with place assimilation). 3 sg. aorist served as the forme de fondation from which the root allomorph *memp- /* mmp- spreads to the rest of the paradigm, giving 3 pl. *(hi)e-mmp-ent, etc.35 On the way
30 For this analysis of Vedic sraddha- see [Jamison 1996: 176-184] who elaborated on Paul Thieme's proposal.
31 See [Weiss 2020: 272, 274] for an explanation of *-s- in *kred-s- and for the demonstration that in Italic and Celtic reflexes of *kreds-dhehi- retained their compound identity.
32 In the discussion of the adverb ¿¡¡ansro^ above, it was suggested that it can be analyzed as a formation based on the thematic aorist stem *s^ane/o-, but as Sergio Neri kindly points out to me, there is another possibility: ¿¡¡an- may go back to PIE compound *en-mn-hip- 'having in the hand' whence 'immediately', cf. Italian immantinente and other semantic parallels cited in fn. 6 above.
33 Root aorist is reconstructed for the base root *hiep- in [LIV2: 237].
34 See [G. Schmidt 1973; Hackstein 2002b]. For this laryngeal loss after a sonorant (*-menhip-t > *-menp-t) one may compare, for instance, the inflectional paradigm of *genhi-mn 'birth', dat. sg. *genhi-mn-ei > *gen-mn-ei > Ved. janmane (with analogical full-grade suffix) or its derivative *genhi-mn-o- ^ *genhi-mne-h2- 'birth, lineage' > *gen-mne-h2- > *genneh2- (aino-rule, see [J. Schmidt 1895: 87-159]) > Gk. (Dor.) yewa [Hackstein 2002b: 2-3]. Another one of Hackstein's examples is Lat. verbum, Hsch. sp0er 908yyexai < *uerhi-dhhi-o-. In principle, one may also posit a laryngeal loss in composition: *m(e)n-hip- > *men-p- > *memp-.
35 I would to thank Stefan Hofler for discussing this scenario with me; in addition, Hofler points out to me that the lack of the laryngeal reflex in Ved. valh- 'to speak in riddles' (cf. Gk. sXeqiaipo^i 'deceive' and Lith. vilbinti 'to lure', [Goto 1995]) may be attributed to the same sound law: *(hi)e-uelhibh-t > *(hi)e-uelbh-t, hence
to Greek this root aorist would be remodeled as a thematic aorist (see [Willi 2018: 344]) , hence *mmp-e/o- > Greek ¡ane/o- 36.
A neat parallel to the proposed development may be found in another verbal governing compound of PIE date, namely, *m(e)ns-dhhi- 'to implement thinking' (cf. OAv. mazda- (f.), Ved. medha- 'wisdom') 37 reflected in Greek as a neo-root *menfh- 'to come to know, to understand' (cf. ¡ev0fpn 'concern') 38 with a zero-grade *mrtfh- that we find in thematic aorist ¡a0e/o-. 39
Abbreviations
1, 2, 3 — 1st, 2nd, 3rd person; adv. — adverb; aor. — aorist; dat. — dative; f. — feminine; pf. — perfect; pl. — plural; pret. — preterite; prs. — present; sg. — singular.
Dor. — Doric; Gk. — Greek; Hitt. — Hittite; Lat. — Latin; Lith. — Lithuanian; Luw. — Luwian; OAlb. — Old Albanian; OAv. — Older Avestan; OCS—Old Church
Proto-Indo-Aryan *ualbh- dissimilated as valh-. [AleW: 1434] tentatively takes the secondary root *uelhb- from a compound *uelhi-bleh2- 'Täuschung sagen'' (cf. Lith. vilti, Latv. vilt 'trick, beguile'): this attractive analysis is structurally quite similar to the one proposed in this paper for *men-hip- ( > *menp- > *memp-).
36 A different way of arriving at the same result would be to posit a present stem *menhip-ie/o- or *menhip-ne/o- in which the laryngeal would likewise be lost by Schmidt-Hackstein's rule; the aorist stem *mmp-e/o- > ¡ans/6- would be back-formed to this present stem (*memp-ie/o- > unattested or *memp-ne/o- > unattested
37 See [Schindler 1975: 266; Scarlata 2001: 256-258; Wodtko, Irslinger, Schneider 2006: 493-496].
38 Possibly with analogical restoration of -n- (*mensti- > *mesti- >> *menti-), see [Hackstein 2002a: 227-228] who also discusses the secondary allomorph *manl (p.av9ccva>).
39 Another parallel may be found in Vedic where denominative verb gopaya- 'to protect' (from compound go-pa- 'cowherd' > 'protector', see [Scarlata 2001: 303-304], cf. Iran. *gau-pana- 'id.', the ultimate source of Slavic *gi>pam> > Polish pan) was reanalyzed as made from a verbal root gop- and forms with zero-grade gup- were back-formed to this neo-root (pf. jugup-, p.p.p. gupta-, pass. gupya-), see [Mayrhofer 1986-2001: 1499-1500].
Slavonic; OE — Old English; OIr. — Old Irish; OHG — Old High German; Osc. —
Oscan; PIE—Proto-Indo-European; PToch. — Proto-Tocharian; SCr. — Serbo-Croatian; Skt. — Sanskrit; Toch. — Tocharian; Ved. — Vedic; YAv. — Younger Avestan.
References
AleW = Altlitauisches etymologisches Wörterbuch. W. Hock, R. Fecht, A. Feulner, H. Helene, E. Hill, D. S. Wodtko (eds.). Version 2.0. URL: https://alew.hu-berlin.de.
Barnes 2011 — T. G. Barnes. Homeric avSpoTfjxa Kai ijßr|v. Journal of Hellenic Studies. 2011. Vol. 131. P. 1-13.
Beekes 2010—R. S. P. Beekes. Etymological dictionary of Greek. Leiden; Boston: Brill, 2010.
Benveniste 1954 — É. Benveniste. Études hittites at indoeuropéennes. Bulletin de la Société de Linguistique de Paris. 1954. Vol. 50. P. 29-43.
Bing 2012 — P. Bing. A proto-epyllion? The Pseudo-Hesiodic Shield and the poetics of deferral. M. Baumbach, S. Bär (eds.). Brill's companion to Greek and Latin epyllion and its reception. Leiden: Brill, 212. P. 177-197.
Blanc 1992 —A. Blanc. La distribution des biens et des maux: sî^apfç et la racine *smer-. Revue des Études Grecques. 1992. Vol. 105. P. 548-556.
Blanc 2018 —A. Blanc. Les adjectifs sigmatiques du grec ancien: un cas de métamorphisme dérivationnel. Innsbruck: Institut für Sprachen und Literaturen der Universität Innsbruck, 2018.
Bowra 1970 — C. M. Bowra. TMiaaai KaxànôXsiç. C. M. Bowra. On Greek Margins. Oxford: Clarendon Press, 1970. P. 27-45 (a revised version of the article originally published in Glotta. 1959. Vol. 38. P. 43-60).
Carling 2000 — G. Carling. Die Funktionen der lokalen Kasus im Tocharischen. Berlin; New York: De Gruyter, 2000.
Curtius 1873 — G. Curtius. Das Verbum der Griechischen Sprache. Leipzig: Hirzel, 1873.
Cop 1964 — B. Cop. Zur hethitischen Schreibung und Lautung. Linguistica. 1964. Vol. 6. P. 37-76.
Egetmeyer 1992 — M. Egetmeyer. Wörterbuch zu den Inschriften im kyprischen Syl-labar. Unter Berücksichtigung einer Arbeit von Almut Hintze. Berlin; New York: De Gruyter, 1992.
Ehrlich 1910 — H. Ehrlich. Zur indogermanischen Sprachgeschichte. Königsberg, 1910.
Frotscher et al. (in press)—M. Frotscher, G. Kroonen, J. Barödal. Indo-European inroads into the syntactic-etymological interface: A reconstruction of the PIE verbal root *menkw- 'lack' and its argument structure. Historische Sprachforschung.
García Ramón 2010 — J. L. García Ramón. Reconstructing IE lexicon and phraseology: Inherited patterns and lexical renewal. S. W. Jamison, H. C. Melchert, B. Vine (eds.). Proceedings of the 21st Annual UCLA Indo-European Conference, Los Angeles, October 30th and 31st, 2009. Bremen: Hempen, 2010. P. 69-106.
Gercenberg 1972 — L. G. Gercenberg. Morfologicheskaja struktura slova v drevnikh indoevropejskikh jazykakh [The Morphological Structure of the Word in the Ancient Indo-European Languages]. Leningrad: Nauka, 1972.
Gercenberg, Kazansky 2005 — L. G. Gercenberg, N. N. Kazansky. Prajazykovaja re-konstrukcija: obshchie problemy [Reconstructing a Proto-Language: General Issues]. VestnikRossijskoj Akademii Nauk. 2005. Vol. 75. No. 12. P. 1077-1088.
Goto 1995 — T. Goto. Griechisch sXsqwípo^ai. W. Smoczyñski (ed.). Kurylowicz Memorial Volume. Part One. Cracow: Universitas, 1995. P. 365-370.
Hackstein 2002a — O. Hackstein. Die Sprachform der homerischen Epen: Faktoren morphologischer Variabilität in literarischen Frühformen: Tradition, Sprachwandel, Sprachliche Anachronismen. Wiesbaden: Reichert, 2002.
Hackstein 2002b — O. Hackstein. Uridg. CH.CC > C. CC. Historische Sprachforschung. 2002. Vol. 115. P. 1-22.
IG — Inscriptiones Graecae. Vol. 1-. Berlin: Mouton de Gruyter, 1873—.
Kösling 1998 — P. Kösling. Die griechischen primären Jotpräsentien untersucht und dargestellt nach Formenbestand, Aktionsarten und Etymologie. Hamburg: Dr. Kovac, 1998.
Kroonen 2013 — G. Kroonen. Etymological dictionary of Proto-Germanic. Leiden: Brill, 2013.
Le Feuvre 2015 — C. le Feuvre. Homeros dysgnöstos: réinterpretations de termes homériques en grec archaïque et classique. Genève: Librairie Droz, 2015.
Lindner 2018—Th. Lindner. Indogermanische Grammatik. Vol. IV/2: Komposition im Aufriss. Heidelberg: Winter, 2018.
Mahlow 1926 — G. Mahlow. Neue Wege durch die Griechische Sprache und Dichtung. Berlin; Leipzig: De Gruyter, 1926.
Malzahn 2010—M. Malzahn. The Tocharian verbal system. Leiden; Boston: Brill, 2010.
Maslov 2011 — B. P. Maslov. The metrical evidence for pre-Mycenaean hexameter epic reconsidered. Indo-European Linguistics and Classical Philology. 2011. Vol. 15. P. 376-389.
Mason 2015 — H. C. Mason. The Hesiodic Aspis: Introduction and commentary on vv. 139-237. D.Phil. Thesis. Oxford University, 2015.
Mayrhofer 1986-2001 — M. Mayrhofer. Etymologisches Wörterbuch des Altindoari-schen. Heidelberg: Winter, 1986-2001.
Meissner 2006 — T. Meissner. S-stem nouns and adjectives in Greek and Proto-Indo-European. A diachronic study in word formation. Oxford: Oxford University Press, 2006.
Nikolaev 2013 —A. Nikolaev. The aorist infinitives in -ssw in early Greek hexameter poetry. Journal of Hellenic Studies. 2013. Vol. 133. P. 81-92.
Neri 2013 — S. Neri. Zum urindogermanischen Wort für 'Hand'. A. I. Cooper, J. Rau, M. Weiss (eds.). Multi nominis grammaticus: Studies in Classical and Indo-European linguistics in honor of Alan J. Nussbaum on the occasion of his sixty-fifth birthday. Ann Arbor; New York: Beech Stave, 2013. P. 185-205.
Opfermann et al. 2022 —A. Opfermann, D. Sasseville, R. Süssenguth. Hethitisch märi-'Sichelschwert(?)' aus archäologischer, philologischer und etymologischer Sicht. Altorientalische Forschungen. 2022. Vol. 49. P. 104-122.
Pinault 2008 — G.-J. Pinault. Chrestomathie tokharienne: textes et grammaire. Leuven: Peeters, 2008.
PMGF — M. Davies (ed.). Poetarum melicorum Graecorum fragmenta. Vol. 1. Oxford: Clarendon Press, 1991.
Puhvel 2004 — J. Puhvel. Vox ex nihilo: Greek sy%oç. D. Groddek, S. Rößle (eds.). Sarnikzel: Hethitologische Studien zum Gedenken an Emil Orgetorix Forrer 19.02.1894-10.01.1986. Dresden: Verlag der technischen Universität Dresden, 2004. P. 531-532.
Radif 1998 — L. Radif. Significato e funzione di alcuni avverbi omerici in -£, -y. U. Rapallo, G. Garbugino (eds.). Grammatica e lessico delle lingue «morte». Alessandria: Ed. dell'Orso, 1998. P. 33-46.
Ranke 1840 — C. F. Ranke. Hesiodi quod fertur Scutum Herculis ex recognitione et cum animadversionibus Fr. Aug. Wolfii. Quedlinburg; Leipzig, 1840.
Rikov 2003—G. T. Rikov. Cornish manal 'sheaf', Latin manus 'hand' and connected problems. Studia Etymologica Cracoviensia. 2003. Vol. 8. P. 149-157.
Ruijgh 1957 — C. J. Ruijgh. L'élément achéen dans la langue épique. Assen: van Gorcum, 1957.
Scarlata 1999—S. Scarlata. Die Wurzelkomposita im Rg-Veda. Wiesbaden: Reichert, 1999.
Schindler 1975 — J. Schindler. Zum Ablaut der neutralen s-Stämme des Indogermanischen. H. Rix (ed.). Flexion und Wortbildung. Akten der V. Fachtagung der Indogermanischen Gesellschaft. Wiesbaden: Reichert, 1975. P. 259-267.
Schmidt 1895 — J. Schmidt. Kritik der Sonantentheorie. Weimar: Böhlau, 1895.
Schmidt 1973 — G. Schmidt. Die iranischen Wörter für "Tochter" und "Vater" und die Reflexe des interkonsonantischen H (a) in den idg. Sprachen. Zeitschrift für vergleichende Sprachforschung. 1973. Vol. 87. P. 36-83.
Schutzeichel 2014 — M. Schutzeichel. Indogermanische Funktionsverbgefüge. Münster: Monsenstein und Vannerdat, 2014.
Schwarz 1932 — F. G. Schwarz. De Scuto quod fertur Hesiodi: quaestiones ad com-positionem et dicendi genus maxime pertinentes. Berolini: Pilz & Noack, 1932. Schwyzer 1939 — E. Schwyzer. Griechische Grammatik. Vol. 1: Allgemeiner Tell.
Lautlehre. Wortbildung. Flexion. München: Beck, 1939. Spatafora 1997 — G. Spatafora. Sul significato di ¡¡ay e dei suoi derivati in eta arcaica e classica. Parnassos. 1997. Vol. 39. P. 247-254. Sitzler 1883 — J. Sitzler. Review of T. Bergk. Poetae lyrici Graeci III, 4th ed. Leipzig
1882. Philologische Rundschau. 1883. Vol. 3. Fasc. 30. P. 920-938. Starke 1986 — F. Starke. Review of Chicago Hittite dictionary, vol. 3, fasc. 2.
Bibliotheca Orientalis. 1986. Vol. 43. P. 157-165. Swift 2019 — L. Swift. Archilochus: The poems. Introduction, text, translation, and
commentary. Oxford: Oxford University Press, 2019. Tichy 1976—E. Tichy. Gr. SsiSsxaxo und idg.*dekti, dektoi. Glotta. 1976. Vol. 54. P. 71-84.
Wachter 2001 — R. Wachter. Non-Attic Greek vase inscriptions. Oxford: Oxford University Press, 2001.
Wackernagel 1914 — J. Wackernagel. Akzentstudien III: Zum homerischen Akzent. Göttinginsche Gelehrte Nachrichten. 1914. Vol. 1014. P. 97-130 (reprinted in Kleine Schriften. Göttingen: Vandenhoeck & Ruprecht, 1955. Vol. 2. P. 11541187).
Weiss 2020 — M. Weiss. The inner revolution: old but not that old. M. Serangeli, Th. Olander (eds.). Dispersals and diversification: Linguistic and archaeological perspectives on the early stages of Indo-European. Leiden; Boston: Brill, 2020. P. 263-288.
West 1990 — M. L. West. Review of Greek Lyric 2. The Classical Review. 1990. Vol. 40. P. 214-216.
Wodtko, Irslinger, Schneider 2008—D. S. Wodtko, B. Irslinger, C. Schneider. Nomina im indogermanischen Lexikon. Heidelberg: Winter, 2008.
