Xian ————— Greek, Roman, and Byzantine Studies 62 (2022) 255–262 Article copyright held by the author(s) and made available under the Creative Commons Attribution License CC-BY https://creativecommons.org/licenses/by/4.0/ Homeric δέχαται (Il. 12.147) Reconsidered Ruobing Xian LIAD 12 DESCRIBES three separate assaults by the Trojan alliance on the Achaean wall. The first is initiated by Asius, who is surprised by the stubborn resistance of two Lapiths, Polypoites and Leonteus, compared to a pair of wild boars awaiting the attack of men and dogs (ἀνδρῶν ἠδὲ κυνῶν δέχαται κολοσυρτὸν ἰόντα, 12.147). In the entirety of the Homeric epics, the verbal form δέχᾰται (“they await,” from δέχοµαι) is attested only in this simile. It is well known that Homeric δέχοµαι has both a perfective meaning (“take/accept”) and a durative one (“await/wait for”).1 The use of δέχαται here is clearly durative.2 As Debrunner 1 See A. Debrunner, “Δέγµενος, ἑσπόµενος, ἄρχµενος,” in H. Kronasser (ed.), Μνήµης χάριν: Gedenkschrift Paul Kretschmer (Vienna 1956) 77–84, at 77, together with detailed bibliography. My central argument here does not depend on the genesis of the Homeric root δεχ-, which is found in place of the original δεκ-. This can be explained either by the influence of ἔχω or by invoking the athematic inflection in the third person plural (cf. O. Szeme- rényi, Syncope in Greek and Indo-European and the Nature of Indo-European Accent [Naples 1964] 173 n.2, with bibliography). In response to Wackernagel’s criticism of the former view (Sprachliche Untersuchungen zu Homer [Göttingen 1916] 24–25), Szemerényi pointed to the importance of the middle forms such as ἀνέχοµαι, ὑπίσχοµαι, and παρέχοµαι, respectively matched by ἀνα- δέχοµαι, ὑποδέχοµαι, and παραδέχοµαι. 2 This is clear from a Homeric parallel. The simile in which δέχαται occurs (12.141–153) is contextually similar to Il. 13.470–477. In the latter, δέχαται is paralleled by µένει, whose durative sense is indisputable: compare ὡς ὅτε τις σῦς οὔρεσιν ἀλκὶ πεποιθώς, / ὅς τε µένει κολοσυρτὸν ἐπερχόµενον πολὺν ἀνδρῶν / χώρῳ ἐν οἰοπόλῳ (13.471–473) to ἀγροτέροισι σύεσσιν ἐοικότε, τώ τ’ ἐν ὄρεσσιν / ἀνδρῶν ἠδὲ κυνῶν δέχαται κολοσυρτὸν ἰόντα (12.146–147). I 256 HOMERIC ΔΕΧΑΤΑΙ (IL. 12.147) RECONSIDERED ————— Greek, Roman, and Byzantine Studies 62 (2022) 255–262 showed,3 the durative meaning of δέχοµαι “is confined to the perfect forms and δεγµ-.”4 Scholarly opinion is divided on the morphological analysis of δέχαται. On the one hand, some ancient critics already interpreted it as a shortened perfect to δέχοµαι.5 δέχαται, on the other, has also been construed as an athematic root-present.6 The evidence for the athematic root-present of δέχοµαι in the Homeric epics is controversial. Leaving δέχαται aside, such a stem is said to be found in some isolated Odyssean imperfects (ἐδέγµην, Od. 9.513, 12.230; ἀνεδέγµεθ᾽, 17.563) and in the par- ticiple δέγµενος (21 times),7 which mainly occurs in compounds 3 Debrunner, in Μνήµης χάριν 78–79. 4 The phrase is borrowed from Szemerényi, Syncope in Greek and Indo- European 172. 5 Aelius Herodian Περὶ παθῶν GG III.2 225.7: τὸ Ἰωνικὸν νενέαται· ἆρον τὴν νε συλλαβὴν “νέαται Πύλου ἠµαθόεντος” [Il. 9.153] ὥστε ὅµοιον εἶναι τῷ δεδέχαται δέχαται. Cf. Debrunner, in Μνήµης χάριν 80–81; E. Risch, Wort- bildung der homerischen Sprache (Berlin 1974) 342. 6 E. Schwyzer, Griechische Grammatik I (Munich 1939) 678; P. Chantraine, Grammaire homérique I (Paris 2013 [1942]) 286: “L’existence d’un présent athématique *δέγµαι est assurée par la forme δέχαται en Μ 147, troisième personne du pluriel athématique, où le κ a reçu l’aspiration d’après l’analogie de formes comme δέχθε (qui n’est pas attesté chez Homère), ou δέχθαι”; G. P. Shipp, Studies in the Language of Homer2 (Amsterdam 1966) 63; J. Narten, “Zum proterodynamischen Wurzelpräsens,” in J. C. Heesterman et al. (eds.), Pratidānam: Indian, Iranian and Indo-European Studies Presented to F. B. J. Kuiper (The Hague 1968) 9–19, at 15 n.43; B. Hainsworth, The Iliad: A Commentary III (Cambridge 1993) 334. According to E. Tichy, “Gr. δειδέχατο und idg. *dḗḱti, *déḱtoi̯,” Glotta 54 (1976) 71–84 (= Kleine Schriften [Bremen 2019] 1–14), δέχαται is an archaism, which, together with the Homeric syntagma ἧσται δέγµενον, preserves the original meaning of the root **deḱ- (“wahrnehmen”), while the epic forms such as δειδέχατ(ο), δείδεκτ(ο), δειδέχαται, δειδίσκετο meaning “welcome” represent the causative reduplicated aorist *δεδέχθαι of the same root. For a criticism of Tichy’s thesis see esp. B. Forssman, “Homerisch δειδέχαται und Verwandtes,” Sprache 24 (1978) 3–24, at 17 n.41. 7 Cf. Chantraine, Grammaire homérique 286, who also took Homeric δέκτο into account (Il. 2.420, 15.88; Od. 9.353). Scholars now agree, however, that RUOBING XIAN 257 ————— Greek, Roman, and Byzantine Studies 62 (2022) 255–262 such as ποτιδέγµενος (15 times) and ὑποδέγµενος (2 times).8 Debrunner, however, argued that δέγµενος was a shortened present participle which replaced the metrically impossible *δεχόµενος, while ἐδέγµην (Od. 9.513, 12.230) and ἀνεδέγµεθ᾽ (Od. 17.563) had been created on the basis of δέγµενος.9 Similarly, Szemerényi suggested that “δέγµην, δέγµεθα would easily follow, once δέγµενος established its place,” although he conceived of δέγµενος as a shortened variant of the well-attested perfect par- ticiple δεδεγµένος.10 In light of Debrunner and Szemerényi, the root-present inter- pretation of δέχαται as an archaism cannot be supported with any reliable Homeric parallels. In contrast, I would like to suggest that δέχαται is an artificially-built present. The creation of such an artificial form could have been motivated by two sources. The first is the durative use of δέγµενος. Irrespective of δέκτο “represents either a root aorist or a pre-alphathematic s-aorist (i.e. δέκτο < *deḱ-s-to),” since δέκτο is semantically equivalent to Homeric (ἐ)δέξατο: A. Willi, Origins of the Greek Verb (Cambridge 2018) 305–306, together with detailed bibliography. 8 A full list of attestations is found in Debrunner, in Μνήµης χάριν 78–79; cf. Szemerényi, Syncope in Greek and Indo-European 172 n.2. Unlike Debrunner, I have limited my counting to the Iliad and the Odyssey, leaving aside the attestations of δέγµενος in Hymn.Hom.Dem. 29 and Hymn.Hom.Herm. 477. 9 Debrunner, in Μνήµης χάριν 79: “δέγµενος ersetzt das hexametrisch unmögliche δεχόµενος, und daran haben sich die vereinzelten ἐδέγµην und ἀνεδέγµεθα angeschlossen als Ersatz für die ebenfalls metrisch unbrauch- baren Formen ἐδεχόµην (δεχόµην wäre möglich, fehlt aber bei Homer) und (ἐ)δεχόµεθ(α)!” 10 Szemerényi, Syncope in Greek and Indo-European 172–173: “It follows there- fore that at one time a poet was led to extract an anomalous δέγµενος from the regular and well-established δεδεγµένος, probably on the strength of the observation that there was no (clear or great) difference between δεδέξοµαι with reduplication and δέξοµαι without.” In the Homeric epics, δέγµενος (Il. 2.794, 9.191, 18.524) is used at verse-initial position, with only one exception (Od. 20.385). One might also note that δεδεγµένος (Il. 4.107, 8.296, 10.62, 11.124, 23.273) always occupies the same metrical position between the trochaic and the bucolic caesurae (|tr δεδεγµένος/δεδεγµένα |buc – ⏑ ⏑ – x#). 258 HOMERIC ΔΕΧΑΤΑΙ (IL. 12.147) RECONSIDERED ————— Greek, Roman, and Byzantine Studies 62 (2022) 255–262 its origin, whether it is ultimately a shortened form of the metrically impossible *δεχόµενος or a shortened perfect participle (δέγµενος < δεδεγµένος),11 δέχαται could have been invented by bards who mistook (ποτι/ὑπο)δέγµενος for an athematic present participle. This interpretation coheres with Debrunner and Szemerényi on how ἐδέγµην (Od. 9.513, 12.230) and ἀνεδέγµεθ᾽ (Od. 17.563) came into being. The case of ἕαται/εἵατο (from ἧµαι “sit”) provides a second possible source of motivation. The similarities between ἕαται (Il. 3.134, 9.628) and δέχαται (Il. 12.147) are striking. Both are in the third person plural, and both have a similar metrical shape (⏑ ⏑ –).12 Most importantly, in the Homeric epics ἕαται/εἵατο is used in close connection with (ποτι)δέγµενοι: αἳ δέ που ἡµέτεραί τ’ ἄλοχοι καὶ νήπια τέκνα / εἵατ’ ἐνὶ µεγάροις ποτιδέγµεναι (Il. 2.236– 237); οἵ που νῦν ἕαται ποτιδέγµενοι (9.628); τοῖσι δ’ ἔπειτ’ ἀπάνευθε δύω σκοποὶ εἵατο λαῶν / δέγµενοι ὁππότε µῆλα ἰδοίατο καὶ ἕλικας βοῦς (18.523–524); Τηλέµαχ’, ἤδη µέν τοι ἐϋκνήµιδες ἑταῖροι / εἵατ’ ἐπήρετµοι, τὴν σὴν ποτιδέγµενοι ὁρµήν (Od. 2.402–403); νῆες ἐΰσσελµοι µένον ἁθρόαι, ἀµφὶ δ’ ἑταῖροι / εἵατ’ ὀδυρόµενοι, ἡµέας 11 Such a dereduplicated perfect form might be paralleled by Hom. ἔρχαται (Il. 16. 481; Od. 10.283) and ἔρχατο (Il. 17.354; Od. 9.221, 14.73), as Schwyzer, Griechische Grammatik 767, has noted; contra E. Tichy, Onomatopoetische Ver- balbildungen des Griechischen (Vienna 1983) 287 n.165. One of the GRBS referees points to the fact that dereduplication after a preverb is a well-known phenomenon in ancient Greek. Cf. Schwyzer 766–767 with n.3, who cites Gortyn Cretan καταϝελµένος (I.Cret. IV 72.X.35 = DGE 179) in contrast to Hom. ἐέλµεθα (Il. 24.662): “die Umschreibung ϝηλ- (aus *ϝε-ϝελ-) […] oder aoristische Auffassung […] sind unnötig.” More examples can be found in D. Ringe, The Perfect Tenses in Greek Inscriptions (diss. Yale 1984), including Cretan προγραµµένον (I.Cret. I XVI 4.53) and ἀνκληµένας (I XVII 2.3), and Cnidian ἀφορισµένους (Ager, Arbitrations 71.I.34; all 2nd cent. BCE). In light of this phenomenon it is possible that δέγµενος (4 times) was extracted from de- reduplicated ποτι/ὑποδεγµενος (17 times), reanalyzed as an aorist participle, and therefore transmitted with an aorist accent. 12 Lex.frühg.Epos s.v. ἧµαι M 3; contra Shipp, Studies in the Language of Homer 25, who, following the lead of F. Bechtel, Die griechischen Dialekte III (Berlin 1924) 46, took ἕαται as disyllabic. RUOBING XIAN 259 ————— Greek, Roman, and Byzantine Studies 62 (2022) 255–262 ποτιδέγµενοι αἰεί (9.544–545).13 In other words, ἕαται could have further contributed to the genesis of δέχαται, not least due to the former’s connection with (ποτι)δέγµενοι, which is the main source of the latter. In contrast to Debrunner’s view that “[a]ls Präsens ist δέχαται ganz singulär: […] nur in reduplizierten Formen kennt Homer -αται -ατο nach Konsonant,”14 the morpho- logically, metrically, and semantically similar forms ἕαται (Il. 3.134, 9.628) and κέαται (from κεῖµαι “lie”)15 could make it easier for δέχαται to be regarded by both the bards and the audience as a present third-person plural. One potential objection might be raised against this inter- pretation. The form ἕαται (⏑ ⏑ –) is not the expected outcome of *ἥαται, which by quantitative metathesis should have given *ἕᾱται.16 To clarify it, Nussbaum has argued that ἕαται (⏑ ⏑ –) might be analogical after “δέχᾰται (Μ 147) etc. and perhaps especially κέᾰται.”17 In my view, it is better to assume that the model for ἕᾰται was κέᾰται, as Nussbaum himself apparently preferred,18 while the hapax δέχᾰται was modelled on ἕᾰται and κέᾰται. To sum up, δέχαται (Il. 12.147) does not represent a real athematic root-present; rather, it is an artificial form created by 13 Cf. also Il. 9.190–191 (Πάτροκλος δέ οἱ οἶος ἐναντίος ἧστο σιωπῇ, / δέγµενος Αἰακίδην ὁπότε λήξειεν ἀείδων) and Od. 23.91 (ἧστο κάτω ὁρόων, ποτιδέγµενος εἴ τί µιν εἴποι). 14 Debrunner, in Μνήµης χάριν 81. 15 ἐν νηυσὶν κέαται βεβληµένοι οὐτάµενοί τε (Il. 11.659 = 11.826 = 16.24); νῦν δ’ οἳ µὲν κέαται δεδαϊγµένοι, οὓς ἐδάµασσεν (19.203). 16 I am grateful to one of the GRBS referees who alerted me to this potential problem. 17 A. Nussbaum, Two Studies in Greek and Homeric Linguistics (Göttingen 1998) 62. 18 Cf. also Nussbaum, Two Studies 62 n.213, who argued that the two attestations of ἕᾰται (# οἳ δἠ νῦν ἕαται |5, Il. 3.134; # οἳ που νῦν ἕαται |5, 9.628), “because of the phonologically recent shape of the verb, might reasonably be expected to have an identifiable model, which might very well be: # νῦν δ’ οἳ µὲν κέαται |5 (Τ203).” 260 HOMERIC ΔΕΧΑΤΑΙ (IL. 12.147) RECONSIDERED ————— Greek, Roman, and Byzantine Studies 62 (2022) 255–262 the mechanism of analogy. I have argued that δέχαται was created through a bardic reinterpretation of δέγµενος, which was thought to be an athematic present participle. As a result, an athematic present stem δεχ-, in opposition to the normal thematic one δεχ-ε/ο, was taken for granted,19 and δέχαται was created. The invention of δέχαται could have also been facili- tated by the similarities between ἕαται/κέαται and δέχαται. The mechanism of analogy can be described as follows: (καθ)ήµενος : (ποτι/ὑπο)δέγµενος = ἕαται : X, where X = δέχαται.20 Admittedly, though δέχαται is an artificial form from a dia- chronic standpoint, it is ‘natural’ for the bards and the audience, who were familiar with forms such as δέγµενος, ἐδέγµην, and ἕαται/κέαται. To my mind, this explanation is preferable to the view that δέχαται is a shortened perfect, not least because the indicative perfect form δέδεκται/δεδέχαται, on the basis of which δέχαται is said to be formed, is not attested in Homer.21 19 Taking δέγµενος as a shortened form of δεδεγµένος, Szemerényi, Syncope in Greek and Indo-European 173, came to basically the same conclusion as mine, without specifying the exact creation process of δέχαται: “And within this framework it was no longer quite so absurd to coin a 3rd plural δέχαται instead of the normal δέχονται.” To the best of my knowledge, Szemerényi’s short comment on δέχαται did not have any effect on scholarship. Further- more, my interpretation differs from Szemerényi’s in my emphasis on the significance of ἕαται. 20 The long vowel in (καθ)ήµενος corresponds to that in *ἥαται (MS. εἵαται), a form which by quantitative metathesis should have given *ἕᾱται (the attested ἕᾰται, as already observed, is an artificial form built upon the model provided by κέᾰται). 21 Apart from the well-attested perfect participle δεδεγµένος, the perfect attestations to δέχοµαι in the Homeric epics are confined to δέδεξο (Il. 5.228, 20.377, 22.340) and δεδέξοµαι (5.238). According to M. Leumann, “Griech. hom. εἰδώς ἰδυῖα und ἐοικώς ἐϊκυῖα, ἀρηρώς ἀραρυῖα,” Celtica 3 (1955) 241– 248, at 242–243 n.5 (= Kleine Schriften [Zurich 1959] 215–258, 252 n.5), δεδέξοµαι (Il. 5.238) is an ad hoc formation built directly upon the imperative form δέδεξο (Il. 5.228). It is possible that #δειδέχατ’ (Il. 4.4, 9.671, 22.435) RUOBING XIAN 261 ————— Greek, Roman, and Byzantine Studies 62 (2022) 255–262 If my main argument presented in this article can be accepted, two additional points might be made. On the one hand, it challenges Narten’s influential thesis that δέχαται as an archaism represents “das erwartete, vollstufig-wurzelbetonte Medium,” which is said to be the counterpart of the long-grade Vedic present dāṣṭi “await” (RV 1.127.4b, 1.127.4c).22 On the other hand, it is important to point out that δέχαται and its cognates (δέγµενος, ἐδέγµην, ἀνεδέγµεθ᾽) as kunstsprachliche alternatives to a thematic paradigm can hardly be isolated cases in Homer. Particularly relevant are γέντο “seized/took” (Il. 8.43, 13.25, 13.241, 18.476, 18.477) and λέκτο “counted” (Od. 4.451), both of which have in more recent scholarship23 been interpreted as and #δειδέχαται (Od. 7.72) are originally perfect/pluperfect forms of δέχοµαι (cf. L. Meyer, “Δειδέχαται, δεικανάοµαι, und δειδίσκοµαι,” Beiträge zur Kunde der indogermanischen Sprachen 2 [1878] 260–264, at 262–263), with a lengthened reduplication-syllable metri causa (due to their verse-initial position), though the majority of scholars, following the lead of W. Schulze, Quaestiones epicae (Gütersloh 1892) 155 (“quam formam [δειδέχαται] non tam perfecto quae est Osthoffi sententia (Perfect 57 sq.) quam praesenti intensivo assigno”), now take them to represent an intensive present stem (*dek-dek-> *dei̯-dek-) to the same Indo-European root *deḱ- (F. Škoda, Le redoublement expressif: un universal linguistique. Analyse du procédé en grec ancien et en d’autres langues [Paris 1982] 174– 175; Tichy, Onomatopoetische Verbalbildungen 296; Willi, Origins of the Greek Verb 163–164). 22 Narten, in Pratidānam 15 n.43, accepted by, among others, H. Rix, Lexikon der indogermanischen Verben (Wiesbaden 2001) 110 with n.8. 23 J. Harđarson, Studien zum urindogermanischen Wurzelaorist und dessen Ver- tretung im Indoiranischen und Griechischen (Innsbruck 1993) 206–207, has argued that γέντο, which always occurs at the same metrical position (γέντο δ᾽ ἱµάσθλην#, Il. 8.43 = 13.25; γέντο δὲ δοῦρε#, 13.241; γέντο δὲ χειρὶ#, 18.476; γέντο πυράγρην#, 18.477), is a “metrisch bedingte Kunstbildung für ein aoristisch verwendetes Impf. *(ἐ)γέµετο (cf. kypr. ἀπογέµω)” (from γέµω “be full of”). Similarly, Willi, Origins of the Greek Verb 322, has pointed to the pos- sibility that some isolated athematic forms to λέγω (from *leǵ- “count, reckon”) such as ἐλέγµην (Od. 9.335) and λέκτο (4.451) might be “altogether artificial.” Αs has been noted by S. West, A Commentary on Homer’s Odyssey I (Oxford 1988) 221, ad Od. 4.451, there is a deliberate word-play between λέκτο “he counted” 262 HOMERIC ΔΕΧΑΤΑΙ (IL. 12.147) RECONSIDERED ————— Greek, Roman, and Byzantine Studies 62 (2022) 255–262 artificial forms.24 December, 2022 Fudan University Shanghai, China ruobingxian@fudan.edu.cn (4.451), which is taken up by λέγε “he counted” (4.452), and λέκτο (4.453, from λέχοµαι “lie down”); both λέκτο (4.451) and λέκτο (4.453) are placed at the same metrical position (λέκτο δ’ ἀριθµόν#, 4.451; λέκτο καὶ αὐτός#, 4.453). 24 I wish to thank GRBS’s anonymous referees for their constructive com- ments and the editorial board for helpful suggestions.