Zachary Rothstein-Dowden (Harvard University)
The verb ἐγρήγορα ‘am awake, am vigilant’ appears in Homer only as 2pl. ipv. ἐγρήγορθε (Η 371 = Σ 299), inf. ἐγρήγορθαι (Κ 67), 3pl. ἐγρηγόρθᾱσι (Κ 419) and the participle ἐγρηγορόων (υ 6). In Attic we find ἐγρήγορα ‘am awake’ (ἐγηγερμένος v.l. Th.7.51, Speus.Fr.28+). The Homeric forms present four problems: 1) the unetymological ρ of the reduplicated syllable. 2) The position of the accent in ἐγρήγορθαι. 3) the use of middle voice in ἐγρήγορθε and ἐγρήγορθαι. 4) the unetymological θ in ἐγρηγόρθᾱσι. The talk will address these four contentious issues.
1-2) Comparison with Vedic jāgā́ra ‘is awake,’ the reduplicated vowel of which is invariably long, paired with the prothetic ἐ- of Greek make it clear that the root began with a laryngeal and that the perfect goes back to *h1ge-h1gór-/*h1ge-h1gr-́. This regularly would have given 3sg. *ἐγήγορ-ε : 3pl. *ἐγήγρ-ατι. It is usually said that the source of the ρ was contamination with the thematic aorist ἐγρέσθαι (GEW I:438, LIV2 246). Α 3pl. *ἐγήγρ-ατι anticipatorily metathesized to *ἐγρήγατι and subsequently contaminated with 3sg. *ἐγήγορ-ε could well have given 3sg. ἐγ[ρ]ήγορ-ε : 3pl. (*)ἐγρηγ[όρ]ασι. The unextended stem ἐγρήγ- (←*ἐγηγρ-) would then be behind ἐγρήσσω (Λ 551+) ‘keep watch’ (like πέπληγα : πλήσσω), a form that has till now lacked an explanation. A stem ἐγρήγ- is further suggested by the “Aeolic” accent of ἐγρήγ[ορ]θαι. The form ἐγηγερμένος is not an archaism. It is simply a replacement of ἐγρηγόρως made productively to pres. ἐγείρω (so Brugmann 1900:260).
3) It is difficult to see why 2pl. ipv. ἐγρήγορθε should have replaced “correct” *ἐγρήγορ-τε, which has the same metrical shape and is morphologically easy to parse. Leeuwen (1918:310) explains ἐγρήγορθε as resulting from a syncope ἐγρηγόρατε(!) → ἐγρήγορθε. Wackernagel (1895:32) suggests that a 2sg. ind. *ἐγρήγορ-θα[ς] produced 2pl. ipv. ἐγρήγορθε through a weakly-motivated analogy. Schwyzer (1939:800n1), Chantraine (1958:429) and Risch (1974:347) maintain that ἐγρήγορθε was formed in imitation of 2sg. ipv. *ἐγρήγορθι. Hackstein (2002:248) posits an undermotivated preform *ἐγρήγορ-σ-τε. None of these explanations is compelling. The morphological makeup of this form becomes apparent from the context:
Η 370—71 νῦν μὲν δόρπον ἕλεσθε κατὰ πτόλιν ὡς τὸ πάρος περ,
καὶ φυλακῆς μνήσασθε καὶ ἐγρήγορθε ἕκαστος·
‘Now have your meal throughout the city as before, and have thought of the defense, and be each of you alert.'
Expected *ἐγρήγορτε has been replaced by middle ἐγρήγορθε under the influence of the preceding imperatives ἕλε-σθε ... μνήσα-σθε ... ἐγρήγορ-(σ)θε. The inf. ἐγρήγορθαι and 3pl. ἐγρηγόρθᾱσι (cf. βεβρίθᾱσι ο 334) both occur in the Doloneia, a book generally considered to be a late addition to the epic. The middle infinitive ἐγρήγορθαι [= ἐγρήγορ-(σ)θαι] occurs in the same position in the verse as ἐγρήγορθε and copies its middle voice.
4) The -θ- of 3pl. ἐγρηγόρθᾱσι resulted from a misparsing of 2pl. ἐγρήγορ-θε ἕκαστος as 2sg. them. ipv. ἐγρήγορθ-ε ἕκαστος due to the singular subject following the verb (cf. thematic pf. ipv. κεχήν-ε-τε Ar. to athematic κεχήν-α). This suggested a paradigm ἐγρήγορ(-)θ-α, -ας, -ε ... -ᾱσι like βέβρῑ(-)θ-α, -ας, -ε ‘is heavy.’