William Watson Goodwin, Syntax of the Moods and Tenses of the Greek Verb

---

[*] 671. In Homer this construction (669) is fully developed in indirect questions: see examples of both indicative and optative in 669, 1 and 2. But in indirect quotations, while the indicative is freely used after both present and past tenses, the change of the indicative to the optative after past tenses had not yet been introduced. In the single case of εἰπεῖν ὡς with the optative, μερμήριξε . . ἕκαστα εἰπεῖν, ὡς ἔλθοι καὶ ἵκοιτ᾽ ἐς πατρίδα γαῖαν, he hesitated about telling him each event, how he had returned, etc., Od. xxiv. 237, ὡς appears only on its way from its meaning how (663, Od. 2) to its later use with the optative as that. We first find the optative in genuine oratio obliqua (with ὡς) Hymn. Ven. 214, εἶπεν ὡς ἔοι. Further, the later principle by which the indicative after past tenses (when it is not changed to the optative) retains the tense of the direct form is almost unknown in the Homeric language. Here a present or perfect indicative of the direct discourse after a past tense is changed to an imperfect or pluperfect; so that I knew that he was planning evil, which in Attic would be ἐγίγνωσκον ὅτι κακὰ μήδοιτο (or μήδεται), in Homer is γίγνωσκον ὅ (= ὅτι) κακὰ μήδετο, Hom. Od. iii. 166. (For examples, see 674.) The aorist indicative, which has no corresponding tense to express its own time referred to the past, was always retained after past tenses; as in γνῶ ὅ οἱ οὔτι ἦλθεν, Hom. Il. xi. 439; so i. 537, Hom. Il. xxii. 445.Likewise the future indicative is once retained, in Hom. Od. xiii. 340, ᾔδἐ ὃ νοστήσεις, I knew that you would return; but elsewhere the past future with ἔμελλον is used, as in Hom. Il. xx. 466, οὐδὲ τὸ ᾔδη ὃ οὐ πείσεσθαι ἔμελλεν, and Hom. Od. xix. 94, Hom. Il. xi. 22.These examples show the need of the later future optative (129). In Hom. Il. xxii. 10, οὐδέ νύ πώ με ἔγνως ὡς θεός εἰμι, and xx. 265 the present expresses a present truth rather than a past fact.

It thus appears that the peculiar constructions with ὅτι and ὡς in oratio obliqua (667, 1, b), which gave such grace and variety to the later language, were not yet developed in Homer; but clauses with ὅτι, ὡς, etc., were still connected with the leading verb by the same looser construction which we use in English (as I knew that he was planning evil), the dependent verb expressing its own absolute time (see 22), as it did in the relative clauses in which these clauses originated, or in the more primitive parataxis. Thus γίγνωσκον ὃ κακὰ μήδετο (above) meant originally I knew as to what he was planning evil; and without ὅ, in a still earlier stage, I knew: he was planning evil (which we can say in English). Even after the more thorough incorporation of the dependent clause was established, by which either μήδεται or μήδοιτο became the regular form, the more primitive imperfect is occasionally found, even in Attic prose (see 674, 2).

The most common Homeric construction in indirect discourse is that of φημί with the infinitive, of which 130 examples occur.¹