Let's say you want to translate the following English sentence to Attic Greek.

We believed that they were good friends, for whithersoever this one went, that one also went.

The first clause calls for indirect discourse with an accusativus cum infinitivo. The latter part, beginning with the conjunction "for", calls for a conditional relative clause.

αὐτοὺς ἐνομίσαμεν τοὺς ἀγαθοὺς φίλους εἶναι, ὁπόθεν γὰρ ὁ μὲν ἔρχεται, ὁ δέ καὶ ἔρχεται.

I feel like my above translation is mostly correct, but there are a few points that leave me unsure.

  1. Is τοὺς ἀγαθοὺς φίλους a correct way of describing good friends? Would the adjective καλοὺς be a better choice here?
  2. Is my choice of tenses correct? I put the main verb ἐνομίσαμεν in the aorist tense, whereas the following verbs are in the present tense, since they are contemporaneous with the main verb.
  3. Should ἔρχεται be repeated twice, the way that "went" is repeated twice in the English sentence? Or is it better style to write it once? If so, where would it go, after μὲν or after δέ?

I hope it's okay to fit three questions into one. The third serves as my title question, although I am interested to hear your feedback on all three.

2 Answers 2


There are several unrelated grammatical points here, which I'll take in the order in which they occur in your Greek sentence.

  1. Position of αὐτούς. The pronoun αὐτ- in its non-emphatic third-person use (as opposed to its emphatic use meaning "himself" etc.) acts basically as a postpositive: it does not stand first in a sentence, nor in a smaller prosodic unit. Its most natural position here is right after the verb.
  2. "We believed" is about an ongoing state rather than a punctual or bounded event, so imperfect is better than aorist here.
  3. τούς is wrong because it's not "we believed them to be the good friends": no need for a definite article here.
  4. The choice of ἀγαθός seems fine to me. There are parallels, e.g. Xenophon, Memorabilia 2.4: ὡς πάντων κτημάτων κράτιστον ἂν εἴη φίλος σαφὴς καὶ ἀγαθός "that of all possessions the best is a sincere, good friend". ΕΤΑ: Alessio suggests in a comment that this may mean "good" in a moral sense rather than "intimate", which may well be right -- this kind of thing is hard to verify without a corpus search. His suggestion of πιστός could work, though that is more about "trustworthiness" specifically than "intimacy" (likewise βέβαιος, which is another possibility). Another option is ἐπιτήδειος, which when applied to people can mean "close, intimate". In my translation below I used the superlative of this adjective, but obviously there are lots of ways you could go here.
  5. ὁπόθεν means "whence, from where". What you need is ὅποι or ἔνθα, "whither, to where".
  6. A μέν ... δέ construction is impossible here because one of the clauses is subordinate to the other. Clauses coordinated by μέν ... δέ need to be syntactically parallel. This means you need something other than ὁ μέν ... ὁ δέ for "the one ... the other". You could go with e.g. οὗτος ... ἐκεῖνος, but since we're talking about two people, the idiomatic choice is ὁ ἕτερος ... ὁ ἕτερος.
  7. You're right that a conditional (a.k.a. indefinite) relative clause is needed here. But the correct type is a past general clause, which has an optative in the protasis and a past indicative in the apodosis.
  8. καί meaning "also" or "even" always immediately precedes what it modifies, so it should not precede the verb but "that one".
  9. For your question about repeating the verb, this is a question of style, but my sense is that most prose authors would avoid this particular repetition by using a synonym the second time. There is actually a fairly close passage in Plato's Phaedo (60b-c): Socrates says ᾧ ἂν τὸ ἕτερον παραγένηται ἐπακολουθεῖ ὕστερον καὶ τὸ ἕτερον "whenever one of the two (sc. pleasure and pain) comes to someone, the other follows later too", using two different verbs.

Putting it all together, we might get something like:

ἐνομίζομεν αὐτοὺς φίλους εἶναι ἐπιτηδειοτάτους· ὅποι γὰρ ὁ ἕτερος ἔλθοι, ἀεὶ ἐπηκολούθει καὶ ὁ ἕτερος.

  • 4. As I pointed out in my answer, "good friends" should not be translated using ἀγαθός, which in your example doesn't mean "good friend" in the same sense in which two people could be good friends each other. 5. ὅποι, as you said, means "whither, to where" and not "whithersoever", which is the intended translation
    – qwertxyz
    Jan 17, 2017 at 20:40
  • 1
    @Alessio: 4. Why do you think ἀγαθός in the Xenophon quote doesn't mean "good friend"? ἐσθλός is a highly poetic word and rare in prose. 5. The "-soever" meaning is, as is normal, conveyed by the construction (optative, or subjunctive with ἄν) -- see LSJ on ὅποι.
    – TKR
    Jan 17, 2017 at 21:20
  • Although your translation of Xenophon is good, I don't think that here ἀγαθός means other than "good" in a moralistic sense, and isn't therefore referred to his friend as "loyal" or "true". In that sense I would rather prefer πιστὸς ἑταῖρος, which is largely attested (Homer, Theognis, Plato and others). I checked the LSJ and I agree with you on ὅποι (writing my answer I look up only the GI)
    – qwertxyz
    Jan 17, 2017 at 21:43
  • 1
    @ktm5124 This is a complicated question, but I would say that as a rule, both imperfect and aorist (or present and aorist, if we're outside the indicative) tend to be possible with verbs of all sorts of meanings, but they force different readings or express different presentations of the event: in this case, for example, using an aorist with either of the verbs you mention might mean "we came to the conclusion that...", versus an imperfect which could mean "we were of the opinion that..."
    – TKR
    Jan 18, 2017 at 2:37
  • 1
    @TKR Ah, I see, so I guess it depends on the English sentence that I am translating. "We believed they were good friends" is more suggestive of the imperfect. But "We judged they were good friends" could possibly go both ways, aorist or imperfect. Perhaps it leans more toward the imperfect. I assume that context can change things as well.
    – ktm5124
    Jan 18, 2017 at 2:48
  1. First of all you need to eliminate the article τοὺς, as ἀγαθοὺς is a predicative adjective with nominal value ("we believed that they were good friends" and not "We believed that they were the good friends"); then, given that "good" in that sense means "true" or "loyal", the correct way to translate is with the adjective ἐσθλός (e.g. S. OT 611: φίλον γὰρ ἐσθλὸν ἐκβαλεῖν ἴσον λέγω / καὶ τὸν παρ’ αὑτῷ βίοτον, ὃν πλεῖστον φιλεῖ);

  2. The correct way to translate "for whithersoever this one went" is with ὅπου and the subjuntctive, as in Pl. Euthyph. 11c: οὐκ ἐθέλει μένειν ὅπου ἄν τις αὐτὰ θῇ, and in order to correctly render the contemporaneity with the main verb you should use the aorist subjuntctive. It is also better to use ἕτερος ... ἄλλος to distiguish the two men rather than the colourless ὁ μὲν ... ὁ δὲ. Furthermore I wouldn't recommend μέν and δέ here, as these particles are canonically used to express antithesis (vd. J. D. Denniston - K. J. Dover, The Greek Particles, Oxford 1954, pp. 369-374), but a single δέ could better fit to connect the two clauses.

  3. You could vary the second clause using a verb like ἥκω ("arrive in a place") using imperfect to give the sense of reiteration of the action, in combination with a dimostrative adverb like ἔνθα ("here") and the modal adverb ἴσως to better render "also".

This is the resulting phrase:

αὐτοὺς ἐνομίσαμεν ἐσθλοὺς φίλους εἶναι· ὅπου γὰρ ἂν ἕτερος ἔλθῃ, ἔνθα δ' ὁ ἄλλος ἴσως ἧκε.

EDIT: I found that expressions like πιστὸς φίλος or πιστὸς ἑταῖρος express better the sense of "good friend". Here are some examples:

  • X. Cyr. 8,7,13: ἀλλ’ οἱ πιστοὶ φίλοι σκῆπτρον βασιλεῦσιν ἀληθέστατον καὶ ἀσφαλέστατον

  • Pl. Phaedr. 233d: οὔτ’ ἂν πιστοὺς φίλους ἐκεκτήμεθα

  • Hom. Il. 17,589: σὸν δ’ ἔκτανε πιστὸν ἑταῖρον

  • Thgn. 1,644-645: Παύρους κηδεμόνας πιστοὺς εὕροις κεν ἑταίρους / κείμενος ἐν μεγάλῃ θυμὸν ἀμηχανίῃ

I also think that ὅποι suggested by TKZ could better express the directional sense of the phrase, as well as the imperfect.

For the word order I propose to follow the example of Lys. 18,26,4: ἄνδρας ἀρίστους ἐνομίζετ’ εἶναι τοὺς ὑπὲρ ὑμῶν ἀποθνῄσκοντας.

Here the reformulated phrase:

πιστοὺς φίλους ἐνομίζομεν εἶναι αὐτοὺς· ὅποι γὰρ ἂν ἕτερος ἔλθοι, ἔνθα δ' ὁ ἄλλος ἴσως ἧκε.

  • 1
    There are a number of errors here, I'm afraid: e.g. ὅπου is "where" (location), not "whither" (direction); the subjunctive would only be appropriate with present reference, not past reference; and the position of γάρ is incorrect (it would be ὅπου γὰρ ἂν ἕτερος.)
    – TKR
    Jan 17, 2017 at 19:41
  • ὅπου as a relative indefinite could be translated as "whithersoever", see e.g. Archim. Aeq. 2.10 ὅπου ἂν ἔρχηται τὸ ἕτερον σαμεῖον. In that case the subjunctive could be appropriate with past reference (e.g. D.L. 5,16 ἀναθεῖναι εἰς Νεμέαν ἢ ὅπου ἂν δοκῇ; Demosth. 35,13: ἐξελόμενοι ὅπου ἂν μὴ σῦλαι ὦσιν Ἀθηναίοις). For the position of γάρ, I agree with you.
    – qwertxyz
    Jan 17, 2017 at 20:35
  • I think the directional use of ὅπου is marginal at best; LSJ (ἔρχομαι) translate that Archimedes quote as "wherever the other point falls", and give no directional sense in their definition of ὅπου. Your quotes with the subjunctive both have future reference in context (the Demosthenes is a decree, the Diogenes is a will).
    – TKR
    Jan 17, 2017 at 21:15
  • The revised version looks better. But I still have reservations: 1. ἄν is inappropriate here (past general). 2. "one ... the other" is generally ὁ ἕτερος ... ὁ ἕτερος. 3. δέ is unnecessary in the last clause (no connective needed and "apodotic δέ" is unusual). 4. ἴσως in Attic prose most often means "maybe", which isn't to say it can't be used to mean "equally", but I don't think it's the most natural choice for "also".
    – TKR
    Jan 18, 2017 at 0:19
  • 2
    @Alessio Thanks for the feedback! It's good to know of the expressions πιστὸς φίλος and πιστὸς ἑταῖρος. Those are some great alternatives. Interesting to consider ἴσως as an alternative to καί. Welcome to the forum. Hope to see more answers from you.
    – ktm5124
    Jan 18, 2017 at 2:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.