E.C. Marchant, Commentary on Thucydides: Book 3

τίς—‘what form of trustworthy friendship or freedom was this turning out to be?’ i.e. how could we put any confidence in a friendship or a freedom like that? (Dion. Hal. Antiq. VI. 78 τίς οὖν ἡ τοιαύτη φιλία καὶ πίστις, ἐν ᾗ παρὰ γνώμην άλλήλους θεραπεύειν ἀναγκασθησόμεθα; seems to have found ἡ, not ἢ, after αὕτη in his MS., and our MSS. have ἡ.)

παρὰ γνώμην—‘without sincerity,’ οὐκ ἀπὸ γνώμης ἀλλ᾽ ἀπὸ γλώσσης μόνον.

ὑπεδεχόμεθα—this word can be used of every kind of reception or welcome afforded to a ξένος, short of rejection: e.g. in Lycurgus § 133 οὐδεμία πόλις αὐτὸν εἴασε παρ᾽ αὑτῇ μετοικεῖν is repeated in the form τοῦτον δὲ τίς ἂν ύποδέξαιτο πόλις;

ὅ τε . . πίστιν βεβαιοῖ, ἡμῖν τοῦτο . . ἐχυρὸν παρεῖχε— passages sometimes cited as parallel to this (as II. 40 δ τοῖς ἄλλοις άμαθία μὲν θράσος, λογισμὸς δὲ ὄκνον φέρει and IV. 125 ὄπερ φιλεῖ μέγιστα στρατόπεδα ἀσαφῶς ἐκπλήγνυσθαι) bear little resemblance to it. If πίστιν, which Stahl bracketed, is sound, we should assume, with Kruger, a conflation of relative with epexegetic clause—ὃ εὔνοια βεβαιοῖ with εὔ. πίστιν βεβαιοῖ. A schol. already takes this view; and unless the text is corrupt, no other explanation can be entertained. (The other scholium is a mere paraphrase, but its conclusion—διὰ τοῦτο ἄπιστος ἦν ἡ φιλία ἀμφοτέρων—shows that the writer discerned that in the whole passage, from τίς οὖν to παραβήσεσθαι ἔμελλον, the underlying idea, not clearly expressed, is that of a πίστις ἄπιστος. Both πιστή above and πίστιν here have been removed by critics more intent, perhaps, on the language than on the thought. ξύμμαχοι ὑπὸ δέους πιστοί are ξύμμαχοι ἄπιστοι.)

παράσχοι—representing ὁποτέρους ἂν παράσχῃ in past sequence. 10

ἀσφάλεια—‘sense of security.’

καί—‘as a consequence’ of the θάρσος.