Skip to main content

Recent work on expressions of obligation in Greek (Ruiz Yamuza 2008, Allan 2013) has refined our understanding of modal verbs like δεῖ and χρή by applying to the problem general linguistic advances in the realm of modality. But while making distinctions such as that between deontic and epistemic usages or setting up a scale of modal strength is a necessary first step towards determining the relative domains of δεῖ and χρή, a satisfactory theoretical account of the difference between the two remains elusive. That is, earlier scholarship (Goodell 1914, Benardete 1965) has established both a general diachronic trend—over the period from Homer to Koine Greek, χρή is gradually replaced by δεῖ—and, in fifth- and early fourth-century Greek, a basic synchronic distinction between χρή as an expression of moral obligation and δεῖ as one of simple necessity, but it is unclear how this can be accounted for by even very recent literature on grammaticalization that blurs the line between the two (“Another modal meaning is obligation (sometimes called ‘necessity’)” (Bybee 2015: 148)). And clarifying the boundary between these two concepts is not just a theoretical desideratum: Goodell states (1914: 102) that the distinction between δεῖ and χρή had become effaced by the time of Aristotle, but one would like rather stronger proof before suggesting that someone so attentive to logic and ethics used verbs of necessity and obligation interchangeably. As an intermediate step towards that goal, the current paper focuses on the use of δεῖ and χρή in Aristophanes for several reasons: first, he lies at a diachronic point where both words are used in near parity with one another, so the distinction between them should be in full force; second, Goodell omitted him from his own study because his “peculiar shifts in tone” (1914: 96) make his usage difficult to assess; third, those very shifts in tone give the linguist a healthy variety of pragmatic contexts to draw on.

When one contrasts the environments in which δεῖ and χρή occur, a couple of features stand out. First, χρή is closely allied to the imperative. At Frogs 354, for instance, the Chorus Leader’s opening words εὐφημεῖν χρή are, as Dover’s commentary notes, essentially equivalent to the imperatives εὐφημεῖτε and εὐφημία ’στω found elsewhere in Aristophanes. Note also Frogs 905, where χρὴ λέγειν occurs in parallel with a “See to it that…” ὅπως-clause. Second, δεῖ often gravitates towards conditional clauses, accounting for three of its four modal occurrences in Frogs, as at line 77: εἶτ’ οὐ Σοφοκλέα πρότερον ὄντ’ Εὐριπίδου | μέλλεις ἀναγαγεῖν, εἴπερ ἐκεῖθεν δεῖ σ’ ἄγειν; (“Then won’t you bring Sophocles back up as Euripides’ superior, if you have to bring someone up from there?”). In such passages, the sequence of thought is of the form, “If p is absolutely necessary, then q,” and the nesting of the modal verb inside the conditional gives it a very different profile from the more imperatival χρή.

Combining these two observations, we can then explain Frogs 1056, which would otherwise be a difficult example: πάνυ δὴ δεῖ χρηστὰ λέγειν ἡμᾶς. Henderson’s translation in the Loeb (“It’s very important that we tell them things that are good”) would lead one to expect χρή here, as the statement seems like the sort of precept that elsewhere takes the more exhortative modal—as happens in fact just three lines above (ἀποκρύπτειν χρὴ τὸ πονηρὸν τόν γε ποιητήν). But one should never ignore δή: as Dover points out, it here has inferential force, marking this line as the logical, inevitable conclusion of the general idea that had initially been presented as an admonition. In Aristophanes, then, logical necessity and moral obligation are kept distinct, and theoretical models of modality should do the same, so that δεῖ can be adduced as evidence for a grammaticalization pathway from necessity to obligation.