Laudator Temporis Acti

Basil Lanneau Gildersleeve, "Brief Mention," American Journal of Philology 35.2 (1914) 227-235 (at 231-232):

Indeed, the relative frequency of χρή and δεῖ in different spheres and in different periods cannot have escaped any careful student of the orators, and I have alluded to it (A.J.P. XXVI 249) lightly as a matter of course, as one would allude to σύν and μετά, as one would allude to ἐθέλω and βούλομαι—familiar ear-marks all.¹ It has long been observed that there is but one δεῖ in Homer,² and there is but one δεῖ in Pindar, as there is but one βούλομαι. A convenient test is furnished by that paraenetic ragbag, the Theognidea. A rapid count reveals nearly a score of 'duty' χρή's, not a solitary δεῖ. χρή gives way to δεῖ in the later Attic orators, and finally δεῖ reigns. Eleusis is merged in Athens. χρή beats δεῖ in Antiphon, χρή beats δεῖ in Lysias, whereas, not being a professional, Andokides, 'the gentleman orator', as I have nicknamed him, is nearer to the later usage. With Isokrates the break begins. It is not necessary to count. A footrule will serve. The orators all follow. There are f.i. four times as many δεῖ's as χρή's in Hypereides, and all the χρή's go without much coaxing into the προσήκει category—the category of moral obligation. Heaven bless the indexes! There is no index to Isaios, so that I have had to count. In Isaios δεῖ draws off gradually, but winds up a good third ahead, a warning against averages made on the basis of segmental reading (A.J.P. VIII 221, footn.).

¹ In the Περὶ πολιτείας attributed to Herodes Atticus, Wilamowitz has noted, as who would not, the strong archaic flavour, which he considers characteristic of Herodes, whereas Drerup, largely on the ground of that very same archaic flavour, has surmised in the little document the hand of a political pamphleteer prior to 404. See his edition, Paderborn 1908. In this performance χρή beats δεῖ five to three, and in the same line of observation λέξαι (bis), λέξει, λέξειε are decidedly old-fashioned (C.W.E.M., A.J.P. XVI 162; XXXI 117). By the way, that Drerup should have retained ἴσως ἄν τις εἴπῃ (30) as an Homeric formula seems to me even more absurd than the retention of ἂν ἐρεῖ (Pind. N. 7, 68) as an Homeric reminiscence, ἴσως would kill any Homeric formula.

² The types made J.H.H. Schmidt (Syn. 3. 702) say that there is only one χρή in Homer. Of course, he meant δεῖ. A.J.P. XXVII 480.

Paraenetic Ragbag would be a good name for a blog, although I hope no one uses it only to abandon it after a few posts (the fate of many a blog).