χρέως, χρέος
(Phryn. Ecl. 371, Moer. χ 7, Philemo [Vindob.] 396.26, [Hdn.] Philet. 146, Poll. 8.31)

D. General commentary

The entries in Atticist lexica A.1, A.2, and A.3 concern the noun χρέος (‘debt’, ‘matter’, ‘need’), prescribing the form with ω, χρέως, as Attic and rejecting the form with ο. Pollux (A.5) includes the form χρέως in his list of lawsuit terminology: it may therefore reasonably be assumed that he too approved this form over that with ο. A different view is reflected in the entry of the pseudo-Herodianic Philetaerus (A.4), where both forms are accepted. Given the marked fluctuation between χρέος and χρέως even among Attic authors and the early spread of the form with short o – as will be shown below – the aim of the Atticist prescriptions was likely to defend a pronunciationPronunciation (and, consequently, a spellingSpelling) that preserved the long vowel, at a time when the loss of quantitative distinctions was causing the two forms to become increasingly homophonous; see Vessella (2018, 255–6).

χρέος is an s-stem neuter noun derived from the ancient substantive χρή (see DELG and EDG s.v. χρή); it presents a variety of forms. In the Homeric poems, the form χρεῖος – a later spelling of the etymological *χρῆος – is the most frequent (e.g. Il. 11.686 = C.1, 11.688; Od. 1.409). χρείως (acc. sing., v.l. χρεῖος) in Od. 8.355 (C.2) is probably an alternative spelling of χρεῖος, with metrical lengthening in longum; see Chantraine (1958–1968 vol. 1, 70–1). Shipp (1972, 30), however, interprets χρείως as a form resulting from a mixture of χρεῖος and χρέως – the latter attested as a varia lectio at Il. 11.686 (C.1) and Od. 8.353 (C.2), see below F.1 – arguing that metrical lengthening at the beginning of the fifth foot is anomalous. The Ionic form χρέος is probably a more recent development from *χρῆος > χρεῖος by abbreviation without metathesis (see Schwyzer 1939, 246; Chantraine 1958–1968 vol. 1, 70), and not the older form from which χρεῖος derives, as Eustathius incorrectly states (B.3). χρέος is already attested in the Odyssey (C.2; 11.479) and is by far the most frequent form in Greek literary texts. The earliest attestations of χρέως occur in the speeches of Lysias and Demosthenes (but see F.1), both in the direct cases (see e.g. C.6) and in the genitive (see e.g. C.5; [D.] 49.18). This form can be explained as the result of quantitative metathesis of the original *χρῆος (see Schwyzer 1939, 245–6), a phonetic development also affecting o-stem nouns of the so-called Attic declensionAttic declension, common in Attic and Ionic (see Schwyzer 1939, 557–8; Chantraine 1961, 43–5). Although ancient sources (see B.4) assimilate χρέως to other nouns exhibiting this type of declension, it in fact represents a different case, since it continues to follow the s-stem declension even in the lengthened form, which, conversely, is not attested in the plural.

With regard to the singular form of the noun, ancient grammatical sources (see B.2, B.3) distinguish between the inflection of χρέος, which follows the pattern of s-stem neuter nouns, and that of χρέως, for which only two inflectional forms are attested: χρέως for the nominative, genitive, accusative, and vocative, and – according to Choeroboscus, B.2 – χρέει for the dative. In fact, while the genitive χρέως is attested (see e.g. C.5, [D.] 49.18) and coexists with χρέους, the dative χρέει occurs in literary and documentary texts only from the late imperial period and may ultimately derive from the inflection of χρέος; however, no clear-cut distinction between the two inflection models can be established. The attested plural and dual forms all belong to the s-stem neuter noun inflection with short ο. The absence of the plural and dual cases in the inflection of χρέως was already noted by ancient grammarians (see B.2, where Choeroboscus explains the defective declension by referring to the grammarian Orus, who likely discussed it in a lost passage of his treatise on orthography (see Dickey 2007, 99–100); Orus pointed to the non-analogical change of ο to ω in this noun, which would make its declension irregular).

As regards the distribution of the forms χρέος and χρέως in Attic authors, it does not seem to fully match the prescriptions of the Atticist lexicographers. χρέος is the only form attested in the tragic triad (Aeschylus 6x, Sophocles 4x, Euripides 10x) and, more strikingly, also in Eupolis (fr. 99.111) and Aristophanes: besides Nu. 30 (C.3), χρέος occurs in Ach. 454 and 455 (C.4). In none of these instances is the short form required by metrical constraints. In addressing Cornelianus, the dedicatee of the Eclogue, Phrynichus (A.1) appears to respond to a reasonable objection to his prescription, and was evidently aware of the Aristophanic use of χρέος, at least in Nu. 30. He justifies it as parodic: the line has a lyric colouring, achieved by using ἔβα instead of ἔβη and by attaching the accusative directly to the verb (ἀτὰρ τί χρέος ἔβα με). Schol. Ar. Nu. 30 cites a line from Euripides which is only preserved in the Aldine edition (fr. 1011: τί χρέος ἔβα δῶμα;, ‘What happened to the house?’; cf. Eur. HF 530: τί καινὸν ἦλθε τοῖσδε δώμασιν χρέος;, ‘What new [calamity] has struck our house?’). Here χρέος does not mean ‘debt’ – an Aristophanic innovation – but rather functions as the expression τί χρέος, with a meaning close to simple τί or τί χρῆμα; see Dover (1968, 96); LSJ s.v. χρέος, II.2; Collard (2018, 60). A tragic model is perhaps also presupposed in Ach. 454–5 (C.4): a scholium (schol. Ar. Ach. 454 [REΓLh]) explains the passage as a parody of Eur. fr. 717 (τί δ’, ὦ τάλας; σὺ [σοὶ Lh] τῷδε πείθεσθαι μέλεις [μέλλεις EΓ, μέλει Lh]; ‘What, wretch? Do you care to heed this fellow?’). Here too Aristophanes gives the noun a different nuance, employing τί χρέος + gen. in the sense of ‘what need for’; see Olson (2002, 192). Given the parodic nature of these occurrences, Aristophanes’ usage does not appear to be representative of the distribution of χρέως and χρέος in Attic authors. Nevertheless, the pseudo-Herodianic Philetaerus (A.4) admits both forms precisely on the basis of the Aristophanic occurrence in Nu. 30 (C.3). As regards Attic prose authors, χρέος appears to be the only form attested in Antiphon (fr. 67 Thalheim = Ath. 12.525b), Plato (Plt. 267a1; Lg. 958b7), and Isocrates (21.14.3). In Lysias, by contrast, the form χρέως at 17.5 and fr. 50.9 Carey represents a correction of the genitive χρέους transmitted in the MSS. In fr. 1.6 Carey (C.5), where it occurs in the oration’s title, χρέως is likewise guaranteed by the manuscript tradition. In the Demosthenic corpus, the form χρέως is predominant (16x [cf. C.6; in 5 cases, χρέος is attested as a v.l.], vs. χρέος 1x: 25.69 [χρέως codd. AF^a]). With the exception of Aristophanes, where the use of χρέος is likely deliberate, the predominance of the short-vowel form in other Attic authors may be attributed to spelling changes in the manuscript tradition, which, however, affected the Demosthenic orations to a lesser extent; cf. Rutherford (1881, 482). Even allowing for such changes, which might account for the absence of the ω-form in authors such as Plato – explicitly cited by Eustathius (B.3) for his use of χρέως – the fluctuation between the two forms in Attic literary texts appears to be confirmed by Eustathius himself (B.3), who relies on earlier grammatical sources, possibly Aelius Dionysius (on the problematic nature of the reference to the Ionic use of χρέως in Eustathius, see F.2). For the interpretation of the label οἱ ὕστερον Ἀττικοί in B.4 in relation to ancient theories of the periodisation of Attic, see F.3.

As regards the distribution of the two forms in Atticising authors, they follow the general trend in koine Greek: the only occurrence of χρέως is found in Favorinus (de Ex. 22.10); otherwise, only χρέος is attested. Interestingly, in documentary sources, χρέως appears as both genitive and accusative in a few non-Attic inscriptionsInscriptions from the Hellenistic period: cf. IG 12,2.135.21 [Delos, 314–302 BCE]: ὑπὲρ τοῦ χρέως, IG 12,4.4044.106 [Kalymna, early 2nd c. BCE]: ἀπὸ τοῦ χρέως, and IG 12 Suppl. 356.4 [Thasos, 300–250 BCE]: τὸ χρέως.

Turning now to the compoundsCompounds of the noun, the forms with ω (ἀξιόχρεως ‘noteworthy’, κατάχρεως ‘burdened with debt’, ὑπέρχρεως ‘over head and ears in debt’) are regarded as Attic by ancient grammatical sources: cf. e.g. [Hdn.] Epim. 207.14–6: τὰ παρὰ τὸ χρέος συγκείμενα, κοινῶς μὲν διὰ τοῦ ο μικροῦ γράφονται· οἷον· κατάχρεος· ὑπόχρεος· ἀξιόχρεος· Ἀττικῶς δὲ διὰ τοῦ ω μεγάλου, ‘The compounds of χρέος are written with ο in the common language, such as κατάχρεος, ὑπόχρεος, ἀξιόχρεος, whereas [they are written] with ω in the Attic manner’; Σ^b α 1558 (= Phot. α 2177, Su. α 2823, ex Σ′; cf. also Ael.Dion. α 151Ael.Dion. α 151, Orus fr. B 26): ἀξιόχρεων· ἐν τῷ ω λέγουσι, καὶ λειπόνεων. καὶ τὰ οὐδέτερα οἱ {γὰρ} παλαιοὶ ὁμοίως. τὸ δὲ ἀξιόχρεον βάρβαρον, ‘ἀξιόχρεων (masc. acc. sing.): They say it with ω, just like λειπόνεων (‘deserting the fleet’, acc. sing.). Ancient [authors use] the same forms for the neuters as well, whereas ἀξιόχρεον [is] barbarous’. In contrast to the distribution of χρέος and χρέως in Greek literary texts, compounds most frequently exhibit the form with ω, as also acknowledged by Eustathius (B.3), who may be drawing on a passage by Aelius Dionysius (χ 17, see B.3, apparatus, and F.2). In B.4, the reference to the nominative plural κατάχρεοι in Polybius (13.1.1) is likely an unfortunate choice on Eustathius’ part to illustrate the alternation between spellings in o and ω, both for the simple form of the noun χρέος/χρέως and for its derivatives, as previously observed for λαγώς and πλέως. One might suppose that the choice of a plural in -οι is intended to evoke a 2nd-declension singular in -ος, even though χρέος follows a different inflectional pattern. In any case, as noted above, in compounds the forms with ω remain the most widespread even in the post-classical period.

E. Byzantine and Modern Greek commentary

In line with the general trend in koine Greek, χρέος is also the most frequently attested form in late antique and Byzantine literary texts. By contrast, the form with ω continues to be used, albeit sporadically, well into the Byzantine age. The spelling χρέως is clearly attested in papyriPapyri until the 7th century CE, appearing in both the direct cases and in the genitive singular (see e.g. P.Dubl. 26.2 (= TM 37476) [Arsinoites, 7th century CE]: τοῦ χρέως ἐμοῦ, P.Lond. 5.1721.9 (= TM 19736) [Thebes, 542/3 or 557/8 CE]: παρʼ ἐμοῦ χρέως); cf. Gignac (1981, 67), who nonetheless notes that in papyri χρέους is the regular form for the genitive singular in papyri. Rather than representing a conscious Atticising choice, the exchange between ο and ω in these cases – at least in Egyptian papyri later than the 2nd century BCE – appears to result from the loss of vowel quantity perception (cf. Mayser, Gramm. vol. 1,1, 73–4; Gignac 1976, 277 with n. 2). As in the direct cases, the genitive singular likely reflects phonetic confusion between ου and ω (cf. Mayser, Gramm. vol. 1,1, 78–9; Gignac 1976, 208–9). However, as far as the use of the long-vowel form is concerned, one cannot rule out a lingering awareness of the coexistence of χρέος/χρέως even in later periods or – more plausibly – the influence of the verb χρεωστέωχρεωστέω (‘to be in debt’), which is frequently attested in documentary papyri throughout the Byzantine era. In literary works, by contrast, it remains difficult to assess the respective roles of the manuscript tradition – where a correction of χρέος to χρέως may be interpreted as an Atticising intervention by a scribe – and of modern editors in determining the distribution of the two forms. In some authors, the long-vowel form appears as a minority variant alongside more frequent occurrences of the short-vowel form. For instance, χρέως is attested twice in Themistius (Or. 16.200b; 34.17), although in both cases it results from modern scholarly emendation. In Libanius (Arg.D. 46.t and 49.t), the form appears in the titles of the corresponding Demosthenic orations, where χρέος is nevertheless transmitted as a varia lectio. Furthermore, χρέως occurs alongside χρέος in the corpus of John Chrysostom, exclusively in spurious works (e.g. In poenitentiam Ninivitarum MPG 64.424.34; Caritatem secundum deum rem esse deo dignam MPG 61.683.21). It is also attested in Constantine VII Porphyrogenitus (four times in De cerimoniis, strictly within acclamatory formulae), in Euthymius Zigabenus (once in Commentary in the Book of Psalms MPG 128.777.10 vs. χρέος in MPG 128.1325.43), and once in Theophylact (Enarrationes in evangelia MPG 124.153.53), as opposed to five occurrences of χρέος. In Modern Greek, only the form χρέος survives, both in the sense of ‘debt’ and in the metaphorical sense of ‘moral obligation’ or ‘duty’ (cf. LKN s.v.).

F. Commentary on individual texts and occurrences

(1)    Hom. Il. 11.686 (C.1); Hom. Od. 8.353 (C.2)

The form χρέως is already attested in the textual tradition of the Homeric poems, namely as varia lectio at Il. 11.686 (C.1) and Od. 8.353 (C.2). In the Iliadic passage χρέως is defended by Aristarchus (cf. schol. (Did.) Hom. Il. 11.686a [A^intT^il]: <χρέως:> οὕτως Ἀρίσταρχος χρέως, ‘<χρέως:>: Thus [wrote/chose] Aristarchus, χρέως’) against χρεῖος, supported by Aristophanes of Byzantium (cf. schol. (Did. | Ariston.) Hom. Il. 11.686b [A]: {τοὺς ἴμεν οἷσι} χρέως ὠφείλετο: […] Ἀριστοφάνης δὲ ‘χρεῖος’ καὶ ‘ὀφείλετο’, ‘‘[that] those should come to whom a debt was owed’: […] Aristophanes [wrote] instead ‘χρεῖος’ and ‘ὀφείλετο’’). While the choice of Aristophanes of Byzantium reflects the use of the noun in the Homeric poems, where χρεῖος is the most common form, the reading adopted by Aristarchus seems to imply an Atticising colouring; cf. Wackernagel (1916, 71). Indeed, following a line of thought popular in antiquity, Aristarchus himself considered HomerHomer an ancient Athenian who spoke a παλαιὰ Ἀτθίς (see below, F.3; cf. [Plu.] Vit.Hom.2 2.1; Vit.Hom.5 247.7–8; see Schironi 2018, 622). On the use of χρέως as an Attic form in Homer, see in particular [Plu.] Vit.Hom.2 12.1: μάλιστα δὲ τῇ Ἀτθίδι διαλέκτῳ κέχρηται· καὶ γὰρ ἐπίμικτος ἦν. καὶ ἐπεὶ λέγεται παρὰ τοῖς Ἀττικοῖς ‘λεὼς’ ὁ ‘λαός’, κατὰ ταύτην τὴν συνήθειαν ἔστι παρ’ αὐτῷ ὁ ‘Πηνέλεως’ καὶ τὸ ‘χρέως’, ‘[Homer] mostly uses the Attic dialect, for he combines different dialects. And since users of Attic say λεώς [instead of] λαός (‘people’), according to this usage Πηνέλεως (‘Peneleus’) and χρέως are found in his [works]’.

(2)    Eust. in Il. 3.549.27–550.4 (B.3)

In the Eustathius passage, Erbse identified a fragment by Aelius Dionysius (χ 17), whose lexicographical work may be referred to by the expression ῥητορικὸν Λεξικὸν (cf. Erbse 1950, 17; more recently, Cantore has proposed that the Rhetorical Lexicon mentioned by Eustathius should instead be identified with the Synagoge: cf. Cantore 2021). Erbse introduced several emendations to Eustathius’ text, the most notable being the correction of χρέως Ἀττικοὶ καὶ Ἴωνες into χρέως· Ἀττικοί, χρέος Ἴωνες. Indeed, the attribution of the lengthened form to the Ionic dialect is not confirmed either by other lexicographic works or by literary evidence. It is therefore more likely that the passage contrasts two usages associated with different dialects. In light of this, one may question whether it would be worth changing the Eustathius’ text as well.

(3)    Eust. in Od. 2.80.17–22 Cullhed–Olson (B.4)

The relatively low frequency of the form χρέως in earlier Attic literature may tentatively be related to the statement in this passage by Eustathius, who speaks of οἱ ὕστερον Ἀττικοί in connection with the form χρέως (and perhaps also with its compounds; on the difficulties relating to the quotation of κατάχρεοι from Polybius as an example of the form in ο, see D.), as well as with other terms (κάλως, λαγώς, and πλέως), which are all explained in modern grammar as part of the so-called Attic declension (see entry λαγώς, λαγός, λαγωός). The label οἱ ὕστερον Ἀττικοί seems to refer to one of the ancient theories concerning the periodisation of Attic. One of these theories divided Attic into two phases: the παλαιὰ Ἀτθίς, extending down to the end of the 5th century BCE and considered closely related to Ionic, and the νέα Ἀτθίς, i.e. the Attic dialect from the 4th century BCE onwards, whose principal representative is Menander. For this periodisation, see D.H. Thuc. 23: οἱ δὲ πρὸ τοῦ Πελοποννησιακοῦ γενόμενοι πολέμου καὶ μέχρι τῆς Θουκυδίδου παρεκτείναντες ἡλικίας ὁμοίας ἔσχον ἅπαντες ὡς ἐπὶ τὸ πολὺ προαιρέσεις, οἵ τε τὴν Ἰάδα προελόμενοι διάλεκτον τὴν ἐν τοῖς τότε χρόνοις μάλιστα ἀνθοῦσαν καὶ οἱ τὴν ἀρχαίαν Ἀτθίδα μικράς τινας ἔχουσαν διαφορὰς παρὰ τὴν Ἰάδα, ‘The historians belonging to the generation preceding the Peloponnesian War, who survived into Thucydides’ lifetime, in general had the same literary aims, whether they chose the Ionian dialect, which was at the height of its popularity at the time, or in Old Attic, which differed but slightly from Ionic’ (transl. Usher 1974, 523–5); D.H. Lys. 2: καθαρός ἐστι τὴν ἑρμηνείαν πάνυ καὶ τῆς Ἀττικῆς γλώττης ἄριστος κανών, οὐ τῆς ἀρχαίας, ᾗ κέχρηται Πλάτων τε καὶ Θουκυδίδης, ἀλλὰ τῆς κατ’ ἐκεῖνον τὸν χρόνον ἐπιχωριαζούσης, ὡς ἔστι τεκμήρασθαι τοῖς τε Ἀνδοκίδου λόγοις καὶ τοῖς Κριτίου καὶ ἄλλοις συχνοῖς, ‘He (i.e. Lysias) is completely pure in his vocabulary, and is the perfect model of the Attic dialect – not the archaic dialect used by Plato and Thucydides, but that which was in general currency in his day, as exemplified in the speeches of Andocides, Critias and many other orators’ (transl. Usher 1974, 23); Su. τ 1049: […] ἡ δὲ παλαιὰ Ἀτθίς ἐστιν, ἧς ἦρχεν Εὔπολις, Κρατῖνος, Ἀριστοφάνης, Θουκυδίδης· ἡ δὲ νέα Ἀτθίς ἐστιν, ἧς ἐστι Μένανδρος καὶ ἄλλοι, ‘Old Attic is [the dialect] which Eupolis, Cratinus, Aristophanes, and Thucydides began [to use]; new Attic is instead [the language of] Menander and the other [authors]’; see Schironi 2018, 621–2 and entries ἀπολλύασιν, ἀπολλύουσιν, and other 3rd person plurals of -νυμι verbs; χολάς, χόλιξ. The fact that χρέως is most frequently attested in Demosthenes may reflect this diachronic development, insofar as οἱ ὕστερον Ἀττικοί could designate Attic authors from the 4th century BCE onwards. This interpretation, however, is at odds with the literary evidence for nouns following the Attic declension, which are attested well before the 4th century BCE (see entries λαγώς, λαγός, λαγωός and νεώς, ναός). It is therefore more likely that Eustathius here uses the label οἱ ὕστερον Ἀττικοί in a broader sense, referring to ‘post-Homeric Attic’ or ‘middle Attic’, in accordance with a linguistic theory that posited a three-stage classification of the Attic dialect. This theory – which Eustathius likely drew from Aelius Dionysius (cf. Eust. in Od. 2.74.25–8 Stallbaum ~ Ael.Dion. η 1Ael.Dion. η 1) – regarded Homer as representative of the earliest phase (this idea is quite apparent in Moeris’ lexicon too: see entries χολάς, χόλιξ; ἀπολλύασιν, ἀπολλύουσιν, and other 3rd person plurals of -νυμι verbs; Moeris, Ἀττικιστής). As shown by Probert (2004) , a similar tripartite division – between old Attic (largely overlapping with the Homeric language), later Attic, and koine Greek – appears to have been adopted also by Herodian, who in turn likely drew on a doctrine dating back to the Hellenistic period. This chronological division is borne out by other passages in Eustathius’ commentary (cf. e.g. in Il. 3.170.7–9: τὸ δὲ ἄξιον ὅτι καὶ ἀντάξιον Ὅμηρος λέγει, προδεδήλωται, οἱ δὲ ὕστερον Ἀττικοὶ ἄξιον καὶ τὸ εὔωνον εἶπον, καθὰ δηλοῖ καὶ ὁ Κωμικός, ‘It is clear that Homer says ἄξιος [in the sense of] ἀντάξιος (‘of equal value’), whereas later users of Attic also used ἄξιος for what is cheap, as the comic poet (i.e. Aristophanes) shows (e.g. Eq. 645)’). As far as χρέως is concerned, this framework is consistent with Eustathius’ own statement (B.3) that Plato used the form, even though it is not attested in his extant works.