νοσσός, νοσσίον, νεοττός, νεοττίον
(Phryn. Ecl. 177, Antiatt. ν 12)

D. General commentary

Phrynichus’ Eclogue devotes an entry (A.1) to the correct Attic form of the word νεοττός/νοσσός ‘chick, nestling, young bird’ (by extension also applied to the young of other animals and humans) and its diminutives, rejecting the variants with a stem νοσσ- in favour of those with νεοττ-. The Antiatticist (A.2), as usual, takes a more inclusive stance, defending νοσσός on the grounds of its use by Aeschylus (C.2). The two variants under discussion are distinguished by two separate phonological features, the presence of the geminate -ττ- in place of -σσ- and the reduction of -εο- to -ο-, although only the latter is discussed by the Atticists. This may be due to the different age and distribution of each phenomenon, as shall be discussed below.

The noun νεοσσός is clearly derived from νέος ‘new, young’, but the second half of the word is difficult to analyse: perhaps it contains the suffix *-t-io-, which is found in deadverbial adjectives such as περισσός ‘beyond the regular size, excessive’ (< πέρι/περί), ἔπισσαι ‘born later’ (< ἔπι/ἐπί), τὰ μέτασσα ‘thereafter’ (< μέτα/μετά) (see Bologna, Dedè 2021, 61), or a second compound member *-ki̯-ó- from the root of κεῖμαι ‘lie down’ (see DELG, EDG s.v.). The two intervocalic sequences *-ti̯- (only when heteromorphemic) and *-ki̯- would have produced the palatalised geminate that appears as -σσ- in Eastern Ionic and most other dialects, but as -ττ- in Attic, Euboean, Boeotian, and Cretan. Although the correspondence -ττ-/-σσ- is one of the key differences between Attic and the koine, the above entries do not focus on this feature, but rather on the omission of -ε- in the koine forms. Indeed, discussion of the correspondence -ττ-/-σσ- is remarkably rare in the main Atticist lexica (for an exception, see Moer. β 25Moer. β 25 on βήττω/βήσσω ‘cough’), despite the fact that Atticist grammarians were clearly interested in the problem of the distribution of -ττ- and -σσ- in different authors and genres (see Ael.Dion. σ 15Ael.Dion. σ 15, reconstructed on the basis of quotations in Eustathius). It is possible that the correspondence -ττ-/-σσ- was so well known that speakers were able to automatically adjust their phonology when they wanted to speak Attic, even without explicit guidance on individual words (but see entry βασίλεια, βασιλίς, βασίλισσα, βασίλιννα for the parodic hyper-Atticism βασίλιττα in Luc. Iud.Voc. 8). On the other hand, lexicographers may have refrained from giving explicit indications precisely because the usage of canonical authors was not consistent (for instance, -σσ- is the norm in tragedy and Thucydides). Doubtlessly less common is the other phonological feature that separates the Attic and non-Attic forms of the word for ‘chick’, namely hyphaeresis, which can be defined as the loss of an unaccented vowel in a sequence of two or more consecutive ones without contraction. In the sequence /eo/, it is usually the first vowel that is lost, probably with synizesis, i.e. loss of syllabicity, as an intermediate step: [neossós] > [ne̯ossós] > [nossós] (see Alonso Déniz 2013, 352). In principle, it is possible that the synizesis of /e/ led to the palatalisation of the preceding nasal, i.e. [nʲo-]; however, if such a phonetic realisation existed, it left no trace in the spelling or in the Modern Greek outcomes, which all have an initial /n/, not /ɲ/, showing that νοσσός was the form inherited by the later stages of Greek (see E.) . Hyphaeresis in the sequence /eo/ before two or more consonants is more systematic in other (especially Doric) dialects, but is restricted to a few isolated lexemes in Attic-Ionic, such as ἑορτή > ὁρτή ‘feast’ (Lejeune 1972, 252‒3; cf. Modern Greek γιορτή, from the same starting point, i.e. synizesis of /e/, which, however, became a glide /j/ instead of being lost). Its comparative rarity may well explain why the Atticists focused on hyphaeresis in their discussion of νοσσός/νεοσσός. The derivation of νοσσός from νεοσσός was known to ancient scholarship: see e.g. B.3, based on a Herodianic source (indeed, later grammarians were concerned with this word because of its anomaly with respect to Herodian’s rule; see F.2). Atticist lexicographers, including Aelius Dionysius and Pausanias, also took note of the additional meaning ‘egg yolk’ attested for νεοττός/-ίον in classical sources (e.g. Men. fr. 40), as shown by B.1 and B.2.

The distribution before the Hellenistic period seems clear enough: νεοσσός is found in non-Attic literary sources, from Homer (C.1) to Herodotus and the Hippocratic corpus, as well as in tragedy (e.g. C.3), while νεοττός is at home in comedy (e.g. C.4) and in Attic prose (e.g. C.6). The unexpected νεοττόν in Archestratus (C.5, ap. Ath. 9.348b) may be due to Atticisation in the manuscript tradition of Athenaeus, rather than an Atticisms in Archestratus’ language (see Olson, Sens 2000, LV‒LVII). Starting from the koine, the variant with the non-Attic consonantism coexists with the Attic one, even within the works of the same author, as in the case of Plutarch, whose corpus amply attests to both stems. As with other instances of -ττ-/-σσ- variation, the more Atticising writers of the imperial period reverted to νεοττός: it is consistently employed by Galen, Lucian (6x), Philostratus, and especially Aelian, whose subject matter leads him to use the term 21 times in the History of Animals (plus 11 occurrences of νεόττιον and 2 of νεοττιά). Later high-register writers, such as Libanius and Synesius, share this preference for the form in -ττ-. While the history just sketched is not significantly different from that of other forms showing -ττ-/-σσ- variation, it is worth noting that the forms rejected by Phrynichus and defended by the Antiatticist, with non-Attic consonantism and hyphaeresis, are almost unattested in literary sources before late antiquity (insofar as we can trust the manuscript tradition). Apart from the Aeschylean usage attested by the Antiatticist (C.2), νοσσός is found in the Septuagint (22x, e.g. C.7), once in the New Testament (Ev.Luc. 2.24, a quotation of LXX Le. 12.8; note that in the NT the spelling νοσσ- is the norm also for the derivatives, cf. Blass, Debrunner 1976, 25), then in the Epistle of Barnabas (11.3b.2) and in SB 5.7814.15 and 19 (= TM 18003) [Oxyrhynchus, 256 CE]. The diminutive νοσσίον already appears in the LXX (Ps. 83.4.2), then in the NT (Ev.Matt. 23.37.4). The variant νοσσιά of the derivative νεοττιά ‘nest of young birds’ (Ar., Attic prose; but note τῶν νεοσσιῶν in Thphr. CP 4.5.7), Ionic νεοσσιή (Archil., Hdt.), has a similar pattern of attestation, being frequent in the LXX; however, an Ionic νοσσίη, with hyphaeresis, is already employed by Herondas (7.72, where it is transmitted with paroxytone accent). The picture becomes more nuanced, however, when other derivatives and compounds are taken into account. Some are attested with νεοττ- in Attic sources, but show the spelling νοσσ- in later Greek: νεοττεύω ‘hatch, build a nest’ (Ar.+) but νενοσσευμένα already in Hdt. 1.159.11, then νοσσ- from the LXX onwards; νεοττοτροφέω ‘rear young birds’ (Ar.+) but νοσσ- in Leon. AP 9.346.2; while νοσσοποιέω ‘hatch, build a nest’ is older (LXX+) than the variants in νεοττ-/νεοσσ-, attested only in the imperial period. While hyphaeretic variants are not attested for the rarer νεοσσοποιία ‘hatching’, νεοσσοκόμος ‘rearing chickens’, and νεοττοκομέω ‘rear chickens’, other derivatives are in fact only attested with the stem νοσσ-. These are νόσσαξ (< *νεόσσαξ) ‘cockerel’ and its diminutive νοσσάκιον, condemned by Phrynichus and first attested in a 2nd-century CE papyrus (C.8, see also F.1), and the adjective νοσσάς ‘young’, attested only in Panyass. fr. 12 πολλὰς δὲ νοσσάδας ὄρνις and com. adesp. fr. 1092.1.13 νοσσάδας.

The extent of the diffusion of νοσσ- becomes even clearer when we look at onomastics. The stem Ν(ε)οσσ-, almost always with hyphaeresis, is well attested in personal namesProper names outside Attica: it appears in the masculine names Νόσσος (with the accent possibly shifting after hyphaeresis occurred in the noun), Νοσσίκας, Νεοσσίων, Νοσσύλος, Νόσσων (see Bechtel 1917, 585) and in the feminine names Νοσσίς, Νοσσυλίς, Νοσσώ (see Bechtel 1917, 591; Alonso Déniz 2018, 535). These names are distributed throughout the Greek-speaking world, with attestations dating back to the 4th century BCE, well before Phrynichus’ time (see LGPN s.vv.). Such names, of hypocoristic origin seem to have been common in aristocratic families, for example in the Doric-speaking Cos, but also in the southern Italian city of Locri, the birthplace of the poetess Nossis (see Cazzaniga 1972). Note that Νεοττίς, the Attic form of Νοσσίς, is also attested as a personal name, probably of hetaerae, in the titles of three comedies by Antiphanes, Eubulus, and Anaxilas. Since such names are not attested in Athenian onomastics, it is likely that Νεοττίς was employed by the comic poets as an Atticised form of the name of non-Athenian female characters (see Konstantakos 2000, 125‒8; Tartaglia 2019, 112).

The overall picture that can be drawn from the evidence presented above is that the hyphaeretic stem νοσσ- was probably more widespread by the time of Phrynichus than the meagre literary evidence would indicate, as is strongly suggested by the use of νοσσός in 3rd-century CE Egypt. One should also bear in mind that, at least in some cases, conservative spellings with <νεοσσ-> may have concealed realisations with [noss-]. If that was the case, Phrynichus’ battle in favour of νεοττ- was doomed to fail in the spoken language, as the outcomes in Modern Greek dialects show (E.), although it may have contributed to the survival of the Attic form in higher-register Greek. The Antiatticist’s defence of the innovative form points to its use by such an early Attic author as Aeschylus, but we can only speculate about the reasons for this (F.3).

E. Byzantine and Modern Greek commentary

During the Byzantine period the learned forms in νεοσσ- continued to be in use alongside those in νοσσ-, which is the more frequent spelling in popular literature (see Kriaras, LME s.vv.), while the stem νεοττ- is restricted to high-register writers. From the latter is derived the learned neologism νεοτ(τ)ιδής ‘young bird’, attested in the 11th (Michael Psellus Epistulae 6.24.24) and 12th centuries (John Syropulus Oratio ad imperatorem Isaacum II Angelum 18). Several forms in Modern Greek dialects continue the hyphaeretic variants debated by the Atticists, thereby confirming that they belonged to the spoken language (see Andriotis 1974 s.vv.; Shipp 1979, 404‒5). In particular, the derivative νοσσιά is continued by νοσσά (Crete, Naxos), νουσσά (Chios), and οσσά (Thera); νοσσίς by νοσσίδα (Euboea) and νουσσίδα (Skiathos); and Pontic preserves forms derived from νοσσάκιον (νοσσάκι, νοσσάκ’, νουσσάκ’, νοσσάκα, νουσσάκα). This last term was also borrowed by the local Turkish dialect of Pontus as nasaka ‘hen that lays eggs for the first time’ (see Papadamou, Papanastassiou 2019, 307). The standard language uses the learned loan νεοσσός, with the phonology of Ionic and the higher koine.

F. Commentary on individual texts and occurrences

(1)    Phryn. Ecl. 177 (A.1)

Of particular interest in this entry is the direct exhortation to the reader of the Eclogue to use the forms in νεοττ- so as to ‘appear like an ancient Attic speaker’. Such a formulation is not without parallels in the Eclogue (cf. 229Phryn. Ecl. 229 ὡς οἱ ἀρχαῖοι, ‘like the ancients’ and especially 371Phryn. Ecl. 371 Ἀττικὸς ἂν φαίνοιο καὶ ἐπιμελής, ‘you would appear Attic and scrupulous’). On Phrynichus’ recommendation to use certain words and expressions in order to ‘appear’ classical and Attic, see now Kim (2023, 142‒3), who stresses the close connection in Phrynichus’ evaluative terminology between ἀρχαῖος (‘ancient’) and δόκιμος (‘approved’ or, as Kim prefers, ‘authentic’, since the adjective is taken from the language of numismatics). By using forms such as νεοττός, the aspiring purist would emulate the diction of classical authors and could, in principle, pass himself off as an ancient Attic speaker. (On numismatic metaphors in the Atticists’ language of ‘authenticity’, see also Lamagna 2004). Also interesting is Phrynichus’ firm rejection of the diminutive νοσσάκιον, which, in addition to the phonological features already stigmatised in νοσσός, also exhibits a double suffixationSuffixes unattested in classical sources. Indeed, our first attestation of νοσσάκιον is roughly contemporary with Phrynichus’ activity, but comes from a clearly substandard source. The noun occurs twice in a magical papyrus (C.8, probably dating to the mid/late 2nd century CE; see Faraone, Torallas Tovar 2022, 66‒7), interestingly enough next to νεοσσόν without hyphaeresis (was a longer word more prone to the reduction of unstressed vowels?). From a morphological point of view, νοσσάκιον is, properly speaking, the diminutive of νόσσαξ ‘chick, cockerel’, itself of late attestation (Dsc. 2.49) and bearing the suffix -ακ-, which may have a hypocoristic value (see Chantraine 1933, 379‒80). One should bear in mind, however, that the widespread Medieval and Modern Greek diminutive suffix -άκι(ν) (which often lost its original diminutive meaning) arose precisely from the reanalysis of koine diminutive formations in -ιον built on stems in -ακ- (see LKN s.v. -άκι, CGMEMG vol. 2, 610 n. 132). The success of this suffix in the later stages of the language suggests that it must have been widespread in the spoken register: Phrynichus’ warning can be seen as an attempt to limit the spread of this innovative morphological feature. Despite the purists’ proscription, both νόσσαξ and νοσσάκιον survived, at least in Pontus, as evidenced by their descendants in the modern Pontic dialects of Greek and Turkish (see E.).

(2)    Epim.Hom. π 113 (B.3)

This entry of the Epimerismi Homerici claims that disyllabic nouns of the form (C)οC₁C₁ος are always paroxytone (note that the unattested noun ὄσσος ‘eye’ is back-formed from the inherited athematic dual ὄσσε < *h₃kʷ-ih₁). Dyck (1995, 616) traces this entry back to the Epimerismi attributed to Herodian, since Eustathius (in Od. 2.55.3–11 = Hdn. Περὶ καθολικῆς προσῳδίας GG 3,1.208-20–209.2), discussing the origin of ὄσσε, explicitly attributes this rule to Herodian and lists some of the same examples (ὄρρος, κόττος) found in B.3, including the exception κομμός. The other apparent exception of νοσσός is justified in B.3 by its derivation from νεοσσός: although Eustathius does not mention it, Lentz (GG 3,1.209.2) attributed it to the same rule.

(3)    Aesch. fr. 113 (C.2)

The Antiatticist defends νοσσός by citing its use in Aeschylus’ satyr play Heralds. Elsewhere Aeschylus only uses νεοσσός, not just in his tragedies (4x, see e.g. C.3) but also in the satyr play Net-haulers (fr. 47a.795; an example of distance from the Attic vernacular according to Dettori 2016, 151‒2). Indeed, this instance of νοσσός would stand isolated in all of Attic theatre were it not for Soph. fr. 219a.80.2, where the letters νοσσων can be read. In the absence of a wider context for the Aeschylean fragment, one may speculate whether the hyphaeretic form was chosen merely for reasons of metrical convenience or whether it served some stylistic purpose, such as the linguistic characterisation of the dramatis personae. Drawing attention to another one-word fragment from the same play that also contains a non-Attic form (Aesch. fr. 112 λογγάσω ‘I will delay’, transmitted by Phot. λ 370 and condemned as φαῦλον ‘bad’ by Poll. 9.136Poll. 9.136), Poli Palladini (2020, 106) argues that ‘it is tempting to imagine that the non-Attic or anyway incorrect words λογγάσω (fr. 112) and νοσσός (fr. 113) characterized Heracles and/or the satyrs as linguistically influenced by dialects spoken in regions other than Attica’ – perhaps, specifically a West Greek one in the light of the Syracusan gloss λογγῶνες ‘stones for mooring cables’ (EM 569.42). The presence of Sicilian DoricDoric features in Aeschylus has been noted since antiquity (see Ath. 9.402c on the Syracusan word ἀσχέδωρος ‘wild boar’ in the Phorkides [fr. 261]) and continues to be an object of discussion in the broader context of Aeschylus’ interactions with contemporary Sicilian culture (on which see recently Smith 2018); several possible Sicilian forms have also been found in the Net-haulers, although Dettori (2016, 7‒8 and passim) downplays the presence of Doricisms in this play. While in non-Attic literature νεοσσός is usually found (including the literary Doric νεοσσώς in Theocr. 14.14, where, however, the metre would rule out νοσσ-), the attestation of the stem Νοσσ- in personal names outside of Attica (see D.) may lend some credence to the idea that in certain (Doric?) dialects νοσσός had already become standard in classical times.