νῆες, ναῦς, νῆας
(Phryn. Ecl. 140, Moer. ν 4, Thom.Mag. 245.1)

D. General commentary

Atticist sources attest to a debate on the admissibility of competing forms of the nominative and accusative plural of diphthongal stemsDiphthongal stems, notably ναῦς (‘ship’), , βοῦς (‘ox’) and γραῦς (‘old woman’), although the latter’s diphthong is not original. The present entry focuses specifically on doctrines concerning the inflection of ναῦς in the nominative and accusative plural.

ναῦς is an ancient root noun from IE *neh₂u-; its inflectional pattern exhibits close parallels with Sanskrit and Latin (cf. Sanskrit nom. náuḥ, Latin nāvis remade from acc. nāvem, cf. Sanskrit nāvam, see further DELG s.v., EDG s.v.). /w/ (ϝ) regularly yields υ before consonantal endings (e.g. nom. sing. *ναϝ‑ς > ναῦς) while disappearing in the intervocalic position without causing contraction; in Attic-Ionic, aside from the demise of ϝ, forms of ναῦς are subject to the change of the stem’s ᾱ into η before a vowel (e.g. nom. pl. Attic-Ionic *ναϝ‑ες > νῆες, nom. sing. Ion. νηῦς). The Attic nom. sing. ναῦς may reflect either *neh₂us, with the intervocalic laryngeal, which does not cause lengthening, or *nāu̯s, with further shortening according to Osthoff’s law (see Sihler 1995, 340). In Attic, the word inflects as follows: nom. sing. ναῦς, gen. sing. νεώς (from νη‑ός and after quantitative metathesis); dat. sing. νηί; acc. sing. ναῦν (modelled on the nom. sing. ναῦς); nom. pl. νῆες, gen. pl. νεῶν, dat. pl. ναυσί, acc. pl. ναῦς (< *ναυ‑νς, modelled on the acc. sing. ναῦν, see Chantraine 1961, 97); nom. and acc. dual νῆε, gen. and dat. dual νεοῖν.

In koine Greek, ναῦς is in competition with the neuter πλοῖονπλοῖον (‘vessel’, ‘ship’) – which is standard, for instance, in the New Testament (e.g. Ev.Matt. 4.21) – but is retained in literary texts and in formal speech (on the synonymySynonyms ναῦς/πλοῖον see B.1, B.2).

Late Attic and Post-classical Greek introduced alternative endings for the nominative plural and the accusative plural, and, alongside νῆες and ναῦς, competing forms came into use – respectively ναῦς (nom. pl.) and νῆας (acc. pl.); the latter, although a result of analogical innovation, comes to coincide with the ancient Ionic form (see below). This phenomenon is observable across the entire class of u-stems, where, for instance, the nom. pl. ἰχθύες is flanked by ἰχθῦς and the acc. pl. ἰχθῦς by ἰχθύας (see entry ἰχθῦς, δρῦς, and AGP vol. 2, Morphology, forthcoming).

The nom. pl. ending ‑ῦς of ναῦς and other stems in diphthong is an analogical innovation that is modelled on the acc. pl. (see Schwyzer 1939, 564; 578), as is the case with the nominatives plural in ‑ῦς of u-stems. Whereas for u-stems, the innovative nominative plural is in use as early as Middle and New Comedy (see AGP vol. 1, 263–4), for stems in diphthong, it appears to be a much later development. The analogical nom. pl. ναῦς appears to enter literary prose in the imperial period, when it is used by Diodorus Siculus (see e.g. C.4), Pausanias (e.g. 6.3.15), Flavius Josephus (e.g. AJ 8.181.2), Chariton (8.54.3), and Aelian (e.g. NA 15.5.13). Phrynichus’ entry (A.1), which criticises this use on the part of sophists of the 2nd century CE (see F.1), is coherent with this picture. Although the innovative nominative ναῦς enjoys some success, it does not replace the Attic νῆες: if anything, it appears to remain subsidiary. Moreover, the use of the analogical forms is also minor in the authors criticised by Phrynichus: in Polemon, the analogical nom. pl. occurs in only one passage (C.6), whereas ναῦς otherwise regularly follows the Attic inflection, as is the case in Favorinus (C.5). It should also be borne in mind that the distribution of the regular and analogical forms may be distorted by the discrepancies in the manuscript tradition, in which misunderstandings or regularisation may reasonably have occurred. As an example, we may cite Polemon’s passage (C.6), which is perhaps one of the loci Phrynichus (A.1) had in mind. In C.6 the analogical plural ναῦς, which agrees with the participle ἐπιδραμοῦσαι, is normalised by most manuscripts into the Attic nom. sing. through the participle’s correction into the singular ἐπιδραμοῦσα. This is a good example of how the manuscript tradition can lead one astray with respect to these forms’ distribution in ancient authors. The innovative αἱ ναῦς is also occasionally used in lexicographical sources, such as Apollon. Lex. 114.34 = Hsch. ν 128: ναύλοχον· εὔδιον, ἐν ᾧ αἱ ναῦς λοχῶσιν, ‘ναύλοχος (‘haven’): A sheltered spot where ships are moored’ (cf. Σ ν 17 = Phot. ν 45, EM 598.36; αἱ ναῦς occurs also in Hsch. ο 154; schol. Arat. 410.2).

Ancient scholarship offers various explanations for the oscillation νῆες/ναῦς, interpreting the form in ‑ῦς as the outcome either of – allegedly – the addition of υ (Orio 14.13–5, cf. Et.Gen. 1409.3–5 = EM 170.55–7) or a ‘contraction of ‑αε‑ into the diphthong αῦ’ (Epim.Hom. in Il. 1.410.b2, cf. Et.Gen. 1409.3–5 = EM 170.5). The latter supports Herodian’s explanation that the nominative plural in ‑ῦς of u-stems is the result of contraction (see schol. [Hdn.] Hom. Il. 24.1; Phlp. Ton. 49.5–11 [= Hdn. Περὶ καθολικῆς προσῳδίας GG 3,1.417.3–8]; Choerob. in Theodos. GG 4,2.335.35–336.13, on these passages see entry ἰχθῦς, δρῦς, with translation and commentary).

Regarding the acc. pl., νῆας is the ancient Ionic and Homeric form, ubiquitous in the Homeric poems (see, for instance, the formula ἐπὶ νῆας Ἀχαιῶν, ‘on the ships of the Achaeans’, e.g. Il. 1.12) and it is the standard form in the epics (cf. Hes. Th. 875). In Attic prose of the 5th and 4th centuries BCE, τὰς νῆας is avoided on the grounds that it belongs to the epic poetic register, while τὰς ναῦς is the norm in historiography (Thucydides, e.g. C.1; Xenophon, e.g. HG 1.1.7), oratory (Isocrates, e.g. 18.59.2; Andocides, e.g. 3.11.8; Lysias, e.g. 12.68.5; Demosthenes, 20.78.7), philosophy (e.g. Pl. Plt. 298e.8) and in comedy as well (Ar. Eq. 1182, Ra. 1465, Cratin. fr. 70). Conversely, νῆας is regularly used in tragedy (Aesch. Supp. 844; Eur. IA 254) and is later retrieved by Hellenistic poets (Theocritus, e.g. 22.17; Arat. 1.38). The Homeric influence determines the clear preference for the disyllabic νῆας in the poetic registerPoetic language and particularly in epic poetry, where it remains the standard throughout the centuries (see, for instance, its systematic use in Quintus Smyrnaeus’ Posthomerica, e.g. 1.95).

A form νῆας – analogically modelled on the nom. pl. νῆες and alternative to the acc. pl. ναῦς – is found in Ionic prose (cf. Hellanic. FGrHist. 4 F 4.11; 142.2; Ctes. FGrHist. 688 F 14.20, 14.23, 14.29) and went on to gain traction in late Attic and Post-classical Greek. The diffusion of νῆας was helped by the fact that it was regarded as the regular form by analogy with the nom. pl. νῆες: this is confirmed by Phrynichus (A.1), who argues that Lollianus’ mistake was induced by the apparent regularity of αἱ νῆες, τὰς νῆας (see F.1). This aligns with the ancient grammarians’ theories, which consider the acc. pl. in ‑ας to be the regular accusative form of u‑stems whose nom. pl. is in ‑ες (see Theodos. Can. GG 4,1.13.16–14.2; Choerob. in Theodos. GG 4,1.234.23–31), whereas, conversely, forms with acc. pl. in ‑ῦς are considered instances of homophony with the nom. pl. (see Choerob. in Theodos. GG 4,1.234.23–31; Choerob. in Theodos. GG 4,2.335.35–336.13; Sophronius Grammaticus GG 4,2.392.4–6, see also entry ἰχθῦς, δρῦς). The analogy with the ending in -ας of thematic nouns of the first declension may also have been instrumental in the diffusion of the innovative acc. νῆας (see E.).

Analogously to the analogical nom. pl., the alternative acc. pl. also remains subsidiary, and authors who vacillate between the two forms mostly use ναῦς – as is the case, for example, for Polybius (ναῦς ca. 80x, e.g. C.2, vs νῆας ca. 12x, e.g. C.3). ναῦς remains the standard in writers of the imperial period, including Diodorus Siculus (ναῦς more than 100x), Plutarch (Them. 8.1), and Cassius Dio, and is the form in which the word is inflected in its unique occurrence in biblical Greek (cf. LXX 1Re 5.6). Again, the vacillations in both manuscripts and the indirect tradition are not representative of the form’s actual distribution; note, for instance, that although in Thuc. 1.48.3 τὰς ναῦς μετεώρους (‘ships on the high sea’) we find the expected Attic acc. pl., Su. μ 773 transmits the phrasing as μετεώρους νῆας, with the analogical form.

The use of both the innovative nom. pl. and acc. pl. remains minor not only in literature but also in epigraphic and documentary sources: Threatte (1996, 228) records a single occurrence of a nom. pl. ναῦς in a decree of the Augustan period, while regular forms invariably occur otherwise. On the inflection of ναῦς and stems in diphthong in papyri, where they are quite rare, see Gignac (1981, 81–2), Mayser, Gramm. vol. 2,27–8.

Atticist lexicographers adopt different stances on the legitimacy of the innovative endings of diphthongal stems. Some opinions are clear: Phrynichus (A.1) discourages both analogical forms, and Moeris (A.2) condemns the innovative nom. pl. However, it is unclear whether Moeris’ prescription originally included the analogical accusative: while the condemnation of the nominative plural in ‑ῦς in A.2 may suggest that Moeris disapproves of analogical forms, the analogical acc. νῆας actually occurs in his lexicon (see Moer. δ 38Moer. δ 38 [cf. Ael.Dion. δ 5Ael.Dion. δ 5, Thom.Mag. 89.15–90.3Thom.Mag. 89.15–90.3]: δελφὶς ὄργανον ἐν ναυμαχίᾳ μολιβοῦν, ὅθεν καὶ Θουκυδίδης νῆας δελφινοφόρους, ‘δελφίς [‘dolphin’] is a leaden instrument [used] in sea-fight, whence Thucydides (7.41.2) [calls] the ships ‘dolphin-bearing’’). Although the possibility remains that a later scribe introduced νῆας into Moeris’ text, one might tentatively suggest that Moeris accepted the form. This would not be an isolated case: see Moeris’ apparent approval of analogical forms of γραῦς and βοῦς (cf. entry γρᾶες, γραῦς, γραῖαι). In the case of the acc. pl. νῆας, its Homeric pedigree may perhaps have been the reason of Moeris’ possible endorsement of the form, since the poet was viewed as a source of proto-Attic forms (on HomerHomer as a linguistic model for Moeris see entry Moeris, Ἀττικιστής). Conversely, it may have been precisely the Homeric and poetic allure of νῆας that rendered it unacceptable for Phrynichus, whose primary interest was the suitability of words to prose and rhetoric.

E. Byzantine and Modern Greek commentary

In Byzantine Greek, the lexeme ναῦς is retained as a learned word (as opposed to πλοῖον) in literary texts in high and middle register regardless of the inflection. The regular Attic nominative and accusative plural prevail, although analogical forms occasionally feature. The regular nom. pl. νῆες is the norm in Atticising writers throughout the Byzantine period up to the 15th century, when it is used by Michael Critobulus (e.g. Historiae 1.42.5.8), Sylvester Syropulus, Gennadius Scholarius, and Laonicus Chalcocondyles. The analogical nom. pl. ναῦς features instead in the chronicle by Theophanes the Confessor (399.14) and the anonymous writings that continue the text, commonly known as Theophanes Continuatus (304.11). The regular ναῦς prevails in the accusative plural, although it coexists with νῆας: several authors fluctuate between the two forms. Anna Comnena uses mainly ναῦς (e.g. Alexias 7.8.5) but sometimes employs νῆας (e.g. 6.10.7), and the same applies to Nicetas Choniates (History, ναῦς e.g. Andron1,pt2 326.12; νῆας e.g. John2 17.21), George Pachymeres (Historia brevis, ναῦς 9.15.7; νῆας e.g. 5.10.15), and Nicephorus Gregoras (History of the Romans, ναῦς e.g. 1.25.13; νῆας e.g. 2.849.4).

An innovative first-declension nom. pl., αἱ νῆαι, is also attested: it is found in Michael Attaliates (11th century, e.g. History 2.223–4.27: νῆαι καρπαθικαί, ‘ships from Karpathos’), and in the Historia Turcobyzantina (e.g. 11.24.10: ἐμπορικαὶ νῆαι, ‘mercantile ships’) by Ducas (15th century). This form partakes of the regularisation that 3rd declension feminine nouns underwent in Medieval Greek, being transferred to the 1st and 2nd paradigm (see CGMEMG vol. 2, 589).

ναῦς is no longer in use in Modern Greek, where the general words meaning ‘ship’ are το καράβι, from καράβιον, diminutive of the ancient κάραβος, ‘light ship’ (cf. EM 490.30–2, see LBG s.v.; LKN s.v.), and το πλοίο, from the ancient πλοῖον (LKN s.v.). Aside from καράβι and πλοίο, Modern Greek has many nouns denoting vessels, among them σκάφος and βάρκα, the latter being a loanword from Latin barca (LKN s.v.), which denotes a small boat and is used mainly in colloquial speech. ναῦς survives only in derivatives and compounds belonging to the technical vocabulary of sailing, such as ναυτικός, ‘nautical’, ‘sailor’ and ναυσιπλοΐα, ‘sailing’.

F. Commentary on individual texts and occurrences

(1)    Phryn. Ecl. 140 (A.1)

Phrynichus reproaches four sophists who are contemporaries of his for their use of the analogical forms of nominative and accusative plural of ναῦς: besides Favorinus of ArlesFavorinus of Arles, who is often the target of Phrynichus’ criticism in the Eclogue (on Favorinus see Philostr. VS 1.8 and Amato 2005, 1–37, on Phrynichus’ polemic against him see entries εἶμι, ἐλεύσομαι, and υἱεύς, υἱέως, υἱέα), he attacks Marcus Antonius Polemon, a certain Sylla, and Publius Hordeonius Lollianus.

Marcus Antonius Polemon, a native of Laodicea, was among the most prominent intellectuals of the Second Sophistic; he was a disciple of Dio Chrysostom’s and a notorious rival of Favorinus (cf. VS 1.8.490–1). His career developed in Smyrna, where he was active as a master of rhetoric and – favoured by the emperor Hadrian – engaged in political activity (on Polemon’s life and career see Swain 2007, 156–76). He is the author, among others, of the treatise Physiognomy, which enjoyed great success in antiquity (see Swain 2007); two of his declamations are extant, edited by Stefec (2016) . Phrynichus refers to Polemon as ‘the Ionic sophist’ (Ecl. 396Phryn. Ecl. 396, ὁ Ἰωνικὸς σοφιστής, cf. Ecl. 424Phryn. Ecl. 424) and mentions him three times in the Eclogue, always as a polemical target: aside from A.1, Phrynichus criticises him for his use of the redundant superlative κεφαλαιωδέστατος, lit. ‘the more principal’, (Ecl. 424) and of the expression κατ’ ὄναρ, ‘in a dream’ (Ecl. 396).

Although his works are now entirely lost, Publius Hordeonius Lollianus, a native of Ephesus, was a prominent figure in the Second Sophistic and held the chair of rhetoric in Athens (on Lollianus’ life, see Philostr. VS 1.23, Schissel 1927, and entry εἰσιέτω, ἰέτω). Lollianus is reproached in two other entries of the Eclogue: Ecl. 141Phryn. Ecl. 141, on the use of the imperative εἰσιέτω (see entry εἰσιέτω, ἰέτω), and Ecl. 152Phryn. Ecl. 152, on the late pluperfect ending -εσαν. Interestingly, the two consecutive glosses Ecl. 140Phryn. Ecl. 140 and Ecl. 141 exhibit similar wording, and both attribute Lollianus’ slips to a mistaken analogy.

SyllaSylla, plausibly a sophist, has not been identified (see Lobeck 1820, 170), nor is he mentioned in other entries of the Eclogue or elsewhere. Phrynichus classifies the use of ναῦς in place of νῆες as a solecismSolecism (σόλοικος); note, however, that the error in question might be better described as a barbarismBarbarism, as it concerns a morphological glitch, whereas in the grammatical tradition the term ‘solecism’ is typically invoked to denote faulty syntax (see Sandri 2020, 19–27).