PURA. Purism In Antiquity: Theories Of Language in Greek Atticist Lexica and their Legacy

Lexicographic entries

Δημοσθένας, Δημοσθένεις
(Moer. δ 48)

A. Main sources

(1) Moer. δ 48: Δημοσθένας Ἀττικοί· Δημοσθένεις τὸ{ν} ἀνάλογον Ἕλληνες.

This entry is only attested in cod. C | Δημοσθένεις <κατὰ> Wendel (1927, 1276; cf. Moer. μ 11), followed by Hansen | τ(ὸν) C (with gratitude to the anonymous referee for this information) : τὸ Hansen.

Users of Attic [employ] Δημοσθένας (acc. plur.). Users of Greek [employ] the analogical [form] Δημοσθένεις.

B. Other erudite sources

(1) Theodos. Can. GG 4,1.7.19–8.3: πληθ(υντικόν). οἱ Δημοσθένεες οἱ Δημοσθένεις, τῶν Δημοσθενέων τῶν Δημοσθενῶν, τοῖς Δημοσθένεσι μόνως, τοὺς Δημοσθένεας τοὺς Δημοσθένεις, ὦ Δημοσθένεες ὦ Δημοσθένεις.

Plural: οἱ Δημοσθένεες οἱ Δημοσθένεις (nom.), τῶν Δημοσθενέων τῶν Δημοσθενῶν (gen.), τοῖς Δημοσθένεσι (dat.) only, τοὺς Δημοσθένεας τοὺς Δημοσθένεις (acc.), ὦ Δημοσθένεες ὦ Δημοσθένεις (voc.).

(2) Choerob. in Theodos. GG 4,1.181.19–24 (= Hdn. Περὶ κλίσεως ὀνομάτων GG 3,2.697.12–6): τοὺς Δημοσθένεας καὶ Δημοσθένεις. κανονίζεται ἡ αἰτιατικὴ τῶν πληθυντικῶν ἀπὸ τῆς αἰτιατικῆς τῶν ἑνικῶν τοῦτον τὸν τρόπον· πᾶσα αἰτιατικὴ ἑνικῶν εἰς α λήγουσα προσθέσει τοῦ ς ποιεῖ τὴν αἰτιατικὴν τῶν πληθυντικῶν, οἷον Αἴαντα Αἴαντας, λέβητα λέβητας, Λάχητα Λάχητας, ἔρωτα ἔρωτας· οὕτως οὖν καὶ Δημοσθένεα Δημοσθένεας.

τοὺς Δημοσθένεας and Δημοσθένεις. The accusative plural is inflected from the accusative singular in this way: every accusative singular ending in -α forms the accusative plural by the addition of ς, as for instance Αἴαντα Αἴαντας, λέβητα λέβητας, Λάχητα Λάχητας, ἔρωτα ἔρωτας. In this way, therefore, [one has] also Δημοσθένεα Δημοσθένεας.

(3) Choerob. in Theodos. GG 4,1.182.17–25 (= Hdn. Περὶ κλίσεως ὀνομάτων GG 3,2.697.16–24): δεῖ δὲ καὶ τοῦτο γινώσκειν, ὅτι οἱ Ἀθηναῖοι ἐπὶ τῶν εἰς ης εἰς ους ἐχόντων τὴν γενικὴν κυρίων καὶ ἐπὶ τῶν παρὰ τὸ ἔτος διὰ τοῦ α ποιοῦσι τὴν αἰτιατικὴν τῶν πληθυντικῶν, οἷον ὁ Δημοσθένης τοῦ Δημοσθένους τοὺς Δημοσθένας, ὁ Ἀριστοφάνης τοῦ Ἀριστοφάνους τοὺς Ἀριστοφάνας, ὁ δωδεκαέτης τοῦ δωδεκαέτους τοὺς δωδεκαέτας, ὁ ἑπταέτης τοῦ ἑπταέτους τοὺς ἑπταέτας· τὴν δὲ εὐθεῖαν τῶν πληθυντικῶν διὰ τῆς αι διφθόγγου ποιοῦσιν ἐπὶ τούτων, οἷον οἱ Δημοσθέναι, οἱ Ἀριστοφάναι, οἱ ἑπταέται, οἱ δωδεκαέται.

One must know this too: concerning proper names ending in -ης with a genitive in -ους and compounds of ἔτος, the Athenians form their accusative plurals by means of α, as for instance ὁ Δημοσθένης τοῦ Δημοσθένους τοὺς Δημοσθένας, ὁ Ἀριστοφάνης τοῦ Ἀριστοφάνους τοὺς Ἀριστοφάνας, ὁ δωδεκαέτης τοῦ δωδεκαέτους τοὺς δωδεκαέτας, ὁ ἑπταέτης τοῦ ἑπταέτους τοὺς ἑπταέτας. They (i.e., the Athenians) form the nominative plural of the aforementioned forms by means of the diphthong αι, as for instance οἱ Δημοσθέναι, οἱ Ἀριστοφάναι, οἱ ἑπταέται, οἱ δωδεκαέται.

(4) Choerob. in Theodos. GG 4,1.190.25–31: ἰστέον ὅτι οἱ Ἀττικοὶ ἐπὶ τῶν εἰς ης εἰς ους ἐχόντων τὴν γενικὴν κυρίων καὶ ἐπὶ τῶν παρὰ τὸ ἔτος εἰς τὴν αι δίφθογγον ποιοῦσι τὴν εὐθεῖαν τῶν πληθυντικῶν καὶ διὰ τοῦ α τὴν αἰτιατικὴν αὐτῶν, οἷον ὁ Δημοσθένης τοῦ Δημοσθένους οἱ Δημοσθέναι τοὺς Δημοσθένας. […] Ἡρακλέας Ἡρακλεῖς καὶ ἀττικῶς Ἡρακλᾶς, ὥσπερ Δημοσθένας.

Concerning proper names ending in -ης with a genitive in -ους and compounds of ἔτος, one must know that Attic speakers form their nominative plurals by means of the diphthong αι and their accusative [plurals] by means of α, as for instance ὁ Δημοσθένης, τοῦ Δημοσθένους, οἱ Δημοσθέναι, τοὺς Δημοσθένας. […] Ἡρακλέας, Ἡρακλεῖς, and Ἡρακλᾶς, in the Attic manner, just like Δημοσθένας.

(5) Choerob. in Theodos. GG 4,1.220.28–35: ἰστέον ὅτι ὁμοφωνεῖ ἡ αἰτιατικὴ τῶν πληθυντικῶν τῇ εὐθείᾳ τῶν πληθυντικῶν, οἷον οἱ Πηλέες οἱ Πηλεῖς καὶ τοὺς Πηλέας καὶ Πηλεῖς· ἐμάθομεν γὰρ ὅτι πᾶσα εὐθεῖα πληθυντικῶν εἰς ς λήγουσα καὶ συναιρουμένη κατὰ τὴν τελευταίαν συλλαβὴν ἔχει τὴν αἰτιατικὴν ὁμόφωνον αὐτῇ δηλονότι καὶ τὴν κλητικήν, οἷον οἱ βότρυς τοὺς βότρυς ὦ βότρυς, οἱ ἰχθῦς τοὺς ἰχθῦς ὦ ἰχθῦς, οἱ βοῦς τοὺς βοῦς ὦ βοῦς, οἱ ἡδεῖς τοὺς ἡδεῖς ὦ ἡδεῖς, οἱ Δημοσθένεις τοὺς Δημοσθένεις ὦ Δημοσθένεις.

One must know that the accusative plural sounds like the nominative plural, as for instance οἱ Πηλέες οἱ Πηλεῖς and τοὺς Πηλέας and Πηλεῖς. Indeed, we know that every nominative plural ending in -ς that is contracted in its last syllable has a homophonous accusative – and also a vocative, as for instance οἱ βότρυς τοὺς βότρυς ὦ βότρυς, οἱ ἰχθῦς τοὺς ἰχθῦς ὦ ἰχθῦς, οἱ βοῦς τοὺς βοῦς ὦ βοῦς, οἱ ἡδεῖς τοὺς ἡδεῖς ὦ ἡδεῖς, οἱ Δημοσθένεις τοὺς Δημοσθένεις ὦ Δημοσθένεις.

(6) Choerob. in Theodos. GG 4,2.336.1–9 (= Hdn. Περὶ παθῶν GG 3,2.320.35–6): ἐν τῷ Δημοσθένεας Δημοσθένεις τὸ ε καὶ α εἰς τὴν ει δίφθογγον συναιρεῖται. […] καὶ ἔστιν εἰπεῖν πρὸς αὐτόν, ὅτι ἐν ταῖς αἰτιατικαῖς ταύταις οὐκ ἐγένετο συναίρεσις, ἀλλ’ ὁμοφωνία ἐστὶ τῆς αἰτιατικῆς τῶν πληθυντικῶν πρὸς τὴν ἰδίαν εὐθεῖαν.

[Herodian says that] in Δημοσθένεας Δημοσθένεις there is syneresis of ε and α into the diphthong ει. Against his view, one can say that syneresis does not occur in these accusatives, but homophony of the accusative plural with the corresponding nominative.

C. Loci classici, other relevant texts

(1) Pl. Smp. 218a.7–b.2: καὶ ὁρῶν αὖ Φαίδρους, Ἀγάθωνας, Ἐρυξιμάχους, Παυσανίας, Ἀριστοδήμους τε καὶ Ἀριστοφάνας.

Ἀριστοφάνας BTW : Ἀριστοφάνεις Y.

I have only to look around me, and there are men like Phaedrus, Agathon, Eryximachus, Pausanias, Aristodemus, and Aristophanes. (Transl. Lamb 1925, 227, adapted slightly).

(2) Plu. Quaestiones convivales 613d: ἂν μὲν γὰρ πλείονας ἔχῃ φιλολόγους τὸ συμπόσιον, ὡς τὸ Ἀγάθωνος Σωκράτας Φαίδρους Παυσανίας Ἐρυξιμάχους καὶ τὸ Καλλίου Χαρμίδας Ἀντισθένας Ἑρμογένας ἑτέρους τούτοις παραπλησίους, ἀφήσομεν αὐτοὺς [μύθῳ] φιλοσοφεῖν.

Σωκράτας… Ἀντισθένας Ἑρμογένας T (Vindobonensis phil. gr. 148), corrected by a scribe to Σωκράτεις... Ἀντισθένεις Ἑρμογένεις.

For if the majority of the guests at a party are learned men, like Socrates, Phaedrus, Pausanias, and Eryximachus at the dinner of Agathon, and Charmides, Antisthenes, Hermogenes, and others like them at the dinner of Callias, we shall let them talk philosophy. (Transl. Hoffleit, in Clement, Hoffleit 1969, 13).

(3) Luc. Nec. 17: πολλοὺς δὲ καὶ ἄλλους ἦν ἰδεῖν ἐν ταῖς τριόδοις μεταιτοῦντας, Ξέρξας λέγω καὶ Δαρείους καὶ Πολυκράτας.

Many others, too, could be seen begging at the cross-roads – your Xerxeses, I mean, and Dariuses and Polycrateses. (Transl. Harmon 1925, 103).

(4) Philostr. Gymn. 1: ἡ μὲν γὰρ πάλαι γυμναστικὴ Μίλωνας ἐποίει καὶ Ἱπποσθένας, Πουλυδάμαντάς τε καὶ Προμάχους καὶ Γλαῦκον τὸν Δημύλου, καὶ τοὺς πρὸ τούτων ἔτι ἀθλητάς.

For the old system of athletic training produced men like Milo and Hipposthenes and Poulydamas and Promachus and Glaucus the son of Demylus, and other athletes. (Transl. König, in Rusten, König 2014, 399–401).

D. General commentary

Moeris’ entry (A.1) discusses two alternative forms of the accusative plural of the personal name Δημοσθένης. One of these (Δημοσθένας) is ascribed to Attic speakers, while the other (Δημοσθένεις) is ascribed to Greek speakers, i.e. the cultivated Greek speakers of Moeris’ own times (see Maidhof 1912, 319–38).

The s-stems-stems acc. plur. -ας was considered typical of the Attic dialect according to Moeris himself and Choeroboscus (B.3), who may have depended on Herodian for this claim. For a possible early instance of an s-stem acc. plur. -ας in Attic, see IG 13.832.3 [490–480 BCE?] φσευδᾶς (= ψευδᾶς, the standard form in prose being ψευδεῖς), about which various hypotheses have been made (see Kaczko 2016, 333, with further bibliography). Nevertheless, Plato (C.1), Plutarch (C.2), Lucian (C.3), and the Atticising writer Philostratus (C.4) show similar forms of personal names (though Philostratus also uses the acc. plur. Γανυμήδεις in VA 3.27, on which see Schmid Atticismus vol. 4, 21; cf. Synes. Calv. 6.1 and Dio. 3.1, in which codd. AVO have Διογένας and Σωκράτας while other MSS have Διογένεις and Σωκράτεις). These forms are probably produced by analogy with masculine nouns in -āā-stems (see Meisterhans, Schwyzer 1900, 137 n. 1197). The analogical influence of masculine nouns in -ā on s-stem nouns is proven by two other phenomena: first, the s-stem acc. sing. -ην (instead of the regular -η), which ‘was the normal form of the accusative by ca. 400 BC’ (Threatte 1996, 174, in discussing Attic inscriptions; see also K–B vol. 1, 512–3; Threatte 1996, 138, 173–7), and second, the nom. plur. οἱ Δημοσθέναι, οἱ Ἀριστοφάναι, οἱ ἑπταέται, οἱ δωδεκαέται reported by grammatical sources (B.3, B.4). This analogical process may also have been influenced by acc. plur. forms of consonantal stemsConsonantal stems such as φύλακας, πατέρας, and δαίμονας, however, as well as by personal pronouns like ἡμᾶς and ὑμᾶς (< ἡμέ-ᾰς, ὑμέ-ᾰς, see e.g. Lejeune 1972, 260), as K–B vol. 1, 433 suggest.

Moeris’ prescription of Δημοσθένας may depend on a lost locus classicus. Moreover, plurals of personal names – a rarity in themselves – can be used as a stylistic meansStyle of magnifying an expression, be it laudatory or contemptuous (see Fraenkel 1950 vol. 3, 679; Finglass 2018, 332, with further bibliography). Moeris’ original entry may thus have been more extended, perhaps including other instances of this declension, such as Ἀριστοφάνας, which is found in Plato (C.1), one of Moeris’ canonical authors (although occasionally with reservations, see entry ἀνθρωπεία, ἀνθρωπίνη).

In the acc. plur. Δημοσθένεις, on the other hand, -εις is modelled on the nom. plur. of diphthongal stemsDiphthongal stems (see e.g. Chantraine 1933, 60). Ancient grammarians appear uncertain in their explanations of acc. plur. -εις. As we learn from B.6, Herodian argued for syneresisSyneresis of ε and α into the ει diphthong. Choeroboscus believed instead that ει was due to homophony with the contracted nom. plur. (B.5, B.6; see also Choerob. in Theodos. GG 4,1.239.5–15, 417.2–3; 4,2.336.33–337.1; K–B vol. 1, 216); it is unknown whether Choeroboscus’ theory depends on previous sources. Moeris himself argued that the form Δημοσθένεις was due to τὸ ἀνάλογον (‘analogy’). This last term must not be confused with what modern linguists would call analogy, however; as Matthaios (2020, 339) explains, ‘the [ancient] analogical method is based on the idea that a grammatical accident, be it inflectional or prosodic, will occur in similar word forms in the same fashion’. Although traces of this method are found in Aristophanes of Byzantium and Aristarchus of Samothrace, the concept was developed more fully by Herodian (2nd century CE), who identified various criteria for correct use of the analogical method and was thus able to formulate ‘prosodic, but also morphological and orthographical rules’ (Matthaios 2020, 339). Indeed, together with other criteria analogy was the touchstone of Hellenismos, i.e. ‘language correctness’ (see Siebenborn 1976, 56–84; Pagani 2015, 832–9). In this context, Moeris’ rejection of Δημοσθένεις – a form that he thought based on analogy – is unsurprising. As Probert (2011, 275–9) notes, the Atticists’ views did not always accept earlier (e.g. Herodianic) criteria for linguistic correctness such as analogy, in relation to which ‘an Attic expression might be found irregular’ (275). Moeris may thus have revisited a preexisting inflectional canon (such as those testified by B.1 and B.2) from an Atticist perspective, comparing said canon with further evaluations on ‘irregular’ Attic forms (such as B.3). A similar situation may have occurred in other entries where Moeris either explicitly mentions or implicitly refers to analogy when discussing forms used by Greek speakers, as at Moer. μ 11Moer. μ 11 (μονομάχης Ἀττικοί· κατὰ δὲ τὴν ἀναλογίαν μονομάχος, ‘Users of Attic [employ] μονομάχης, while μονομάχος [is said] by analogy’; see entry μονομάχης, μονομάχος) and μ 17Moer. μ 17 (Μειδίου Ἀττικοί· Μειδία τό τε ἀναλογικὸν καὶ τὸ ἑλληνικόν, ‘Users of Attic [employ] Μειδίου [gen. sing.]. Μειδία is the analogical and the Greek form’). Finally, Moer. μ 10Moer. μ 10 (μάντεων τὴν πρώτην ὀξυτόνως Ἀττικοί· τὴν δευτέραν ὀξυτόνως Ἕλληνες, ‘Users of Attic [employ] μάντεων with an acute on the first syllable. Users of Greek [employ it] with an acute on the second syllable’) probably implies a reference to the analogical rule according to which gen. plur. forms like πόλεων should be paroxytone (see Probert 2011, 278–9).

E. Byzantine and Modern Greek commentary

Due to the gradual merger of masculine and feminine nouns of the 1st and 3rd declensions (Horrocks 2010, 121, 286–88; CGMEMG vol. 1, 253), s-stem nouns in Medieval Greek are virtually limited to cultured literature. In particular, s-stem acc. plur. -ας of personal names is sporadically found in poetry and prose writing, possibly as a learned alternative to the prose standard ending -εις; George of Pisidia has Ἱπποκράτας (Hexameron 1497), but Ἱπποκράτεις is found in Tzetzes (H. 7.155.966), while Michael Psellus uses both Σωκράτας (Orationes hagiographicae 8.322 Fisher) and Σωκράτεις (Opusculum 6.20 and 6.96 Duffy).

F. Commentary on individual texts and occurrences



Chantraine, P. (1933). La formation des noms en grec ancien. Paris.

Clement, P. A.; Hoffleit, H. B. (1969). Plutarch. Moralia. Vol. 8: Table-Talk. Books 1–6. Translated by P. A. Clement, H. B. Hoffleit. Cambridge, MA.

Finglass, P. J. (2018). Sophocles. Oedipus the King. Edited with Introduction, Translation, and Commentary. Cambridge.

Fraenkel, E. (1950). Aeschylus. Agamemnon. 3 vols. Oxford.

Harmon, A. M. (1925). Lucian. Vol. 4: Anacharsis or Athletics. Menippus or The Descent into Hades. On Funerals. A Professor of Public Speaking. Alexander the False Prophet. Essays in Portraiture. Essays in Portraiture Defended. The Goddess of Surrye. Translated by A. M. Harmon. Cambridge, MA.

Kaczko, S. (2016). Archaic and Classical Attic Dedicatory Epigrams. An Epigraphic, Literary, and Linguistic Commentary. Berlin, Boston.

Lamb, W. R. M. (1925) Plato. Vol. 3: Lysis. Symposium. Gorgias. Translated by W. R. M. Lamb. Cambridge, MA.

Lejeune, M. (1972). Phonétique historique du mycénien et du grec ancien. Paris.

Maidhof, A. (1912). Zur Begriffsbestimmung der Koine besonders auf Grund des Attizisten Moiris. Würzburg.

Matthaios, S. (2020). ‘Greek Scholarship in the Imperial Era and Late Antiquity’. Montanari, F. (ed.), History of Ancient Greek Scholarship. Leiden, Boston, 260–372.

Meisterhans, K.; Schwyzer, E. (1900). Grammatik der attischen Inschriften. 3rd edition. Berlin.

Pagani, L. (2015). ‘Language Correctness (Hellenismos) and Its Criteria’. Montanari, F.; Matthaios, S.; Rengakos, A. (eds.), Brill’s Companion to Ancient Greek Scholarship. Vol. 2. Leiden, Boston, 798–849.


Andrea Pellettieri, 'Δημοσθένας, Δημοσθένεις (Moer. δ 48)', in Olga Tribulato (ed.), Digital Encyclopedia of Atticism. With the assistance of E. N. Merisio.
DOI: https://doi.org/10.30687/DEA/2974-8240/2023/02/018

This article provides a philological and linguistic commentary on the accusative forms Δημοσθένας and Δημοσθένεις discussed in the Atticist lexicon Moer. δ 48.

AnalogyDeclension metaplasmHerodianProper names