ἀνοητία, ἀνοησία
(Phryn. PS 35.8–9, Poll. 2.227–8, Poll. 3.140, Poll. 4.9, Moer. α 58, Moer. α 69)

D. General commentary

Phrynichus (A.1) and Moeris (A.5, A.6) prescribe the abstract nounsAbstract nouns ἀνοητία (only attested in Aristophanes: see C.1) and ἀκρατία (first attested in Plato, see C.3) instead of their assibilated equivalents ἀνοησία and ἀκρασία. Phrynichus also recommends the use of other nouns in -τία, implicitly rejecting the equivalent forms in -σία as well as other competitors (see also B.1 and B.2, based on Atticist sources). Pollux (A.2, A.3, A.4) seems to accept both -τία and -σία, although his evaluation of ἀνοητία/ἀνοησία raises some problems: see F.2. It is conceivable that the focus of some of our entries is morphological; therefore, for a better evaluation of the Atticist sources and their preference for -τία, a preliminary analysis of the distribution of this suffix is needed.

Abstract nouns in -τία (with the corresponding assibilated form -σία, found in Mycenaean and Attic-Ionic) are mostly compoundsCompounds – many of them with the privative ἀ- as the first component – often corresponding to uncompounded abstracts in -σις. Nouns in -τία/-σία are often derived from forms in -της and -τος (e.g. ἀνοητία/ἀνοησία < ἀνόητος ‘unknown’) and/or verbs in -ζω and -αίνω. However, -τια/-σια gradually spread to other categories by analogy (for further details on the history of the suffix, see Lobeck 1820, 505–12; Fraenkel 1912, 36–7; Fraenkel 1913; Chantraine 1933, 83–6). Although the assibilated -σία tends to take over, -τία is also found in recent derivatives, especially when the base form has -τ- – but analogy with older abstracts in -τία must also have played a role (an example of a recent derivative in -τία is PSI 10.1103.9 (= TM 13828) [Krokodilopolis, 192–194 CE] διεραματία ‘lading corn’ – a liturgy).

Most importantly, -σία spread widely in the koine and eventually supplanted -σις formations (see the impressive list in Buck, Petersen 1945, 147–9, 157–63, 169; see also E.). This is probably the reason why the forms in -σία seem to be discredited by stricter Atticists. In this respect, other entries may help to define a clearer picture. The first parallel is Moer. ν 15Moer. ν 15: νουθέτησις Ἀττικοί· νουθεσία Ἕλληνες, ‘Users of Attic [employ] νουθέτησις (‘admonition’), users of Greek [employ] νουθεσία’. This entry can be compared with Poll. 9.139–40Poll. 9.139–40: νουθεσία καὶ ὡς Πλάτων νουθετεία· φαῦλος γὰρ ὁ Μενάνδρου νουθετισμός (‘νουθεσία and νουθετεία, as Plato [Comicus?] [says]: for Menander’s νουθετισμός [fr. 629] is ordinary’) and Phot. ν 272Phot. ν 272: νουθετίαν καὶ νουθετησμὸν λέγουσι (‘[Users of Attic] say νουθετία and νουθετισμός’). Note that νουθεσία is also attested in Aristophanes (Ra. 1009)Ar. Ra. 1009: it is conceivable that Moeris condemned it because it was isolated, but one cannot exclude that a negative appraisal of the suffix -σία contributed to its rejection.

A second parallel is provided by Phryn. Ecl. 52Phryn. Ecl. 52: ἱκεσία καὶ τοῦτο ἀδόκιμον, ἱκετεία δὲ λέγε, ‘ἱκεσία (‘supplication’): this too is unapproved, say rather ἱκετεία’. This entry can be compared with Phryn. PS 77.1–2Phryn. PS 77.1–2: ἱκετεία. διὰ τοῦ τ, οὐ διὰ τοῦ σ (‘ἱκετεία. With τ, not σ’; on the apparent difference between -εία and -ία, see Vessella 2018, 80: ‘[s]pellings with <ει> and with <ι> are treated as interchangeable when the only reference is to the length of the vowel, the digraph automatically implying a long vowel, the simple iota a short vowel’; cf. also the entry ἱκεσία, ἱκετεία).

For a final parallel, see Phryn. Ecl. 84Phryn. Ecl. 84: θερμότης λέγε, ἀλλὰ μὴ θερμασία, ‘Say θερμότης (‘heat’), not θερμασία’ (see Rutherford 1881, 198 and entry θερμότης, θερμασία).

In sum, it is conceivable that forms in -σία were frowned upon by stricter Atticists, at least when an alternative form was found in canonical authors. Such a dislike for -σία may have been determined by its abundant use in the koine.

Be that as it may, the forms in -τία prescribed in A.1 (a series of nouns originally found in Old Comedy, according to Kaibel in PCG vol. 3,2, 381) and A.5 (see also B.1, B.2) did not meet with any favour. ἀνοητία is attested only by Aristophanes (C.1) and Nilus of Ancyra (Ep. 1.78.8). ἀκρατία, used by Plato (Grg. 525a.4 = C.3), has only a few occurrences in later philosophical writings. νουθετία in A.1 is probably the same as νουθετεία, which nonetheless is only attested in Phld. Hom. 10.9 Dorandi; but see the aforementioned Ar. Ra. 1009: νουθεσίας. Furthermore, φιλοποτία is never attested in the extant literature (the same holds true for μοιχοληπτίαμοιχοληπτία), while φιλοποσίαφιλοποσία is also found in Xenophon (2x) and Plato (1x). λυχνοκαυτίαλυχνοκαυτία is only found in Cephisodorus (fr. 11, see F.1). Finally, ἀθλοθετία only occurs in Aristophanes (C.2), while ἀθλοθεσίαἀθλοθεσία is found in an Attic inscription dating from before 178 BCE (IG 2².1368.131–2).

The picture sketched so far seems to point to an Atticist tendency to promote alternatives to -σία even when they appear poorly attested in canonical Attic authors. In this respect, it cannot be ruled out that -σία was considered foreign to the Attic dialect. A hint may be provided by Et.Gud. 554.18Et.Gud. 554.18: εὐεργεσία· ... ἐκ τοῦ εὐεργέτης εὐεργετία, ὡς ἱκέτης ἱκ⟦ε⟧τεία, καὶ τροπῇ Ἰωνικῇ τοῦ τ εἰς σ εὐεργεσία, ‘εὐεργεσία (‘well-doing’)· ... from εὐεργέτης [one has] εὐεργετία – just as [from] ἱκέτης [one has] ἱκ⟦ε⟧τεία – and, with the Ionic change of τ to σ, [one has] εὐεργεσία’. Conversely, several grammatical sources characterise the phonetic change of σ into τ as typically Attic, see e.g. Orio 154.10, Ep.Hom. in Il. 1.8.B.2.A.1, and Ael.Dion. σ 15Ael.Dion. σ 15 (in Eust. in Il. 2.95.29–96.11). According to Aelius Dionysius, sigmatism was rejected by the poets of Old ComedyOld Comedy, who considered it a ‘tragic’ trait (see Pl.Com. fr. 29 and Eub. fr. 26, both reacting against Euripides’ alleged fondness for sigma – see Eur. Med. 476 and fr. 129), and was even avoided by Pericles. More generally, sigma was perceived as a ‘graceless’ and ‘unpleasant’ sound (to put it with Dion.Hal. Comp. 14.80) – something which emerges here and there throughout the history of ancient Greek literature (for a survey, see Clayman 1987; cf. also Willi 2002, 125; Pirrotta 2009, 104–6). Thus, it cannot be excluded that the Atticist preference for -τία over -σία was influenced by the general aversion to sigmatism, in addition to the reasons mentioned above.

E. Byzantine and Modern Greek commentary

As noted above, the suffix -σία spread widely in the koine and supplanted -σις formations, so that it is impossible here to give a detailed account of all its uses throughout the history of Greek. Suffice it to say that -σία survives in Medieval and Modern Greek as a suffix for the formation of abstract nouns (especially derivatives of verbs and agentive nouns, as well as adjectives), attached to bases in -τ-: see e.g. κυριολεξία (‘literality’, < κυριολεκτώ ‘to speak literally’) and αδιαλλαξία (‘intransigence’, < αδιάλλακτος ‘intransigent’). It still plays an important role, e.g. in medical vocabulary, forming termini technici which refer to pathological conditions: see LKN s.v. -σία.

F. Commentary on individual texts and occurrences

(1)    Phryn. PS 35.8–9 (A.1)

Among the forms quoted by Phrynichus, λυχνοκαυτία is found in Cephisodorus, a poet of Old Comedy. Ath. 15.701a–b, the source of Cephisodorus’ fragment, seemingly presents the form in an ‘Atticist’ way: λυχνοκαυτίαν δὲ ἣν οἱ πολλοὶ λέγουσιν λυχναψίαν Κηφισόδωρος ἐν Ὑί, ‘Cephisodorus in The Pig [says] λυχνοκαυτία, which the multitude calls λυχναψία’. Olson’s translation of this passage is misleading: according to him οἱ πολλοί means ‘many authorities’ (Olson 2012, 209) – but λυχναψία is not found in literary sources until the imperial era, while it has many occurrences in documentary papyri from the 2nd century CE onwards (cf. also Phot. λ 493: λυχνοκαυτίαι, οὐχὶ λυχνοκαΐαι, possibly from an Atticist source; see also PCG vol. 4, 67; Orth 2014, 357 n. 593), a distribution which confirms that it is a common Atticist expression used to refer to the language of the masses (see Matthaios 2013). It is conceivable that Athenaeus’ source was an Atticist one. Furthermore, Phrynichus himself may have found λυχνοκαυτία in Cephisodorus and accepted it on the basis of the presence of other similar forms in canonical authors. This need not be inconsistent with the fact that Phrynichus’ main comic sources in the PS are Aristophanes, Eupolis, and Cratinus. Indeed, as Photius says, Phrynichus also used other comic poets ‘as long as they speak good Attic’: as Tribulato (2024)    shows, these are minor poets of Old Comedy – among whom Cephisodorus may well have found his place.

(2)    Poll. 4.9 (A.4)

This passage is problematic. First of all, it is not clear what σκληρά means exactly. Secondly, MSS have the alternative readings ἀνοησία and ἀνοητία. Pollux’s use of the evaluative term σκληρός is addressed by Bussès (2011, 65–6); Franciò (2011, 79); Conti Bizzarro (2014, 45–8); Radici Tavernese (2016–2017, 201–3); Conti Bizzarro (2018, 17–34). Although σκληρός often seems to refer to euphonic issues, it does not seem to have a univocal meaning. To take one example, in Poll. 7.127Poll. 7.127 ὁμοιωτική is accepted, while ὁμοιωτής is said to be ‘harsh’ – according to Pollux, one should rather use μιμητής. This does not seem to be a matter of euphony: rather, ὁμοιωτής (a hapax) is probably considered an unusual and strange word compared to the long-established μιμητής (also well attested in Plato). See also Poll. 3.119Poll. 3.119, 6.125Poll. 6.125, 6.209Poll. 6.209, 8.81Poll. 8.81, where the label σκληρός is tacitly opposed to another evaluative term, εὐτελήςεὐτελής (‘mean, worthless’), so that σκληρός probably has a diaphasic meaning, referring to words that are perceived as unusual (see also Poll. 9.7Poll. 9.7, where σκληρός is probably opposed to κοινόςκοινός, ‘common, ordinary’). Be that as it may, it is conceivable that Pollux rejected ἀνοησία/ἀνοητία because an authoritative, better-attested synonym like ἄνοια was available (Poll. 4.9Poll. 4.9). Even so, the textual problem remains unresolved. Indeed, it is not clear whether ἀνοησία, the form also rejected by Phrynichus and Moeris, should be read or not. First of all, other words classified by Pollux as σκληρά are found in canonical authors such as Thucydides and Plato (see e.g. Poll. 5.126Poll. 5.126, who describes νόμισις, ‘belief’, as ‘harsh’: note that the word is found in Thuc. 5.105.1). Moreover, although Pollux apparently accepts ἀνοητία in A.2, there are cases where he simply mentions a form and later labels it as ‘harsh’, see Poll. 7.190Poll. 7.190: γελωτοποιεῖν (‘to provoke laughter’), and 9.148Poll. 9.148: σκληρότερον γὰρ τὸ γελωτοποιεῖν (‘γελωτοποιεῖν is somewhat harsh’). Thus, the reading ἀνοητία cannot be excluded.

(3)    Thom.Mag. 2.3–5 (B.2)

ἀκρατία and ἀνοητία are not attested in Thucydides, where only ἀδυνασία is found (Thuc. 7.8, 8.8): see already Poppo (1821, 243). It is impossible to know why Thomas Magister apparently fails to give the correct information. The entry as edited by Ritschl may be the final outcome of a process of reworking Thomas’ original lexicon (on this issue, see e.g. Gaul 2007; Nuti 2013). In this respect, it cannot be excluded that the entry was originally a shorter one, aimed at discussing a form actually attested in Thucydides, and that later scholarship on similar matters was tacitly added. For a similar, even if not identical, case, see Thom.Mag. 329.3–330.14Thom.Mag. 329.3–330.14, a ‘stratified’ entry which may have originally dealt with a Thucydidean use.