κυνίδιον, κυνάριον
(Phryn. Ecl. 151, Antiatt. κ 87, Phryn. PS 84.22)

D. General commentary

Ecl. 151 (A.1) concerns two diminutive forms, κυνίδιον and κυνάριον. Both may indicate a puppy or a little dog, although both can also be used generally to refer to any dog (see some of the translations above and the discussion below). Phrynichus overtly recommends using κυνίδιον and then mentions that κυνάριον is attested ‘only once’ in a comedy by Theopompus, seemingly – albeit not openly – rejecting it (on this and the apparently different prescription of the entry in the PS, A.3, see below). Since both words occur in 5th-century Attic literature, the issue at stake in this entry of the Eclogue is not just the suitability of either form, but also the question of which literary models should be considered canonical. Although Phrynichus does not cite any locus classicus for κυνίδιον, he may have comedy in mind: in the whole of 5th-century Attic literature, κυνίδιον is an exclusively comic word and we must wait for Xenophon (Oec. 13.8), Plato (C.9 and Euthd. 299a), and Aristotle (HA 612b, etc.) for more widespread attestations. Similarly, in seeming defence of κυνάριον, the Antiatticist (A.2) quotes Alcaeus Comicus. Both lexica may perhaps depend on an earlier source that was focused on comedy.

The different attitudes of Phrynichus and the Antiatticist are interesting for several reasons: the role of the canonCanon in Phrynichus’ appreciation of linguistic correctness, the use of hapax legomenaHapax in classical literature, and the Atticists’ attitude towards features that, albeit already classical, developed significantly inKoine koine Greek. To begin with the point about theCanon canon, Phrynichus’ preference for κυνίδιον reflects the fact that it occurs in two of the three major authors of the Old Comedy triad: Eupolis (C.3) and Aristophanes (C.4, C.5, C.6, C.7). The Eclogue preserves some twenty-nine direct references to Old Comedy, with Aristophanes, Cratinus, and Eupolis together garnering twenty-one mentions. Other Old Comedy authors seldom appear in the Eclogue and rarely without some expression of caution on Phrynichus’ part. In Ecl. 151, Theopompus’ lesser authority, paired with the information that κυνάριον is a hapax in his works, contribute to making the word suspicious (on Phrynichus’ carefulness in dealing with Attic forms which are hapax legomena, see Monaco 2024 and entries βρέχει and πρόσφατος). Alcaeus Comicus, to whom the Antiatticist instead resorts, is also not an author whom Phrynichus favours. Phrynichus mentions him only once inPhryn. Ecl. 231 Ecl. 231 in reference to βασίλισσα ‘queen’, a form that he rejects for lack of sufficient evidence in Attic sources. Clearly, the authority of Alcaeus Comicus alone is not sufficient to redeem the word: the same situation likewise applies to κυνάριον.

Turning to the linguistic aspect of this question, two further factors must have influenced Phrynichus’ rejection of κυνάριον. As aforementioned, both κυνίδιον and κυνάριον are apposite Attic words in chronological terms: first attested in comedy, they continue to be used by 4th-century Attic authors (κυνίδιον: Xenophon, Plato, Lysias, Aristotle; κυνάριον: Xenophon, Plato). The diminutive suffixes -ίδιον and -άριον are among the strategies used to recharacterise the primary suffix -ιον which, owing to its wide semantic range, was not sufficiently expressive (see Chantraine 1933, 68). It may also be noted that the bleaching of diminutives regularly leads to double formations in -ιδάριον, -ισκάριον, -ιλλύδριον, etc. (see Chantraine 1933, 73–5). However, while -ίδιον is very productive already in the classical period, even outside the class of diminutives (Petersen 1910, 212–40, Chantraine 1933, 68–9), -άριον is much less frequent in classical Attic (diachronic overview in Petersen 1910, 260–71). Peppler (1902, 11) counts only thirty diminutives in -άριον in the period up to Aristophanes, but twenty-one in Aristophanes alone. However tentative these figures may be, they provide a lens through which to make sense of Phrynichus’ attitude.

While already classical, -άριον becomes much more productive in koine Greek, often at the expense of -ίδιον. In the New TestamentNew Testament, a text where diminutives abound to such an extent that later Atticising scribes seem to have tried to eliminate them (cf. Elliott 1970), -άριον and simple -ιον replace -ίδιον (Chantraine 1933, 71–2; see Mayser Gramm. vol. 1,3, 38–44 on the presence of these suffixes in the papyri). Concerning κυνάριον and κυνίδιον in particular, it should be noted that the New Testament contains κυνάριον but not κυνίδιον. The word occurs only in two parallel passages that relate the history of the Canaanite woman (ΝΤ Ev.Matt. 15.26–7, C.10, and ΝΤ Ev.Marc. 7.27–8). In both cases, κυνάριον is used in the generic meaning ‘dog’ (see Elliott 1970, 394). Many later attestations of the word are in exegetical texts dealing with this famous New Testament pericope. Moreover, although both words are common in post-classical literature, not only is κυνίδιον more frequent, but it also features in authors who never use κυνάριον (see especially Galen and Lucian), a fact which confirms that the higher frequency of the latter in the lower koine must be behind Phrynichus’ uneasiness with κυνάριον.

Another linguistic factor that may have played a role in Phrynichus’ judgment is the fact that in the context of Greek-Latin bilingualismBilingualism -άριον ends up crossing paths with the LatinLatin neuter suffix -ārium. Although the two suffixes are unconnected (the short quantity of the /a/ in -άριον seems to be original; the neuter -ārium is formed with an adjectival suffix going back to -āsios, continued in Oscan: see Leumann 1977, 297–300), many Latin words in -ārium had been borrowed into Greek by Phrynichus’ period (see below re. κοχλιάριον), and the concomitant loss of vowel quantity may have contributed to making flawless Greek forms in -άριον appear as Latinisms (Cavenaile 1952, 201–2 discusses the impact of these forms on the language of the papyri, while at 195–7 he collects Greek diminutives derived from Latin words).

The rarity of κυνάριον in good classical Attic, together with the suspicious status of many forms in -άριον, must have settled the matter for Phrynichus. Notice that in Ecl. 398Phryn. Ecl. 398 too, Phrynichus recommends a form in -ίδιον (λιθίδιον) against one in -άριον (λιθάριον), while in Ecl. 292 he disapproves of the Latinism κοχλιάριον ‘spoon’, incongruously recommending λίστρον ‘shovel’ because it is used by Aristophanes (see the discussion in Dickey 2012, 61–2). Other entries of the Eclogue show how forms in -ίδιον were perceived to be acceptable (e.g. Ecl. 50Phryn. Ecl. 50 on κορίδιον ‘[young] girl’, Ecl. 61Phryn. Ecl. 61 on νοίδιον ‘little thought’ and βοίδιον ‘little cow’, Ecl. 223Phryn. Ecl. 223 on ῥοίδιον ‘small pomegranate’), sometimes in opposition to other forms (see Ecl. 362Phryn. Ecl. 362 on στηθίδιον vs. στηθύνιον as diminutives of στῆθος ‘breast’). It should be noted that many of these words are comic formations in which the suffix frequently lacks a concrete diminutive force (as in νοίδιον ‘little thought’) but is used rather to signify affection (as in the case of βοίδιον, which at Acharnians 1036 refers not to small oxen but to animals which are dear to Dercetes; for diminutives of endearment, see Peppler 1902, 17–23, Chantraine 1933, 54 and 70).

The Antiatticist’s reaction to Phrynichus’ proscription of κυνάριον is expected, given that κυνάριον is perfectly classical (though not as classical as Phrynichus requires). Less easy to settle is the opposing doctrine that is apparently transmitted by the PS (A.3). On one level, we could take this entry in the PS as an example of Phrynichus’ more tolerant take on correct language and the canon in this work. However, the evidently abridged status of the entry should induce some caution. Even Ecl. 151 – as it survives, at least, in most branches of the textual tradition – might be read as an ambiguous entry that, while overtly prescribing κυνίδιον, also makes a concession to κυνάριον. It is telling that the manuscripts of the q family show a fuller version that may have been devised to make the text of Ecl. 151 clearer to readers (this is also the family of the oldest dated manuscript of the Eclogue, N, cod. Ferrara, Biblioteca comunale Ariostea 155, 1337 CE; description in Fischer 1974, 17–23). In the q family, Ecl. 151 reads κυνίδιον λέγε, ἀλλὰ μὴ κυνάριον, κἂν Θεόπομπος ὁ κωμῳδὸς ἅπαξ τούτῳ χρῆται, a formulation that influenced Thomas Magister (A.4), who read the Eclogue in copies belonging to this branch (see Fischer 1974, 31). In conclusion, there is room to doubt that the original entry of the PS (A.3) was a prescriptive lemma approving of both κυνίδιον and κυνάριον. More probably, it simply registered that both forms occurred in Attic (comedy), while subsequent epitomisation turned the information into a prescription.

One final issue needs clarifying. The discussion above has treated κυνίδιον and κυνάριον as uncontroversial instances of diminutives. In fact, already in some of its earliest attestations κυνίδιον has the generic meaning (‘dog’), which is part of the generalising functions of the suffix -ίδιον (see Petersen 1910, 227–8 and 236; the same applies to ‑ιον and -άριον). In Eup. fr. 220 (C.3), κυνίδιον is not a clear diminutive: it could refer to puppies, small dogs, or dogs in general. In his interpretation of the fragment, Olson (2016, 238–40) plumps for ‘puppy-dogs’, adducing the parallel of X. Oec. 13.8X. Oec. 13.8, which describes how young dogs learn to run around and turn somersaults. But the emphasis on young dogs is not at all evident in Eupolis: the diminutive here is rather a way to underline a household’s familiarity with its own dogs (τείχια are the walls the surround the house), as is implied in fact by Olson himself (2016, 240: ‘a homely diminutive’). Turning to Aristophanes, κυνίδιον is purposely used in Acharnians 541–3 (C.4) to highlight the negligible importance of Seriphus and hence the futility of some pretexts for war (see Lilja 1976, 69 and Olson 2002, 214 for an interpretation of the image); in fr. 209 (C.5) too, κυνίδιον probably refers to a puppy. However, in the two Pax passages (C.6, C.7), which deal with hungry dogs, κυνίδιον surely refers to dogs in general. Although the image of hungry dogs may certainly apply to puppies, the fact that both similes are used in a dramatic political context suggests that the reference is to adult dogs (this is especially evident in C.7, where the hungry dogs preparing to tear their prey to pieces function as a simile for jurors: see Olson 1998, 203 for a discussion and parallels).

Unlike κυνίδιον, κυνάριον indubitably refers to puppies in all its earliest occurrences (see C.8, C.9 – though for the latter, see the alternative interpretation of Petersen 1910, 264 n. 1) or to small dogs (cf. the Μελιταῖον κυνάριον, ‘Maltese dog’ in Thphr. Char. 21.9; on this type of lapdog, a breed mostly referred to as Μελιταῖον κυνίδιον, see Busuttil 1969). It is not until the New Testament (see C.10 and above) that we find the generic meaning ‘dog’. It is regrettable that we do not have a context for the only three occurrences of κυνάριον in Greek comedy (C.1, C.2, and as the title of a play by Timotheus) because this may have helped to clarify whether the issue at stake in the Eclogue entry was also semantic. In the text transmitted in cod. Par. Coisl. 345, the Antiatticist lemma apparently informs us that κυνάριον was used by Alcaeus Comicus in a comic fashion (κωμικῶς), a reading accepted in the editions of both Bekker and Valente. The correction of κωμικῶς into κωμικός variously proposed by van Dam, Lobeck, and Meineke (see apparatus of A.2) may well be correct. However, if κωμικῶς is authentic, this would suggest that κυνάριον was used in a metaphorical or symbolic fashion, or as a nickname for a character (see the brief discussion in Orth 2013, 137), perhaps to express contempt (for this function of -ιον formations in general, see Petersen 1910, 113–30; on comic diminutives, see Peppler 1902, 23–8). It is unfortunate that nothing is known about the plot of Timotheus’ Κυνάριον, from which we have only a three-line fragment concerning a dinner party transmitted by Athenaeus (6.243c–d = Tim.Com. fr. 1).

E. Byzantine and Modern Greek commentary

Both κυνίδιον and κυνάριον are used in Byzantine texts, often in exegetical discussions of the New Testament expression βαλεῖν τοῖς κυναρίοις (see C.10). While they both occur in high-register prose, κυνίδιον gradually becomes the less common form in the lower register, with κυνάριον gaining ground in both high- and low-level texts (the latter includes works such as Digenis Akritis, the Life of St. Andrew the fool, and hagiographies). Kriaras, LKN, does not register either form: the common word for ‘dog’ in Medieval and Modern Greek is σκύλος (with derived forms), with κύων surviving only in the extremely rare (probably one-off) form κύνας, as testified by its absence in Modern Greek dialects.

F. Commentary on individual texts and occurrences

N/A