οἰκόσιτος, αὐτότροφος, οἰκότριψ, οἰκογενής
(Phryn. Ecl. 174, Moer. ο 25, Thom.Mag. 265.10)

D. General commentary

Phryn. Ecl. 174 (A.1) discusses four compounds, rejecting two of them – i.e. αὐτότροφος (literally ‘who feeds himself’, and thus, ‘who brings one’s own provisions’) and οἰκογενής (‘house-born’) – and recommending οἰκόσιτος (‘who brings food from home’) and οἰκότριψ (literally ‘the one who frequents the house’, and thus, ‘house-slave’). The underlying reason for discussing these four compounds in the same entry is not immediately clear: it is possible that the main aim of Phrynichus’ entry was to proscribe αὐτότροφος and to prescribe οἰκόσιτος, both of which are generic terms referring to the act of bringing food – and thus providing – for oneself, and that the lexicographer chose to discuss οἰκογενής and οἰκότριψ, two words more specifically related to slaves, in the same entry because of the presence of the same first constituent οἰκο-. Moeris (A.2), followed by Thomas Magister (A.3), also recommends οἰκότριψ, stating that its koine equivalent was οἰκοτραφής (‘raised at home’).

αὐτότροφος is a possessive compound from the pronoun αὐτός (‘self’) and the noun τροφή (‘nourishment’ or ‘food’; for an interpretation of -τροφος compounds, see Tribulato 2015, 89–93, 310). Compounds ending in -τροφος are already found in Homer and are well attested in the classical period (cf. e.g. ἱπποτρόφος ‘horse-feeding’ in Xenophon and Demosthenes; μηλοτρόφος ‘sheep-feeding’ in Herodotus and Sophocles; παιδοτρόφος ‘nourishing children’ in Aeschylus, Euripides, and Sophocles; and σύντροφος ‘raised together with, living with’ in Thucydides, Euripides, Sophocles, Herodotus, and Aristophanes), although most of these are post-classical (for a list, see Buck, Petersen 1945, 404–5). However, what sets αὐτότροφος apart is that it has only one other attestation outside of A.1 (for the only occurrence in Byzantine literature, cf. E.), in Favorinus’ On Exile (C.4). Here, the term describes the Greeks who followed Agamemnon to war without being paid by him, instead providing for themselves. Remarkably, in C.4, αὐτότροφος is preceded by a similar-looking compound, i.e. αὐτόμισθος ‘paying for oneself’ (from αὐτός and μισθός ‘pay, wage’), which is a hapaxHapax in Greek literature. Overall, given that the only other attestation of αὐτότροφος (besides A.1 and C.4) is in a Byzantine work (cf. E.), it is difficult not to consider A.1 as Phrynichus’ indirect attack on Favorinus’ use of the word. Given Phrynichus’ tendency to openly criticise his contemporary by name (cf. e.g. Phryn. Ecl. 140Phryn. Ecl. 140, with entry νῆες, ναῦς, νῆας; Phryn. Ecl. 161Phryn. Ecl. 161, with entry εἶμι, ἐλεύσομαι; Phryn. Ecl. 215Phryn. Ecl. 215, with entry διδόασι, διδοῦσι; and Phryn. Ecl. 209Phryn. Ecl. 209), it is conceivable that a direct reference to Favorinus was originally included in the entry and was subsequently lost due to epitomisationEpitome (cf. the apparatus of A.1).

The term that Phrynichus recommends in place of αὐτότροφος is οἰκόσιτος, a compound of οἶκος ‘house’ and σῖτος ‘grain, bread’ and, more generally, ‘food’. While better attested than αὐτότροφος, οἰκόσιτος is not found in authors whom Phrynichus considered reputable. Indeed, the term first appears in the 4th century BCE in comedies by Anaxandrides (C.3, in reference to a son) and Antiphanes (fr. 198, in reference to an assembly member), and it is used three times by Menander (fr. 98, fr. 340, and Cith. fr. 6, in reference to a bridegroom, a wedding feast, and an audience, respectively). All of these passages are quoted in direct succession by Athenaeus in a section dedicated to -σιτος compounds (Ath. 6.247e–248a). A further occurrence in Middle Comedy is in Anaxil. fr. 38 (where the adjective is used in reference to the persona loquens). Regarding documentary texts, οἰκόσιτος occurs in two 4th-century BCE inscriptions (i.e. IG 2².1672 [Attica, 329–8 BCE] and IG 2².1673a [Attica, end of the 4th century BCE], several times in both inscriptions), where the term indicates ‘slaves hired by the state from their master, who nevertheless undertakes to provide their food’ (Millis 2015, 126). Overall, the sense gained is that οἰκόσιτος became a specialised term referring to slaves in the official language of inscriptions, but was used more generically in literature, and this is likely the usage intended by Phrynichus, who promotes it against αὐτότροφος. The lack of papyrus attestations suggests that the term may not have been adopted in koine Greek, and this, together with the literary occurrences, may have made οἰκόσιτος preferable to αὐτότροφος for Phrynichus.

The second term proscribed by Phrynichus in A.1 – i.e. οἰκογενής ‘house-born [slave]’, a verb-final compound, from οἶκος and the full grade of γίγνομαι ‘to be born’ – was surely widespread in koine Greek. Indeed, after two attestations in Classical Greek – namely in Aristophanes (Ar. Pax 788, where it refers to a domestically bred quail; a similar use is in Arist. NA 558b.20) and Plato (C.1, where it applies to a slave) – the term remains in use in Post-Classical literary Greek (e.g. Plutarch 5x), spreading to koine texts. This is evident by its many occurrences not only in the Septuagint (11x) but also in documentary texts. Indeed, οἰκογενής is the technical term indicating a house-born slave (in the Septuagint it is often opposed to ἀργυρώνητος, lit. ‘bought with money’, cf. e.g. LXX Ge. 17.12). It is thus extremely frequent in manumission inscriptions (particularly in the Delphic manumissions corpus; cf. e.g. F.Delphes 3.1.567.1–3 [ca. 125 BCE]: ἀπέδοτο Ξένων Δη[μη]τρίου τῶι Ἀπόλλωνι τῶι Πυθίωι παιδάριον οἰκογενὲς ὧι ὄν[ομα] Στράτων, ‘Xenon [son] of Demetrius gave to Pythian Apollo a house-born boy, named Strato’) and in papyrus documents concerning slave sales, including P.Köln 4.187.19 (= TM 3177) [Herakleopolis, 146 BCE], P.Oxy. 31.2582.6 (= TM 16897) [Euergetis, Kynopolites, 51 CE], and P.Col. 8.222.19 (= TM 17632) [Oxyrhynchus, 160–1 CE] (on house-born slaves in general, see Forsdyke 2021, 65–8, with further bibliography). Clearly, Phrynichus aimed to proscribe this specific use of οἰκογενής in reference to the house-born slave, a usage that had only one antecedent in Classical Greek (i.e. in Plato, C.1, while in Ar. Pax 788, the term had a different meaning, cf. above) and was instead extremely common in the koine. It is likely that, owing to the presence of οἰκογενής in Aristophanes and Plato, a later compiler added the sentence μήποτε δὲ καὶ τῷ οἰκογενὴς ὡς δοκίμῳ χρηστέον (‘but perhaps one should also use οἰκογενής as acceptable’) at the end of Phrynichus’ entry. That this is a later addition and not part of the original entry is evident from its content, which is in clear opposition to Phrynichus’ proscription of οἰκογενής (see already de Pauw in Lobeck 1820, 201). However, though later than Phrynichus, this addition must certainly predate the q redaction of the lexicon, where the sentence is rephrased as ἴσως δὲ καὶ τὸ οἰκογενὴς οὐκ ἀδόκιμον (‘but perhaps οἰκογενής too is not inacceptable’; on the phrase οὐκ ἀδόκιμον, cf. Fischer’s proposal τὸ μὲν κοιτὼν <οὐκ> ἀδόκιμον in Phryn. Ecl. 222Phryn. Ecl. 222 and entry ἱστών and other place-nouns in -(ε)ών). The terminus ante quem for the q redaction of the Eclogue is in the early 1330s (i.e. the estimated period for the composition of Thomas Magister’s Eclogue, which depends on the q version of Phrynichus’ lexicon for no less than 69 glosses; see the entry Thomas Magister, ’Ονομάτων Ἀττικῶν ἐκλογή).

Pollux’s (B.1) treatment of οἰκογενής is notable in that the term does not occur in the main list of words relating to slaves, but rather appears in the following sentence in which he states that the female servant σηκὶς καλεῖται ἢ οἰκογενής (‘is called σηκίς or οἰκογενής’). The term σηκίςσηκίς is extremely rare and is confined to ancient comedy (Epicharmus 1x, Pherecrates 1x, Aristophanes 1x) and it is not clear whether ἢ οἰκογενής means that Pollux considered οἰκογενής to be an alternative to σηκίς for the female slave or if οἰκογενής was simply a gloss for this rare literary word. The fact that several lexica gloss σηκίς with οἰκογενής (cf. the list in the apparatus to B.1) makes this second interpretation more plausible, and this is an indirect confirmation – in addition to the fact that the term does not feature in the main list of slave-related words – that Pollux, like Phrynichus, did not recommend οἰκογενής (but was still willing to use it as a commonly understood word to explain a much rarer one like σηκίς).

The term that Phrynichus prescribes in place of οἰκογενής is another verb-final compound, i.e. oἰκότριψ ‘house-slave’ (from οἶκος and the root of the verb τρίβω, literally ‘to rub, to grind’ and, metaphorically, ‘to spend [time]’). οἰκότριψ belongs to the root compounds category, an ancient and rare compound type that was already scarcely used in Archaic Greek (for an overview with further bibliography, see Tribulato 2015, 86–8). The extant classical attestations for this term are limited to one case in Aristophanes (C.2), where it is used as a disparaging appellative for Euripides. A further attestation that Phrynichus may have considered is in the Demosthenic speech On Organisation, first identified as spurious by modern scholars (cf. Sealey 1993, 235–7; Trevett 1994) – [D.] 13.24[D.] 13.24: νῦν δ᾿, ὦ ἄνδρες Ἀθηναῖοι, φθόρους ἀνθρώπους οἰκοτρίβων οἰκότριβας, τιμὴν ὥσπερ ἄλλου του τῶν ὠνίων λαμβάνοντες, ποιεῖσθε πολίτας (‘But now, men of Athens, you make citizens of pestilential fellows, slaves born of slave parents, taking a price as you would for any other goods for sale’, transl. Trevett 2011, 235–6). In addition to these two cases, the adjective occurs twice in Menander (Asp. 176, Sic. 78; cf. Gomme, Sandbach 1973, 640) and is found several times in non-Atticising post-classical authors, including Philo (12x), Plutarch (6x), and Origenes (15x). Therefore, as seen for οἰκόσιτος, the acceptability of the term οἰκότριψ does not appear to rely on its literary attestations. Consequently, it is conceivable that Phrynichus regarded -τριψ compounds as a whole as typically Attic. However, this hypothesis does not appear to reflect the extant evidence since the -τριψ compounds in Attic authors amount only to θησειότριψ ‘one who is always in the Theseum’ i.e. ‘a runaway slave’ (Ar. fr. 475) and ἀστύτριψ ‘one who always lives in the city’ (Critias Diels–Kranz 88 B 72). Nevertheless, in Phryn. Ecl. 390Phryn. Ecl. 390, the lexicographer condemns πορνοκόπος (‘fornicator’), prescribing the synonym πορνότριψπορνότριψ, stating that it was used by ‘the ancient Athenians’ (cf. entry πορνοκόπος, πορνοκοπέω). An Attic attestation of πορνότριψ is not extant; however, the compound was probably used in Old Comedy (it is, in fact, edited as com. adesp. fr. 551; similarly, a term only attested in grammatical texts, i.e. χοιρότριψ ‘one who lives with pigs’ – although a sexual meaning cannot be ruled out given that χοῖρος may indicate the female sexual organ – is considered by modern scholars as com. adesp. fr. 935). Moreover, in Phryn. PS 17.15–20Phryn. PS 17.15–20, Phrynichus mentions the words ἄτριψ (‘not rubbed’) and, again, πορνότριψ, along with another root compound, i.e. βοῦκλεψ ‘thief of oxen’ (on this entry, see Monaco 2025, 267). To this should also be added Phryn. Ecl. 308Phryn. Ecl. 308 – where Phrynichus proscribes ἐπίτοκος (‘near childbirth’) and prescribes ἐπίτεξ – as well as Phryn. PS 18.11–12Phryn. PS 18.11–12 and Phryn. PS 100.3Phryn. PS 100.3, where he mentions ἄζυξ (‘unyoked’) and περίζυξ (‘yoked together’). In view of this, it may be reasonably assumed that Phrynichus’ prescription of οἰκότριψ is part of his more general preference for the root compound category, a compound type that he may have perceived as rare and thus refined (on compounds in general in Phrynichus’ Eclogue and Praeparatio sophistica, see Monaco 2025).

A possible further element in Phrynichus’ preference for οἰκότριψ may have been the ancient grammarians’ interest in the term. Indeed, there appears to have been a debate regarding the word’s etymology (cf. Slater 1986, 109), which some linked to the verb τρέφω (‘to nourish, to feed’; traces of this are found in Pseudo-Ammonius’ entry on οἰκότριψ and οἰκέτης, B.3), while others – most notably Aristophanes of Byzantium (fr. 328, quoted by Eustathius, B.11) – correctly traced it to τρίβω, understanding οἰκότριψ to be a metaphorMetaphors for a slave being ‘worn out’ in the house for a long time. An entry in the Synagoge (B.4) and another in Photius (B.9) consider οἰκότριψ, identifying the term as a synonym of οἰκογενής (and, in the case of B.4, also of οἰκέτης, unlike Β.3). B.4 and B.9 have been identified as coming from Aelius Dionysius (ο 9)Ael.Dion. ο 9 and Pausanias (o 8)Paus.Gr. ο 8, respectively; if these identifications are correct, Phrynichus may have also based his preference for οἰκότριψ on the presence of the term in these lexica.

The koine counterpart of οἰκότριψ presented in Moeris’ entry (A.2, followed by Thomas Magister, A.3) is not, as could be expected based on A.1, οἰκογενής, but οἰκοτραφής. This is also a verb-final root compound: namely, from οἶκος and the zero-grade verbal stem of τρέφω ‘to nourish, to feed’. The term is almost only found in erudite texts; however, an attestation in a papyrus document on a slave sale dating to the imperial period ‒ i.e. P.Oxy. 86.5567.9–10 (= TM 976578) [290 CE] ‒ confirms that οἰκοτραφής was in fact in use in Moeris’ time to denote a house-born slave. Actually, οἰκοτραφής follows οἰκότριψ in part of the manuscript tradition of Pollux as well (Poll. 3.76, B.1) but the word is not in the text by Bethe, seemingly because the term is proscribed by Moeris (cf. Bethe 1900, 178); yet, it is possible that Pollux – given his less strict Atticist tendency – allowed for οἰκοτραφής alongside οἰκότριψ regardless of Moeris’ stance. Moreover, the likelihood of οἰκοτραφής originally being in Pollux’s text is increased by the fact that the word is found in CHeid. Pal. gr. 375 (which is the oldest extant manuscript of the Onomasticon) as well as in BPar. gr. 2647 and EMatr. 4625 (the latter is a 14th-century manuscript that was largely neglected by Bethe 1900, XII, but may have relied on ancient sources; for a detailed analysis, see Cavarzeran 2025, 76–92).

E. Byzantine and Modern Greek commentary

Following its one-time use by Favorinus (C.4) and Phrynichus’ criticism (A.1), αὐτότροφος disappears completely, only to resurface a final time in Byzantine literature: namely, in a heading in Euthymius Zigabenus’ commentary on Paul’s Second Epistle to the Corinthians (11.1). οἰκόσιτος, in contrast, is primarily found in scholia, lexica, and grammatical texts; however, a few literary attestations are also extant, where οἰκόσιτος describes a soldier who brings his own provisions (John Zonaras 1x, Theodorus Metochites 1x, and Georgius Gemistus 2x).

οἰκότριψ, while not particularly frequent, remains in steady use throughout late antiquity and the Byzantine period, possibly owing to the fortune of the expression οἰκοτρίβων οἰκότριψ ‘house-slave born of house-slaves’ (originally in the spurious Demosthenic speech [D.] 13.24, cf. above), which is reused e.g. by Basilius of Caesarea 2x, Joannes Damascenus 2x, and, as late as the 19th century, Neophytus Ducas 2x. Alongside οἰκότριψ, the forms οἰκοτριβής ‘house-servant’, οἰκοτριβέω ‘to stay in a house, to stay at home’, and οἰκοτριβία ‘household’ are also attested (cf. LBG s.vv.).

οἰκογενής is clearly the most widespread term among the four in Byzantine authors, and is mainly used to mean ‘house-born slave’ (see e.g. Nicetas Stethatus 2x, Georgius Cedrenus 4x, Johannes Scylitzes 3x, Theodorus Metochites 2x, and Nicephorus Gregoras 2x; cf. also the noun οἰκογέννητον). The late medieval substantive οικογένεια ‘household’, derived from οἰκογενής, is attested in literary texts from the 16th century onwards (e.g. Damascenus Studites 1x) and became the most commonly used term for ‘family’ in Modern Greek (cf. LKN s.v.).

F. Commentary on individual texts and occurrences

N/A