στρωματόδεσμον
(Antiatt. σ 7, Phryn. Ecl. 380, Poll. 7.79, Moer. σ 44, Thom.Mag. 332.9–11)

D. General commentary

The Antiatticist entry (A.1) discusses the gender of the word στρωματόδεσμον/-ος, a rare and technical termTechnical language for ‘bed-sack’ or, more precisely, ‘a leathern or linen sack in which slaves had to tie up the bedclothes’ (LSJ s.v.; see Whitehorne 2001, 1289 for a list of occurrences in Greek literature). In this entry, both the masculine and the neuter are deemed correct, with reference to the comic poet Ameipsias (C.3). στρωματόδεσμον/-ος is also discussed in other Atticist lexica, which contrast its use with that of στρωματεύς, attested in Classical Greek and typically translated as ‘coverlet’: its use in the sense of a ‘bed-sack’ is lamented as more recent (A.3, B.2 – in the latter only if we read, with Erbse, στρωματεῖς instead of στρωματόδεσμα: see apparatus and F.2) and explicitly proscribed (A.2, A.4, A.5). It is noteworthy that sources mentioning both masculine and neuter variants of στρωματόδεσμον/-ος label both as Attic: as already observed by Matthaios (2013, 122 n. 247; see also Fiori 2022, 29 n. 35), this agreement among individual lexicographers suggests that the idea that Phrynichus’ approach differed starkly from that of other lexicographers should not be overemphasised.

Only the neuter form is well attested in classical Attic, where it is found in Old Comedy (C.1, C.2) but also in prose (C.4, C.5, C.6, C.7). Among the prose writers of the imperial age, Plutarch unambiguously uses masculine forms (C.8: on the precise meaning of στρωματόδεσμος in this passage, see Whitehorne 2001), as do Artemidorus (1.74.31–2, 4.1.98) and Hippolytus (Haer. 6.27.2.2, 6.27.3.2–3), although the neuter form also persists in use (see, e.g., Lib. Ep. 234.1.5, Stob. 1.49.59). Based on this distribution, it is likely that Ameipsias is cited by the Antiatticist as having used the rarer masculine form (as cautiously suggested by Orth 2013, 338) or, at most, because he used both forms. A similar situation is encountered in the Antiatticist's entry devoted to the noun σκότος (see entry σκότος), which states that the word can be both masculine and neuter, with a similarly elliptical reference to Ameipsias.

In contrast to the case of ὁ/τὸ σκότος, for which the change in gender also implied a metaplasmDeclension metaplasm from the second to the third declension, στρωματόδεσμον/-ος remained a thematic noun throughout. Possibly, στρωματόδεσμον underwent the influence of the simplex δεσμός ‘bond’, which itself oscillates in the plural between masculine οἱ δεσμοί and neuter τὰ δεσμά (for attestations in Attic inscriptions, see Threatte 1996, 274–5): the notion that speakers created a plural στρωματόδεσμοι in addition to στρωματόδεσμα on the analogy of δεσμοί in addition to δεσμά and that the singular στρωματόδεσμος was back-formed therefrom (see further F.1) is an attractive one. As an additional reason for the gender shift, it should also be noted that most compoundsCompounds with a second member -δεσμο- are masculine (37), while only a marked minority is neuter (7). Finally, in light of its meaning, the noun στρωματόδεσμον/-ος, for pragmatic reasons, might be expected to occur most frequently in oblique cases (in which the masculine and neuter are indistinguishable) or in the accusative (in which, in the singular, masculine and neuter would be distinguishable only in the presence of the article) rather than in the nominative.

Similarly to their treatments of σκότος, it appears that lexicographers were not driven to distinguish between the use of στρωματόδεσμον/-ος in the neuter, which should have been labelled as the only authentically ‘Attic’ form, and in the masculine, which should have been rejected by stricter Atticists, such as Phrynichus (A.2) and Moeris (A.4), at least; meanwhile, it is unsurprising that the Antiatticist (A.1) admits both forms. This distinction, even if it was recorded, was likely perceived as less important than that between στρωματόδεσμος/-ον and στρωματεύς. The latter form (on which see Perpillou 1973, 293–4; Vitale 1980) is first attested at the turn of the 4th century BCE, in Middle and New Comedy (Antiph. fr. 40, Alex. fr. 120.3, Apollod.Car. fr. 2, Apollod.Gel. fr. 5) and in Theophrastus (HP 4.2.7). Although its proper meaning was ‘coverlet, blanket’, it apparently encroached on the meaning of στρωματόδεσμος/-ον: strict Atticists condemned this semantic extension as not proper to classical usage (see Arnott 1996, 335).

E. Byzantine and Modern Greek commentary

As a term for ‘bed-sack’, στρωματόδεσμος/-ον disappears almost entirely from literary texts after late antiquity (for which, see e.g. Lib. Ep. 234.1). The rare exceptions in the Byzantine period include John Tzetzes (H. 4.594) and Theodorus Metochites Σημειώσεις γνωμικαί 63.5.5 Wahlgren (14th c. CE): the former uses the word as a neuter, while the gender cannot be ascertained in the latter. In the Byzantine era, this noun produced its only attested derivative: the denominal verb στρωματοδεσμέω ‘to make the bed’, attested only in Michael Psellus (Or. 8.182, Ep. 77.41–2 = 130.41). These should all be regarded as learned revivals: in common language, it was necessary to gloss στρωματόδεσμος/-ον with the later στρωματεύς but possibly also with terms borrowed from Medieval Latin (see F.2).

F. Commentary on individual texts and occurrences

(1)    Hsch. δ 705 (B.1)

The gender oscillation in the plural of the simplex δεσμός ‘bond’ was discussed by lexicographical sources who unanimously agreed that δεσμά should be labelled as Attic and was thus preferable to δεσμοί (see Ael.Dion. δ 8Ael.Dion. δ 8, Moer. δ 28Moer. δ 28); Thomas Magister (79.15–8)Thom.Mag. 79.15–8 explicitly warns that this distinction is valid only for the plural form, given that, in the singular, only masculine forms are used (ἐπὶ δὲ τῶν ἑνικῶν τὰ τοῦ ἀρσενικοῦ μόνα ἐν χρήσει). However, the oscillation in the plural may have resulted in the creation of a neuter singular τὸ δεσμόν: this entry in Hesychius, taken at face value, indeed appears to imply that the singular δεσμόν is neuter in Attic, and it is interpreted in this way by LSJ s.v. δεσμός. Nonetheless, the entry may have been subject to abridgement and may have originally stated that while the singular is masculine, the accusative plural is neuter in Attic. One may compare the somewhat confusing formulation in Philemo (Laur.) 358 Philemo (Laur.) 358, δεσμά, μὴ δεσμόν ἐρεῖς· οὐ γὰρ δεῖ ἀρρενικὸν ποιεῖν (‘You shall say δεσμά, not δεσμόν, for one must not make [it] masculine’), in which the nom./acc. neut. pl. δεσμά is contrasted not with the nom. masc. pl. δεσμοί or the acc. masc. pl. δεσμούς – as would reasonably be expected based on the comment that immediately follows – but with the acc. masc. sg. δεσμόν. At any rate, τὸ δεσμόν is attested at least once, in a Delphic inscription (CID 2.59.15 [336/335 BCE]): if it is not merely a simple error for τὸ<ν> δεσμόν, it may attest to the same kind of analogical mechanism that, mutatis mutandis, produced στρωματόδεσμος in addition to στρωματόδεσμον.

(2)    Schol.Aeschin. 2.99.217 (~ Ael.Dion. σ 39) (B.2)

The text of this scholion on Aeschines’ On the False Embassy (C.3) is transmitted in a corrupted form. Dilts put the words πόφελλα καὶ ἀβελτάρια between cruces, but while the first is indeed a vox nihili, the latter may be, as suggested by DGE s.v., a derivative in -άριον of ἀβέλτης, attested as a variant of ἀβέρτα ~ ἀβερτή, a loan from Latin averta ‘portmanteau, saddle-bag’, in turn derived from the Greek ἀορτή ‘knapsack’ (see LBG s.v. ἀβερτή). This lemma is among the scholia on Aeschines that were included as marginal glosses in cod. Par. Coisl. 345Par. Coisl. 345 (10th century CE) – cod. B of the Synagoge tradition, and thus printed by Bachmann (1828, 372.14) – but that do not belong to the Synagoge (Cunningham 2003, 66–9). Bachmann’s text has τὰ φόλλα καὶ ἀμεντάρια for the final, problematic words. The first form has been explained as a borrowing from the Latin follis ‘bag, sack’ (see LBG s.v. φόλλον, based on this single attestation). This scholion was identified by Erbse as a fragment of Aelius Dionysius based on the discussion in Eustathius (B.3) that quotes Aelius Dionysius on the gender variation in several Attic nouns, including στρωματόδεσμον. Erbse also supplied the prescription οὐδετέρως Ἀττικοί, for which, however, as discussed above (D.), no parallel is known in extant Atticist treatments of στρωματόδεσμον. Comparing the discussion of στρωματόδεσμον in Pollux (A.3), Erbse also corrected the perplexing second mention of στρωματόδεσμα (maintained by Bachmann, albeit with the note ‘sic’) to στρωματεῖς. However, he omitted the final part of the gloss.