ANTIATTICIST η 6 VALENTE (ἡσύχιος): A PROPOSAL FOR A DIFFERENT TEXTUAL ARRANGEMENT

The Antiatticist lexicon has a short lemma on the adjective ἡσύχιος ‘silent, quiet’ which has suffered heavy shortening and possibly interpolation. The present article argues that the text of this lemma should be edited as it stands in cod. Coisl. 345 and should not be changed. The entry and the mutual relationship between the words it includes should be assessed in the light of the Antiatticist's approach to classical Greek, and of the Byzantine reception of the lexicon and its contents.

ἡσύχιος⋅ ἡσυχώτερον duas gl. contractas esse vid. Bekker (vd. ad Antiatt. η 7), sed le. ἡσύχιος corrupt. pro ἡσυχαίτερον (cf. Thom. Mag.) vel ἡσυχιώτερον (cf. Marc. Aur. 4.3) vel ἡσυχιώτατον (ap. Plat. Charm. 160a9 -ος) esse mihi videtur. 2 In what follows I shall review these hypothetical reconstructions and argue that they do not improve our understanding of the lemma. This entry and the mutual relationship between the words it includes should rather be assessed against their textual arrangement in the manuscript and in the light of the Antiatticist's approach to Classical Greek.

Should ἡσύχιος be corrected to ἡσυχαίτερον?
In proposing that ἡσυχαίτερον may lie behind the corrupt ἡσύχιος, Valente makes reference to a lemma in Thomas Magister's lexicon, which identifies the Attic comparative ἡσυχαίτερον as the correct formation: ἡσυχαίτερον, οὐχ ἡσυχώτερον. Θουκυδίδης κτλ. ('[use] ἡσυχαίτερον, not ἡσυχώτερον, as Thucydides, etc. ',p. 173 Ritschl). This hypothesis implies that the Antiatticist lemma had a prescriptive focus, recommendingjust like Thomas Magisterthe use of the Attic form in place of the morphologically regular but apparently much rarer ἡσυχώτερος. 3 It should be noted in passing that our perception of the higher frequency of the Attic formation may often result from the vagaries of textual transmission, during which ἡσυχώτερος may have been replaced by ἡσυχαίτερος, considered to be more prestigious. 4 However, the presence of a prescriptive lemma advising readers to use ἡσυχαίτερος instead of ἡσυχώτερος seems unusual to me in a lexicon such as the Antiatticist, which tends to oppose the rigidity of Atticist normativism and to promote linguistic variety. 5 The study of the Antiatticist's approach to the category of comparatives and superlatives also suggests that the correction of the transmitted lemma to ἡσυχαίτερον should be approached with caution. Starting from Philoxenus' Περὶ συγκριτικῶν, the morphological variations of comparatives and superlatives attracted the attention of ancient and 2 Valente (2015) 177.
3 On the basis of modern editions, ἡσυχώτερος seems to have featured only in Hippocrates' Epidemics (4.1.30, 6.4.3), in a spurious homily attributed to Clemens Romanus (4.9.1) and finally in commentaries on the Epidemics (Gal. In Hippocr. Epid. 17b.124, Palladius Comment. in Hippocr. 2.106). Neither ἡσυχώτερος nor ἡσυχαίτερος is attested in documentary papyri. 4 See, for instance, the feminine comparative at Soph. Ant. 1089. All manuscriptsincluding cod. Laur. 32.9 (L), the oldest Antigone manuscript (mid-ninth century)transmit ἡσυχωτέραν, which is the form printed in the old editions by Dindorf and Jebb. The variant ἡσυχαιτέραν crops up in cod. Laur. 32.2 (Zg, fourteenth century), as well as in its apograph cod. Laur. 31.1 (Zs, fifteenth century). These manuscripts contain scholia attributed to Thomas Magister and it is uncertain whether ἡσυχαιτέραν represents his correction or perhaps a rare ancient variant. The more recent OCT editions of Sophocles by Lloyd-Jones and Wilson choose it against ἡσυχωτέραν, probably in consideration of the importance of the codici recentiores for Sophocles' constitutio textus, as established in Turyn (1952) and Dawe (1973). It might be that ἡσυχαιτέραν is Thomas' correction, because his Eclogue (p. 173 Ritschl) recommends ἡσυχαίτερος as the correct Attic form, albeit with reference to Thucydides (this is coherent with Thomas' practice of privileging the model of prose authors: cf. the breakdown in Gaul (2011) 144). However, the question cannot be settled with certainty.
Byzantine grammarians alike. 6 With no less than thirty entries on comparatives and superlatives, the Antiatticist is no exception. Consistently with its more open take on the classical canon, 7 it shows a preference for formations that go against Attic usage: analogical comparatives in -εσ-such as ἀwθονέστερον (α 74 Valente) and ἀρχαιέστερον (α 75), where the suffix is imported from sigmatic stems; 8 neuter comparative adverbs in -ως such as ἀληθεστέρως (α 39), ἐχθροτέρως (ε 21), καταδεεστέρως (κ 48) and μειζόνως (μ 13); reduplicated superlatives and comparatives such as μάλιστα ὁμοιότατος (μ 29), μᾶλλον μᾶλλον (μ 21) and μεῖζον μεῖζον (μ 22). Despite this wealth of information, however, the Antiatticist never comments on Attic comparatives and superlatives in -αι-, the category to which the form suggested by Valente, ἡσυχαίτερον, belongs. The only typically Attic superlative in the lexicon is βράχιστον, which is glossed with βραχύτατον (β 27 Valente). In consideration of all this, it seems unlikely to me that the original lemma of the Antiatticist was ἡσυχαίτερον.

Should ἡσύχιος be corrected to ἡσυχιωτερον?
In his apparatus Valente also considers a second hypothesis: that the original lemma was ἡσυχιώτερον. In this second scenario, the Antiatticist would be focusing on two regular comparatives: the first from ἡσύχιος, the second from ἥσυχος. As noted by Valente, ἡσυχιώτερος is currently attested only in Marcus Aurelius' Περὶ ἑαυτου̃(4.3.1.4), in a passage where, together with ἀπραγμονέστερος, the adjective identifies the soul as the quietest place in which man can take refuge (οὐδαμου̃γὰρ οὔτε ἡσυχιώτερον οὔτε ἀπραγμονέστερον ἄνθρωπος ἀναχωρεῖ ἢ εἰς τὴν ἑαυτου̃ψυχήν, κτλ. 'in no quieter and more peaceful place can a man retire than in one's own soul') The reading ἡσυχιώτερον in the Περὶ ἑαυτου̃is certain and confirmed by the best manuscripts. 9 There are no other certain attestations of the comparative and superlative of ἡσύχιος. The superlative ἡσυχιώτατος, which Valente adds as a third possibility for the corrupt lemma with a reference to Plato's Charmides (160a.9), is itself a modern correction for the 6 Cf. Philox. fr. 337 Theodoridis (ex Et. Mag. s.v. αἰδοιέστατος;cf. Eust. in Od. 1,p. 92.20), which defines analogical comparatives and superlatives in -εσ-as 'Ionic' and those in -ισ-and -αι-as 'Attic'. The forms ἡσυχαίτερον and ἡσυχαίτατα are cited in the last category, without direct reference to any author; however, ἡσυχαίτατα occurs after other forms attributed to Plato and this leads Theodoridis to identify its locus classicus in Pl. Chrm. 160a. On this Platonic passage see also below, n. 10. 7 The Antiatticist seems to have based its defence of non-Attic usages on Alexandrian sources, especially Aristophanes of Byzantium: see Slater (1976)  transmitted reading, ἡσυχώτατος. 10 In any case, out of Valente's three proposals for emendation ἡσυχιώτατος is the least likely: the superlative would be inconsistent with the interpretamentum ἡσυχωτερος, which Valente correctly considers to be sound. With regard to the comparative ἡσυχιωτερος, we are unable to tell whether this form was used in a lost classical text which served as a model for the Antiatticist. It is improbable that the author of the lexicon had Marcus Aurelius as a reference point, since the Περὶ ἑαυτου̃is an almost contemporary work and, moreover, its circulation before the Byzantine age seems to have been limited. 11 Therefore, if Marcus Aurelius is behind this lemma, the most likely conclusion would be that the entry does not belong to the original version of the Antiatticist, but represents a later addition instead.
This scenario leads us into the Byzantine age, and more precisely to the lifetime of Arethas of Caesarea (ca 850-935). The Byzantine scholar, who owned the oldest known manuscript of the Περὶ ἑαυτου, was familiar with this work and quoted it several times in his scholia to Lucian and Dio of Prusa. 12 There is no trace of the Περὶ ἑαυτουb efore Arethas' lifetime: it is never quoted in the Synagoge or in Photius and even the references in the Suda seem to be based on a collection of excerpta. 13 To sum up, correcting ἡσύχιος to ἡσυχιωτερον is undesirable: there is no known locus classicus which could have served as a model for this lemma, nor would its aim in the context of the Antiatticist be clear.

Could ἡσυχώτερον be a later addition?
If ἡσύχιος is sound, what are we to do with its incoherent gloss ἡσυχώτερον? Hypothetically, this form is a good candidate for a later, specifically Byzantine, accretion. Arethas uses this form of the comparative (not ἡσυχιώτερον) when quoting the same passage of the Περὶ ἑαυτου̃in one of his scholia to Dio of Prusa. In commenting on Dio's statement that 'there is no better and more profitable retirement than retirement into oneself and 10 ἡσυχώτατος is already attested in cod. Bod. Clarke 39 (dated to 895 AD), fol. 286 r . This reading is accepted in Bekker (1816) 316. The correction into ἡσυχιώτατος was first proposed by Cobet (1878) 40, on two grounds: (1) the whole Platonic passage uses ἡσύχιος and ἡσυχιότης, but not ἥσυχος, which therefore is out of place; (2) ἥσυχος is an older variant, typical of poetry, but not of prose: the prose attestations of ἥσυχος are corruptions of ἡσύχιος. Cobet's first argument is correct, but one may advance two objections: that only a little earlier in the same chapter (160a.5) Plato uses the superlative ἡσυχαίτατα, which derives from ἥσυχος, not ἡσύχιος; and, more generally, that Plato makes ample use of both ἥσυχος and ἡσύχιος (cf. Ast (1836) 39-40). Cobet's second argument is faulty: ἥσυχος is not more ancient than ἡσύχιος and its post-classical attestations (which cannot all be dismissed as corruptions) show that it was a common alternative. Specifically on this last point see below, section 4. 11 Brunt (1974) 1 and Ceporina (2012b) 47, with references. For the dating of the Antiatticist to the second century AD see Latte (1915) and Valente (2015) 5. 12 For the manuscript, which is the archetype for the medieval transmission of Marcus Aurelius, see Areth. Epist. 44, p. 305 Westerink;Dalfen (1987) v-vi;Cortassa (1997); and Ceporina (2011). The presence of references to Marcus Aurelius is often used as a criterion for attributing to Arethas some exegetical material transmitted in the margins of manuscripts: cf. Wilson (1996) 127.
attending to one's own concerns' (μὴ οὖν βελτίστη <ᾖ> καὶ λυσιτελεστάτη πασῶν ἡ εἰς αὑτὸν ἀναχώρησις καὶ τὸ προσέχειν τοῖς αὑτου̃πράγμασιν, Dio 20.8), Arethas annotates οὐδαμου̃γὰρ οὔτε ἡσυχώτερον κατὰ τὸν αὐτοκράτoρα Μάρκον οὔτε ἀπραγμονέστερόν τις ἀναχωρεῖ ἢ εἰς τὴν ἑαυτου̃ψυχήν ('according to emperor Marcus there is no quieter and more peaceful place into which one can retire that one's own soul'). 14 The scholium can be read in cod. Vat. Urb. Gr. 124 (tenth century), one of the oldest testimonies of Dio's speeches, copied from a codex owned by Arethas, and the archetype of Dio's medieval tradition. 15 The reading ἡσυχώτερον is certain. 16 The replacement of ἡσυχιωτερον with its synonym ἡσυχώτερον is clearly the result of Arethas' quoting Marcus Aurelius by heart, as shown also by the paraphrase of Marcus' ἄνθρωπος ἀναχωρεῖ with τις ἀναχωρεῖ. Since Dio of Prusa was a very popular author at Byzantium and the object of renewed exegetical interest at the time of the so-called 'Renaissance' of the ninth century, 17 it is not improbable that Arethas' scholia on Dio circulated in the erudite circles of this period. 18 We know for sure that some of Arethas' exegetical material ended up in the slightly later (mid-tenth century) cod. Coisl. 345, as shown by the selection of Lucian's λέξεις transmitted in its fols. 178 v -186 r . 19 Arethas' direct involvement in the production of this manuscript, posited by Kougeas (1913), is now discarded, but his influence has been noted by several scholars. 20 In conclusion, the presence of ἡσυχώτερον in the medieval copy of the Antiatticist may perhaps go back to Arethas' scholium: a scholium, that is, in which ἡσυχώτερον is used in place of ἡσυχιώτερον, the comparative of the original lemma of the Antiatticist, ἡσύχιος.
In the following section I shall defend the authenticity of ἡσύχιος by considering the textual organisation of the lemma in cod. Coisl. 345 and the scribe's graphic habit. I shall then come back to the linguistic interpretation of the sequence ἡσύχιοςἡσυχώτερον in order to evaluate its meaning in the light of the history and usage of these adjectives in both classical and post-classical Greek. The consideration of the Byzantine stage is crucial for the understanding of what cod. Coisl. 345 has transmitted and of why it may not be as corrupt as it appears.

The arrangement of the lemma in the manuscript
In cod. Coisl. 345 ἡσύχιος and ἡσυχώτερον (erroneously written ἡσυχότερον) occur in the same line of text (Fig. 1), separated by two vertical points (dicolon). 21 The manuscript was entirely copied by the same scribe, who usually employs the dicolon to separate a lemma from its interpretamentum, to introduce quotations or reference to ancient authors, and finally to mark the end of a lemma and separate it from the following one. 22 After ἡσυχότερον (sic) there is another dicolon, followedon the next line of textby another word, ἡσύχιμον, which is provided with a locus classicus (Πίνδαρος Ὀλυμπιονίκαις) introduced by another dicolon. In his editio princeps Immanuel Bekker (1814) kept the manuscript's punctuation, treating the three words as three different lemmata (and thus implicitly assuming that the first two have lost their locus classicus): Antiatt. p. 98.18-20 Bekker ἡσύχιος : ἡσυχώτερον : ἡσύχιμον. Πίνδαρος Ὀλυμπιονίκαις.
By contrast, as we saw above, Valente thinks that the first two words are part of the same lemma, in which however ἡσύχιος would be corrupt. 23 Like Bekker, he then edits ἡσύχιμον as a separate lemma, in which the interpretamentum has been dropped (not an infrequent case in the Antiatticist): Antiatt. η 7 Valente ἡσύχιμον⋅ Πίνδαρος Ὀλυμπιονίκαις (2.32). ἡσύχιμον: Pindar in the Olympian Odes. There might be a different way to explain the distribution of these three words in Coisl. 345. The original author of the lexicon may have produced a synonymic-onomastic lemma devoted to various forms in ἡσυχ-, of which only the last locus classicus (Πίνδαρος Ὀλυμπιονίκαις), specific to ἡσύχιμον, survives. 24 Other lemmata of the Antiatticist display the same kind of structure. An example is β 4, which Valente himself quotes to illustrate this type of entry: 25 Antiatt. β 4 Valente βλάξ, βλακεύειν, βλακεύεσθαι καὶ βλάκες καὶ βλακικῶς⋅ Πλάτων Γοργίᾳ (488a8), ὁ αὐτὸς Εὐθυδήμῳ (287e2), Ἀριστοwάνης Πλούτῳ (325). βλάξ, βλακεύειν, βλακεύεσθαι and βλάκες and βλακικῶς: Plato in the Gorgias, Plato again in the Euthydemus, Aristophanes in the Plutus.
The loci classici quoted in Antiatt. β 4 provide exact parallels only for βλάξ (Pl. Grg. 488a.8: βλᾶκα) and βλακεύεσθαι (Ar. Plut. 325, though the form used by Aristophanes is actually the adverbial compounded participle κατεβλακευμένως). 26 By contrast, Plato's Euthydemus is not a locus classicus for any of the words directly included in the lemma: the dialogue uses the feminine form βλακεία, which is not included in the Antiatticist's selection. 27 All of this shows that the Antiatticist may contain lemmata in which the locus classicus does not necessarily refer to all of the forms quoted in the main entry.
In the entry devoted to βλάξ the different forms of the word are separated by simple stops, whereas the dicolon is used to introduce the loci classici (Fig. 2). 28 By contrast, the three adjectives ἡσύχιος, ἡσυχώτερον and ἡσύχιμον are separated by the dicolon. Although less frequent, the use of the dicolon in this function (where one would instead expect a simple dot or comma) is not unparalleled in Coisl. 345. An example is provided by the lemma Antiatt. α 25 Valente: ἀστοργία, wιλοστοργία, στοργή⋅ Ἀντιwῶν ἐν β ′ Περὶ της ῥητορικης τέχνης (fr. 73). In the manuscript (Fig. 3) a dicolon occurs after the first word, as if only this word were the main lemma. A simple dot sets wιλοστοργία apart from στοργή and no sign is employed to introduce the locus classicus; finally, a dicolon marks the end of the line. Another telling example is Antiatt. ε 47 Valente: ἐπ' ἐκεῖνα, ἐπὶ τάδε⋅ Πλάτων Περὶ ψυχης. Here the scribe seems to have wrongly interpreted the lemma and has used the dicolon to separate its two parts, as if the entire sequence ἐπὶ τάδε Πλάτων Περὶ ψυχης were the interpretamentum (Fig. 4).
A similar use of the dicolon features in some of the other lexica in Coisl. 345, particularly those which, like the Antiatticist, present the lexical material in a succinct way, with a short interpretamentum immediately following the lemma. Let us consider for instance the entry on πηκτίς in the Ἡροδότου Λέξεις (version A), collected in fols. 165 v -167 v of the manuscript (Fig. 5). 29 If one were to judge this sequence of words superficially, it would be easy to think that πηκτίς is followed by its interpretamentum (ὄργανον ψαλτήριον 'stringed instrument') and that the following word, αὐλός, is a new lemma. However, this is obviously not the correct interpretation, as is shown by the version of this lemma in Stein's edition: Glossae in Herodotum 1.3 Stein πηκτίς. ὄργανον ψαλτήριον, αὐλός, εἶδος ἀκολάστου σχήματος. πηκτίς: stringed instrument, flute, a kind of licentious figure of speech.
In the lemma on πηκτίς, therefore, the dicolon is used to separate different parts of the interpretamentum (and is replaced by commas in Stein's edition). A similar arrangement is found in the entry ἑδώλοισιν, where again the dicolon separates its various intepretamenta (Fig. 6).  Glossae in Herodotum 1.7 Stein ἑδώλοισιν. ὑποστρώμασι νηός, ζυγαῖς, καθέδραις. ἑδώλοισιν: the lower parts of a ship (rower's benches?), pairs, seats.
In Moeris' lexicon, which occurs immediately after the Ἡροδότου Λέξεις in the manuscript (fols. 167 v -175 v ), the scribe uses the dicolon to separate the two main parts of each lemma, which in the typical style of this lexicon consist of the usage of the Ἀττικοί and that of the Ἕλληνες. Out of many possible examples, let us consider Moer. α 4 Hansen, contrasting the dual and plural forms of the expression 'ageless immortals': ἀθανάτω ἀγήρω Ἀττικοί⋅ ἀθάνατοι ἀγήρατοι Ἕλληνες. In the manuscript (Fig. 7), the dicolon occurs before and after Ἀττικοί, and then again after Ἕλληνες.
To conclude, the above examples show that the scribe of Coisl. 345 sometimes resorts to the dicolon to separate different elements within the same lemma, much as he might do with a simple dot. 30 Such usage is more frequent in those lexica -Antiatticist, Ἡροδότου Λέξεις, Moeriswhich are part of the 'secondary miscellany' contained in fols. 150 r -186 r of the codex. 31 These parallels allow us to confirm that the sequence ἡσύχιος : ἡσυχώτερον : ἡσύχιμον. Πίνδαρος Ὀλυμπιονίκαις, as presented in fol. 161 r of the manuscript, may have constituted one lemma of the onomastic type. In order to pinpoint its function, I shall now turn both to the classical   attestations of these words and to the Byzantine context of their reception, which is central to the understanding of the lexical selection copied out in Coisl. 345.

In defence of the interpretatio Pindarica
The first question to consider is the provenance of the words contained in the lemma. Since ἡσύχιμος is a hapax, one would expect a different organisation of the entry in the Antiatticist, with ἡσύχιμος as the lemma and ἡσύχιος as its gloss. Indeed, this is the strategy adopted by one of the scholia uetera on this line (schol. uet. Pi. Ol. 2.58f), which explains ἡσύχιμον ἁμέραν with ἡμέρα δὲ ἡσύχιος, ἡ του̃θανάτου⋅ ἐπεὶ ἐν αὐτῇ θανόντες ἡσυχάζομεν 'the peaceful day, the day of (one's) death; because in this day, by dying, we find rest' (ἡσύχιος is also used in schol. 58f, while schol. 58c glosses ἡσύχιμος with ἥσυχος, ἀπράγμων, εὐτυχής). 32 The Pindaric scholia provide a glimpse of an exegetical context in which ἡσύχιμος and ἡσύχιος were discussed together and which may have inspired the author of the Antiatticist in his creation of the hypothetical lemma ἡσύχιμον⋅ ἡσύχιον: the original order would have become corrupt during transmission. 33 However, the authenticity of the preserved order (with ἡσύχιος as the main lemma) can be defended with two further pieces of evidence. First, the same scholia on Ol. 2.32 are behind an entry in the Rhetorikai lexeis (121 Naoumides) in which the lemma is not 32 Schol. 58c and 58g have the Doric variant ἁσύχιμος which finds no parallel in the textual tradition of the Epinicians.
It has usually been thought to be a pseudo-Doric form (cf. e.g. LSJ s.v.), but more recently Forssmann (1966) 48-54 and Beekes s.v. have considered it to be authentic. 33 The cross-fertilisation of hypomnemata and lexica is examined by Ucciardello (2006) 54-6 and is often behind the frequent agreement between scholiastic corpora and lexica. It is likely that the author of the Antiatticist derived most of the Pindaric lemmata (see note below) from hypomnemata or other forms of Pindaric exegesis: see also Ucciardello (2006) 62 n. 117.
ἡσύχιμον as expected, but ἡσύχιον. This shows that the interpretamenta of the scholia (or their exegetical ancestors) could be extrapolated to constitute the main entry of lexicographical lists, as a result of simplification. 34 Secondly, Valente (2015) seemingly overlooks the fact that ἡσύχιος itself is a Pindaric word, used in Pythian 9: . . . ἦ πολλάν τε καὶ ἡσύχιον βουσὶν εἰρήναν παρέχοισα πατρῴαις κτλ.
(Pind. Pyth. 9.22-3) [Cyrene, the nymph] providing much peaceful security for her father's cattle . . . It is therefore highly probable that this synonymic-onomastic lemma of the Antiatticist was specifically devoted to Pindaric words. This would not be surprising, because Pindar is the most frequently quoted lyric poet in the Antiatticist, with seven entries, all provided with direct references to Pindaric loci. 35 It is also noteworthy that in two of these seven lemmata the reference is not to the Epinicians, but to Threnoi and Hymns, which confirms the pre-Byzantine provenance of the information provided in these entries. In conclusion, it cannot be ruled out that the original lemma was, for example, ἡσύχιος (or ἡσύχιον, with lemmatisation in the accusative as in the Pindaric model) καὶ ἡσύχιμον⋅ Πίνδαρος Ὀλυμπιονίκαις, perhaps with the sole erroneous attribution of both words to the Olympians (which may either be original or the result of the loss of the first locus classicus).
As for ἡσυχώτερον, the matter is more complicated. In principle, it cannot be ruled out that it may have featured in a lost Pindaric work and hence that the whole lemma was concerned with Pindar's language. Alternatively, as suggested above, ἡσυχώτερον could have been interpolated when the copy of the Antiatticist in Coisl. 345 was assembled. In this case, it may be that the antigraph displayed a columnar alphabetical layout which aided the addition of new material. 36 This hypothetical scenario is suggested not only by the occurrence of the comparative in the Arethas scholium considered above, but also by the development of ἡσύχιος and ἥσυχος in post-classical Greek, which may have influenced the reception of the original Antiatticist lemma in the Byzantine linguistic context. 34 Naoumides (1975)  Valente (2015) 59), and hence must also have been characteristic of its later copies. For alphabetisation in the Greek world see the classic Daly (1967).

The role of semantic development in the reception and perpetuation of the lemma
In the preceding section I mentioned the hypothesis that the entry on ἡσύχιος in the Antiatticist might have originated in Pindaric exegetical material reused for the kind of descriptive and onomastic purposes typical of lexicography. In contrast to other lemmata of the Antiatticist, that on ἡσύχιος eludes a closer definition of its aims (be they of a polemical, purist or prescriptive nature). In this section I shall briefly look into the use of ἡσύχιος and its synonym ἥσυχος throughout the history of Greek in order to better define how the original Antiatticist lemma may have been received in the Byzantine linguistic and cultural milieu behind the creation of its codex unicus, Coisl. 345. ἡσύχιος is attested from Homer onwards and is also frequently used in Attic texts. ἥσυχος is similarly ancient (its first attestation being Hes. Th. 763) and enjoys a considerable popularity at all chronological stages of the language. The two adjectives are synonyms and it is hard pinpoint their semantic differentiation in different contexts, not least becauseas mentioned above apropos of their comparativesthe manuscript tradition shows that they were often confused and exchanged. An important stage, however, is represented by biblical and New Testament Greek. Here ἡσύχιος, ἥσυχος, ἡσυχία and ἡσυχάζω are employed with a vast semantic extension, whereby the usage of these terms is initially associated with the semantic field of rest and peacefulness and then develops into the expression of religious virtue. 37 Consider also the popularity of the personal name Ἡσύχιος, which begins its life as a Judaic name (corresponding to Hebrew Noah) and soon becomes popular in Christian contexts as well.
The biblical use of ἡσύχιος became a model for Christian Greek vocabulary down to the late Byzantine age, often under the influence of biblical passages that attained the status of Christian loci classici. 38 Among these is Is. 66.2, a passage very frequently quoted in Christian exegesis, and one in which ἡσύχιος and ταπεινός feature as the attributes of the quiet and humble man who fears God's word (καὶ ἐπὶ τίνα ἐπιβλέψω ἀλλ' ἢ ἐπὶ τὸν ταπεινὸν καὶ ἡσύχιον καὶ τρέμοντα τοὺς λόγους μου; 'to whom shall I look if not the quiet and humble man who fears my word?'). Another popular expression is ἡσύχιος βίος, which is already attested in the classical age but becomes the definition of the Christian way of life, devoted to humbleness and peacefulness. An influential text that employs this expression is St Paul's Second Letter to Timothy (2.2): ἵνα ἤρεμον καὶ ἡσύχιον βίον διάγωμεν ('[let us pray] so that we can lead a quiet and peaceful life'), which again is often quoted in later texts. 39 37 Spicq (1978) 358 and 363. 38 In some cases ἥσυχος seems to have been employed specifically to refer to the quality of being silent (a meaning that the adjective already had in classical Greek) rather than being generally quiet: cf. Sir. 25.20 ἀνάβασις ἀμμώδης ἐν ποσὶν πρεσβυτέρου, οὕτως γυνὴ γλωσσώδης ἀνδρὶ ἡσύχῳ ('like a sandy hill for an old man is the nagging wife for a silent man'); Book of Wisdom 18.14.1 ἡσύχου γὰρ σιγης περιεχούσης τὰ πάντα κτλ. ('for when a peaceful silence encompasses everything, etc.'). 39 See Spicq (1978) 362 n. 1.
The centrality of ἡσύχιος and ἥσυχος in the Greek Christian lexicon might explain why the Antiatticist lemma devoted to these forms may have been interesting for Byzantine readers. The Pindaric model provided these common adjectives with an ancient pedigree, which was an integral part of the Byzantine mediation between Christian identity and the classical past. In principle, such a Judeo-Christian dimension may have been already present in the original version of the Antiatticist, since the language of the Septuagint and the New Testament is a repository of lower koine usages which the Atticists may have wished to discuss in their works for various reasons. 40 However, the interest of the Antiatticist entry on ἡσύχιος for Byzantine readers could also have resided in the fact that the adjective, while common in Byzantine literary language, does not seem to have been used in the medieval vernacular. Texts in this language variety employ ἥσυχος instead (cf. Kriaras s.v.) and this is the only form that has survived in modern Greek, where ἡσύχιος exists only as a personal name. It seems, therefore, that ἡσύχιος was a marked term in Byzantine Greek, typical of Christian language but not necessarily common in everyday communication. This sheds light on the context in which the original Antiatticist lemma was received, transmitted and perhaps altered at Byzantium: the entry placed a marked term, typical of Christian Greek, in continuity with classical usage and at the same time preserved memory of its rare synonym, ἡσύχιμος.

Conclusions
The study of the graphic habit of the scribe responsible for cod. Coisl. 345 allows us to defend the hypothesis that in the manuscript the sequence ἡσύχιος : ἡσυχώτερον : ἡσύχιμον. Πίνδαρος Ὀλυμπιονίκαις constitutes a single entry. The linguistic study of these words in their classical context further confirms that it is not necessary to correct or obelise any part of the entry, which can be explained as a synonymic-onomastic lemma based on Pindaric exegesis. The epitome preserves just one locus classicus, which is correct only for the last term of the entry, ἡσύχιμον. Since ἡσύχιος is attested in Pythian 2, it may be that its locus classicus has been lost in transmission. However, if the original purpose of the entry was to provide a list of synonyms, it is also possible that the ancient lexicographer thought of adding a direct reference only for ἡσύχιμονa hapax, and thus a marked form. In this scenario, ἡσυχώτερον should not be considered a (faulty) gloss of ἡσύχιος, but another form which the author of the Antiatticist (or perhaps, as suggested above, a medieval interpolator) added for its morphological interest. Thus there is no reason to separate ἡσυχώτερον from the preceding ἡσύχιος and the following ἡσύχιμος.
The original purpose of the entry may have been to illustrate the classical usage of ἡσύχιος, a common adjective in koine and Christian Greek, in the light of its occurrence in Pindar. In later ages interest in these Pindaric forms may have crossed paths with biblical exegesis, contributing to the shaping of their reception in the Byzantine age. This Byzantine outlook on the selection of the Antiatticist (and of many other ancient lexica of which we have medieval reworkings) should be a necessary starting point for any analysis of the linguistic meaning of its lemmata. More broadly, we should cautiously investigate not only how the original material was transmitted and hence possibly rearranged, but also how much new (i.e. medieval) material may be lurking behind ancient lemmata. This bigger question exceeds the scope of the present article and would require a careful monographic study, which should specifically look into the interpolation of biblical glosses or into the biblical recasting of ancient ones in order to pinpoint the different paths of use and reuse of classical material in Byzantine lexicography.

Works cited
Abbreviations