φροίμιον, φροῦδος
(Moer. φ 5, Philemo [Vindob.] 396.20, Moer. φ 6)

D. General commentary

Two consecutive lemmas in Moeris’ lexicon (A.1, A.3) prescribe the forms φροίμιον (a variant of the more common προοίμιον, ‘prelude, preamble, proem’) and φροῦδος (‘gone away, perished’) as Attic. The first prescription finds a parallel in Philemon’s Atticist lexicon (A.2) and is echoed in Hesychius (B.2) and other later erudite sources. Indeed, the adjective φροῦδος, as will be discussed in greater detail below, is a favourite of the Attic dramatists that was revived in later prose. Attic tragediansTragic language also favour the noun φροίμιον (9x in Aeschylus, see e.g. C.1; 10x in Euripides, see C.2, C.3, C.4, C.5), but it is absent from prose until the imperial period, whereas προοίμιον (Pi.+) is the only form that occurs in prose but is rare in tragedy (Aesch. 1x, Eur. 3x): ‘this distribution suggests that φροιμ- was a spoken variant; both variants, however, appear to have had a somewhat lofty ring (note their absence from Aristophanes, except Knights 1341–44 [C.6])’ (Maslov 2012, 194). The verbal derivative προοιμιάζομαιπροοιμιάζομαι (‘to make a prelude, to say as a premise’) likewise has the variant φροιμιάζομαιφροιμιάζομαι: the latter is the only form found in drama (twice in Aeschylus and once in Euripides), but it is also employed by Aristotle (EN 1905a.12, Metaph. 995b.5, Poet. 1460a.10, Pol. 1323b.17, 1325b.33; at Rh. 1415b.21 φροιμιάζῃ is a quotation from Eur. IT 1162) in addition to the uncontracted form (EE 1217a.18, Rh. 1415b.24, 1416b.34). A similar situation is encountered in the Aristotelian Rhetorica ad Alexandrum now ascribed to Anaximenes of Lampsacus, which boasts two instances of φροιμιάζομαι and two of the verbal adjective φροιμιαστέον in addition to a single instance of προοιμιάζομαι; in the same text, the noun is spelled προοίμιον 16 times and φροίμιον only once (31.3.3). While these data suggest that oscillation between contracted and uncontracted formsContraction persisted into the 4th century BCE, the Hellenistic age presents a different picture: the contracted forms disappear from prose writings with the exception of isolated attestations in Josephus (AI 18.221.6 φροίμιον) and Plutarch (Apophthegmata Laconica 224c.6 φροιμιάζῃ); verse occurrences of φροίμιον are found in a dactylic fragment of Callimachus (C.12) and in AP 9.236.4 (by Lollius Bassus, 1st c. BCE – 1st c. CE). Among Atticising writers, only Lucian uses φροίμιον (5x) and φροιμίαζομαι (1x: see Schmid, Atticismus vol. 1, 351) and even then considerably less frequently than προοίμιον (20x) and προοιμιάζομαι (2x). Later, φροίμιον is used twice by Synesius (Regn. 5.1.3, Provid. 1.6.48).

The word’s etymology has variously been explained as a compoundCompounds of either οἶμος ~ οἷμοςοἶμος ~ οἷμος ‘path’ or of οἴμηοἴμη ‘song’ (see DELG, EDG s.v.). Maslov (2012) argues for the former derivation and for the original meaning of ‘proper speech act preceding any undertaking’ (literally: ‘what one says before setting out on one’s way’), only secondarily specialised in the sense ‘extended poetic composition performed before an epic performance’. The etymologyEtymology of προοίμιον was already the subject of discussion in antiquity; a possible instance of wordplay in Plato’s Phaedrus (266d.7–9)Pl. Phdr. 266d.7–9, where the word is implicitly analysed as though derived from πρῶτον οἶμαι (‘I think, first of all’), may allude to a contemporary etymological debate (see Sansone 2007, who argues that Plato himself preferred οἴμη as the ‘real’ etymon, while Aristotle may have preferred οἶμος). The variant φροίμιον pleads in favour of a second compound member with an initial aspiration (attested for οἷμος, but not for οἴμη), as it can only be explained based on *προ-οἵμιον with crasis between the prefix and the second member, accompanied by anticipation of aspiration of the initial consonant cluster (on the phonetic nature of this change, see Jatteau 2016, 49–53; Sayeed 2019, 167–8; Batisti 2022: it should, perhaps, be distinguished from the prehistoric metathesis of anticipation of the kind *ēsmai > *ēhmai > ἧμαι). However, the aspiration in οἷμος itself is only marginally attested (e.g., by Hdn. GG 3,1.546.17) and has no clear etymological justification. As such, this argument alone is not sufficient to resolve the etymological debate: as an anonymous reviewer suggests, the aspiration in both οἷμος and φροίμιον may be analogical after ὁδός ‘road’.

The same phonological change occurs in a handful of other compoundsCompounds – φροῦδοςφροῦδος (‘gone away, perished’; < *προ- + ὁδός), φρουράφρουρά (‘watch’) and φρουρόςφρουρός (‘watcher, guard’; < *προ-ὁρά, *προ-ὁρός) – and the verb registered in modern etymological dictionaries as ‘πίφρημι’ (‘to let in; to intrude’), which is actually attested only in prefixed formations (with ἀπο-, δια-, εἰσ-, ἐκ-, ἐπεισ-) and primarily in tenses other than the present. Although they were products of the same change, these words’ individual histories– and their reception by the Atticist movement – may differ considerably.

The adjective φροῦδος is common in tragedy (Aesch. only Suppl. 863, in a textually difficult passage, but 15x in Soph. and around 40x in Eur., see e.g. C.4) and comedy (Eup. fr. 169; Ar. 21x; Men. DE 18, Dysc. 776), but rare in classical prose (C.9, Arist. GC 318.17, Mete. 340a.2, 353a.1). Its defective attestation (basically restricted to the nominative, singular and plural: cf. LSJ s.v.) may be attributed to its use in predicativePredicative constructions with verbs of being or motion. The word, which had an expressive connotation (‘gone’, ‘vanished’, ‘ruined’), belongs to the ‘poetic vocabulary’Poetic language (Olson 2016, 78; see also Rau 1967, 134: ‘echtes Tragödienwort’; Gomme, Sandbach 1973, 121: ‘a favourite of tragedy’), and Aristophanes employs it frequently in paratragic or lyric passages (see e.g. C.7, discussed in F.2). Like φροίμιον, φροῦδος was perceived as typically Attic: Moeris (A.3) contrasts it with the synonyms ἄφαντος and ἀφανής, which, while they were not innovations of the koine (the former is attested since Homer, the second since Aeschylus and Herodotus), were not restricted to classical Attic. Prose writers take it up again from the late 1st century CE, arguably under the influence of Atticist theories: the word is common in Atticising writers, including Aelius Aristides (see Schmid, Atticismus vol. 2, 210), Lucian (see Schmid, Atticismus vol. 1, 351), Dio Chrysostomus, Aelian (see Schmid, Atticismus vol. 3, 225), Philostratus (see Schmid, Atticismus vol. 4, 336), and Alciphron.

On the other hand, φρουράφρουρά was significantly more typical (attested since Aeschylus, it is common in Attic poetry and prose but also in Herodotus, and lived on in the koine) and had no competing non-Attic variants; as such, it was not of special interest to purists. In neither case is the uncontracted form attested as a mere phonological variant, unlike προοίμιον in the case of φροίμιον. The morphologically more transparent adjective πρόοδος (X.+) is attested at a later date than φροῦδος and has the more literal meaning ‘going before’, but it appears as a formal variant of φροῦδος only in erudite texts that discuss the formation of the latter. More common is the noun ἡ πρόοδος ‘progression, advance’, also attested since Xenophon. φρουράφρουρά, φρουρόςφρουρός and their cognates (especially φρούριον ‘fort’ and φρουρέω ‘to keep watch’) are the only forms found throughout the entire Greek literary corpus (dialectal forms with an unaspirated stop, such as Ionic and Thessalian προυρός and Cyrenaic πρωρός, are attested only epigraphically); while the verb προοράωπροοράω (‘to see before, foresee’) does exist, no *προορά or *προορός forms are attested.

Again, the case of the compound verbsCompounds in -φρῶφρέω differs. Willi (2020) recently elucidated their complex history, demonstrating that they ultimately originated in the aorist subjunctive προ-ὧ of προΐημι, which replaced the simplex ὧ due to the excessively slight Wortumfang (‘word body’) of the latter. Forms with two consecutive o-sounds (1st sing. προ-ὧ, 1st pl. προ-ὧμεν, 3rd pl. προ-ὧσι) then underwent crasis, yielding φρῶ, φρῶμεν, φρῶσιν. With such forms serving as a pivot, a new stem -φρ- first entered the future tense; two competing presents (thematic -φρέω and athematic -πίφρημι) and aorists (sigmatic -φρησα and kappatic -φρηκα) were then created by analogical means. These innovative forms, first attested in Euripides (C.2), are ‘exclusively Attic’ (Willi 2020, 310) and appear in 5th- and 4th-century-BCE Attic authors across all genres. They almost disappear in literary koine (only Plb. 21.27.7) before being revived by 2nd-century-CE Atticising writers (Lucian, Alciphron) and other prose writers of the Roman period (Philostratus Iunior, Heliodorus, Diogenes Laertius). No discussion of these verbal forms has survived in the main Atticist lexica, but other lexicographical sources do comment on them, leaving no doubt that they were regarded as proper to classical Attic. See, for instance, Harpocration’s (B.1) observation that εἰσφρέωεἰσφρέω, used by the orators Antiphon (C.10) and Demosthenes (C.11), also occurred frequently in Old Comedy (see C.8 for an Aristophanic example). The grammarian PhiloxenusPhiloxenus (1st century BCE) correctly recognised that the prefix προ- underwent the same mutation in the verb (-)φρῶ as in the nominal forms φροίμιον, φρούρα, and φροῦδος (F.1). His doctrine is preserved in a scholium to Euripides (B.3) and in Byzantine scholarship (B.5). This exemplifies a (sporadic) sound change that was characteristic of classical Attic but did not affect any other literary dialects. Purist lexicographers observed this distribution and consequently prescribed the φρ-forms as Attic – albeit, of course, only where a concurrent form (either a phonological variant or a synonym) was in use.

E. Byzantine and Modern Greek commentary

Uncontracted προοίμιον is the prevalent form in Post-classical Greek, likely due to its greater morphological transparency and the fact that in classical Attic it already predominated in prose; however, sporadic occurrences of φροίμιον may still be found in high-register texts from the medieval period (Michael Ephesius in Arist. PA 104.4, Iohannes Beccus Ad Constantinum 2.361.15, Manuel Philes Carmina 5.29.11) along with a previously unattested derivative of φροιμιάζομαιφροιμιάζομαι such as φροιμίασις (‘address’) (Nicetas Choniates Historia 336.28). Modern Greek has the learned form προοίμιο, meaning ‘opening of a text, proem’ in accordance with other modern European languages (where it was mediated by the Latin loan prooemium: cf. English proem), and metaphorically also ‘beginning’. For its part, φροῦδος continued to be used throughout the medieval period in the high register; in Modern Greek it survives only in the fixed expression φρούδες ελπίδες ‘vain hopes’ (see LKN s.v.); the collocation is already found in Eur. Ion 866Eur. Ion 866 φροῦδαι δ᾿ ἐλπίδες and reappears in late ancient and Byzantine literary writers.

F. Commentary on individual texts and occurrences

(1)    Et.Gen. AB s.v. διαφρῶ (= Philox.Gramm. fr. 87) (B.5)

Philoxenus’ doctrine concerning the derivation of the verb φρῶφρέω may be reconstructed on the basis of a scholium on Euripides (B.3) and the parallel entries in Orion’s lexicon and the medieval Etymologica regarding the words φρέαρ (Orion Et. 160.38 ~ Et.Gen. AB s.v. φρέαρ = Philox.Gramm. fr. 28) and ἀφρός (Orion Et. 21.13 ~ Et.Gen. AB α 1486 ~ Et.Gud. 247.22 = Philox.Gramm. fr. 56). Theodoridis also ascribed to Philoxenus the entries on διαφρῶδιαφρέω (B.5) and φρόνιμος (Orion Et. 162.2 ~ Et.Gen. B s.v. φρόνιμος = Philox.Gramm. fr. 194). The former entry is of particular interest, where the cases of προοίμιον > φροίμιον and προορός > φρουρός are adduced as parallels to the development προ-ὧ > φρῶ. This explanation essentially anticipates that advanced by modern linguistics, although, as Willi (2020, 303) observes, Philoxenus apparently derived φρῶ from a present subjunctive προ-ἱῶ and not from an aorist subjunctive. As the entries’ lemmas suggest, the verbal form φρῶ abstracted from compounds like διαφρῶ was exploited for the etymological analysis of nominal forms containing the sequence (-)φρ- (φρέαρ, ἀφρός, φρόνιμος). The derivation of φροῦδος from πρόοδος is explained along similar lines in the Etymologica (B.6), quoting Apollodorus of Athens (1st century BCE), and in Pseudo-Arcadius’ epitome of Herodian’s Περὶ καθολικῆς προσῳδίας (B.4).

(2)    Ar. Nu. 717–22 (C.7)

This is the second of two tirades (the first at lines 709–15) that are uttered by Strepsiades, who laments his woes in the style of a tragic threnos, and consequently show paratragic traits in their language and metre. These verses’ object of parodyParody has been identified as a passage (C.4) of Euripides’ Hecuba (parodied elsewhere in this play: cf. Ar. Nu. 1165–70 and Eur. Hec. 171–4) that is likewise characterised by the polyptotic anaphora of φροῦδος (according to Olson 2021, 156, however, ‘the parallel is not very close, and Strepsiades’ language may just as well be intended to sound generically tragic’). This instance is typical of Aristophanes’ use of φροῦδος, which is prevalently – though not exclusively – paratragic, as also in Ach. 470, Th. 691, Ra. 94, 1343b; examples of non-paratragic use are Th. 794, Ach. 208–10, Pax 197, Ra. 305, Ec. 311, 341, 950, Lys. 106 (Rau 1967, 134, 190; Dover 1968, 189; Austin, Olson 2004, 243). Interestingly, the anaphora is interrupted by the allegedly proverbial expressionProverbs φρουρᾶς ᾄδων – literally ‘singing while on guard’ (cf. Apostol. 17.95) – which referred to ‘passing the time in tedious or uncomfortable situations’ (Dover 1968, 189). The choice of this phrase was likely suggested by the punning assonance between φρουρᾶςφρουρά and the preceding and subsequent instances of φροῦδος: this entire passage exhibits an unusual degree of repetition and assonance (Dover 1968, 188). Eup. fr. 169Eup. fr. 169 φροῦδον τὸ χειρόνιπτρον ‘gone is my hand-washing basin!’ may come from a paratragic speech comparable to this Aristophanic passage (Napolitano 2012, 221–2 n. 600).