γρηγορέω, ἐγρήγορα
(Phryn. Ecl. 88, Antiatt. ε 109)

D. General commentary

Entries in Phrynichus’ Eclogue (A.1) and in the Antiatticist (A.2) deal with the verb γρηγορέω, an innovationInnovative forms formed on the basis of ἐγρήγορα ‘to be awake’, the perfect of ἐγείρω ‘to awaken’. In keeping with each lexicon’s characteristic stance, Phrynichus’ entry has an explicitly proscriptive purpose (later followed by Orus, B.1), whereas the Antiatticist defends the innovation.

The active intransitive perfect ἐγρήγορα, with present meaning and showing the so-called Attic reduplication, is a formation of Indo-European origin, reconstructable as *h₁ge-h₁gor- (from the root *h₁ger-, cf. ἐγείρω < *h₁ger-i̯e/o-) on the basis of the Indo-Iranian cognates (cf. Sanskrit jāgāra, Young Avestan jaγāra ‘s/he is awake’; on the form’s origins see now Jasanoff 2023; Rothstein-Dowden 2024, 30–4). The additional -ρ- in the Greek reduplication syllable was perhaps taken from other paradigmatic forms such as the Homeric middle aorist ἔγρετο (Od. 6.117), or may result from a metathesis *egēgr- > *egrēg- (Rothstein-Dowden 2024, 33). Forms of ἐγρήγορα are attested in the Iliad, albeit with an element -θ- of uncertain origin (see Schwyzer 1939, 540 n. 4 and 800 n. 8; Chantraine 1958–1963 vol. 1, 429; Rothstein-Dowden 2024, 35–40): the imperative ἐγρήγορθε (Il. 7.371 [C.1] = 18.299), the infinitive ἐγρήγορθαι (10.67), and the 3rd-pers. plur. indicative ἐγρηγόρθασι (10.419). In Attic-Ionic, at any rate, the perfect ἐγρήγορα and its pluperfect ἐγρηγόρη are the norm in both poetry and prose (see Veitch 1887, 210), with numerous occurrences in authors such as Hippocrates (10x), Heraclitus (3x), Aeschylus (2x), Aristophanes (3x, e.g. C.2), Xenophon (9x, e.g. C.3, C.4), Demosthenes (6.19.1), and especially Plato (26x, e.g. C.5) and the corpus Aristotelicum (68x). Notably, the pluperfect was as a rule unaugmented (see K–B vol. 2, 27). Although Photius (B.2, an entry without parallels in the lexicographical tradition) points out that Menander (C.7) used both ἠγρηγόρειν and ἐγρηγόρειν, this remains our only attestation of the augmented form in the whole of Greek literature. It was precisely from the reinterpretation of the 3rd-pers. sing. (unaugmented) pluperfect ἐγρηγόρει as an imperfect that the present γρηγορέω was probably derived (Schwyzer 1939, 768).

The innovative γρηγορέω first occurs in the later stages of Attic (Demad. fr. 72 De Falco = C.6) and becomes common in Hellenistic prose, starting from the Septuagint (9x, vs. 2x ἐγρήγορα); in the New Testament γρηγορέω, occurring 22 times (e.g. C.8), has completely replaced ἐγρήγορα (Blass, Debrunner 1976, 58). In the literary koine, by contrast, the inherited perfect largely endured (Plb. 3x; D.S. 2x; Dion.Hal. 2x; Ph. 2x; Ios. 3x; Plu. 27x). A slightly different innovative present, ἐγρηγορέω, is attested perhaps even earlier (Diocl. fr. 182, Hp. Vict. 86, Arist. Pr. 877a, etc.); see Debrunner (1929) and, more recently, Rothstein-Dowden (2024, 42), who analyses ἐγρηγορέω as a denominative verbDenominative verbs from an adjective *ἐγρηγορός ‘wakeful’ (distinct from the later ἐγρήγορος, on which see below), possibly also underlying the isolated Homeric participle ἐγρηγορόων (Od. 20.6). Incidentally, while the Atticist debate focuses on the more common γρηγορέω, ἐγρηγορέω features in several grammatical sources discussing the derivation of ἐγρήγορα and related forms (cf. Hdn. Περὶ παθῶν GG 3,2.287.13–8 = Et.Gen. AB s.v. ἐγρηγορῶ, EM 312.21–7, [Zonar.] 611.1–5; Eulogius in Et.Gud. ε 396.5–8; Choerob. in Theodos. GG 4,2.113.5–15). Both innovations may be seen as part of a broader tendency to replace perfects with present meaning – an archaism – with morphological presents, thereby eliminating a perceived mismatch between form and meaning (cf. e.g. Hom. ἄνωγα > ἀνώγω ‘to command’; cf. Rutherford 1881, 200–1). It is also significant that the attestations of ἐγρήγορα are skewed towards the non-finite forms from the outset, especially the perfect participle ἐγρηγορώς ‘awake’: an online TLG search of the occurrences up to and including Aristotle yields a total of 38 infinitives and 78 participles, against just 22 indicatives, two imperatives (the already-mentioned Homeric forms), and a single subjunctive (X. Cyn. 5.11 ἐγρηγόρῃ). Therefore, γρηγορέω and its aorist ἐγρηγόρησα may also have provided speakers with finite forms to replace those of the perfect, which were only sparsely used. It should further be noted that γρηγορέω underwent a semantic shiftSemantic shift towards the meaning ‘to be alert, to keep watch’ and became especially common in Biblical Greek and other religious contexts – e.g. in the exhortation γρηγορεῖτε καὶ προσεύχεσθε ‘watch and pray!’ (NT Ev.Marc. 14.38 = C.8), and in the imperative γρηγόρει ‘stay awake!’, addressed to the deceased on Palaeo-Christian funerary monuments (see DGE s.v.).

The innovative γρηγορέω is rejected by Phrynichus (A.1) and, perhaps less predictably, by Orus (B.1); the latter’s entry also prescribes the other correct forms of the paradigm of ἐγρήγορα, including the pluperfect, imperative, and subjunctive. The Antiatticist (A.2), on the other hand, attributes to Xenophon the use of the 3rd-pers. sing. ἐγρηγόρησεν, i.e. the aorist of γρηγορέω, although this attribution is arguably based on a faulty MS reading (see F.1). The likely aim of the Antiatticist’s extremely abridged entry was to assert the acceptability of the innovative form as an alternative to the classical ἐγρήγορα, rather than to promote it exclusively at the latter’s expenses.

Practitioners of literary Atticism avoided γρηγορέω, adhering instead to the usage of the higher koine: the main Second Sophistic authors all employ exclusively ἐγρήγορα (D.Chr. 14x; Luc. 6x; Aristid. 2x; Favorin. 1x; Ael. 2x; Philostr. 8x), while a mildly Atticising novelist such as Achilles Tatius (4.17.3 = C.9) offers a single occurrence of the innovative form. This distribution shows that γρηγορέω was widely perceived as a markedly unacceptable form from an Atticist viewpoint, as is further suggested by the agreement between Phrynichus and the usually more inclusive Orus (on the anti-Phrynichean tendency of the latter’s lexicon, see entry Orus, Ἀττικῶν λέξεων συναγωγή). Pollux likewise recommends forms of ἐγρήγορα several times in the Onomasticon (1.43Poll. 1.43; 1.155Poll. 1.55; 3.120Poll. 3.120; 3.121Poll. 3.121), while he makes no mention of the innovative variants. In addition to its lack of attestations in canonical authors (despite the Antiatticist’s appeal to Xenophon’s authority), γρηγορέω may have suffered from its distribution, largely restricted to the lower koine, and possibly also from its association with Christian language (on the significant overlap between New Testament terms and the forms proscribed by the Atticists, see Lee 2013, 303). The vitality of γρηγορέω in spoken Greek is further confirmed by its reflexes in several Modern Greek dialects (see E.).

E. Byzantine and Modern Greek commentary

Although the perfect disappeared from spoken usage in post-classical times, with only a few relic forms surviving in non-literary sources (see CGMEMG vol. 3, 1765–6), ἐγρήγορα remained in use throughout the Byzantine period in texts written in the higher register, with the same preponderance of non-finite forms already noted for earlier periods (see D.). At the same time, the innovative γρηγορῶ (with the variants ἐγρηγορῶ, γληγορῶ, γρηγοράω) was used in the lower registers of Medieval Greek, with the transitive meaning ‘to bring back to consciousness’, in addition to the intransitive ones ‘to stay awake, to wake up, to regain consciousness, to hurry’ (see Kriaras LME; LBG s.vv.). As a biblical term, γρηγορῶ was also employed in religious texts (such as homilies and theological treatises) even by authors (Photius, Psellus) who otherwise had a perfect command of the high register.

As for the present ἐγείρω, its meaning shifted from ‘to awaken’ to ‘to raise’ in Post-classical Greek, as is already apparent from New Testament usage. In Modern Greek, the reflex of ἐγείρω is γέρνω, with a further semantic shift to ‘to tilt, to lean forward’ (for a possible explanation of this, see Hatzidakis 1934, 131), but ἀνεγείρω and (ἀνα)γέρνω partly preserve the meaning ‘to raise’ in some dialects (see Andriotis 1974 s.vv.; Shipp 1979, 75–80). In the sense ‘to wake up (trans.); to be awake (intrans.)’, on the other hand, ἐγείρω was replaced by the verb ξυπνάω (< Ancient Greek ἐξυπνάω). Modern Greek εγείρω ‘to raise’ is a learned borrowing from Ancient Greek, employed in certain fixed phrases (see LKN s.v.). Similarly, Modern Greek γρηγορώ ‘to haste, to be alert’, from Hellenistic γρηγορέω, is a learned term, especially in literary or ecclesiastical expressions (see LKN s.v.); however, direct reflexes of (ἐ)γρηγορέω occur in the dialects (e.g. Kerkyran γρηγορώ, Pontic α(γ)ληγορώ: see Andriotis 1974; ILNE s.v.). The Ancient Greek adjective ἐγρήγορος ‘wakeful’ (unlikely to continue a prehistoric *ἐγρηγορός, as argued above), derived from the perfect stem ἐγρήγορα, lies at the root of several other Medieval and Modern Greek forms, such as γρήγορος ‘quick, fast’, γρήγορα ‘quickly’, and γρηγοράδα ‘quickness’ (also with less literary variants in γληγορ-, showing a dissimilation /r… r/ > /l…r/; see Shipp 1979, 226–7; CGMEMG vol. 1, 214).

F. Commentary on individual texts and occurrences

(1)    X. An. 4.6.22 (C.3)

The MSS of the Anabasis transmit here the 3rd-pers. plur. ἐγρηγόρησαν (i.e. the sigmatic aorist of γρηγορέω), but since Porson editors have corrected it to ἐγρηγόρεσαν (i.e. the unaugmented pluperfect of ἐγρήγορα), which is required by the sense (note the coordination with the imperfect ἔκαιον, and see Lobeck 1820, 119). According to Alpers (1981, 173), Porson’s conjecture is confirmed by Orus’ prescription of ἐγρηγόρεσαν (B.1), since he presumably had the same locus classicus in mind. Indeed, Xenophon (C.4) is most probably also the authority behind Orus’ prescription of the 3rd sing. pluperfect ἐγρηγόρει, for which he provides the only attestation before the Hellenistic period. It should be noted, however, that Orus explicitly cites Aristophanes (C.2) in support of ἐγρηγόρεσαν. Be that as it may, the Antiatticist (A.2) was arguably misled by a copy of Xenophon that already contained the same error as our MSS, and therefore taking the author of the Anabasis to be an early champion of the post-classical γρηγορέω (on disagreements between Xenophon’s direct tradition and forms quoted by the Atticist lexica, see Favi 2024). Cobet (1873, 41) proposed restoring ἐγρηγόρησαν as the Antiatticist’s lemma, but Sicking (1883, 110–1) rightly judged this unnecessary, pointing out that the lemma may reflect not the exact form found in the classical source, but rather the form attested in some later author whose usage the Antiatticist was defending (see also Valente 2015, 170). Finally, it should be noted that the Antiatticist quotes a form in the 3rd person singular, whereas C.3 has a 3rd person plural.