κάθησο, κάθου, κάθησαι, κάθῃ
(Antiatt. κ 1, Antiatt. κ 2, Moer. κ 49, [Hdn.] Philet. 90, Orus fr. A 55.19, Orus fr. A 57.1‒9)

D. General commentary

Two entries in the Antiatticist (A.1, A.2), one in Moeris (A.3), one in the Philetaerus (A.4), and two in Orus (A.5, A.6) deal with the verb κάθημαι ‘to be seated’, focusing on two main problems: (1) the correct form of the 2nd-person singularEndings, 2nd-person singular present imperative (i.e. κάθησο or κάθου, cf. A.1, A.3, A.4, A.6), and (2) the correct form of the 2nd-person singularEndings, 2nd-person singular present indicative (i.e. the ‘long’ form κάθησαι or the ‘short’ one κάθῃ, cf. A.2). The latter issue is also discussed by Atticist sources in relation to the verbs δύναμαι ‘I am able’ and ἐπίσταμαι ‘I know’: see entry δύνῃ, ἐπίστῃ. A third problem that seems to be implied by Orus’ discussion of κάθημαι is the use of internal vs. external augment in the imperfect (cf. A.5 and below). Other sources (not analysed here, but see B.2) also discuss the semantic difference between κάθησο and κάθισον (i.e. the aorist imperative of καθίζω ‘to sit down’), explaining that the former is used to prevent someone who is sitting from standing up, whereas the latter is used to have someone who is standing sit down (cf. Luc. Sol. 11, [Ammon.] 262, Thom.Mag. 211.11‒8, Schol. rec. Ar. Pl. 724b, and entry καθίζω, καθέζομαι, καθιζάνω).

The verb ἧμαι (‘to be seated’, first attested in Homer 138x), mostly in its preverbated form κάθημαι (already in Homer 36x), is an athematic present going back to *h₁eh₁s-, which ‒ as explained by Willi (2018, 205 n. 179) ‒ is ‘not truly a root, but reduplicated *h₁es- ‘be’’, since ‘‘sitting’ can be a temporally bounded form of ‘being’’ (cf. also LIV 232 and Willi 2018, 151). Given its basic meaning, the verb is extremely common not only in literature (for the classical period cf. e.g. Aristophanes 57x, Thucydides 12x, Plato 38x, Xenophon 53x), but also in documentary texts, thanks to its use in relation to judges, assembly members, etc.; cf. e.g. Syll.³ 37–8.B.31‒2 [Teos, after 479 BCE], SEG 26.72.5 [Athens, 375/4 BCE], IG 4².121.111 [Epidaurus, ca. 350‒300 BCE], UPZ 1.1 (= TM 65797) (4x) [Memphis, ca. 350‒300 BCE], and P.Cair.Zen. 1.59107.6 (= TM 757) [Alexandria (?), 257 BCE].

As already mentioned, the Atticists’ main concerns regarding the paradigm of κάθημαι are the two competing forms of the 2nd-person singular present imperative (κάθησο and κάθου) and indicative (κάθησαι and κάθῃ). As far as the imperative is concerned, the 2nd-person singular middle ending is *-so, which is retained as -σο in the athematic conjugation, while in the thematic conjugation, after the loss of the intervocalic /s/, the ending undergoes contraction with the thematic vowel (-εσο > -εο > -ου; cf. e.g. Sihler 1995, 605; for a detailed discussion of IE imperative endings, see Forssman 1985). Homer has only κάθησο (in C.1 and Hom. Il. 2.191), which is the expected form since κάθημαι is athematic. However, the form κάθου – clearly resulting from the development of athematic κάθημαι into thematic κάθομαι, cf. AGP 1, 313–4 – appears already in the classical period, with one attestation in Aristophanes (C.3, while there are five cases of κάθησο in the playwright’s extant works, i.e. Ar. Ach. 49, Ec. 144, 169, 554, and Pl. 724; for other -ου/-ω contract imperatives of athematic verbs in classical authors, cf. Lautensach 1918, 87–8). Later occurrences of κάθου are mainly in Middle and New Comedy (cf. Anaxandr. fr. 14 and Millis 2015, 91; Alexis, C.7; Diphilus, C.8; Men. Dysc. 931), in the Septuagint (5x), in the New Testament (2x), and in religious texts, mostly commenting on the scriptural passages in which κάθου occurs. Evidently, the proliferation of κάθου in koine texts led stricter Atticists to reject the form, and the isolated attestation in Aristophanes (C.3) did not constitute sufficient evidence to support its usage (specifically on Phrynichus’ scepticism towards hapax legomena, cf. entry βρέχει). Indeed, κάθου is labelled as κοινόνκοινός by Moeris (A.3; on the meaning of this label in this and other entries in the lexicon, see Moeris, Ἀττικιστής and Monaco, forthcoming) and proscribed by the Philetaerus (A.4; cf. also Thomas Magister, B.3), whereas the Antiatticist (A.1) and Orus (A.6) accept it (the former in view of its occurrence in Alexis, C.7, and Diphilus, C.8, the latter on the basis of    its occurrence in Aristophanes, C.3). Herodian (Περὶ παθῶν GG 3,2.299.12–3 quoted in EM 483.36–40, cf. Epim.Hom. κ 82, Et.Parv. κ 3) appears to have explained the form as a development of κάθησο, which, in turn, was considered the truly Attic form (for κάθησο and κάθου in Homeric exegesis, cf. schol. (ex.) Hom. Il. 2.191a1 (T), schol. (ex.) Hom. Il. 2.191a2 (BCE³), Eust. in Od. 2.171.7–9).

Moving on to the 2nd person of the present indicative of κάθημαι, the form κάθησαι is the ancient one, preserving the primary middle ending -σαι (Proto-Greek *-sai or *-soi, cf. Sihler 1995, 476; Willi 2018, 559‒60), as happens in Attic Greek for all athematic verbs, while thematic verbs lose the intervocalic /s/ with the ensuing contraction with the thematic vowel (-ε-σαι > -εαι > -ῃ: K–B vol. 2, 68; Schwyzer 1939, 668; on ancient interpretations of this form, cf. also Choerob. in Theodos. GG 4,2.176.20‒177.16, Et.Gud. 290.40‒291.10, EM 484.20‒35, and entry δύνῃ, ἐπίστῃ). The earliest literary occurrences of κάθησαι are in Xenophon (3x). The verb is then found in poetry (Theocritus 1x, Callimachus 2x, Menander 1x), in the Septuagint (5x), in Atticising authors of the imperial period such as Lucian (3x), Aristides (2x), and Libanius (1x), and in religious texts (several of which refer directly to the Septuagint passages where κάθησαι occurs). On the other hand, the form κάθῃ, analogical to the thematic conjugation, is first attested in Hyperides (C.6, quoted by the Antiatticist, A.2; note that Cratinus, C.2 – despite the accentuation κάθῃ in A.6, which preserves the fragment – has the subjunctive καθῇ, as evidenced by the negation μή). After Hyperides, analogical κάθῃ reappears sporadically in the Septuagint (1x), the New Testament (1x), and Epictetus (1x, who however also uses κάθησαι 1x), before becoming widespread in religious texts commenting on (or being influenced by) Scripture (for κάθῃ in documentary texts, cf. P.Oxy. 8.1160.24–5 (= TM 31726) [3rd–4th century CE]: ἔγραψές μοι δὲ ὅτι κάθῃ ἐν Λεξάνδριαν, ‘you wrote to me: ‘you are staying in Alexandria’’; see also Gignac 1981, 414).

In order to fully reconstruct the Atticist debate over κάθησαι and κάθῃ, one needs to compare the evidence provided by the Antiatticist (A.2) and Orus (A.6) with Phrynichus’ stance on another middle athematic verb, δύναμαιδύναμαι (‘I can’). Indeed, in Ecl. 336Phryn. Ecl. 336, Phrynichus condemns the use of the short form δύνῃ in the indicative and prescribes instead δύνασαι as the only correct indicative form (cf. entry δύνῃ, ἐπίστῃ). Although he may never have devoted an entry to κάθημαι, we may assume that Phrynichus held the same view with regard to κάθησαι and κάθῃ. If this were the case, the Antiatticist’s statement κάθῃ· ἀντὶ τοῦ κάθησαι (A.2) can be interpreted as a response to this alleged prescription: based on the attestation in Hyperides (C.6), the Antiatticist would accept κάθῃ in both the indicative and the subjunctive (the same interpretation applies to Antiatt. ε 28Antiatt. ε 28: ἐπίστῃ· ἀντὶ τοῦ ἐπίστασαι, cf. entry δύνῃ, ἐπίστῃ). Orus’ perspective (A.6), however, appears to be different. Indeed, after quoting Cratinus (C.2), he states that the playwright forms the subjunctive (that this is the mood in C.2 is made clear by the negation μή) ‘from κάθημαι, κάθῃ, κάθηται in the indicative’. In other words, Orus seems to regard κάθῃ as the standard 2nd-person singular indicative form (completely overlooking, it seems, the ancient form κάθησαι) and to think that Cratinus also used κάθῃ as a subjunctive. This stance may have resulted from a misunderstanding of an earlier source, possibly the same one from which Orus drew Hyperides’ quotation (C.6, where κάθῃ is indeed the indicative form) and on which the Antiatticist (A.2) also relied, namely Aristophanes of Byzantium (cf. Alpers 1981, 108). Another factor may have been that in Orus’ time κάθῃ was in fact the most common form for the 2nd-person singular indicative of κάθημαι (in literature dating between the 4th and the 6th century CE one finds κάθῃ 47x and κάθησαι 21x).

Regardless of Orus’ opinion on the matter, it is plausible that in the imperial period the stricter Atticists allowed only the form in -σαι in the middle indicative of athematic verbs, whereas the Antiatticist (A.2) rejected such a prescription and accepted the ending -ῃ alongside -σαι for both κάθημαι and ἐπίσταμαι (cf. above and entry δύνῃ, ἐπίστῃ).

Other forms of κάθημαι are listed by Orus (A.5, A.6). Except for the optative καθοίμην (A.6), the remaining forms all belong to the imperfect indicative: καθήμην (1st pers. sing.), καθῆστο (3rd pers. sing.), and καθῆντο (3rd pers. plur.) are mentioned in A.5, while in A.6 Orus briefly touches upon the 1st person plurals καθήμεθα and ἐκαθήμεθα. A.5 is the last line of Orus fr. A 55, a long entry primarily devoted to the future καθεδοῦμαι (‘I will sit down’) and various forms of καθίζω (‘to sit down’; cf. entry καθίζω, καθέζομαι, καθιζάνω). In all likelihood, the entry is strongly epitomised and what now appears as a mere list of forms at the end of it was originally a more articulated discussion. Unfortunately, parallels with earlier lexicography are lacking. Nevertheless, the comparison with Thomas Magister (B.3) may at first suggest that one of Orus’ concerns was to recommend forms with the internal augment καθῆστο (also καθῆτο, cf. below) and καθῆντο against their counterparts with the external augment ἐκάθητο and ἐκάθηντο. In A.6, however, the lexicographer appears to allow both καθήμεθα and ἐκαθήμεθα (in B.1, Pollux notes that Demosthenes, C.5, used καθήμεθα instead of ἐκαθήμεθα).

Externally augmented forms of verbs with a preverb arose already in the classical period, when speakers stopped perceiving the prefixPrefixes as a separate unit and began to consider it as part of the root (cf. entries ἀνέῳγoν, ἤνοιγoν and δεδιακόνηκα, διηκόνηκα, ἐδιακόνουν, διηκόνουν). In the case of κάθημαι, only the forms with internal augment are attested in Homer (cf. C.1 and Hom. Od. 4.628; for the ancient grammatical explanations of these Homeric forms, cf. Epim.Hom. κ 106, Epim.Hom. 569a, Et.Gen. AB s.v. καθῆστο, Et.Gud. 291.49–292.11, Eust. in Il. 1.229.26–230.14). Although they are well attested in classical authors (e.g. καθῆστο: Euripides 1x, Aristophanes 1x, Plato 3x, cf. C.4; cf. also καθῆτο in Demosthenes 2x; καθῆντο: Aristophanes 2x, Demosthenes 1x), their counterparts with external augment are already found in the 5th century (e.g. ἐκαθήμην: Aristophanes 1x; ἐκάθητο: Thucydides 1x, Xenophon 4x; ἐκάθηντο: Thucydides 3x, Xenophon 4x, Plato 1x). Internally and externally augmented forms of κάθημαι coexist in Post-classical Greek, with some authors using both, e.g. Plutarch καθῆστο 11x (καθῆτο 1x), ἐκάθητο 2x, καθῆντο 2x, ἐκάθηντο 2x; Lucian καθῆστο 1x, ἐκάθητο 1x, ἐκάθηντο 1x; Aristides καθῆστο 5x (including C.10), καθῆντο 1x = C.9 (but the reading is uncertain, cf. the apparatus), ἐκάθηντο 1x. Nevertheless, internally augmented forms appear to have been the exception in the koine (e.g. they are absent from the Septuagint).

In sum, the relative rarity of forms of κάθημαι with internal augment may have been the main reason why strict Atticists rejected their externally augmented counterparts, which were standard from the Hellenistic period onwards (cf. Gignac 1981, 252; see also E.). Still, there is no trace of such a prescription for the paradigm of κάθημαι in imperial lexica and the only evidence of an overt preference for internally augmented forms is in Thomas Magister (B.3). Nevertheless, the existence of a more ancient doctrine is suggested not only by Thomas Magister’s entry, but also by the fact that Orus (A.5) seems to take the opposite stance by accepting both καθήμεθα and ἐκαθήμεθα.

E. Byzantine and Modern Greek commentary

The Byzantine survival of the imperative κάθησο is almost entirely limited to scholia, lexica, and other grammatical texts, with the exception of a few sporadic literary occurrences dating no later than the 15th century (cf. Laonicus Chalcondyles 1x). The thematic form κάθου, instead, is much more broadly attested in Byzantine literature and is found as late as the 18th century (cf. Blasius Atheniensis 1x, Nectarius Terpus 1x).

The situation is different for κάθῃ and κάθησαι. Indeed, the short form is not attested after the 16th–17th century (cf. Meletius Pegas 1x), while κάθησαι remains in use up until the 19th century (cf. e.g. Blasius Atheniensis 1x, Nicodemus Hagiorita 1x, Neophytus Ducas 2x). In Modern Greek, the conjugation of κάθομαι (cf. CGMEMG vol. 3, 1384) is as follows: κάθομαι, κάθεσαι, κάθεται, καθόμαστε, κάθεστε/καθόσαστε, κάθονται.

When it comes to the forms with internal and external augment, the former survive in the Byzantine period almost exclusively in lexicography and scholia (with the exception of a few attestations in high-register texts, e.g. Symeon Metaphrastes 1x, Theodorus Metochites 1x, Demetrius Cydones 2x). Meanwhile, the externally augmented forms become standard in Byzantine literature and remain in use until the early modern period. In Modern Greek, the imperfect of κάθομαι is inflected as follows: καθόμουν(α), καθόσουν(α), καθόταν(ε), καθόμαστε/καθόμασταν, καθόσαστε/καθόσασταν, κάθονταν/καθόντανε/καθόντουσαν.

F. Commentary on individual texts and occurrences

N/A