šawtar-
‘watcher’
- Language
- Lydian
- Class
- Derivative
- Grammar
- subst, c.
- ID
- 260
1. Individual Anatolian Languages
1.1 Lydian
šawtar-
‘watcher’1.1.1Transmission
This lexeme is attested in the inscription LW 14 that has not been properly dated yet.
[D.S.]
1.1.2Forms
| nom. sg. | šawtaars | LW 14.2 (Sardis) | |
| dat. sg. | [šaw]ṭaarλ | LW 14.3 (Sardis) |
1.1.3Graphic Features
The dative form [šaw]ṭaarλ in LW 14.3 is reconstructed by Brandenstein 1931a:23. His reconstruction is persuasive, since the nominative form occurs in the previous sentence. It is taken over by Gusmani 1964a:193.
[D.S.]
1.1.4Meaning ‘watcher, guard (vel sim.)’
The meaning ‘watcher, savior’ for šawtaar- is suggested first by Haas 1962a:177 based on dubious assumptions. He based his claim on the healing cry of the Bacchanalians “εὐοῖ σαβοῖ, εὐαί σαβαῖ” (Demosthenes 18.260) and the foreign deity Σαβάζιος. He connected etymologically the root σαβ- (= Lyd. šaw-) with Gr. σαός ‘safe’ and retrieved further philological evidence pertaining to the semantics of the root σαβ- from Greek and Latin sources. Although his approach was hazardous, independent evidence for the semantics of Lyd. šawtaar- confirms his hypothesis. Melchert 1992a:40 suggests the meaning ‘watcher’ by analyzing šawtaar- as the verbal root *šaw- ‘to see’ + the agent noun suffix -taar- (cf. Hitt. -tar- + -a- in LÚweštara- c. ‘shepherd’, LÚakuttara- c. ‘drinker’). The meaning of šawtar- c. is supported by its verbal derivative šawtaari-(d) ‘to protect’, whose semantics are posited on contextual grounds.
[D.S.]
1.1.5Stem
The substantive šawtaar- is a non-mutated consonantal stem. The absence of i-mutation is proven by the lack of palatalization of the nominative singular ending -s.
[D.S.]
1.1.6Derivatives
2. Reconstruction
2.1Reconstruction
The substantive šawtaar- c. appears to be derived from the verbal stem *šaw- ‘to see’ (< *sHegu̯-) by means of the suffix -taar- (< *-tor-), Melchert 1992a:40, Gérard 2005a:89, 107 n. 727. However, since the root is reconstructed with *h3, it becomes difficult to account for the palatalization of the initial š-, which requires the presence of the vowel *e; for a solution see Part 2: Proto-Indo-European under Morphological Reconstruction.
[D.S.]
For the derivational base see Lyd. *šaw-.
3. Proto-Indo-European
3.1Semantic Reconstruction
Lyd. šawtaar- ‘watcher’ as well as YAv. aiβii-āxštar- ‘overseer’ and Gk. ὀπτήρ ‘spy, scout’ are merely synonymous (‘±watcher’). In Greek, the meaning became narrowed down to ‘who is looking for secrets’. The Latin noun secūtor, however, means ‘follower’. Accepting the semantic connection of ‘to follow’ and ‘to see, watch’ (cf. the root etymology s.v. PAnat. *sHógu̯-o-/*sHgu̯-´), the discrepancy between ‘follower’, on the one hand, and ‘watcher’, on the other, can be disregarded. Thus, the underlying formation is reconstructed here as PIE *sh3éku̯‑tor- ‘who attaches oneself/one’s eyes’.
3.2Morphological Reconstruction
Lyd. šawtaar- can be projected back to *sh3eku̯‑tor‑,Sub PAnat. Reconstruction above, it is mentioned that the palatalization of *s- > Lyd. š- actually only happens before *‑e- and not before *‑o- (< *‑h3e‑). Thus, the loss of *‑h3- after *s- could be posited, which, however, is not proven, cf. corresponding Core IE *seku̯- ‘to accompany’ < *sh3eku̯‑.YAv. (aiβii‑)āxštar- to *h3Eku̯‑s‑ter‑,On the sibilant, cf. the root etymology s.v. PAnat. *sHógu̯-o-/*sHgu̯-´.Gk. ὀπτήρ to *h3(E)ku̯‑tér- and Lat. *sector (→ secūtor) to *séku̯‑tVr‑.The original suffix vowel of the Latin lexeme cannot be determined because Latin has always ‑tor, never ‑ter, albeit in kinship terms, Weiss 2011a:308f.From the phonological point of view, Lydian and Latin share an isogloss, (acrostatic) *s(h3)éku̯‑tor- ‘(habitual) follower (Lat.), (habitual) watcher (Lyd.)’ (cf. Gk. δώτωρ, ‑τορος ‘habitual giver’).On the two types of agent nouns in *‑ter-, cf. most prominently Tichy 1995:376–379.On the other hand, Gk. ὀπτήρ, ‑τῆρος shows reflexes of hysterokinetic inflection: *(s)h̥3ku̯‑tér- (cf. Gk. δοτήρ, ‑τῆρος ‘occasional giver’).Cf., e.g., the meaning ‘eye witness’ and ὀπτῆρες ‘(occasional) watchers’ in A.Suppl.185.YAv. aiβii-āxštar- shows the Common Indo-Iranian extension of the root by *‑s- (PIIr. *ak‑š‑) and may thus be einzelsprachlich. Furthermore, it is closely connected with and maybe derived from the verb YAv. aiβii-āxšaiia- ‘to watch (over)’.
In view of the Avestan situation and the deviant hysterokinetic type in Greek, it is recommended to exclude the Greek and Young Avestan lexemes from the comparison. Thus, only on the basis of Lydian and Latin evidence an acrostatic agent noun can tentatively be reconstructed: PIE *sh3éku̯‑tor‑/*sh3éku̯‑tr- ‘who attaches oneself/one’s eyes habitually’.However, Lat. secūtor ‘follower’ may, like the Avestan and Greek lexemes, be created einzelsprachlich, too. In this case, Lyd. šawtaar- ‘watcher’ would have no direct cognates.
[A.O.]
4. Transmission
(?)YAv. aiβii-āxštar- m. ‘overseer’Cf. also the verb YAv. aiβii-āxšaiia- ‘to watch (over)’ s.v. PAnat. *sh3ógu̯-o-. On the long root vowel in both formations, cf. de Vaan 2003a:33, 103, NIL:382 n. 86.
Bartholomae 1904:97, Tichy 1995:39
Lat. secūtor m. ‘follower, attendant (also epithet of Mars) (Apul.+); (type of Gladiator) (Juv.+)’Lat. secūtor replaced Lat. *sector ‘follower’ (homophonous with Lat. sector ‘cutter’, Pl.+), cf. LHS:150, EDLIL:555f. etc. Similarly and probably not independently, also Lat. PPP secūtus ‘followed’ is a replacement of *sectus (reconstructable from sectārī ‘to run after’ etc.). Lat. secta ‘course (to be followed)’ (Naev.+, OLD²:1895) is the only relic form of the old participle. A parallel is Lat. loquī ‘to talk, speak’ : PPP locūtus ‘told, spoken’ (← *loctus) : locūtor ‘chatterer’ (← *loctor), cf., e.g., EDLIL:349.
EDLIL:—, WH II:—, EM:—, OLD²:1900

