alarm-

This noun is attested in inscriptions of various genres, all of which were found in Sardis. The inscription LW 3 is early Hellenistic, LW 80 is datable to the 4^th century BCE, whereas the other texts can be dated only more approximately within the broad range of 5^th - 4^th century BCE.

[I.Y.]

nom. sg. c.	alarms	LW 12.6 (Sardis)
acc. sg. c.	alarmn	LW 10.17 (Sardis)
dat. sg.	alarmλ	LW 12.2 (Sardis)
acc. pl. c.	alarms	LW 3.2 (Sardis)
	alarmas	LW 13.1 (Sardis)
poss. adj. acc. sg. c.	alarmlν	LW 14.5 (Sardis)
case unknown	alarms	LW 16.4 (Sardis)
	[a]larms	LW 80.1 (Sardis)
	ararms	LW 14.11 (Sardis)

The grammatical interpretation of the accusative plural forms depends on the analysis of Lyd. wsta-. In the instance of LW 80.1 [a]larms, the position at the beginning of the text suggests the default interpretation of nom. sg. c., but positive evidence is lacking.

[I.Y.]

Given that the form ararms occurs only once and the possessive adjective alarmlν occurs in the same inscription LW 14, it is probably easier to take ararms as reflecting a scribal error of anticipation, rather than genuine phonetic assimilation.

[I.Y.]

It seems reasonably clear that alarm- has the syntactic distribution of an adjective, agreeing with the head-noun in number and case. Indicative of this function are LW 10.17 alarmn kastaλcν, LW 12.6 alarms τesas, and the collocation alarm(a)s … wsta(a)s (see above). Its putative status as an adjective speaks against the meaning ‘founder’ (Meriggi 1935a:102f.), but would be compatible with the meaning ‘self’, offered in Vetter 1959a:34f. and tentatively accepted in Gusmani 1964a:60. Subsequently, this interpretation has generally been followed on the assumption that the stem alarm- appears in funerary inscriptions in those cases where the sponsor of the tomb designated it for himself, and not only for his relatives (Yakubovich 2005b:85).  

The challenge to the established wisdom comes from two separate observations. On the one hand, if wsta- is to be interpreted as a designation for some sort of transaction, the idiom alarm(a)s … wsta(a)s can no longer refer to the sponsor of the tomb. On the other hand, and more crucially, if kastaλc(i)- represents the derivative of the word for ‘bones’, it probably refers to funerary remains, which can hardly possess agency within the context of a burial inscription. The context of LW 10 suggests an alternative solution: one can interpret the opposition between alarmn kastaλcν and ẽnas amãs as ‘(my) own remains’ and ‘(those of) my ancestors’. The new translation is also conducive to interpreting alarm(a)s … wsta(a)s as ‘personal expenses’, or something similar. Presumably, this idiom appears in those instances where the author of the inscription invests in his own grave, a situation that remains substantially similar to the one reconstructed by Vetter 1959a:34f., although the inscriptions emphasize sources of funds rather than beneficiaries of the tombs. At the same time, one must acknowledge that the new proposal does not automatically elucidate the remaining attestations of alarm-, which occur in obscure or fragmentary contexts.

The suggested interpretation associates alarm- with the same meaning that had previously been assigned to its possible cognate arlil(i)-, which, in turn, obtains the new translation ‘belonging to the estate, household’ on independent grounds.

[I.Y.]

The form alarmas suggests that we are dealing with an etymological a-stem, which became affected by syncope of the final syllable, for which see Gérard 2005a:49f. The change affected all the contexts except for one poetic passage.

[I.Y.]