________________
214
Colette Caillat
pāṇe ya näiväejjā,
which (curiously) does not precede, but follows the recommendation not to permit the killing of living beings,
na hu pāṇa-vaham anujāne (ib. 8.8 JAS 216).
In Suy the phrase is clearly recognizable (in spite of the apocope affecting the preverb), e.g. in 1.1.1.3a (JAS 3a), where it refers to the first part of the threefold mahavvaya, whereas forms of HAN (or of a cognate word), "to strike", are used for the complementary second and third parts of the prohibition:
sayam tivāyae pane aduva annehi ghayae
haṇantam vâṇujāṇāi...
"if a man kills living beings, or causes other men to kill them, or consents to their killing them..." (Jacobi).
It is seldom that the verb occurs without the modifier pane. According to the JAS reading, it is so used in Suy 1.5.1.22d (JAS 321d):
...tikkhāhi sülähi tivāyayanti,
"they destroy with sharp pikes 45; but the usual reading 'bhitavayanti would here seem preferable (= 1.5.2.10a= JAS 336).
Because of the apocope the commentators are prompted to propose two etymologies: they admit that ti- can be the MIA counterpart either of OIA tri-, "three", or of the OIA preverb ati-.46 In fact, in spite of this ambivalent gloss, and even though the preverb ati- is of limited and comparatively archaic use in OIA,47 they certainly equate Amg (a)tivaei with the Sk causative atipatayati (cf. PW 4.396, s.v. PAT [ati-, caus. 3 "hinraffen"]). But perhaps they were taking it in a more or less specialized (technical ?) sense, "to make fall beyond (recovery)", "to utterly destroy, kill".48 On the other hand, perhaps at one stage, it was felt desirable to disjoin the two meanings: 1) "to let fly beyond", "to shoot (an arrow) through", and 2) "to destroy, injure"; hence the canonical Pa tendency to specialize atipateti in the first