G74 J 1 ii 36’ i-ši OV(l) – J has the wrong inflected verbal
T 1 ii 26 i-šu ing, or a possible difference in pronuncia-tion. (^550)
(^) G75 C ii 3’ (^) ana (^) OV – The preposition ana is written logo-
W 1 ii 29 ˹ a˺-na graphically in C.
G76 C ii 3’ ana lìb-bi giš MÁ SV(2) – C has an additional clarifying
W 1 ii 29 a-na giš MÁ noun. 551
G77 J T 1 ii 37’ ap-te-ḫi OV(l) – Possible difference in pronuncia-tion. (^552)
W^11 ii 27 ii 29 a]p-ap-˹˹tite˺-˺ḫ-ḫi i
(^) G78 C ii 4’ (^) [ ḫ]i-i (^) OV(l) – Possible difference in pronuncia-
W 1 ii 30 a-na pe- ḫe-˹e˺ tion. 553
G79 C ii 4’ pe-zu-ur ḫi-i šá GIŠ.MÁ d[ m pu- SV(1) – The preposition proper noun in W. ana precedes the
(^550) A.R. George, Gilgamesh, 429, counts this among “minor differences in words or expressions,” but this
could equally be considered as a peculiarity of Kuyunjik orthography (cf. A.R. George, Gilgamesh, 441,
type (s), “wrong vowel in inflected verbal ending”). Otherwise this may be a rare attestation of the allo-
phone [ü] as described in M. Luukko, 551 Neo-Assyrian, 87-88, appearing when /u/ is in contact with a sibilant.
The construct form in C clarifies the action, ērub ana libbi elippi, “I entered into the midst of the boat,”
expanding on the phrase in W, 552 ērub ana elippi, “I entered into the boat.”
The spelling aptiḫi in W against apteḫi in the other sources could reflect a pronunciation difference
where /i/ > /e/, cf. note above. W. Von Soden, Akkadisches Handwörterbuch (Wiesbaden: Harrassowitz,
1981) 853, has √paḫû as themed ‘i/i,’ so here the spelling in W is to be expected while the other sources
may reflect Assyrian orthography. Such seems to be the case in the orthography of both the 3ms verbal
preformative and medial ‘i’ class vowels in Neo-Assyrian letters according to M. Worthington, "Dialect
Admixture of Babylonian and Assyrian in SAA VIII, X, XII, XVII and XVIII," Iraq 68 (2006). However,
the generally inconsistent picture that emerges regarding the representation of ‘i’ class vowels in the Neo-
Assyrian texts does encourage the view that the phonemes /i/ and /e/ are in free variation, at least in this
period, and this is certainly the view expressed in M. Luukko, Neo-Assyrian, 40-42, and 87. Luukko makes
the novel suggestion that a solution to the problem of spellings with /i/ > /e/ in Neo-Assyrian might be to
assign “high index number to some of the signs, e.g. te = ti 7 , iš = eš 15 ” (M. Luukko, Neo-Assyrian, 41). This
was kindly pointed out to me by L.R. Siddall (personal communication). 553
A similar phenomenon possibly underlies this orthographic difference as has been noted above, cf. note
and. This variation is noted in A.R. George, Gilgamesh, 425, as a lexical interchange, but as both sources
preserve an ‘i’ class vowel a difference in pronunciation of the same lexeme seems equally plausible.