88 holger gzella
Singular dual plural
masc. abs. /Ø/ (-yn /ayn/) -n44 /īn/
masc. cstr. /Ø/ -y /ay/ -y /ay/
masc. emph. -ʾ / āʾ/ (> /ā/) — -yʾ/ayyāʾ/ (> /ayyā/)
fem. abs. -h /ā/ (< /(a)t/)45 (-tyn /tayn/) -n /ān/46
fem. cstr. -t /at/ -ty /tay/ -t /āt/
fem. emph. -tʾ /tāʾ/ (> /tā/) — -tʾ /-ātāʾ/ (> /ātā/)the gender of a noun in Semitic can be known from its agreement with
adjectives and verbs; not all nouns that behave like feminines in concord
have the corresponding ending though: unmarked feminines include nbš
/nabš/ ‘self ’ (Kai 222 B: 39) and many place names; judging from later
and comparative evidence, body parts that naturally come in pairs, like
yd /yad/ ‘hand’, are also feminine, but direct evidence from the earli
est texts is lacking. Sometimes singular and plural take opposite gender
endings, e.g., mlh /mellā/ ‘word’, plural mln /mellīn/ (Kai 224: 2), or šnh
/šanā/ ‘year’, plural šnn /šanīn/ (Kai 222 a: 27). a few nouns have differ
ent bases in singular and plural (some of them by means of expansion or
reduplication), e.g., ʾb /ʾab/ ‘father’, cstr. plural ʾbhy /ʾabahay/ ‘fathers of ’;
br /bar/ ‘son’,47 cstr. plural /banay/ ‘sons of’ (Kai 222 a: 2 and elsewhere);
by(t) /bay(t)/ ‘house’,48 cstr. plural bty /bāttay/ (< */bayatay/?) ‘houses of ’
(Kai 202 B: 9); rb /rabb/ ‘great’, plural rbrbn /rabrabīn/ ‘great ones’
(Kai 216: 10 and elsewhere). at times this coincides with distinct gen
der marking, as in the cstr. plural nšy /nešay/ ‘women of’ (Kai 222 a: 41;
but the absolute plural nšwn /nešawān/ with the corresponding feminine
44 the tell fekheriye inscription has some instances of a plene spelling {yn} with this
ending (e.g., ʾlhyn /ʾelāhīn/ ‘gods’ in Kai 309: 4 or šʿryn /śaʿārīn/ ‘barley’ in l. 19, but šʿrn
in l. 22), although defective spelling of the absolute masc. plural remains dominant even
in official aramaic.
45 a remnant of the older ending /(a)t/ seems to survive in šʾt /θa ʾt/ ‘sheep’ (Kai 222
a: 21; otherwise spelled šʾh in Kai 215: 6, 9), but the context is broken. another alleged
example, mrmt ‘treachery’ (Kai 224: 22), by contrast, could be better analyzed as an abso
lute fem. plural (see hoftijzer – Jongeling 1995: 694, s.v.).
46 exceptionally, a vestige of older /āt/ may have been preserved in mln lḥyt ‘evil
words’ (Kai 224: 2), perhaps because this was a formulaic expression.
47 the vowel differs in cuneiform transcriptions and may originate from an indistinct
central vowel due to syllabic /r/ (unless these spellings reflect different words).
48 it remains unclear whether by in by ṭb ‘a fine house’ (Kai 216:16) is an early attesta
tion of the shift /bayt/ > /bay/ of the absolute form of this word in aramaic (so, among
others, Beyer 1984: 530) or a socalled “sandhi writing” for expected byt ṭb due to assimila
tion of dentals in a stressunit beyond wordboundaries (degen 1969: 43). however, the
only clear example for sandhi writing is the personal name brkb /Barrākeb/ in Kai 215:
19 instead of the usual brrkb. comparable instances appear, again with names or filiations,
in the earliest phoenician inscriptions (Kai 6: 1; 7: 3), hence this phenomenon may have
been inherited from there.