By Alice Faber, Gainesville
The two major inscriptions in Yaudi Aramaic (KAI 214, 215) contain
between them 8 clearcut instances of the word mt.
1. wmt yqhw mn ydy 'and they took from my hands' (H12)^
2. mt nM y'mr 'and [if] he utters his oath' (H28-9)'
3. whnw mt yS' ydyh I'lh 'bh 'and if he raises his hands to the god of
his father' (H29)
4. plw ntn hdd mt'* 'and may Hadad give' (H13)
5. pbnyt mt 'and I built' (H 14)
6. wbymy 'by, pnmw Sm mt b''ly kpry wb'^ly rkb 'and in the days of my
father, Panamuwa, he (em)placed chiefs of villages and chiefs of
cavalry' (PIO)
7. wmh 'S'l mn 'Ihy mt ytr^ ly 'and what I asked from the gods they
gave me' (H12-3)
8. wytrh mt ml ' msgrt ' and with their remnants he filled the prisons '
(P4)
They syntax ofthe word is unclear (Dion 1974: 298) and none ofthe
"obvious" Semitic cognates ('man,' 'dead', 'country') provides any
help: Kutscher (1976: 58) refers to it as an "orphan," and Gibson
suggests that it is unattested outside of Yaudi. The item is generally
interpreted as an "emphatic" meaning 'certainly, surely' (Dupont-Som-
' I am grateful for a National Endowment for the Humanities Summer Sti¬
pend (1983), which provided partial support for the research reported on in this paper. This paper is an expanded version of a paper read at the 1985 joint ses¬
sion of the American Oriental Society and the North American Conference on
Afroasiatic Linguistics. 1 would like to thank all those who commented on the
oral version and to absolve them of responsibUity for any shortcomings in my
presentation.
^ I use H(adad) for KAI 2U and P(anamuwa) for KAI 215.
^ iL4/treat this sentence as an implicit conditional, an interpretation that is supported by the foUowing eontext.
" The end of this sentence is unclear.
^ Gibson reads here ytnw. While the verbs here are clearly prefix conjuga¬
tion, the context is unambiguously past. See below for discussion.
MER ap^^d Rosenthal 1967; Dion 1974; KAI), which, as Gibson (1975: 72) notes, is a "colourless translation which naturally fits every
context." As an altemative, Gibson adopts the rendering 'always,''
noting a possible etymological link with the common Semitic interroga¬
tive ""matay 'when' and temporal conjunctions like Syriac 'emat 'when¬
ever.' It is the thesis of this paper not only that the 'certainly' interpre¬
tation cannot be supported but that the 'always' interpretation, far from
being merely the best of a bad lot, receives confirmation from a wide
range of data.
The 'certainly' interpretation, although not explicitly given an etymo¬
logical grounding, seems based on a subliminal connection with forms
like Ugaritic imt 'in tmth' and Hebrew 'emedJ To the extent that this
subliminal connection survives the light of day, the 'certainly' interpre¬
tation of Yaudi mt is strengthened. But, if the subliminal connection
caimot be maintained, then attempts to interepret Yaudi mt must expli¬
citly ignore its resonances.
The basis of any attempt to link mt with a proto-form *'amin-t lies in
establishing the plausibility of the phonological derivation that must be
posited. And it is in precisely this attempt that the 'certainly' interpre¬
tation founders. For, while loss ofthe pre-consonantal /n/ is normal for
Yaudi Aramaic (Dion 1974: 107), loss ofthe initial /'/ is not: Like
other Aramaic dialects, Yaudi contains words with prosthetic /'/,
unmotivated on etymological grounds. We find 'Sm 'name' (H16, 21),
'zh 'that' (P2) and, alongside gm,* we find 'gm (P5). Furthermore, the
form hd 'one' cannot be used as a potential parallel for loss of /'/ (pace
Dion 1974: 324), since it cannot be determined on intemal grounds
whether hd stems from '*'hd or from '*whd. This demonstrated implausi-
bility of deletion of /'/ in the derivation of mt decreases the plausibility of interpreting mt as 'certainly.'
The status of a derivation from *matay 'when,' although more com¬
plex (and although confused more than supported by a Mari Akkadian
parallel) is nonetheless less problematical. Throughout the Semitic lan¬
guages we find instances of ""matay-ma' being used as a temporal ad-
' DISO attributes the 'always' interpretation to Poebel.
' This connection is, to my knowledge, made explicit only by Gibson (1983:
60).
' Dion (1974: 176) notes that, while gm is not otherwise attested in Aramaic, it is not found in Phoenecian or Ugaritic either, and, so, must be a retention.
' For indefinites in-ma, see Barth (1967: 170-172) and Hummel (1957: 86,
88). The etjTnological origin of this particular use ofthe enclitic mem is dis¬
cussed in Fabbb (1984).
verb 'always,' or, with a negative, 'never.' Thus, Akkadian matima,
while glossed 'sometimes' in the dictionaries, is generally more appro¬
priately translated 'never' or 'always.' An additional parallel is found in
the Phoenician Karatepe inscription (KAI2Q). Gibson (1983: 60) inter¬
prets mtm following a negative in A II 16-17 of the inscription as
'never.' Linguistically closer to Yaudi we find Xanthos mtwm 'ever' (in
a conditional sentence: 'p hn 'yS mtwm yhn^l mn kndws 'Ih' 'w mn kmr'
nhwy' 'Furthermore, if anyone ever appropriates from King the God or
from the current priest . . .')'" and Syriac mto:m 'ever(lasting), never.'"
Given that the presence of a back vowel in the Xanthos and Syriac
forms can be treated as assimilation to the labial /m/,'^ the only ob¬
stacle to relating these forms to Yaudi mt is the absence ofthe final /m/
in the latter. But this is not, in fact, an obstacle. One ofthe unique fea¬
tures of Yaudi Aramaic is that, while it retains a distinction in mascu¬
line plurals between nominative /-u:/ and oblique /-i:/, these suffixes
do not have the nasal (/m/ or /n/) found in other Semitic dialects, in¬
cluding more prototypical dialects of Aramaic (Dion 1974: 144). Dion
(1974: 336) posits a loss of/m/ in these endings in Yaudi Aramaic, and
I would like to suggest that any such loss would inevitably have affected
the /m/ of '''mataym(a) as well. The presence of orthographic w and y
representing the long vowels in the plural affixes," as contrasted with
the lack of any such indication in mt can be explained either as indica¬
tive of a difference in vowel quality between the final vowel of mt and
those ofthe plurals or as the result ofthe differing functional load ofthe
two sorts of terminations, that is, the contrast between the masculine
plurals /-u:/ and /-i:/ compared with the absence of contrast affecting
the termination of mt.
Ironically, the one piece of evidence that provides the most support
for the interpretation of Yaudi Aramaic mt as 'always, continually' also
provides major difficulties for the chronology implied by the above
account. The relevant piece of evidence is the single use in Mari Akka¬
dian of mati (rather than matima) to mean 'always. ' Because the word is
used three times in a single sentence (mati mati aStasi mati ARM X:
Xanthos 19-22; Dupont-Sommeb apud Metzgeb et al. (1979).
" Rosenthal (1978: 86) notes the puzzling status ofthe Sjrriac form.
The discovery ofthe Xanthos inscription post-dates the attempt by Gins¬
berg (1933: 135) to use the back vowel in the Sjriac form to argue for a Phoe¬
nician source for mlo:m, on the basis of the *a: > /o/ change in that language, and renders it unlikely.
" Kaufman (1982: 155);cf Range (1971:49) who sees here /uw/ and/iy/, with semi-consonantal offglides.
39.7-8, translated without comment as 'Toujours, toujoiu-s je n'ai cesse de prior, toujours'), the possibility ofa scribal error can be discounted.
It might be tempting to see in this form yet another example of West
Semitic influence on Mari Akkadian. Yet given the treatment of Yaudi
mt as an intra-Aramaic development and generally accepted chronolo¬
gies for dialect differentiation in Aramaic and West Semitic (see, e.g.,
Dion 1974: 334), it is unlikely that the Yaudi and the Akkadian forms
could be related. The ultimate resolution of this question must await,
among other things, reanalysis ofthe early Aramaic dialect situation in
the light ofthe recently discovered Tell Fekherye inscription.'"
Although Dion (1974: 298), as noted previously, finds no pattem to
the syntax of mi, one pattem worthy of note emerges from a careful exa¬
mination ofthe data. The verbs ofthe clauses in which mt occurs, vrith
one exception to be dealt with shortly, divide into two groups: unambi¬
guous prefix conjugation forms {yqhw 'they took' H 12, y'mr 'he'll say'
H29, yS' 'he'll raise' H29) and unambiguous suffix conjugation forms
(win '(that) he give''^ H 13, bnyt 'I built' H 14, Sm 'he placed (he named?)'
P 10). There is, in addition, one /y/-initial form, ambiguous as to conju¬
gation (yir" 'was bestowed'" H 12). All three ofthe prefix conjugation
forms are preceded by mt, and all three of the suffix conjugation forms
are followed by it.'* If the ambiguous ytr in H 12 is interpreted as a pre¬
fix form, it, too, is consistent with the generalization: it is preceded by
mt.
There is an additional instance of mt, wytrh mt ml' msgrt (P4; (8)
above), that, at first blush, appears to counter the generalization just
proposed. Here, mt appears to precede the sufiix conjugation verb ml'
'he filled.' While the reading presented in KAIa,nd adopted by Dion,
'with its remnants he filled the prisons,' is plausible, the sentence fol¬
lows a long damaged section at the end ofthe previous line. It is impos-
'" Kaufman (1982: 146n) posits three dialect groups for Old Aramaic: Nor¬
thern (Yaudi), Eastern (Tell Fekherye) and Western ("standard" Old Aramaic).
See also Lipinski (1983) for possible aflfmities of Yaudi Aramaic and Palmjrene among the West Aramaic dialects.
The precative force is provided here by the preceding Iw (cf. pIktSh 'and
may he break him in pieces' in H31).
The Dion and KAI, the latter noting the morphological ambiguity. Gibson reads ytnw, with no change in interpretation. If this reading is correct, the clause including it can safely be grouped with the prefix conjugation forms.
" See Degen (1969: 66) for intemal passives in Standard Old Aramaic.
It is interesting, but, given the limited number of examples, surely not sig¬
nificant, that Phoenician mtm follows a suffix conjugation verb in Ä^/26 A II
16-17: wbl kn mim I . . . 'and the . . . never had . . .'
sible to determine whether the optimal parsing of an ideal, nondegener-
ate text would start a sentence precisely where our degenerate text
picks up. In any case, Gibson (1975) notes the awkwardness of an
'always' interpretation in this particular sentence. Pending a satisfacto¬
ry explanation, which may be precluded by the degenerate state of the
text, P4 cannot be taken as a counterexample to an otherwise valid
generalization.
Now, if the correlation between relative order of mt and its accompa¬
nying verb and the conjugation ofthat verb is not fortuitous, a different
question arises: Whence such a correlation? Any answer to such a ques¬
tion is necessarily speculative and premature. I suspect, but cannot
prove, that the answer is to be found in a study ofthe uses ofthe sufiix
conjugation in Yaudi Aramaic. The correlation is with the form of the
verb: both prefix and suffix conjugation forms occur with past reference,
and it seems to be irrelevant in this matter whether a particular prefix
conjugation form is preterite {yqhw 'they took') or jussive (y'mr 'he'll
say'). Investigation of this possibility is a topic for another paper.
In the interest of completeness, two additional possible instances of
mt must be noted." Dion restores mt in H 12: WByMYhLbTfy m]t YHb
I'lhy 'in my reign one gave to the gods,' but the reading is dubious at
best. KAIre&d WByMYhLbT-t-YHb, with space after the t, which they
do not associate with any word,^" for an additional letter. The verb yhb
is also unclear, but, despite KATs treatment of it as suffix conjugation,
parallelism with the clearly prefix conjugation yqhw 'they took' in the
next clause suggests that it, also, is to be interpreted as prefix conjuga¬
tion. The context, primarily the first person genitive suffix on ydy 'my
hands,' as well as the following prefix conjugation form 'S'l 'I'll ask,'
suggests that yhb should also be first person. Yet, neither the suffix con¬
jugation nor the prefix conjugation interpretation of yhb lends itself
readily to a first person interpretation.^' Of course, the sentence, while
it may under Dion's interpretation support the generalization
advanced here, may ultimately be irrelevant to it. Finally, the last in¬
stance of mt (in H 13) occurs in a context so degenerate that no verb can
be associated with it. Thus, it is unclear whether this token is at all
relevant to the thesis of this paper.
" Af< in P 16, although indexed by KAI&s an instance of our mt, clearly means 'he died,' and is so translated in all treatments of the text, including KAI.
Gibson reads dt, which, if correct, renders further discussion in this con¬
text superfluous.
^' The aftemative interpretation in which yhb is construed as a passive form with an understood indefinite (but third person) subject is equally awkward.
In any case, if the analysis suggested in this paper is valid, Yaudi mt
can no longer be treated as a mysterious, randomly occurring word,
bereft of etymological connections. It means 'always, continually' and
has a clear Semitic etymology in the indefinite adverb "^matay-ma, from
which it developed through unexceptionable phonological processes.
And, although the elucidation of the reasons for its particular distribu¬
tion vis ä vis verb conjugation types must remain an open question for
further inquiry, the fact of its distribution is clear; mt precedes prefix
conjugation forms and follows suffix conjugation forms. As this particu¬
lar "orphan" finds its root(s), we should bear in mind that absence ofa
perceived pattem does not mean absence of an actual pattem; and, we
should plan our investigations accordingly.
References
Arm X. 1978. Correspondance Feminine. Georges Dossin, ed. Paris: Geuthner.
(Archives Royales de Mari. 10.)
Bange, L. A. 1971. Study in the Use of Vowel-Letters in Alphabetic Consonantcd Writing. Munich: UNI-Dr.
Barth, Jacob. 1967. Die f^onominalbildung in den semitischen Sprachen. HUdes¬
heim: Olms, (reprint of 1913 Leipzig ed.).
Degen, Rainer. 1969. Altaramäischer Grammatik der Inschriften des 10-8. Jh.
r. Chr. Wiesbaden: Steiner. (AKM. 38, 3.)
DISO = Charles-F. Jean and Jacob Hoftijzeb. 1965. Dictionnaire des in¬
scriptions semitiques de l'cmest. Leiden: BrUl.
DioN, Paul-Eugene, 0. P. 1974. La langue de Yaudi. Waterloo, Ont.: Editions S. R.
Fabeb, Alice. 1984. Negative Polarity Items in Northwest Semitic. Paper read at the Annual Meeting ofthe North American Conference on Afroasiatic Linguis¬
tics.
Gibson, J. C. L. 1975. Textbook of Syrian Semitic Inscriptions. Vol. 2: Aramaic Inscriptions, Ineluding Inscriptions in the Dialeet of Zenjirli. Oxford: Clarendon Pr.
Gibson, J. C. L. 1983. Textbook of Syrian Semitic Inscriptions. Vol. 3: Phoenecian Inscriptions. Oxford: Clarendon Pr.
Ginsbeeg, H. L. 1973. Ugaritico-Phoenecia. In: Joumal of the Ancient Near
East Society 5, 131-147.
Hummel, H. D. 1957. Enclitic mem in Early Northwest Semitic, Especiaüy
Hebrew. In: JBL 76, 85-107.
KAI = Donnee, Hebbeet and Wolfgang Rollig. 1971—. Kanaanäische und
aramäische Inschriften. 3rd ed. Wiesbaden: Harrassowitz.
Kaufman, Stephen A. 1982. Reflections on the Assyrian-Aramaic Bilingual
from Tell Fakhariyeh. In: Maarav 3, 137-175.
Kutscheb, E. Y. 1976. A History of Aramaic. Jemsalem: Akademon. (in
Hebrew).
20 ZDMG 137/2
Lipinski, E. 1983. The God Arqu-Rashap in the Samallian Hadad Inscription. In:
Arameans, Aramaic and the Aramaic Literary Tradition. Michael, Sokolof, ed. Ramat-Gan, Israel: Bar Ilan Univ., 15-22.
Metgeb, Henri, Emmanuel LaRoche, Andrä Dupont-Sommer, and
Manfred Mayrhofer. 1979. La stele trilingue du Letoon. Paris: Klincksieck.
(Fouilles de Xanthos. 6.)
Rosenthal, Franz, ed. 1967. An Aramaic Handbook. Wiesbaden: Harrasso¬
witz.
Rosenthal, Fbanz. 1978. Aramaic Studies During the Past Thirty Years. In:
JNES 37, 81-92.
Von Klaus Beyer und Alasdair Livingstone, Heidelberg
Zuletzt hat R. Degen: Die aramäischen Inschriften aus Taima' und
Umgebung. In: Neue Ephemeris fiir Semitische Epigraphik 2 (1974), S.
79-98 + Tafel Vllf., die bis dahin bekannten 15 reichsaramäischen
(Degen Nr. 1-10.12.13.15.17.18: um 400 v.Chr.) und vier nabatäi¬
schen (Degen Nr. 11.14.16.19: 2.-1. Jh. v.Chr.) Inschriften aus dem
Gebiet von Taima bis al-'Ula zusammengestellt. In den Jahren 1979 und
1982 sind nun in Taima bei Ausgrabungen des General Department of
Antiquities and Museums Riyadh (Kingdom of Saudi Arabia) weitere
Inschriften zutage gekommen, sechs reichsaramäische, zwei naba¬
täische und eine undatierbare. A. Livingstone hat diese neuen
Inschriften in: Atlal 7 (Riyadh 1983), S. 104-112 -I- pl. 87-97 veröffent¬
licht. K. Beyer hat sie darauf aufbauend noch einmal untersucht. Wir
legen nun hier das Ergebnis unserer gemeinsamen Bemühungen vor.
Die Zeichnungen und die Schrifttafel hat A. Livingstone angefertigt.
Alle genannten Inschriften stammen von Arabem. Das Aramäische
diente ihnen nur als Schriftsprache.
ATTM K. Beyer: Die aramäischen Texte vom Toten Meer samt den In¬
schriften aus Palästina, dem Testament Levis aus der Kairoer Genisa, der Fastenrolle und den alten talmudischen Zitaten. Aramaistische Einleitung, Text, Übersetzung, Deutung, Grammatilc, Wörterbuch, Deutsch-aramäische Wortliste, Register. Göttingen 1984.
Bulletin J. Teixidor: Bulletin d'epigraphie semitique. In: Syria 44 (1967)- 56 (1979), Naclidruck mit Nachträgen und Register. Paris 1986.
Cantineau J. Cantineau: Le Nabateen. 2 Bde. Paris 1930. 1932, Nach¬
druck 1978.
CIS Corpus Inscriptionum Semiticarum. Bd. 2. Paris 1889 ff.
DISO C. F. Jean — J. Hoftijzeb: Dictionnaire des inscriptions shni-
tiqrues de l'ouest. Leiden 1965.
Harding G. Lankester Harding: An Index and Concordance of Pre-
Islamic Arabian Names and Inscriptions. Toronto 1971.
Höfneb M. Höfner in: Götter und Mythen im Vorderen Orient. Hrsg. von
H. W. Haussig. Stuttgart 1965, S. 407-552.
KAI H. Donner und W. Röllig: Kanaanäische und aramäische In¬
schriften. 3 Bde. 3. Aufl. Wiesbaden 1971-1976.
O'Connor M. O'Connor: The Arabic Loanwords in Nabatean Aramaic. In:
JNES 45 (1986), S. 213-229.
20«