The Patterning of Root Morphemes in Semitic

Abstract

It is an obvious, though little-noted fact, that the characteristic triconsonantal verb morphemes of Semitic languages (the traditional triliteral roots) do not ordinarily contain identical first and second consonants. On the other hand, a pattern of identical second and third consonants is of frequent occurrence, con-stituting the well-known geminate subtype of Semitic verb. Thus, while sequences such as *mmd are virtually non-existent in Semitic languages, Arabic mdd 'to stretch', frr 'to flee' etc. are representatives of a common Semitic type. The existence of this degree of patterning led to the present investigation of the over-all patterning of the triconsonantal verb morphemes of the Semitic languages, particularly Arabic. The most general conclusions are stated here by way of anticipation in order to orient the reader in the detailed discussion which follows. 1. In the first two positions, not only identical but homorganic consonants are excluded. For example, no Semitic language has triconsonantal verb morphemes beginning bm-, since this would involve two labials, or gk-, since such a form would contain two velars in the first and second positions. The lack of gemination in 1-11 is therefore seen to be a special case under the rule referring to homorganic consonants in general. 2. Homorganic consonants are likewise excluded in positions two and three, though not quite as rigorously as in the first two positions. The rule for positions two and three does not preclude identical consonants, as W!'J have seen, so that it should be rephrased as referring to homorganic but not identical consonants. Thus in Arabic, we have skk 'to split' but *skg, containing nonidentical velars in positions two and three could not occur. The geminate type is thus clearly an anomaly in terms of the overall patterning of Semitic verbal roots. 1 3. In positions one and three there is marked, but less rigorous exclusion of homorganic, including identical consonants, than in other combinations of po-sitions. Thus there are few instances of the general type of Arabic qlq 'to be disturbed', with identical first and third consonants or Syriac prm 'to tear' with homorganic, in this case, labial, consonants in these positions. The concept homorganic requires some consideration at this point. The ar-ticulatory positions arrived at inductively by the present study of pattern phe-nomena agree well with the statements of the early Arab grammarians regarding the maxrag (plural maxa:rig) literally 'place of egress'. Perhaps the most strik-ing instance is the classification of r, l, and n as homorganic on the basis of the 1 This evidence can be considered relevant to the traditional controversy regarding the former biconsonantal form of the Semitic triconsonantal roots. That during the Pre-Semitic period, the device of gemination was one of the methods of forming triconsonantal roots from former consonantals is probable. On the other hand, the existence of geminates in Egyptian suggests that some geminates go back to the Hamito-Semitic period; cf. also such etymologies as Egyptian tmm equals Proto-Semitic *tmm 'to finish'. 162