(p.281) Appendix 3. Imala in Zamaxshari
(p.281) Appendix 3. Imala in Zamaxshari
It is a general and important issue to determine the degree to which post-Sibawaihian grammarians added significantly to the phonological, morphological, and syntactic data base of Classical Arabic. In this issue I agree with the observation of Carter (1999), that the data base was largely closed after Sibawaih, or shortly thereafter.
The extent to which later grammarians depended on the description of Sibawaih is, however, an empirical question which, as always, needs to be worked out on a case-by-case basis. While studies such as Alhawary (2003), which summarizes the method ology of how early grammarians worked on the basis of reports compiled by later grammarians are interesting, an essential metric is a comparison between the material found in Sibawaih (or Farraʔ or other early grammarians) and later ones. Alhawary (2003: 14), for example, in reporting on Ibn Jinni's (d. 392/1002) elicitation techniques would imply that Ibn Jinni was actually extracting new information. The example he gives, however, an elicitation frame built around d̥arabtu ʔaxaa-ka ‘I hit your brother’, clearly cannot add information about Arabic which by Ibn Jinni's day was not already well known. Ultimately, the only way to know the extent to which Ibn Jinni added new interpretations about Arabic based on new facts is to compare his examples and his analyses thereof, with those of his predecessors.
In this short appendix, I make such a short data comparison, based on the example of imala, discussed in Ch. 7. Phonological phenomena have a physical basis which, in the perspective of this work, can be given a concrete articulatory interpretation, even on the basis of phonetic descriptions from the classical period. It will therefore be apparent whether later grammarians merely mimicked the phenomenon as described by Sibawaih on the basis of his written description, or whether they reWned them and added their own interpretations based on actual aural observations.
In these terms, a comparison between Sibawaih's description of imala and that of Zamaxshari (d. 538/1154) clearly indicates that the former is the case. Zamaxshari in this instance adds little to Sibawaih's observations, and in fact it may be suspected that he based his analyses on written philology rather than on first-hand aural observations, which was a hallmark of Sibawaih's methodology. A summary will make this clear.
Zamaxshari divides his description (pp. 335–8) into fourteen subcategories. That Zamaxshari basically takes over Sibawaih's description is first of all apparent in the description of imala conditioning factors, even allowing for the fact that in the two and a half pages Zamaxshari clearly cannot treat the phenomenon in anywhere near the detail of Sibawaih's fifteen pages. Zamaxshari, for instance, singles out in separate subsections imala in suffixes (such as 3FSG-haa), context-determined imala in bi-baabihi ‘with his door’, the harmonic nature of imala in ʕimaadaa ([ʕimiedie]) (p.282) ‘support’, the imala-inhibiting effects of raised consonants, the effects of /r/ on imala, imala of short [a], and the imala in hollow verbs such as xaaf ‘he feared’. There is only one class, the particles, singled out by Zamaxshari which Sibawaih did not render prominent as a class. All in all Zamaxshari's fourteen subcategories give an adequate overview of imala, though aspects of imala prominent in Sibawaih are Wltered out of Zamaxshari's description. For instance, there is no mention of imala as an individual variational phenomenon. If Zamaxshari was interpreting written words rather than oral signals, this is not surprising.
Zamaxshari's dependence on Sibawaih is furthermore clear in his choice of lexemes illustrating imala. He cites about 110 individual lexemes, most with long [aa], four with short [a]. Of these, twenty-three are not found in Sibawaih, the rest already cited by Sibawaih. Eighty per cent of the actual words used to illustrate imala are therefore identical. I assume that Zamaxshari's naaqif is a printing mistake for Sibawaih's (II: 286) naafiq ‘marketable’. Eight, or one-third of the non-cited total, are found in the longest set of lexemes, comprising forty-four examples in all, namely the context of raised consonants which inhibit imala, examples such as ʕaarid̥ ‘obstruction’ and ullaab ‘students’. Another seven not found in Sibawaih are in the class of particles (e.g. ʕalaa ‘on’, ʔiðaa ‘if’), as noted above not singled out as a separate imala subcategory in Sibawaih. In identifying this imala subcategory Zamaxshari probably follows Mubarrad (d. 285/898; III: 52).
Zamaxshari is therefore clearly dependent upon Sibawaih for his general phonological and lexical description of imala. The one issue of interest is that noted at the beginning of Ch. 7, namely that Zamaxshari perhaps gives imala a different phonetic value from Sibawaih. If this is the case, it could be because Zamaxshari was dependent on Sibawaih's (or other earlier grammarians’) written description, which Zamaxshari interpreted as [ai] (if this is correct; see discussion in sect. 7.1). Alternatively it could be that Zamaxshari in fact transmitted a twelfth-century imala norm which was [ai], or perhaps [ee]. Zamaxshari was from the eastern Arabic region (Iran), closest to today's qultu dialects with the imala value [ee]. It may be that what is interpreted in this book as the original imala value [ia] had died out in the eastern region, and that Zamaxshari was following local norms. As noted in sect. 7.3 (Fig. 7.5), already by Sibawaih's day different, competing phonetic reflexes of imala were already present. Whatever the explanation, Zamaxshari may be ignored for purposes of Old Arabic reconstruction on this issue.