Morphological Stem: a Separationist Concept

Morphological stem is defined as the form of a lexeme which plays the role of the base in morphological process in the lexeme-based morphology framework introduced by Aronoff (1994). According to this model, the base is a lexeme and the stem is some form of the lexeme. Indeed, there are many bound morphemes considered as bases/stems in Persian morphological processes which do not denote specific meaning and are not registered in the lexicon, however they belong to a lexeme. The question then arises: “what are they”? The present study tries to put Aronovian framework to the test in order to examine how much this theoretical framework can account such morphological facts in Persian. From this perspective, morphological stem is considered as a separationist entity because it acts only as a phonetic form without regard to its meaning and syntax in morphological processes. In this article, this separationist aspect of the stem is studied in order to show if the separation of the stem from its semantic and syntactic aspects can explain a number of mismatches between form and meaning of the stems in inflected and derived words. The absence of a systematic one-to-one correspondence between form and meaning in the elements of complex words is one of the linguistic facts which challenges Morpheme-based Morphology. In creating complex words, while morpheme-based­ approach maintains a concatenation of elements into a linear sequence on the syntagmatic axis (with a one-to-one correspondence between form and meaning pertaining to every element), the lexeme-based approach, however, insists on the separation of the form from meaning. The analysis of data in this article has been carried out in line with lexeme-based morphology; a number of Persian derived and inflected words as well as a group of verbal compounds were included in data for this purpose. After analyzing every word, it was observed that the bases of many Persian derived and inflected words as well as some compounds, are not real lexemes; actually they turn up in morphological processes without which these processes cannot operate. Some instances of this type of stem include the morphological stem ''xasteg'' which appears during the derivation of ''xastegi'' (boredom) by means of adding suffix –i to the stem; the stem ''xasteg'' is not an existing Persian lexeme but it belongs to the lexeme ''xasteh'' (bored); ''xasteg'' does not have anytransparent meaning, but it occurs as a phonetic form or one of the morphological stems of this lexeme in this derivational process due to phonological context. The other example of morphological stem is ''săx'' which appears in inflection when past tense marker suffix (–t) is added to the stem, the output of which is ''săx-t'' (made/ constructed). The inflectional stem ''săx'' is not a real free morpheme, but its presence is inevitable in inflection because of phonological assimilation with the following past marker –t. In a considerable number of verbal compounds, the same form can denote several meanings, that is, several syntactic/semantic features maps on to one morphological form or morphological function. This function is neither syntactic nor semantic but rather purely morphological; morphology by itself as Aronoff names it (1994); he calls the level of such purely morphological functions ''morphemic''. For example, some Persian verbal compounds including ''danesh ămuz ''(student/agentive) ,''dast ămuz ''(pet/past participle( and "bad ămuz " (cause to teach a bad habit/ causative) ,all of which derived from one verb stem in present tense, the same form of the verb "ămuz'' (teach) combines with its arguments to build a construction which refers to several semantic and syntactic properties including agentive noun/adjective (so called “sefat-e fa'eli” in Persian), past participle (so called “sefat-e maf'uli” in Persian) as well as causative concepts. Indeed, exactly the same form for agentive nouns/ adjectives may correspond to other concepts like past participle or causative meanings. So, the form "ămuz'' is a purely morphological stem regardless of its syntactic and semantic features in word formation processes in Persian. Therefore, the presence of such forms in a lot of Persian words shows that these elements are not haphazard as they may seem, but they should be related to a level of language system at which phonetic forms of the elements in derivation, inflection and compounding are important, rather than their semantic and syntactic aspects; this level is called '' morphological spell out'' in Aronoff's lexeme-based morphology, these types of stems are ''morphemes'' which belong to spell out level. Based on the separationist hypothesis, the morphological spell out of this type of stem in morphological processes is independent from its semantic/syntactic aspects. In sum, this separation can account for morphological asymmetries such as one-to-many and many-to-one relationships between form and meaning in stems in a significant number of Persian derived and inflected words.