Synthetic compounding: a syntactic or morphological processing? A corpus-based study

The present article studies the structure of synthetic compounding to discover the process or processes creating this productive category. Compounding has played crucial and leading roles in creating words along with progress in various fields and it has the ability to meet the needs to produce words for new phenomena in most languages, particularly Persian. In other words, compounding is a very productive and important word formation process. Compound words are of two root and synthetic types. Root compounds are made of two or more morphemes which have meanings in isolation. Moreover, syntactic head of root compounds are not derived from verb (Xabbaz, 2006). Synthetic compounding is formed by the incorporation of argument, adjunct, adjective or syntactic phrases into a verb stem, and the syntactic head of this structure is derived from the verb. The non- head element may be an argument, an adjunct, an adjective or a syntactic phrase. Verb-stem constituent as head, non-head constituent and derived affixes are three essential elements of synthetic compound structures. A derived affix is of two overt and covert types.
Previous studies on the interactions between compounding on the one hand, and morphology and syntax on the other hand have discussed three approaches. Some linguists maintain that compounding is a lexical process and a compound is comprised of two or more simple words. This view is related to a period of generative grammar in which morphology and syntax were completely detached from each other. Other linguists claim that compounding is a syntactic category made of applying transformational rules to deep structure. The approach that was in contrast with the above paradigms was introduced by linguists such as Xabbaz (2006), Giegerich (2009) and Lieber (1992 & 2010). They believe that compounding is placed in the interaction of two morphology and syntax modules. That is, this category possesses both syntactic and morphological properties. So far, no research has been carried out on synthetic Persian compounds and their relation with morphology and syntax. Therefore, regarding Persian corpus, it is essential to examine these three approaches to clarify whether compounding is a syntactic or morphological process.
In order to know whether synthetic compounding is syntactic or morphological, this category is examined in terms of 16 properties; 8 properties are syntactic and 8 are morphological. Synthetic compounds behaved syntactically in 8 cases, morphologically in 5 cases and both syntactically and morphologically in 4 cases. Regarding the nature of synthetic compound's behavior, it mostly holds syntactic properties, although it possesses morphological characteristics as well. In other words, this productive category is arranged on a continuum which is composed of two syntactic and morphological extremes and synthetic compounding inclines to the syntax extreme. As synthetic compounds share more common properties with syntax, their behavior is mostly similar to syntactic phrases, and this is the main reason that they cannot be easily distinguished from phrases.
Comparing 8481 synthetic compounds of this article with 3974 root compounds of  Khabbaz (2006), both being extracted from Soxan dictionary of Anvari (2002),  shows that this category is two times more productive than root compounding. The investigation of syntactic and morphological properties of synthetic compounding indicates that it cannot be merely a syntactic or morphological process. Studying synthetic compound words according to the morphological and syntactic criteria (non-pronominal reference, word formation process, irregularity, recursiveness, constituent structure, constituents’ relations, coordination and gaping, wh-word, productivity, aspect, morphological and syntactic integration, constituent expandability and non-expandability, semantic opacity and transparency) demonstrates that synthetic compounding is a morpho-syntactic process in Persian which possesses more syntactic characteristics. Moreover, transparent, opaque, expandable and nonexpendable compounds and attaching inflectional morphemes to both the whole and the first constituents of compounds prove that this category has both morphological and syntactic structures simultaneously. Regarding these features, synthetic compounding in Persian holds the following properties:
1-Synthetic compounding is non-pronominal reference. That is, pronoun can't be replaced by compounding constituents.
2- No constituents of synthetic compounds can be fronted.
3- The output of compounding process is word formation.
4- Compounding is recursive.
5- The relationship between compounds' constituents is similar to the relation which exists between clauses' constituents.
6- Compounds' constituents can be coordinated. The constituents can be related to each other with conjunctive words.
7- Synthetic compounds possess aspect such as clauses.
8- Not all synthetic compounds have morphological and syntactic integration.
9- Synthetic compounds are both semantically opaque and transparent.
10- Synthetic compounds are extendable and non- extendable entities.