Object Omission and Verb Classes in Persian

An alternation is a pair of sentences with identical structures and meaning. Transitivity alternation, on the other hand, is a type of grammatical alternations in which the verb is used with a different number of arguments without a significant change in meaning. Participating in specific alternations is considered as a criterion for categorizing verbs in semantic classes. For investigating this hypothesis, and in order to obtain more accurate information about the syntactic behavior and semantic properties of the verbs which can omit their objects, 435 cases of context independent object omission extracted from different sources such as books, magazines, movies, series, and everyday conversations were studied. The findings show that the verbs which participate in specific alternations, have identical semantic properties and can be categorized in one semantic class. Accordingly, the semantic behavior of the verb can be predictable based on its syntactic behavior. An interesting fact about the verbs which participate in same alternations and belong to same semantic class is that, the kind of objects these verbs take, have common features as well. The syntactic behavior and semantic characteristics of the investigated verbs and their objects reveal that, a verb can omit its object, when the object is general, indefinite, unspecific, and its meaning is predictable for the hearer.



1. Introduction

The notions of “verbal-valency” and “valency-changing processes” have been investigated in many reserches in recent years. Transitivity, is connected to the notion of valency that refers to the number of arguments required for a given verb to form sentences. Natural languages have developed means to go from transitive to intransitive and from intransitive to transitive con‌structions. Understanding the way languages encode alternations in the transitivity of verbs is a key to understand numerous and wide-ranging phenomena in the syntactic, semantic and morphological behaviour of languages (Kageyama and Jacobsen, 2016, p.1).

Object omission construction is a valency-reduction process in which a transitive verb appears without its object and is used as an intransitive one. However, the addressee is able to understand the meaning of the objectless sentence.

Transitivity alternations, on the other hand, are dual syntactic structures that change the transitivity of verbs. In these alternations, the verb will be used with different number of arguments without significant changes in meaning.

Transitivity has traditionally been explained based on the number of syntactic arguments a verb takes. It is clear that omitting one of the arguments of the verb will change the degree of transitivity and transformes the transitive verb into an intransitive one. Therefore, this type of alternation is called “transitivity alternation”. In the current research, different types of transitivity alternations that are related to the object omission construction will be investigated.



2. Literature Review

Many linguists like Goldberg (2001), Velasco and Munoz (2002), Luraghi (2004), Tsimpli and Papadopoulou (2006), Zyzik (2006), Liu (2008), Graf et all. (2015), Cennamo (2017), Eu (2018) and Glass (2020) have considered context-independent object omission in different languages. These studies reveal that the object of activity verbs which is unspecific and indefinite, especially when is used in generic or iterative contexts, or in the framework of any types of structural omission, and when the interlocuters have the same knowledge of the word, can be omitted.

Levin's research, on the other hand, can be considered as the most comprehensive research on transitivity alternation, which has presented different semantic categories for English verbs based on the alternative behavior of each verb. However, her research was accompanied by shortcomings, among which we can point out the lack of adaptability of the verb classes presented by Levin in other languages, the determination of syntactic criteria for the classification of verbs, and the lack of semantic coherence of the examples. Accordingly, her findings, especially in typological studies, were not sufficient and more studies were needed in order to be able to examine the topic under discussion from a semantic point of view and provide the cross-linguistic comparisons (Comrie et al. 2015, p. 5).



3. Methodology

Many researchers have tried to provide semantic classifications for verbs based on transitivity alternations, and Levin's research (Levin. 1993) can be considered as one of the most comprehensive one. In her research, Levin has presented 79 verbal alternations and 49 semantic classes for English verbs. What is examined in the present research under the title of “object omission” is introduced in Levin's classification under the title of “unexpressed object alternation” and includes subclasses.

In this study, 1010 transitive verbs were extracted from Dehkhoda, Moin and Amid which are three comprehensive and well-known Persian dictionaries. Then, the possibility of omitting the object in each case was examined. Ultimately, a database including 435 cases of object omission construction was obtained. After selecting the appropriate cases of context-independent object omission in Persian, the verbs which can participate in this construction were classified based on some related transitivity alternations.



4. Results

The findings of this research show that the verbs that were classified into separate semantic classes based on their participation in the investigated transitivity alternations have common semantic components and their objects have common characteristics. Hence, the common semantic features of the verbs of each class can be considered as a factor for predicting their syntactic behavior. This classification of verbs, make it possible to get familiar with the verbs of Persian language more, and predict their syntactic behavior. This classification of verbs helps us to study object omission construction, and also take a step towards providing computer models and creating the verb network (verbnet) of Persian language.