Statistical Evidence for Mood Functions of ‘be-’ in New Persian: A Diachronic Study

So far in the mood system of New Persian, the inflectional element ‘be-’ has had three grammatical roles: subjunctive marker, imperative marker, and future marker, all of which having in common the feature of unrealized occurrence of the event (as opposed to its realized occurrence for indicative mood). This investigation, based on a sampled corpus of 55 texts, selected from 10^th to 20^th centuries, follows the changes of these functions in New Persian diachronically. The results show that in a period of eleven centuries the frequency of ‘be-’ as a marker of subjunctive has increased three times, and as the imperative marker twice-and-half, while it has been productive in its third role (future marking) mostly in 10^th to 12^th centuries, and this function gradually disappeared later. Furthermore, since 19^th c., ‘be-’ which was formerly used with simple verbs almost exclusively, began to be used with complex predicates as well. However, there is some amount of irregularity in the increasing frequency of ‘be-’ in the whole period, and the author attempts to explain it with such hypotheses as grammatical archaism in some texts, and the possible change of grammatical forms in the process of copying the texts



1. Introduction

A precise survey into the gradual development of grammatical markers is to follow the changes in their frequency patterns statistically, relying on the number and percentage reports. This paper attempts to fulfil this task for the marker ‘be-’ in the mood system of New Persian (from 10^th century onwards). The goal the research, then, is to show the grammaticalization path of the marker and its mood functions throughout the intended period up to the present time. A restriction of the research, however, is that the source of data has been limited to written texts, even for contemporary Persian.



2. Literature Review

The first distinction in the mood systems is that of indicative vs. imperative. This distinction is morphologically shown in Persian by imperative zero agreement marking (in the singular form) as opposed to overt marking of indicatives (for the same form), and also by the indicative ‘mi-’ as opposed to ‘be-’ that its emergence is discussed in this paper.

The third distinctive form in the mood system of Persian is subjunctive which is controversial among grammarians, both in its definition and instances. Natal Khanlari (1986: 2/306-339) includes in the category all present and past subjunctives as well as some forms suffixed with ‘-i’, and excludes conditionals and optatives from the category. By contrast, Ahmadi Givi (2001) subcategorizes conditionals and optatives as subjunctive, to the exclusion of past forms with ‘-i’ as well as past progressives. On the other hand, Mazaheri et al. (2004) define the subjunctive category so broadly that it also includes forms such as ‘bayæd mi-ræft’ (he/she had to go) and ‘šayæd ræfte æst’ (he/she may have gone).



3. Methodology

The corpus of this research includes 77 thousand verbs, extracted by the author from 55 texts from 10^th to 20^th centuries. To the sampling purpose, five texts from each century were selected, and then, two parts of each text were adopted randomly, each one amounting to 700 verbs (1400 verbs from each text, altogether). These verbs were labelled for tense, mood, and the structure of the verb (simple, prefixed, or complex), determining whether they had ‘be-’ for mood marking or not.

This research follows the definition given for the subjunctive in Ahmadi Givi (2001), and the kinds of subjunctive included by Darzi and kwak (2015) and Ilkhanipour (2017). Accordingly, all instances of irrealis (including past progressives with ‘hæmi/mi-’, and past forms with ‘-i’) were excluded from the statistics, and all of the semantic types of present subjunctive (probability, obligation, optative, conditional, etc.) were included.



4. Results

The results show a gradual increase in the imperative and subjunctive functions of ‘be-’, while the future marking function is mostly active in 10^th to 12^th centuries, and disappears gradually afterwards. Also, as shown in Figure 1, the imperative marking function of ‘be-’ has always been, and still is, ahead of its subjunctive marking function.



Figure 1

Grammatical functions of ‘be-’ by century





Furthermore, the results show an imbalance with regard to the morphological structure of the verb: the verbs taking ‘be-’ before 19^th century were almost exclusively of a simple root. Simple positive verbs reach a rate of 80% and 98% for subjunctives and imperatives, respectively, in 20^th century. Complex verbs, however, began to take ‘be-’ increasingly in the last two centuries (19^th and 20^th), reaching the average rate of 28% and 30% for subjunctives and imperatives, respectively, in the last two texts of the corpus. Prefixed verbs are still very low in their rate of taking the marker.



5. Discussion

The first issue to be discussed here is that in the whole corpus there are no instances of ‘be-’ detached from the verb. This shows that from the very beginning of New Persian period, ‘be-’ is functioning as a bound morpheme, having started its grammaticalization process well before 10^th century. The second observation in a holistic approach is that ‘be-’ in the territory of complex predicates is a relatively recent phenomenon, having increased since 19^th century. This late development can be due to the competing nature of the elements that occupy the preverbal position (i.e. non-verbal elements, derivational prefixes, and other inflectional markers) which puts them in a complementary distributional relationship. Thirdly, the status of complex predicates in contemporary Persian can probably be explained with resort to the partial syntactic independence of non-verbal element which allows for the insertion of ‘be-’. This can be compared to derivationally prefixed verbs for which the prefix resists ‘be-’ insertion, with some exceptions that can be signs of an emerging phenomenon.



6. Conclusion

According to the statistics, ‘be-’ has been in a grammaticalization path in the whole New Persian period, being still expanding its territory in the past decades (i.e. increasing its frequency for non-simple verbs). In other words, the generality and obligatoriness of ‘be-’ for its mood functions has been increasing in the whole period under discussion, and it is still making progress