Abstract

The tense driven asymmetry of the Pashto clause is analyzed from the perspective of the minimalist framework. The study proves that the split ergativity in Pashto is tense based and does not have the aspect driven features proposed by Roberts (2000). The study argues that the object is assigned a theta role by the V and the subject is assigned a theta role by the little v. The accusative case is assigned by the little v but the nominative and ergative cases are assigned by T. It claims that the T head assigns multiple cases as the split ergativity is tense driven. It highlights the syntactic effects of the possible phonological processes in combining some of the closely adjacent words and making a single phonological word. The study also discusses clitic placement and prosodic inversion to refute the assumption that perfective feature is a strong feature in Pashto.

Key Words

Tense Driven Asymmetry, Split Ergativity, Compound verb, Perfective and Imperfective aspects.

Introduction

The present study aims to investigate the Tense driven asymmetry in the Pashto Language. All world languages have transitive and intransitive clauses, and in intransitive clauses, there is a verb and a single NP, but the transitive clause has two NPs (Tallerman, 1998). These NPs are known as the core arguments of transitive and intransitive verbs. The word order of these can be distinguished according to their core arguments. Most of the world languages have a basic constituent order which is unmarked. The basic constituent order of the English language is SVO. In the English language, the function of the constituents is known by the placement of the constituents in the word order. 

1. Ahmad gave her an apple (S V IO DO) Ahmad gave-PST 3SG.F an apple Ahmad gave her an apple.

2. She gave Ahmed an apple (S V IO DO) F.3SG gave-PST Ahmad an apple She gave Ahmad an apple.

It is clear from the above examples that the function in the English constituent order is determined by the place of the constituent in the clause. The NP constituents in (1) and (2) have different semantic roles inside the clause. According to Tallerman (1998), there are three ways in which the relationship of the NP arguments, with their verbal predicate can be determined. The first has been discussed in (1) and (2), where the English core NP arguments have a strict NP position according to their function within a clause. On the other hand, 

According to Babrakzai (1999:13), the order of the basic and neutral constituents in a Pashto sentence is Subject – Adverb - Indirect Object - Direct Object – Verb – Aux. The Pashto word order within a clause is very flexible, and variation of the order of the constituents in a sentence depends on functional constraints, the speaker’s attitude, and the possible available syntactic system. The possible NP arguments function with the verbal predicate inside the clause has been given in the following (4) sentences. 
I. [Ahmad] [Men -e la] [saib][warkro] (S IO DO V)
II. [Ahmad] [saib] [Men-e la][warkro] (S IO DO V)
III. [Men -e la] [Ahmad] [saib] [warkro] (DO S IO V)
IV. [Men -e la] [saib][Ahmad] [warkro] (DO IO S V)
V. [Saib] [Men -e la] [Ahmad] [warkro] (DO IO S V)
VI. [Saib] [Ahmad] [Men -e la] [warkro] (DO S IO V)
In Pashto, the relationship of Pashto NP constituents with its verbal predicate is determined by the case marking system; it has nothing to do with the NP position in the clause. Pashto has six possible word orders which have the same meaning of the NP arguments with the verbal predicate. Examples in 4 (I - II) represent the canonical orders or unmarked constituents’ orders, and the rest of the sentences are good examples of marked constituents' orders. 
Roberts (2000) argues that free constituent orders are very frequent in the past tense. The main reason for this free variation is the ergative case system in the past tense. The following sentences (5) and (6), cited in Roberts (2000), are the same in meanings and function, but exhibit a different NPs position within the sentence:
1. Spi pisho khug kr -a dog.OBL.M cat.DIR.F hurt do-PST.PFV -F.3SG
SOV: The dog hurt the cat (John Muhammad, P.C)
2. Pisho spi khug kr n -a cat.DIR.F dog.OBL.M hurt do-PST.PFV -3.F.SG
OSV: The dog hurt the cat (John Muhammad, P.C)
But even the same constituent in the present tense cannot move freely inside the clause. The main constraint is the case system, where in the past tense, there were two case systems: oblique and direct, but in the present tense, it has the same case system. The following example was cited in Robert (2000). Compare (6) with (7) to see the difference that OSV constituents’ structure is ungrammatical:
3. *pisho spay hug-aw -I cat.DIR.Fdog.DIR.M   hurt-PRS.IPFV-3SG
SOV: The dog is hurting the cat/OSV *the cat is hurting the dog
A violation of the movement of the NP constituents in the present tense with the same case resulted in the ungrammatical sentence. In Pashto, tense and case at the same time are constraints on the constituent’s placement in the clause with its verbal predicate. The word order in example (3) is flexible, but Pashto remains strictly a verb-final language. 
 There are three absolute tenses such as the present, past, and future. A language can distinguish between these three tenses by using separate morphology for every tense. It is also possible that a language may use the same form for two or more tenses. There is a future/non-future opposition when a language combines past and present tense, for example, Mao Naga (Bhat, 1999). If there is a combination of present and future tense in a language, called past/non-past opposition such as Kannada (Bhat, 1999). If a language presents a combination of all three tenses, then it is a tenseless language.
When event time overlaps with the moment of speech then the present tense is used. It is hardly possible for event time and speech time to become identical if it is about performative verbs, otherwise, event time and speech time normally overlap but they are not identical. According to Haan (2012), the most unmarked tense is present tense cross-linguistically. It has the meanings which are not temporal in nature in a strict sense. For instance, the habitual aspect as presented in the example given by Comrie (1985).
John goes to work at eight o’clock (every day).
The moment of speech and event time in the above example does not have to overlap here. However, it can be said that such types of sentences refer to habits and as such, they are true at the situation time. The past habitual aspect would not be true at situation time. For instance, the universal sense, universally true statements such as ‘violets are blue, and roses are red’. Such kind of statements includes the situation time.
The past tense can refer to many interpretations, so, it is a bit complicated. It refers to events or actions that take place before the moment of speech, however, there are other aspects of past tense as well. First, it is about the truth, such as the past is fixed, so, it can be thought that the past events are certain but this is not the case. An example of English can be taken in this regard like many other languages where a past tense can be used for modal notions. Another matter regarding the past tense is if the action completely took place before the moment of speech and so no longer holds at or even after the situation time. Conventionally, this is left open and is dealt with as an implicature. In some languages, the past events do or do not hold for the present, so the presupposition seems to be grammaticalized. Comrie (1985) has given examples of certain Bantu languages in this regard. It is important to note that it is different from the case where the past actions have relevance for the present. It means that the event or action may be over but its effects or outcomes can still be felt. In many languages, even minute differences can be made by adding degrees of remoteness in the tense system. Remoteness distinctions are usually made in the past rather than in the future by a language and they are around two or three usually. There are no remoteness distinctions in the present tense. However, some languages have more distinctions regarding the past tense such as the dialects of Bantu language Bamileke (Comrie,1998).

Pashto Verb

Morphologically, the Pashto verb is more complex in nature than other Pashto grammatical categories. The Pashto verb is marked for tense, aspect, mood, transitivity, gender, person, and number. The verb in the Pashto language also provides information about nominative and accusative arguments and has a placement for clitics (Babrakzai, 1999). In Pashto, the agreement of the verb with its NPs depends on tense. In the present tense construction, the verb shows an agreement with its subject, and in the past tense, the verb shows its agreement with the object. In compound verb construction, the participle form of the verb is conjugated with the auxiliary for tense, aspect, and agreement. There are two auxiliary verbs in the Pashto language, intransitive keg, and transitive kaw. When the intransitive auxiliary keg is used as a full verb it means ‘become’, and the transitive auxiliary kaw as a full verb means ‘do’. When these auxiliaries function as full verbs, they take ‘be’ form of the verb in the present tense, and in the past tense; the keg is changed into the ked.




Conjugation in Pashto Verb

In Pashto, a participle form of the verb takes a copula to carry the inflection features of tense, aspect, and agreement. This conjugation is limited only to the main verb and its inflection with a copula, as shown in (8): 

Intransitive verb kinast?l  ‘to set’

a. Past tense

z? kinast?l -?m I was sitting down.

t?kinast?l -e you were sitting down.

mung kinast?l -u we were sitting down.

taso kinast?l -y you were sitting down.

da kinast?l -a she was sitting down. 

day kinast? he was sitting down

dui kinast?l they were sitting down

kinast?l -e they (F) were sitting down




Nominal Verb

In Pashto, the nominal verb, or verbal noun, which is equal to the English gerund, rather than infinitive, is the dictionary form taking al marker at the end (Babrakzai, 1999). It is regarded as basic stem (augmented), based on the past, and all other forms are derived from it. 

These nouns can function as an argument of the verb. It may function as subject, direct object, and indirect object in a sentence. In certain constructions, it can be marked for the case (Babrakzai, 1999).  It may be seen in (9):

Khath lekal ‘to write a letter’

a. khath likal nim mulaqat wi letter writing.NOM half meeting.ACC COP.PRS.IPFV.3SG Letter writing is like half meeting chay skal ‘to take tea’

b. z? chay skal gwar-am1SG.NOM tea drinking.ACC want.PRS.IPFV.1SG I want to drink tea (Babrakzai, 1999)




Verb Classification

Tegey (1996) has classified the verb according to its three different classes. He distinguished the three classes as a simple verb, derivative verb, and doubly irregular verb. He described the verbs according to perfective and imperfective aspects. In Babrakzai (1999), the verb has been described according to its transitive and intransitive nature. Roberts (2000), in his study on Pashto clitics, has focused on the verb structure according to aspects, stem variation, and in the complex predicate, its function within sentence.




Complex Verb

Pashto language, like other Indo-Iranian languages such as Urdu and Hindi, takes complex predicates. In Tegey (1996), the compound verb is known as derived verb where adjectives and nouns are combined with the transitive and intransitive auxiliaries to make a compound verb. In Babrakzai (1999), the verb has been divided into transitive and intransitive verb according to its function and thematic role in the sentence. He has classified the different form of the verb as light verbs and inchoative verbs. He has discussed in detail how the adjectives and nouns form compound verbs with the transitive and intransitive auxiliaries. 

Babrakzai (1999) has differentiated between intransitive ‘Inchoative verb’ and ‘light verb’. The inchoative verbs are derived from stative or adjectival stems, where the subject is affected by the event. The aspect plays a role to show the change in the subject with the help of the intransitive auxiliaries:




Inchoative Verb

Imperfective

a. gwaha pah-eg-I Meat cook-PRS.IPFV.3SG The meat is cooking.




Perfective 

b. gwaha pah-a shw -a meat cook -F become.PRS.PFV –F.3SG The meat has cooked  




Light Verb – Inchoative

A light inchoative verb is made up of a verbal element or nominal element. Babrakzai (1999) has differentiated two types of light verbs. In the first group, the nominal element functions as ‘subject’ of the auxiliary, and the second group it functions as object to the intransitive auxiliary.  If a predicate triggered another argument, then it would be in the oblique form

Dltha d?  footbal lube keg -I ADV.PROX of   footbal.OBL play.NOM become.PRS.IPFV.3SG Here soccer is played 

(Babrakzai, 1999, p. 134)

In the second group of the light inchoative verb, the verbal element and the intransitive auxiliary make a compound verb. In this group, it is functioning as single predicate, and that is why it takes another noun which functions as subject:

Ghnam rebale ked -al wheat.NOM thresh become.PST.IPFV –M.3PL Wheat was being threshed. (Babrakzai, 1999, p. 134)




In (12), the verb agreement is triggered on the subject ghnam ‘wheat’, the other two elements, rebale ‘thresh’, and kedal ‘to become’, function as compound verbs. In the first group, as in (11), the relationship between a nominal verb and the intransitive auxiliary is not closer as in (12). 




Transitive Verb

The transitive verb in Pashto is of great importance because the verbal agreement is triggered by the tense with the NP arguments. In the present tense, the verbal agreement is marked on the nominative subject, but in the past tense, the agreement is marked on the absolutive object. 


a. ahmad kitab -una gur -I ahmad.NOM   book  -PL.ACC see -PRS.IPFV.3SG Ahmad is reading books.
b. Ahmad kitab -una kat -al ahmad.ERG book  -PL.ABS see -PST.IPFV.3PL Ahmad was reading books.
In (14a), the agreement is triggered by the inflectional –i, and the Nominative subject in the present tense. On the other hand, in (14b), the verbal agreement is marked on the direct absolutive object kitabuna, ‘books’. Here, the tense also shows the split ergativity as the transitive subject is marked as ergative case in the past tense, and the subject of the present tense in (14a) is marked with nominative case. 
In Pashto, there are two types of transitive verbs: one is known as simple transitive and the other as derived transitive. The simple transitive verb is that type of verb which consists of a finite verb and an agreement marker and is not derived from the stem. The past tense is marked with the past tense marker al, except for the third person singular masculine (Babrakzai, 1999).
a. Mine hapal las wenz -o meena.ERG own hand.ABS wash.PST.IPFV.M.3SG Meena was washing her hand.  (Babrakzai 1999)
b. Mine hapala hpa wenz –al  – a meena-ERG   own. F   leg.F.ABS wash -PST.IPFV -F.3SG Mena was washing her leg.
(Babrakzai, 1999, p. 137)


Derived Transitive Verb
In Pashto language, the process of causativization is a rich source of new entries of verbs. Most of the transitive verbs are derived from stative or adjectival stems (Babrakzai, 1999). It has the same function of deriving verbs from the adjective as the intransitive verbs are derived. For example, with the adjective ‘difficult’ gran, when the intransitive ed auxiliary is suffixed, then we have an intransitive verb gran-ed-al,’ to become difficult’. When the same adjective stem gran functions with the transitive suffix aw, then the derived verb becomes like gran-aw-al, ‘make something difficult’. The derived transitive verb tells about the change or the effect on an object, that is why, very often it takes the inanimate object:
a. Ustaz -a lahth -ay mat-aw -I teacher -PL stick -PL break –do.PRS.IPFV -M.3PL the teachers are breaking the sticks.
b. Ustaz -an -o lahth -ay mat -aw -ale teacher -PL -ERG stick -PL break –do.PST.IPFV -3PL The teachers were breaking the sticks.

Pashto Clause Structure

The clause consists of three different layers, the thematic layer, the CP layer and the inflectional layer (Rahman, 2014). The inflectional layer shows the functional projection in the clause like TP, DP, Agrp and CP etc. The thematic layer of the clause shows the thematic (semantic) aspect of the clause, built around VP of the lexical verb. The third layer, the CP domain is the highest layer of the clause which is also known as left periphery. The present study describes all the three layers of the clause as Pashto has clitics which are used in complementary distribution with the full NPs and in some environment co-occur with these NPs. This clitic placement sometimes affects the structures of the clause and the wh-word in relative clause for example invert minimally to the left of the complementizer and shows complementizer as head final in the CP. But this left periphery is not movement as such but minimal inversion of a word which is a common feature in Pashto in other constructions too. But it will not be discussed in detail as it is not the focus of the present study, see Robert (2000) for a detailed discussion of left periphery in Pashto.

Pashto unlike other languages of the region shows tense driven asymmetry. The verb in Pashto agrees only with the nominative NP and not with ergative NP. Ergativity in languages is related to several conditions including tense, aspect, mood and the semantic natural of the verb or the NP. The verb in Pashto does not agree with ergative marked NP either in past or non-past. This analysis is different from Roberts (2000) where he is of the view that ergativity in Pashto has features for aspect driven ergativity like other Indo-Aryan language likes Hindi/ Urdu. But the following examples show that ergativity is tense driven not aspect driven in Pashto (Rahman & Bukhari, 2014). 

a) Ta xat like 2.SG. Nom letter.SG. Nom write PRES.SG.NOM You are writing a letter.

b) Za Kirkay Matom 1.SG. Nom windo.SG. Nom break.PRES.NOM I am breaking the window

c) Ma Kirkay Matawala 1.SG.ERGwindow.SG.NOM. brack. PST. IMPF.NOM I was breaking the window.

 d) Taa ba Kirkay Matakaryi we 2.SG. Erg Fut window SG.NOM break.PST.NOM be.PST.PF You would have broken the window. 

Pashto Ergative case is structural because it depends on the tense as it is tense, and transitivity driven not by the thematic role that much. But the use of dative case instead of Ergative for experiencer role too shows it to be inherent to an extent. Bobaljik (1993) considers that Ergative case in south Asian languages is structural and according to him the ergative case depends on the perfective aspect while in Pashto it depends on tense only. But as far as the inherent, nonstructural property of the ergative is concerned the ergative has an inherent case (Ergative).

Pashto Complex Verb and Tense Driven Asymmetries

The T assigns a nominative case to its subject in the non-past which is the unmarked direct case in Pashto. The little v assigns the theta role to the subject which if it is agent role in the non-past then it is assigned the nominative case and if it is agent role in the past then takes the ergative case. But the case is assigned by T not by ASP head unlike Urdu Hindi and Gojri because ergativity in Pashto is not aspect driven but tense driven. The accusative case in the non-past and absolutive in past is assigned not by little v but by T as agreement for both subject and object is tense driven and not aspect driven but the little v can assign the thematic role to the object which indicates whether the object will get accusative or nominative case later in the derivation. So, instead of using (+IMPF) for Pashto, (±Past) for Pashto case assignment should be used.  So, when the T has the value (-Past) the NP in spec vP is assigned nominative and when [+past] then it is assigned ergative caseThe nominative and ergative cases are assigned in the following way. 



The verb merges with the object and form VP. The VP then merges with the little v to form v?. This v? merges with the subject to project vP. The little v here assigns the theta role to the external argument (subject).  If the theta role assigned by the little v is agent then the subject is assigned either nominative or ergative case because both may be the subject. If the theta role is the experiencer role for example, the subject may be assigned dative case (by putting a preposition or postposition with the NP which itself is always in oblique form for dative). The vP merges with Asp to project Aspp. The AspP may have the (±PERF) features but as the nominative and ergative case marking is not aspect driven because the past imperfective for example bear ergative case, so the condition for assignment are not met here and the derivation proceeds further. The Asp merges with T, which has the features [+past] and if the verb is in the past and the NP has agent role, it is assigned ergative case. But if it is [-past] and the NP is agentive, it is assigned nominative case. And if the NP for example, has the experience role and either it is past or non-past, the NP is assigned dative case. So, the subject gets its unvalued [case] facture valued from T either ergative, nominative or dative. The rules of agreement and case assignment are given below. 

1. If the little v assigns agent role to its specifier and the tense is [+past] the subject is valued ergative by T.

2. If the little v assigns agent role to its specifier and the tense is [-past] the subject is valued nominative by T. 

3. If the little v assigns a role other than agent to its specifier, dative case is assigned to the NP. 

4. Accusative case is assigned by the little v to those subject NPs, which are marked [+Hum] and /or [+DEF].

5. The NP that values the phi-features of the T is assigned Nominative case. 

6. If a clause has more than one nominative NP, the verb agrees with the highest NP (Under the locality of agree).

7. If a clause does not have nominative NP which can value the phi-features of the T, the verb takes the default marker (?) for past imperfective and (?) for perfective.

Conclusion

Marking Tense and aspect on verb is complex. The Pashto verb is marked for agreement with subject in the present tense and on object in past tense. Tense plays an important role to determine the agreement of the verb with its arguments (NPs) at subject and object position. Here, the tense also shows the split ergativity as the transitive subject is marked as ergative case in the past tense, and the subject of the present tense is marked with nominative case. The Pashto compound verb in the imperfective aspect works as a single unit but in perfective aspect it functions as two unites. The study has focused the structure of Pashto clause in the light of the minimalist program. The study reveals the case assignment and agreement feature in the language. It shows that the ergativity in Pashto is tense driven not aspect driven. The ergative case is mostly structural with inherent characteristics of agent role but not necessarily a volitional reading on the part of the agent. The nominative and accusative cases are assigned by the T and the accusative by v. The simple and complex predicates behave similarly in terms of case and agreement of the verb with the object under the general consideration of the tense for split ergativity in the language. 

References

• Babrakzai, F. (1999). Topics in Pashto syntax. (Unpublished doctoral dissertation). University of Hawai'i at Manoa: Manoa, HI.
• Bobaljik, J. D. (2003). Eccentric agreement and multiple case checking. In J. A. Massam, D. and Ndayiragije, J. (Eds.), Ergativity: emerging issues, Dordrecht: Springer.
• Bukhari, N. H. (2009). The Syntax of Serial verb Construction in Gojri. (Unpublished doctoral dissertation). Newcastle University: UK.
• Bhat., D. N. S. (1999). The prominence of tense, aspect and mood (Studies in Language Companion Series 49). Amsterdam/Philadelphia: John Benjamins.
• Bybee, et al, (1994). The evolution of grammar: Tense, aspect, and modality in the languages of the World Publisher: Chicago: University of Chicago Press.
• Comri, B. (19985). Tense. Cambridge: CUP
• Haan, F., D. (2012). Evidentiality and Mirativity. Oxford Handbooks.
• Khan, A. A. Khalid, A., Rahman, G. (2020). The Integration of Embedded Language Islands in PashtoEnglish Bilingual data: Hamdard Islamicus, ISSN: 0250-7196, Vol. 43. No. 1 (2020), 439- 459.
• Khan, A, A. Anees, M. Rahman, G. (2020). The Patterns of Code-Switching in Pashto-English Bilingual Data. Global Social Science Review; Vol. V, No. I (Winter 2020), p- ISSN: 2520-0348; e-ISSN: 2616-793X; ISSN-L: 2520-0348; Page: 123 ‒ 133
• Khan, A.A. Bukhari, H.N., Khalid, A., Abbas, N.S. (2019). The prediction of morphemes distribution at surface and abstract level: the case of Embedded Language nonfinite verbs Pashto-English Bilingual data, Kashmir Journal of Language and Research: ISSN-1028-6640. Volume 22 N
• Khan, A.A. Bukhari, H.N. & Khalid, A. (2018). Pashto-English Bilingual data: Testing the Diagnostic Features for the Patterns of Codemixing. Kashmir Journal of Language and Research: ISSN- 1028-6640. Volume 21 NO. 1. 39-59
• Khan, A. A. Khalid, A. (2018). Pashto-English Code switching: Testing the Morphosyntactic Constraints of the MLF Model. Lingua-Elsevier 201, 78 - 91
• Khan, A. A. & P. Muysken. (2014) Strategies for incorporating nouns and verbs in code-mixing: the case of Pashto-English bilingual speech. Lapudum 18 (special issue on bilingualism and language contact), P. 97- 137.
• Khan, A.A & Bukhari, H.N. (2014). Descriptive Study of Compound Verbs and Light Verbs Constructions in Pashto-English Code Switching. Kashmir Journal of Language and Research: ISSN- 1028-6640. Volume 17 NO. 2. 2- 14
• Rahman, G, (2014). Verbal Clitics in Pashto. (Unpublished doctoral dissertation). Azad Jammu and Kashmir University: Muzaffarbabad, Pakistan
• Rahman, G. & Bukhari, N. H. (2014). Case System in Pashto: Kashmir Journal of Language Research, 17(1), 173-193.
• Roberts, T. (2000). Clitics and agreement. Cambridge, MA: MIT dissertation.
• Tegey, H. (1975). A Study of Pashto clitics and implications for linguistic theory. Studies in the Linguistic Sciences 5(1), 154-196.
• Tallerman, M. (1998). Understanding syntax. London: Arnold.
• Tegey, H. (1977). The grammar of clitics: Evidence from Pashto and other languages. (Unpublished doctoral dissertation). University of Illinois at Urbana: Champaign, USA.