Case Valuation in Transitive Clauses: A Comparative Study of Punjabi and English Syntax

Case is a morphological realization on a noun phrase (NP) to represent the NP's grammatical relationship with the main verb of the clause. With respect to case, languages, in many cases, can be broadly divided into two alignment systems, i.e., ergative-absolutive and nominative-accusative. In the former type of languages, e.g., Punjabi, the subjects usually receive an ergative post position in transitive clauses (with perfective aspect) while in the latter type of languages, e.g., English, the subject, i.e., in nominative case receives, no post position. There has been a widespread controversy on whether ergative is a structural case or a lexical/inherent case and how the arguments are, i.e., subject and objects valued case in case of ergative clauses. With this ongoing debate in the background, this study aims to compare the marking of case on the arguments, i.e., subjects and objects in the transitive clauses of English and Punjabi. The study is conducted under the minimalist framework of Chomsky (2008), who emphasized on Strong Minimalist Thesis (SMT): language provides the best possible solution to the interface conditions imposed by other systems of the human mind, i.e., related to meaning and sound, which interact with language through their interfaces Conceptual Intentional (C-I) and Sensori-Motor (SM) respectively. In this framework, a feature valuation mechanism is induced by the probes, i.e., C and v*. The study finds that in split ergative languages (the languages which take both case patterns, i.e., nominative and ergative) like Punjabi, the EA, i.e., subjects of perfective transitive clauses are assigned the ergative case by the functional heads v* at [Spec-v*] while the IA, i.e., objects are valued accusative case by the same functional head v* under Agree operation. A consequence of this finding concludes that T has default agreement in such languages, which is possible because Punjabi (like its other South Asian counterparts, e.g., Urdu-Hindi, Bengali, and Kashmiri) is a pro-drop language. So, it is easy to assume that EPP and Agree features of T are an option


Introduction
Case can be defined as 'the morphological marking on a noun phrase (NP) that reflects its grammatical relationship to the main verb of the clause' (Baker, 2013: 608). This is reflected by the different forms NPs assume in syntactic constructions. For instance, English pronouns assume different forms in subject and object position. In two mono-clausal constructions: 'I told him.' and 'He told me.' The first person singular 'I' and 'me' and third-person singular 'he' and 'him' assume two different morphological forms. This variation is due to their different case positions i.e. nominative and accusative. The study of case is not new to the study of grammar. Butt and King (2004: 153) highlight that in ancient Greece, the actants of a clause were identified concerning their semantic status as 'accusative 'accused/affected', dative 'giving', or vocative 'called/named'. They further mention that in the Paninian tradition of ancient India, the relationship between the arguments of a clause and their case realization was contemplated according to the semantic roles: 'actor' and 'undergoer'.
These ancient notions appear to persist with us even today as the languages are divided in subject vs. object marking. This distinction divides languages into two broader types i.e. ergative-(absolutive) and (nominative)-accusative (Nayudu, 2008: 61). In split ergative languages, e.g. Punjabi, subject of a transitive clause (in perfective aspect) is assigned case differently from subject of an intransitive clause, subject of a non-perfective transitive clause, and object of a transitive clause. On the other hand, in nominative-languages e.g. English, the subjects of both transitive and intransitive clauses are marked similarly and object of transitive clause is marked differently. Dixon (1994: 9) presents the example of Italian as a nominative-accusative language as evident in the following intransitive construction (a) and (b): a. 'domin-us veni-t' (The master comes) and 'b. domin-us serv-um audi-t' (The master hears the servant). The subject in (a) and (b) bear the same nominative case morphologically realized by the suffix '-us', but the object 'servum' bears the accusative case realized by suffix '-um'. Bobaljik and Branigan (2003) provides example of 'Chukchi' as an ergative language where transitive subjects are marked in one way and intransitive subjects and transitive objects are marked in another way. Nayudu (2008), seeking examples from different language, illustrates that ergativity can be divided in two types i.e. syntactic and morphological. In the former type, certain grammatical constraints treat transitive subject in a different way from intransitive subject and object. Dyirbal is an example of such ergative system. Whereas, in morphological ergativity arguments (subject and object) are assigned case inflections (Latin), adpositions of particles (Tongan) or verbal coreferencing indexes (Sawahili). However, he views that languages cannot be tightly compressed in the compartments of ergative and accusative types. There are languages which show a splitergative system; that is, they utilize both the systems. This split system is determined by different factors i.e. semantic nature of NP, tense, mood, aspect etc. Most of the Indo Aryan languages e.g. Hind, Kashmiri, Gujrati, and Marhati etc. show a split ergative system based on tense/mood/aspect of the clause. In split ergative system, these languages are nearest to Punjabi which is most relevant here.
This study identifies that current approaches to ergativiy, as maintained in some recent works of Woolford (1999Woolford ( , 2017, Legate (2008Legate ( , 2017, Mahajan (2012Mahajan ( , 2017, Bobaljik (1993), Bobaljik and Branigan (2003), and Baker and Bobaljik (2017) among others, do not fit in the more recent developments of minimalist research as explored by Chomsky in last two decades i.e. based on Strong Minimalist Thesis (SMT): the faculty of language provides an optimal solution to the conditions imposed by other systems of human mind i.e. related to meaning and sound which interact with language through their interfaces Conceptual Intentional (C-I) and Sensori-Motor (SM) respectively. The faculty of language is a perfect system if it satisfies the conditions imposed by the interface (Chomsky, 2008: 3). The SMT has simplified the mechanism of narrow syntactic derivation by eradicating many redundant concepts of Government and Binding (G&B) Theory and early Minimalism. A particular problem with most of the recent studies on ergativity is that they still incorporate some redundant concepts which have been eliminated because of their theoretical and empirical insignificance. For instance, Mahajan's recent studies (2012Mahajan's recent studies ( , 2017 still apply case assigning mechanism which is nearer to G&B, if not, it is at least not near to current minimalism. A test case for strengthening the current argument is that if two conflicting but significant studies of Legate (2017) and Baker and Bobaljik (2017) are incorporated in the current case valuation mechanism as stipulated in Chomsky (2004Chomsky ( , 2008, their findings may find it hard to provide a satisfactory solution to some core problems.

Significance of the Study
Identifying some gaps between some recent developments in Minimalist Syntax and the current approaches to ergativity, this study aims to contribute a satisfactory solution to the problem of structural case of objects i.e. Internal Arguments (IA) (an issue which has caused interesting debate in recent years with interesting results) in a complex ergative case system. For this purpose, it intends to compare the case marking mechanism in transitive clauses of English: a West Germanic language of the Indo European family i.e. currently dominant in Britain, America, Ireland, Australia, New Zealand, Canada, and different Islands of Caribbean Sea and Pacific Ocean (Potter & Crystal, 2020) and Punjabi: one of the Indo-Aryan Languages (Shackle, 2017) i.e. mainly spoken in Punjab provinces of Pakistan and India. The rationale for this comparison comes from the fact that Punjabi exhibits a split ergative behavior. In non-perfective clauses it has a similarity with canonical nominative languages like English, while in perfective clauses where the subject is in third person, it allows ergative marking on subjects i.e. External Arguments (EA). This split behavior has been posing problems and controversy among the researchers on the issue of structural case features of IA. Hence, this study aims to explore the phenomenon of ergativity seeking answer to the following research questions under the mechanism stipulated by Chomsky (2008)

Research Questions
 What is the nature of case valuation in English and Punjabi transitive clauses?  What is the mechanism of case valuation in Punjabi ergative clauses?  How are Internal Arguments assigned case in ergative clauses?

Literature Review
In the two systems i.e. ergative and nominative, the former has been more problematic for the researchers of case phenomenon. The reason is that there is no contradictory opinion regarding the status of nominative being structural case. However, in the UG and MP paradigm, researchers are divided on ergative case: whether it is structural or inherent case. The available literature on ergativity can be divided into inherent case approaches and structural case approaches. Woolford (1997Woolford ( , 2006, and many others (Anand & Nevins 2006;Legate 2008Legate , 2017Massam 2006;andMahajan 2012, 2017) view ergative as an inherent case assigned to EA by little v at Spec-v position which is a similar position where the argument is assigned θ-role. These studies, however, differ in further details of case marking to other arguments of the clause particularly to IA. On the other side are the studies of Bobaljik (1993), Chomsky (1993), Bittner and Hale (1996), Bobaljik and Branigan (2006) consider ergative as structural case which finds its locus in CP/TP domain. Following discussion reviews some leading trends of the two approaches briefly. Bobaljik (1993) maintains that ergative-absolutive are structural cases which are assigned in similar way as nominative-accusative by functional heads AgrS and ArgO (Bobaljik labels them as Agr 1 and Agr 2 respectively). According to this mechanism, in transitive clauses Nom/Erg are assigned at Spec Arg 1(S) to subject NP's and Acc/Abs are assigned at Spec 2(O) under the same mechanism. It is in intransitive clause where the ergative-absolutive languages differ from nominative-accusative languages. Obligatory Case Parameter (OCP) helps in deciding whether the subjects in intransitive clauses should be assigned nominative or absolutive. More in line with Chomsky (1995), Bobaljik and Branigan (2003) modify Bobaljik by assuming that in nominativeaccusative languages case features of subject are checked by T, so subject NP moves to Spec-T position. The case features of objects are checked by v. In ergative languages, v can't check the case features of object because it assigns ergative to the subject, so the object moves to Spec of a higher functional head i.e. T.
Another group of scholars (Marantz 1991;Baker 2014Baker , 2015Baker & Bobaljik, 2017), within structural case approach, views ergative as dependent case. They propose that ergative assignment on an NP/DP depends on some other DP/NP, present in the same domain, which has not already been assigned lexical case. This study criticizes the structural and dependent case approach for over simplifying ergative case which involves a cluster of factors. Most of these factors e.g. aspect and transitivity are centered in little v (Chomsky, 2008). Hence, in spite of some problems in overall mechanism, the core argument of inherent case theorist appears sounder. Following is a brief review of literature on inherent case approach. Woolford (1999Woolford ( , 2006 argued in favor that ergative is inherent case, and case can't be separated from agreement. Woolford (2006), providing evidence from Hindi, assumed that transitive v (light verb) assigns case to subject NP at Spec vP position. As soon its case features are checked it becomes unavailable for Agree. In this situation, T (Tense) has to agree with the object i.e. the other nearest NP of the transitive clause. In intransitive clauses, the little v does not assign case to subject NP which remains active for agreement. Hence, this NP moves to Spec T position to get its case features checked and Agree with T. As T licenses case of the intransitive subject NP, T assigns a similar case to it which it assigned to object NP in transitive clause i.e. Absolutive. Massam (2002) favors a similar argument based on evidence from Niuean (a Tongic subgroup oceanic language with VSO word order). Legate (2008) viewed that the IA is assigned case by little v, and the accusative marking on IA is dative instead of accusative. This proposal is problematic in that if v assigns ergative to EA and accusative to IA, how are the 'agree' features on T valued? Mahajan (2017) counters Legate's proposal by arguing that object (IA) is not assigned accusative case, rather it receives nominative case value from T. This study attempts to answer the problematic question how are the unvalued ϕ features of little v satisfied in such a framework where it is assumed that v assigns ergative to EA and T assigns Nominative to IA? The problem lies in that EA, falling outside the domain of v, cannot value the agree features of v.
Butt (2017) maintains that despite being a widespread language, there is dearth of work on Punjabi. She maintains Bhatia (1993) is a standard grammar and Akhtar (1999) is a significant contribution. However, both these works existed before the recent developments is Minimalist syntax wherein more crucial and subtle problems were raised with reference to SMT. More in line with the recent minimalist framework, Khan and Kausar (2019) have attempted to study Punjabi in comparison with English; however, the study was limited to non-finite Tdef constructions. As Tdefhas an incomplete set of ϕ feature, it cannot value the structural case features of an argument NP falling in its domain; hence, that study could not address the issue, which has been more challenging for UG researchers, of valuation of structural case features of EA and IA and unvalued ϕ feature of T and v* (v* is the v with a complete set of ϕ features) in transitive clauses with ergative subjects. On the basis of some strong empirical evidence found from the data, this study attempts to address this more challenging issue.

Methodology
The present qualitative study borrowed methodological framework from Chomsky's (2008) 'On Phases' (Oph) which is based on Principles and Parameters (P&P) approach: languages have, among them, both common universal principlesand varying parametric choices; and 'Strong Minimalist Thesis' (SMT): language provides an optimal solution to the conditions imposed by interfaces. The current framework assumes that features parametrically assemble into lexical items (LIs) which serve as atoms for derivational procedure of language i.e. a computational system which takes 'n' Syntactic Objects (SOs) already formed and convert them into new SOs. Language interacts with the other systems of human mind through interface: C-I and SM which are responsible for interpretation and production of language respectively. These language external but mind internal systems put some conditions which language as a perfect design must meet to fulfill the SMT.
'Oph' provides a simplified mechanism wherein as soon a syntactic derivation reaches a certain 'phase'(i.e. a level at which a syntactic derivation may be transferred to the interfaces: C-I and SM), it is transferred to the interfaces. C (Complementizer) and v* are the phase levels because they own their properties as functional heads while T is not because its properties are borrowed from C. The primary operation of derivational procedure of language is 'Merge' which takes two syntactic objects already formed and constructs new objects from them. 'Merge' is the only operation which comes for free. At the time of 'Merge' some features of the LI's are uninterpretable at the interfaces, hence these features enter into derivation unvalued. For a derivation to converge, these features must be valued. For valuation of these features, the Agree operation is actuated when some functional head: v* or T i.e. enter into derivation with the un-interpretable ϕfeatures, the functional heads become the 'probe' to search for some goal with matching interpretable ϕ-features. The goal is active because it has some uninterpretable structural case feature which gets its features valued from the probe. It is important to note that only a probe having a complete set of ϕ features has the strength to value the structural case features of the goal, so only T or v* can value the structural case features of a nominal. If v* values the structural case of a nominal, the case is labeled as accusative. If T values the structural case features of some nominal, the case is labeled as nominative. The composite operation 'Internal Merge (IM)' is induced by the EF or EPP of functional heads. Under the framework stipulated above, this study intends to comparatively analyze the transitive constructions of Punjabi and English. The data for analysis has been taken from authentic grammars of both languages.
Adopting the mechanism outlined above, this study compares the case valuation mechanism in English and Punjabi. For ergative case parameter, this study assumes the core argument of the inherent case theorists like Legate (2017) and Woolford (2017), among others, that ergative is assigned by little v. However, it intends to provide solution to some problems which their frameworks on ergativity face while providing details of case marking in ergative clauses.
The data consists of transitive clauses with complete argument structure i.e. Internal and External arguments (IA and EA). However, the data is limited to transitive structures which only include argument; the structures involving the phenomenon of adjunction do not fall in the domain of this study. The Punjabi data is collected from Bhatia (1993). The English data is collected from Colloins and Hollo (2000) and Swan and Walter (2003).

Analysis and Discussion
This section presents the comparative analysis of case valuation in English and Punjabi transitive constructions. The major concentration rests on how transitive clauses converge at semantic interface after valuation of un-interpretable/unvalued features under Agree operation induced by C, T and v* probes which search for the nominal goals in their domains. Before the induction of Agree operation, it is assumed that merge comes for free and all constructions are constructed by successive cyclic Merge which takes two already formed syntactic objects SOs to form a new syntactic object SO. Derivation completes in phases i.e. C and v*. T borrows (receives?) its phase like characteristics from C; hence T itself is not a phase head. (1) The schematic descriptions in (1) (a) and (b) taken from Baker and Bobaljik (2017, p.118) (where v o means v*) illustrate the case alignment difference in nominative and ergative languages as shown in (1a) and (1b) respectively. It is evident from (1a) that the 'merge' of V and IAforms VP which merges with v* results in v*P. The EPP features of v*allow the merge of EAat Spec-v*. This merger results in c.

c. v * P[EA v * VP [V IA]]
The further merger ofv*P and T results in TP. It is important to notice that θ-roles are assigned to arguments by the heads they initially merge with. It implies that θ-roles are assigned at the time of merge so there is no need of some extra mechanism for θ-role assignment, and it does not make any difference for nominative and ergative languages. However, (1a) shows that the IA and EA are not assigned their respective case by the same heads they initially merged with at the time of merger. It implies a separate mechanism for derivation to proceed if case features are to be valued for the nominals, i.e. IA and EA. The uninterpretable features of v* probe for the nearest goal in its domain which is only IA in a transitive construction. Although, the EA seems nearer than IA to v* on the left side, but EA does not fall in the structural case marking domain of v* which finds IA as the nearest goal. In nominative languages, like English, v* probe Agrees with IA goal to value the unvalued probe; hence the goal receives accusative case value as structural case. The goal EA Agrees with the probe T to value its unvalued features of T and the goal i.e. EA is valued Nominative structural case. Configuration (1b) demonstrates a very significant difference: the EA receives ergative case from its θ-role assigning head i.e. v* at the θ-position. Being inactive for case features, the EA does not come in the way of T to induce long distance Agree operation with IA to value the unvalued features on both sides. Hence, the IA receives nominative case.

3) C[[They T [They v* VP[grow sugarcane]]]
In construction (2), schematically descried in (3), the verb 'grow' merges with the nominal (IA) 'sugarcane to form VP. According to (1a), the V assigns θ-role to the IA at this stage. The v*P is formed by the merger of VP with v* which allows the merge of EA i.e. 'they' at Spec-v* by virtue of its EPP features. This can be seen in (3a) below.

(3a) v * P [They v* VP [V grow sugarcane]]
The un-interpretable ϕ features of v* are unvalued which must be valued before the derivation is mapped to C-I interface. The unvalued features of v* probe some goal with matching ϕ features to value the unvalued features. In its domain it finds the IA 'sugarcane' as an active goal which has unvalued structural case features. Hence, the functional head v* induces Agree operation with the goal IA for obtaining the feature valuation mechanism. In this Agree, the probe values its ϕ features and the goal values its structural case features. It is important to note that the EA 'they' cannot Agree with v* as the former does not come in the domain of the latter after Chomsky's (2001) stipulation that there is no m-command. Therefore, the Spec of a head is not in the domain of the head. The v*P so formed merges with T which merges with C to form a transitive CP. As in the domain of v*P, the Agree operation was induced by v*, in the higher domain i.e. CP, the locus of operations will be the phase head C. T derives the Agree features from C and having a complete set of ϕ features searches for a goal to value all the features. The EA 'they' comes under the domain of T, hence the Agree is induced between probe and goal to value the unvalued feature on both sides. To satisfy the EPP features, the EA is raised to Spec-T under a composite operation. In this way, all features are valued and the derivation is ready to be interpretable at C-I interface.

5) C [I T [I v*[phone you]]].
All structures with complete argument structure which possess a v*-v with complete set of ϕ features are formed in the way described for (2). Another instance of mechanism which holds for transitive clauses in languages like English can be seen in (4), schematically described in (5). The difference between (4) and (2) is that the former contains an overt T auxiliary while the latter does not. In (4), like (2), a probe and goal relation is established between v* and the IA 'you'; and between T and EA 'I' valuing all unvalued features. The IA 'you' is valued accusative case by v*while the EA 'I' is valued as nominative by T.
Unlike uniform nominative marking in transitive structures of English, Punjabi exhibits split behavior regarding case. There are grammatical situations in which Punjabi observes canonical nominative behavior (1a), but in perfective aspect it observes ergative case alignment pattern (1b). There are two phenomena which affect ergative case marking i.e. aspect and person. In perfective clauses Punjabi allows ergative case marking on subjects; however, this is restricted to third person subjects. There is no overt realization of ergative on first and second person subjects. Whether it has a covert realization is an important question yet to be explored. Construction (6) and (7) exhibit that in usual cases the perfective clauses allow ergative marking on subject while non-perfective clauses do not. For non-perfective clause it is obligatory that they do not allow ergative case on subjects; however, perfective clauses may or may not allow ergative on the subjects. Apart from aspect, agentivity, and person, the selection of light verb v* plays a crucial role in determining ergativity on subject. There are certain light verbs which bar the occurrence of ergative on the subject. For instance, the perfective clause (6) allows ergative on subject. It is evident that (6) contains no overt v*. The same construction (6) can occur with different light verbs. Constructions (8) to (13) below instantiate this phenomenon.
Construction (8) and (9) select a transitive light verb litta (from the root 'lena' to take) while construction (12) and (13) select an intransitive light verb bathiaa (from the root 'bathnaa' to sit). In case of the former light verb ergative is allowed; however, in case of the latter light verb, ergative causes ungrammaticality, as it can be noticed in (13). A clarification is needed in the case of (10) and (11) (thanks to an anonymous reviewer of this paper for raising this question) that 'Os' is not similar to 'O' as a subject. It is evident from (8) to (11) that ergative is obligatory in the case of 'O, but non-obligatory in the case of 'Os' which allows both cases. For this reason, it is assumed that Os is not in the nominative case even if it does not allow an overt ergative case. This leads to the possibility of a null ergative realization which may be addressed in detail in a separate study; however, the possibility of both cases on 'Os' is enough here for stipulation that it is not similar to a nominative subject 'O' in Punjabi. This difference in case alignment as manifested in (8), (9) and (12) (13) strengthens Legate's (2008Legate's ( , 2017 stipulation:v* is the locus of ergative.There are some v*s which allow ergative while the other do not. The constructions (14) to (17)  Hence, ergative is an inherent case marked at the θ-position by the functional head v* at the moment of merge. This mechanism of case marking is different from valuation of accusative and nominative (structural cases) which are not assigned at the θ-position and need the stipulation of a different mechanism. Construction (14) and (16) also exhibit the split Ergative and Nominative alignment, respectively, in Punjabi.
In (18), the IA (NP) 'xat' merges with 'likh' forming VP which merges with v* 'litta' forming v*P which merges 'O-ne' at Spec-v*. 'O-ne' is the EA. As the light verbs selection requires ergative marking on the EA, according to (1b) both EA and IA are assigned θ-roles of agent and theme by their respective heads at the moment of merge. However, at θ-position only EA can receive ergative case by v*: the head which also assigns θ-role. The unmarked IA still has structural case features unvalued. How they are valued is an important question for this study. There are two significant proposals from inherent case theorists. Legate (2008) proposes that IA is assigned case by v*. This argument is countered by Mahajan (2017) who provides evidence from Hindi data that IA is not valued case by the v*, rather it is valued nominative by T. In the current mechanism, it is possible to assume that T values the structural case features of IA under long distance Agree operation. After the raising of EA to Spec-T position for the satisfaction of the EPP features of T, there is no intervention caused by the EA to hinder the long distance Agree relation between the T and IA. It is easy to assume this mechanism for (18), but the dispute of Nom/Acc for IA remains to be settled for the Punjabi data which includes structures where the IA bears overt accusative marking. The case marking in (19) is similar to canonical Nom/Acc alignment of English according to (Ia). So, it is easy to assume without any confusion that EA and IA are valued their structural cases by the heads T and v* respectively. In the following discussion the mystery of case valuation of IA in ergative constructions is further probed into for structures where IA overtly manifests accusative case. Ergative construction (20), schematically described in (21), exhibit an overt accusative marking on IA 'mariiz-nu' which poses problem in simplified assumption that IA is assigned Nominative by long distance Agree operation. According to Legate (2008), the IA 'mariiz-nuu' is in dative instead of accusative case. Under current minimalist framework Legate's assumption poses problems. If IA is also in inherent case, how can the unvalued features of T and v* valued. Before stipulating any solution to this problem, it is important to note that the ergative case is not assigned by v* under the usual Agree relation by which the unvalued features of the probes: T and v* and the goal NP's are valued. The ergative case is marked to EA argument which does not fall in the domain of v*, so it should be more near to θ-role assignment, in mechanism, than structural case assignment. A very significant question arises that whether the assignment of ergative case values un-interpretable features of v* or not. In Oph framework, it is not possible because EA is not in the domain of phase head v*. It has a further consequence that the unvalued features of v* must be valued by some other head. The analysis of (20) makes it easy that v* values its case features by Agree with IA valuing the structural case features of goal nominal 'mariiz-nu' as accusative. In light of this evidence, this study proposes (22).
The functional head e.g. v* is not rendered inactive for assigning structural case features under Agree operation with some nominal goal.
A question arises, as a consequence of (22), about the Agree features of T., Chomsky (2001) suggests that in case of quirky (dative) case, it may be assumed that T has default agreement if a particular language allows it. This can be combined with the assumption of Ura (2006) that T may have default agreement in case of pro drop languages like Hindi. Both the proposals can be applied to the case of Punjabi which is a pro-drop language. According to Butt (2001: 2) generally languages spoken in South Asian possess pro-drop feature i.e. they may drop any (or) all arguments. She particularly provides evidence form Urdu-Hindi, Punjabi, Bengali, and Kashmiri. In this observation, Ura's proposal is also applicable to Punjabi. In case there is no goal for T to Agree with, it is not difficult to assume that the T has default agreement. In case of T, not only the Agree but EPP features are also optional in South Asian languages like Punjabi. When there is no unmarked argument, as evident in the case of (20), it is assumed that T has default agreement. A more visible example of default agreement may be seen in (21). The rationale of optional EPP feature comes from Butt's observation that South Asian languages, including Punjabi, never take expletive subject as they have the ability to drop the subject. So, EPP and Agree features are optional for T.
In light of the empirical evidence, provided in this section, this study rejects Mahajan's (2017) and Woolford's (2006) claim that IA is assigned/valued case by the functional head T. The claim of Legate (2017) that accusative marking on little v is in fact dative rather than accusative is also not compatible with current case valuation mechanism stipulated by Chomsky (2008). If it is assumed that the object in mono-transitive construction is in Dative instead of Accusative, the uninterpretable ϕ features of v would remain unvalued and the derivation would not converge at C-I interface. It can be assumed that T has default agreement in light of evidence provided by the data analysis of this study and some already available stipulations in literature e.g. Ura (2006) and Chomsky (2000) which allow default agreement on T in case of quirky case marking on arguments i.e. ergative or dative.

Conclusion
The study concludes that because of split ergative alignment Punjabi behaves like English in nonperfective transitive clauses with just the natural difference of preposition in English and postposition in Punjabi. In such Punjabi clauses, the IA Agrees with v* which values accusative case features of the former and unvalued ϕ features of the latter. The EA Agrees with T which values nominative case features of the former and unvalued ϕ features of the latter. However, this study rejects a widely assumed claim, e.g. in Mahajan (2017) and Woolford (2006), that in ergative clauses of Punjabi the IA is valued case features by T under long distance Agree operation, rather, on the basis of empirical evidence to strengthen the phenomenon, it assumes that the IA Agrees with v* which values accusative case features of the former and ϕ features of the latter.
This study also concludes that the EPP and Agree features of T, which are derivative from C, are not obligatory. The optional nature of these features is derived from two facts: One is that pro-drop would not be possible if the functional head T had obligatory EPP features; secondly, it is assumed that agree features are also default because the unvalued ϕ features of little v are valued by the IA. The EA already assigned ergative case is also not active for being a goal. Hence in the presence of evidence like pro-drop, absence of expletive, and absence of an active goal in the domain to satisfy Agree features, it is assumed that in Punjabi T has default Agree features in Punjabi like most of it South Asian counterparts e.g. Hindi and Urdu.