My Thoughts on Case Theory (December 2025)

Below are my current thoughts about Case Theory as of November 2025. These thoughts result from my reviewing the relevant literature during Fall 2025 in order to teach the subject in Syntax I. My thoughts on Case Theory are evolving and they are not set in stone.

I thank Lydia Grebenyova for discussions of these issues.

1. History

There are four historical trends in Case Theory in the last 50 years:

(i) Vergnaud’s 1977 letter developed subsequently by Chomsky in LGB,

(ii) The minimalist theory of case as a reflex of phi-feature agreement 

(Chomsky 2000, 2001, usually presented as a development of (i)), 

(iii) Dependent Case Theory, developed by Marantz and subsequently by Baker and others, 

(iv) The idea that case heads a syntactic projection KP which can be stacked with other KPs. 

The theory is in a state where these trends have not led to any natural synthesis.

2. DPs need to be licensed.

There are many examples where DPs appear in positions that are either theta-positions or in a chain with theta-positions, but are unacceptable. The examples are easy to generate, and fall squarely in classical Case Theory. 

For example, consider the sentence: ‘I wonder whether to leave.’ This is synonymous with ‘I wonder whether I should leave.’ But it is impossible to have an overt subject in the embedded clause: *‘I wonder whether Mary to leave.’ What is the problem with this sentence? Why doesn’t it mean ‘I wonder whether Mary should leave.’? Here, ‘Mary’ could easily be interpreted as an argument of ‘leave’, but the sentence is not acceptable. A similar kind of example is the obligatoriness of the prepositional complementizer in ‘*(For) him to leave would be nice.’ More distantly related examples include: ‘I am proud *(of) John’. Here the ‘of’ is necessary, but does not seem to play any semantic role in the sentence. Under the theory of the passive in Collins (2024), the by-phrase is in a theta-position (Spec vP), and ‘by’ is semantically vacuous, but it cannot be omitted: ‘The book was written *(by) John.’ All of these examples are ruled out straightforwardly by the Case Filter (see below).

These examples all have the following form:

(i) The relevant DP occupies a theta-position or is in a chain with a theta-position.

(ii) The sentence has a plausible semantic interpretation.

(iii) The DP is not in a position standardly associated with case 

(e.g., nominative, accusative, genitive).

(iv) The unacceptability of the sentence can be explained in terms of the position of the DP in a non-case position.

It is quite easy to generate such examples. The ones presented above are just the tip of a very deep iceberg. A fascinating area of current investigation is the extent to which these constraints hold cross-linguistically (e.g., in Bantu).

3. Morphological Case

In some languages (e.g., Latin, Russian, Japanese), DPs require overt morphological case markers.

4. Licensing and Morphological Case

The fundamental insight of Vergnaud is that points (2) and (3) are related in some way. A DP needs to be licensed, and it is licensed by having case, and this case shows up overtly in some languages. I take this brilliant insight to be on the right track. See Pesetsky and Torrego 2011 for a recent defense.

For this reason, I draw no distinction between abstract Case (upper case letter) and morphological case (lower case letter). For uniformity, the word is always in lower case in these notes (except in expressions like ‘Case Filter’, ‘Case Theory’ and ‘Dependent Case Theory’).

5. Case Filter

All phonologically overt DPs need case. 

Some covert DPs do not have case (e.g., the implicit argument of the passive, see Collins 2024). I assume PRO needs case, because of well-known facts in Icelandic. For nouns like ‘house’ in ‘house building’ or ‘house builder’, if they are not DPs, they do not need case.

6. Merge and Case

Case morphemes are introduced syntactically by Merge. 

The idea that a head such as DP can be assigned an optional case feature which is then realized as overt morpheme (in some languages) runs contrary to MaS (Morphology as Syntax), since morphemes are only introduced by Merge.

No morphemes, including case morphemes, are introduce post-syntactically. More generally, there is no such thing as dissociated or ornamental morphemes (concepts from Distributed Morphology). They are all introduced syntactically with Merge.

7. Case morphemes head KPs.

As in Caha’s famous treatment of syncretism in Armenian (Caha 2010, 2013), I take Case morphemes to head KPs. The hypothesis that case morphemes head KPs has not been adequately explored. It could lead to very rich consequences for parametric variation.

8. Lexical Case is a KP c-selected by a Lexical Item

Lexical case (including quirky case) is a KP c-selected by a particular lexical item in a particular syntactic configuration (e.g., Spec-Head). For example, quirky subject case in Icelandic is a KP c-selected by the light verb v that introduces it. At least in some instances, ergative case is a lexical case licensed in Spec vP (as in Legate).

In some languages, quirky case KP and structural case KP might be stacked with only the lower quirky case KP being realized phonologically. See Collins 2025a on the actually mechanism of spell-out.

9. Agreement and Case

In Minimalist syntax (e.g., Chomsky 2000, 2001), case is assigned as reflexive of agreement in phi-features. For example, when finite T agrees in phi-features with a DP, that DP has its nominative case feature valued as a ‘reflex’.

While this works well for finite T, there are many kinds of case in English that do not seem to be accompanied by agreement in phi-features (e.g., accusative case assigned by the preposition ‘for’). 

Rather, it may be that case assignment itself is a kind of agreement between a head and a DP (see also Grebenyova forthcoming). From Pesetsky and Torrego 2011: “P&T suggested that the process otherwise called ‘case assignment’ is in fact nothing but Agree applying so as to value an otherwise unvalued T-feature on a nominal. When nominative case is assigned to a nominal expression, they argued, what is actually happening is the valuation of an unvalued T-feature on the nominal by a valued counterpart on Tense itself…”

10. Nominative Case 

Nominative case is the result of a syntactic relation between a DP/KP and finite T.

11. Genitive Case

 Genitive case is the result of a syntactic relation between a DP/KP and D.

12. Accusative Case

Accusative case is the result of a syntactic relation between DP/KP and P or transitive little v.

13. Default Case

In some languages (e.g., English), accusative case is also a default. That means that it can be licensed by little v or P, but in some exceptional cases is also licensed as a default, as in coordination: ‘Me and Kenny left early.’ In this sentence, there does not seem to be any little v or P that can license the accusative ‘me’. Similarly, the accusative case forms are used as single word answers to questions. 

Default case is still a mystery and needs to be rethought completely.

14. Burzio’s Generalization 

Burzio’s generalization: If a verb assigns accusative case, then it takes an external argument.

Burzio’s generalization follows from independent principles. There is no reason to stipulate it. 

15. Agreement and Case (Revisited)

I propose that agreement occurs when phi-features are merged into the syntactic structure and enter into Form Chain with interpretable phi-features of a DP. Here is the process (see Collins 2025):

(i) Form phi-feature bundle (similar to pronoun or clitic).

(ii) Merge bundle into syntactic object.

(iii) Form Chain applies between the phi-bundle and the interpretable features of DP.

It follows that agreement exist, and there is no need to stipulate an Agree operation, nor is there any need for unvalued features. Rather, agreement follows from the basic building blocks of minimalist syntax (Merge, Form Chain).

I hypothesize that case exists for the same reason that agreement exists. An explanation of one will count as an explanation of the other. Case assignment/checking occurs when case features are merged into the syntactic structure and enter into form chain with the features of a syntactic head (e.g., T). It follows that case assignment/checking exists, and there is no need for any further stipulations.

From this perspective, the reason why nominative case assignment often accompanies subject-verb agreement is that nominative case assignment is also a form of agreement (between finite T and DP). Subject-verb agreement in English is agreement for phi-features and case at the same time.

16. Dependent Case Theory 

The main claim of Dependent Case Theory is that if two DPs occupy a certain domain, then one of them needs to marked with dependent case (the choice of DP varies parametrically).

This mechanisms of comparing the morphological forms of two DPs is beyond what is allowed in the SMT ‘Strong Minimalist Thesis’, which we can state as follows: The operations of I-language are simple, and they are constrained by third factor economy conditions. 

Certainly, the comparison of DPs and the subsequent case marking of one of the DPs is not an instance of either internal or external Merge. It also not Agree. It has nothing to do with the CI (LF) interface, in the sense that it is completely blind to the interpretation of the DPs. Rather, it seems to be an additional process characterizing only the mapping to the SM (PF) interface. 

But even at the SM-Interface, it does not resemble the mechanisms independently needed there (ordering of words and phrases, determining whether occurrences are spelled-out). The ordering of words and phrases at the SM-Interface is necessary, since we pronounce words in a particular linear order. No similar SM-Interface Condition seems to motivate the presence of case marking, so it is not natural to postulate that case marking takes place at the SM-Interface.

Similarly, the spell-out of occurrences is motivated by economy (third-factor). If one syntactic object has two occurrences then it is more economical to spell-out only one of the occurrences. There does not seem to be any similar economy motivation for case marking. So once again, from this perspective, there is no reason to postulate that case marking takes place at the SM-Interface.

Let us break down dependent case assignment a bit more:

(i) In a domain (phase), find the highest DP (DP1).

(ii) If DP1 does not have any kind of case marking, then search its c-command domain for another DP (DP2).

(iii) DP2 cannot be a copy of DP1 (the are not related by internal Merge).

(iv) If DP2 does not have any kind of case marking, and the language assigns dependent Case downward, the replace DP2 with DP2’ where DP2’ differs from DP2 in having the domain appropriate dependent case. 

This process requires two search actions (to find the two DPs), then a complex DP replacement operation (contextually defined by the phase, CP or DP). This process is very different from anything else found at the PF-Interface. The normal operations at the PF-Interface convert a syntactic structure into a linear ordering of terminals. But the dependent Case operation is actually altering the syntactic structure, replacing one DP with another (mapping one syntactic representation to another).

Furthermore, all versions of Dependent Case Theory have an assumption like the following (Marantz 1991/2000: 24): “Unmarked case may be sensitive to the syntactic environment; for example, in a language GEN may be the unmarked case for NPs inside NPs (or DPs) while NOM may be the unmarked case inside IPs.” But syntactic environment is uniquely determined by syntactic heads and their projections. This observation strongly suggests that the unmarked Cases are assigned by syntactic functional heads.

Lastly, Dependent Case Theory has deep problems handling examples such as ‘For him to leave would be unacceptable.’ ‘for’ cannot assign lexical case, because ‘him’ is not theta-marked by it. ‘him’ cannot be dependent case, because there is no competitor. ‘him’ cannot be default case here, because then there is no explanation for the dependence of ‘him’ on the presence of ‘for’. 

Selected References:

Baker, Mark and Vinokurova. 2010. Two Modalities of Case Assignment: Case in Sakha. Natural Language and Linguistic Theory 28, 593-642.

Caha, Pavel. 2010. The Parameters of Case Marking and Spell Out Driven Movement. Linguistic Variation Yearbook 10, 32-77.

Caha, Pavel. 2013. Explaining the Structure of Case Paradigms by the Mechanisms of Nanosyntax: The Classical Armenian Nominal Declension. Natural Language and Linguistic Theory 31: 1015-1066.

Collins, Chris. 2024. Principles of Argument Structure: A Merge-Based Approach. MIT Press, Cambridge.

Collins, Chris. 2025a. A Syntactic Approach to Case Contiguity. Continua.

Collins, Chris. 2025b. Eliminating Agree. Ms., NYU.

(https://ordinaryworkinggrammarian.blogspot.com/2025/01/eliminating-agree.html)

Grebenyova, Lydia. Forthcoming. Syntax of Case and Agreement. Cambridge University Press, Cambridge.

Lasnik, Howard. 2008. On the Development of Case Theory: Triumphs and Challenges. In Robert Freidin, Carlos P. Otero and Maria Luisa Zubizarreta (eds.), Foundational Issues in Linguistic Theory: Essays in Honor of Jean-Roger Vergnaud. MIT Press, Cambridge.

Legate, Julie. 2024. On Theories of Case and Universal Grammar. In Christina Sevdali, Dionysios Mertyris, Elena Anagnostopoulou (eds.), The Place of Case in Grammar. Oxford University Press, Oxford.

Levin, Theodore and Omer Preminger. 2015. Case in Sakha: Are Two Modalities Really Necessary? Natural Language and Linguistic Theory 33, 231-250.

Marantz, Alec. 1991. Case and Licensing. Proceedings of ESCOL, 234–253.

Cornell Linguistics Club. Republished in Reuland 2000, 11–30.

Pesetsky, David and Esther Torrego. 2011. Case. In Cedric Boeckx (ed.), The Oxford Handbook of Linguistic Minimalism, 52-72. Oxford University Press, Oxford.

Preminer, Omer. 2024. Taxonomies of Case and Ontologies of Case. In Christina Sevdali, Dionysios Mertyris and Elena Anagnostopoulou (eds.), The Place of Case in Grammar. Oxford University Press, Oxford.

Schafer, Florian and Elena Anagnostopoulou. To Appear. Case and Agreement in Distributed Morphology. A. Alexiadou, R. Kramer, A. Marantz and I. Oltra-Massuet (eds.), The Cambridge Handbook of Distributed Morphology. Cambridge University Press, Cambridge.