Basque unergatives, Case-competition, and ergative as inherent Case

Omer Preminger (omerp@mit.edu) January, 2010

1. Background 1.1. Two theories of ergative Case 1.1.1. Case-competition (Marantz 1991)
(1)

disjunctive m(orphological)-case hierarchy

lexically-determined Case dependent Case unmarked Case (incl. quirky Case, inherent Case)

dependent Case: assigned to a DP when there is a distinct as-of-yet-unmarked DP (i.e., another DP without lexically-determined Case) within the local domain of the same V+I complex note: another instance of the same DPi.e., a trace/copy of the a given DPdoes not count as a distinct DP the ergativity parameter: nom-acc languages: dependent Case is assigned by V+I downwards (to object)

dependent Case accusative

erg-abs languages: dependent Case is assigned by V+I upwards (to subject)

dependent Case ergative

prediction: no erg-marked DP without an abs-marked DP in the same V+I domain parallel prediction for nom-acc languages (not relevant for this talk): no acc-marked DP without a nom-marked DP in the same V+I domain

1.1.2. erg as inherent Case (Legate 2008, Woolford 1997; on Basque, Laka 2006)
external arguments are introduced by a dedicated head (v0 ) separate from the lexical verb (V0 ) (following Bowers 1993, Chomsky 1995, Collins 1997, Kratzer 1996, Marantz 2001)

My thanks to Andrew Nevins, Norvin Richards, Milan Rezac, and Marcel Den Dikken, for very helpful discussions and comments. While these individuals have generously shared their time and insights with me, their mention here should not be taken as an unqualied endorsement of the proposal advanced in this talk.

LSA Winter Meeting

Baltimore, MD January, 2010

cross-linguistically, it is the case that given heads can assign a particular Case-marking to the argument they introduce ladierent prepositions idiosyncratically selecting for dierent Case-markings on their nominal complements it is conceivable for v0 , in a given language, to be an assigner of such lexically-driven Case to the argument that it introducesnamely, the external argument call this Case-marking ergative, and what you have is a Split-S (or active) ergative language subjects of unaccusative intransitives will pattern with objects of transitives: abs subjects of unergative intransitives will pattern with subjects of transitives: erg predictions: I. if there exist instances of V0 that assign quirky Case to the object (as exist in Icelandic, for example):

we expect to nd instances of an erg-marked DP with another DP in the same V+I domain which is marked with a Case-marking other than abs I will not be exploring this prediction in this talk (though we can certainly talk about it in the discussion period)

II. if there exist instances of V0 that introduce no arguments:

we expect to nd instances of an erg-marked DP with no abs-marked DP in the same V+I domain

1.2. Unergative Verbs

(2) a. Jon-ek dantza egin d--u-. Jon-erg dance do 3.abs-sg.abs-have-3sg.erg Jon danced. b. Jon-ek dantzatu d--u-. Jon-erg dance-prt 3.abs-sg.abs-have-3sg.erg Jon danced. (3) Hale and Keysers (1993) conjecture: All unergative verbs have direct objects (at least underlyingly)

thus, sentences like (2b) contain an implicit (but syntactically real!) direct object
Apparent support for (3): the auxiliary in (2b) (du 3.abs-sg.abs-have-3sg.erg) has the same form as it does in (2a) This sameness of the auxiliary has been described in dierent ways: (i) both sentences (2a) and (2b) contain the transitive auxiliary constructed from *edun(/ukan) (have), rather than izan (be)

Basque unergatives, Case-competition, and ergative as inherent Case

Omer Preminger (omerp@mit.edu)

(ii) the auxiliaries in both sentences (2a) and (2b) exhibit absolutive-agreement in particular, 3rd-person (singular) Regarding (i): the formulation in (i) uses transitivity to regulate the choice between *edun(/ukan) (have, a.k.a. the transitive auxiliary) and izan (be, a.k.a. the intransitive auxiliary) but the choice between the two auxiliaries might also be derivable simply on the basis of: whether there is an erg agreement target (Laka 1996) or whether the agreement-complex contains erg morphemes (Arregi 2004)

Arregi (2004): the erg-based characterization of *edun(/ukan) -vs.-izan distribution fares

better than the transitivity-based characterization with respect to phenomena such as allocutive agreement (Eguren 1995, Oyharabal 1993) and absolutive displacement (Arregi 2004, Rezac 2007) as well as the impossibility of non-nite *edun in the synchronic grammar notice: Under most circumstances, of course, being transitive and having erg agreement-morphology are co-extensive; but not always! Consider, for example, allocutive agreement (Arregi 2004, Eguren 1995, Oyharabal 1993): In this construction, the auxiliary exhibits agreement-morphology that co-indexes the addressee (roughly speaking) alongside the otherwise expected agreement-morphology, co-indexing nominal arguments This additional agreement-morphology can be erg or dat, depending on what morphological slots are not already occupied one can nd erg agreement-morphology with verbs that are unambiguously intransitive: (4) Jon- eror-i d--u- k . Jon(abs) fall-prf 3.abs-sg.abs-have- 2sg.erg Jon has fallen. (2sg allocution)

[Arregi 2004:(11a)]

Crucially, the auxiliary in such cases must be an *edun(/ukan) auxiliary (have, a.k.a. the transitive auxiliary)

As for simplex unergatives: they have an erg argument, and their auxiliary has erg
agreement-morphology; that much is uncontroversial Under the erg-based approach: the appearance of an *edun(/ukan) auxiliary (have, a.k.a. the transitive auxiliary) with simplex unergatives is expected and doesnt bear (directly) on the question of whether or not simplex unergatives have an implicit object

LSA Winter Meeting

Baltimore, MD January, 2010

Regarding (ii), repeated here: (ii) the auxiliaries in both sentences (2a) and (2b) exhibit absolutive-agreement in particular, 3rd-person (singular) The observation in (ii) can only be taken as evidence for an implicit object if we accept the linking hypothesis in (5): (5)

absolutive-agreement linking hypothesis absolutive agreement-morphology can only come about as a result of the agreementprobe successfully establishing an agreement relation with an abs target

Preminger (2009): the hypothesis in (5) is untenable

consider the data in (6ab), from the same substandard variety of Basque discussed by Etxepare (2006): (6) a. [ Miren-entzat harri horiek altxa-tze-n ] probatu Miren-ben stone thosepl (abs) lift-nmz-loc attempted d- it -u-zte. 3.abs- pl.abs -have-3pl.erg They have attempted to lift those stones for Miren.
(subject is [pro -3pl.erg])

b. [ Lankide-e-i liburu horiek irakur-tze-n ] probatu colleague-artpl -dat book thosepl (abs) read-nmz-loc attempted d- /*it -u-(z)te. 3.abs- sg.abs/*pl.abs -have-3pl.erg They have attempted to read those books to the colleagues.
(subject is [pro -3pl.erg])

[Preminger 2009:(29, 32)]

abs agreement-morphology can arise even when the relation to the putative controller of agreement (liburu horiek book thosepl (abs)) has been disrupted in this case, by an intervening dat DP its just that when agreement has been disrupted, the agreement-morphology reects default (3rd-person singular) -features

We are left with no concrete argument in favor of simplex unergatives having an

implicit object except, perhaps, for a uniformity-style argument, seeking to unify light-verb constructions of the kind in (2a) with simplex unergatives like (2b) Of course, none of this constitutes evidence that simplex unergatives lack an implicit object but the stage is set for us to look for such evidence

Basque unergatives, Case-competition, and ergative as inherent Case

Omer Preminger (omerp@mit.edu)

2. The Absence of an Implicit Object 2.1. Lexically Impossible Implicit Objects

Laka (2006) notes the existence of a class of unergatives with the following properties: syntactically, they behave like the simplex unergative in (2b) they take a single overt DP argument, marked with erg Case the auxiliary is an *edun(/ukan) auxiliary (have, a.k.a. the transitive auxiliary) this auxiliary exhibits 3rd-person singular abs agreement-morphemes crucially, however, these verbs have no nominal counterparts in the language that could function as the implicit object (7) a. Klara-k ondo eskia-tzen d--u-. Klara-erg well ski-impf 3.abs-sg.abs-have-3sg.erg Klara skies well. b. Eguzki-a-k disdira-tzen d--u-. sun-artsg -erg shine-impf 3.abs-sg.abs-have-3sg.erg The sun shines. function as the implicit object of (7ab) These verbs alone might be enough to convince you that there are unergative verbs that lack any (explicit or implicit) object

[Laka 2006:(8a)]

[Laka 2006:(9b)]

there is no nominal [N0 eskia ] or nominal [N0 disdira ] in Basque, that could

A potential alternative account of cases like (7ab), however: these are merely gaps
in pronounceability suppose that the Basque lexicon contains lexical entries [N0 eskia ] and [N0 disdira ], which are syntactically no dierent from [N0 dantza ] except that these lexical entries, unlike [N0 dantza ], lack phonological content

these lexical entries would be unable to function as the N0 head of a

pronounced noun-phrase while being perfectly capable of functioning as an unpronounced implicit object

LSA Winter Meeting

Baltimore, MD January, 2010

2.2. The Iterative/Repetitive construction

Etxepare (2003): many of these light-verb predicates are able to appear in a dierent variation of the light-verb construction where the complement of the light-verb is locative/adverbial, rather than nominal (resulting in an iterative reading) (8) a. Dantza(n) egin d--u-te. dance-loc do 3.abs-sg.abs-have-3pl.erg They danced (repeatedly). b. Laster(ka) egin d--u-te. d. Oihu(ka) egin d--u-te. run-adv do aux scream-adv do aux They ran (repeatedly). They screamed/yelled (repeatedly). c. Borroka(n) egin d--u-te. ght-loc do aux They fought (repeatedly). e. Errieta(n) egin d--u-te. dispute-loc do aux They disputed (repeatedly).
[Etxepare 2003:(117)]

Crucially, abs agreement-morphology on the auxiliary persists, whether the complement

of the light-verb is nominal or not If the source of abs agreement-morphology in (2b) were, in some form or another, an implicit object (dantza dance): the persistence of abs agreement-morphology in examples such as (8ae) would remain unexplained since the nominal in question is now the complement of an adposition note: it was shown in Preminger (2009) that the adposition / -n / does not constitute a locality-boundary for Agree however, crucially, its complement is not treated by the grammar as an accessible noun-phrase for the purposes of agreement (idem) If, on the other hand, simplex unergatives systematically lack an implicit object, then the persistence of abs agreement-morphology in examples such as (8ae) is predicted: both in (2b) and in the locative/adverbial versions of (8ae), there is no abs nominal to be foundand therefore, Agree has failed, much like it has in (6b) the auxiliaries in both constructions bear the hallmark of failed Agree namely, default (i.e., 3rd-person singular) abs agreement-morphemes

Basque unergatives, Case-competition, and ergative as inherent Case

Omer Preminger (omerp@mit.edu)

2.3. Agreement with low absolutives in LDA-unergatives

(9) altxa-tze-n ] probatu d- it -u-zte. [ Harri horiek stone thosepl (abs) lift-nmz-loc attempted 3.abs- pl.abs -have-3pl.erg They have attempted to lift those stones.
(subject is [pro -3pl.erg])

[Etxepare 2006:(85a)]

The matrix verb in (9) is very similar to the simplex unergatives discussed earlier but in addition to the erg subject (pro -3pl.erg, in (9)), it selects an adpositionallyheaded embedded clause

crucially, it selects no abs argument

If simplex unergatives had an implicit object that is, if selecting an erg subject were contingent on the presence of an implicit abs object then: it would be entirely surprising that the abs agreement-morphemes on the auxiliary in (9) are available to co-index the abs DP inside the embedded clause rather than bearing 3rd -person singular agreement with the aforementioned tacit object, like their counterparts in (2a) If, on the other hand, simplex unergatives systematically lack an implicit object, then this behavior is predicted: since the projection of an erg argument is not contingent on projecting an implicit abs argument, there is no abs argument in the matrix clause the abs agreement-morphemes on the matrix auxiliary are free to co-index the embedded abs DP A potential objection: According to Marantz (1991), erg(dependent) Case is contingent on the presence of an abs(unmarked) DP within the same domainnot necessarily the same clause maybe the entire construction in (9) constitutes a single Case-competition domain in other words, perhaps both the matrix erg the (embedded) abs DP in (9) are governed by the matrix V+I

Alas, this view is untenable:

If DPs within such an embedded adpositional clause counted as Case-competitors for DPs in the matrix, it would be impossible to get an abs DP both in the matrix and in the adpositional clausecontra to fact: (10)

pro -3pl.abs. [ Liburu-a irakur-tze-n ] saiatu dira :::::::::::: ::::::::: bookartsg (abs) read-nmz-loc tried 3pl.abs.be ::::::::::::::::: They tried to read the book.

[Etxepare 2006:(53a)]

Instead, the dierence between (9) and (10) really seems to boil down to the lexical items in question namely, it is a property of probatu vs. saiatu (more on this below)

LSA Winter Meeting

Baltimore, MD January, 2010

3. Interim Conclusion
Basque simplex unergatives constitute an instance of erg being assigned in the absence of an abs Case-competitor showing that a Case-competition theory (Marantz 1991) cannot be true for erg Case in Basque Does this mean that Woolfords (1997) and Legates (2008) account of erg Case extends to Basque? in other words, does it mean that erg in Basque is inherent Case?

What we have shown, so far, is that erg in Basque is a lexically-conditioned phenomenon

in that it is dependent on the particular V0 present in the derivation recall probatu , in (9), vs. saiatu , in (10) !! this does not automatically make it inherent Case inherent Case is theta-dependent Case i.e., Case that is tied not just to any properties of its licensor, but specically to its thematic properties we should investigate whether erg in Basque is thematically conditioned

4. erg in Basque: Conditioned thematically, or otherwise?

Weve already seen that there exist agent arguments in Basque that are not marked with erg Casefor example, the agent argument of a verb like saiatu (try) it is not entirely clear that this is a problem for the view that erg Case is thematically-conditioned the fact that a particular lexical item L1 (e.g., v *) assigns lexically-determined Case to arguments that it theta-marks with A (e.g., agent), does not mean that any other lexical item L2 that theta-marks its argument with the same A must assign (the same) lexically-determined Case in fact, both patterns are attested: (i) in languages with inherent dat Case, being a goal argument entails dat Casemarking (Woolford 1997, 2006) (ii) on the other hand, experiencer theta-roles in Icelandic are sometimes associated with quirky dat Casebut not always: there are assigners of an experiencer theta-roles which do not assign dat (see Thrinsson 2007, and references therein) Woolford to posit these di erent behaviors are, in fact, the kind of thing that lead the distinction between lexical and inherent Case

Basque unergatives, Case-competition, and ergative as inherent Case

Omer Preminger (omerp@mit.edu)

The more interesting question, then, is the following:

Are there non-agent arguments that are marked with erg Case? to which the answer, as far as Basque is concerned, is quite clearly yes: (11) a.
! Ni-k

ura irakin d--u-t. 1sg.erg water(abs) boil 3.abs-sg.abs-have-1sg.erg I boiled the water.

b.

Ura-k irakin d--u-. water-erg boil 3.abs-sg.abs-have-3sg.erg The water has boiled.

[Holgun 2007:(24ab)]

(12)

Eguzki-a-k disdira-tzen d--u-. [=(7b)] sun-artsg -erg shine-impf 3.abs-sg.abs-have-3sg.erg The sun shines.

[Laka 2006:(9b)]

To this fact, there can be two responses, as far as I can see: I. Loosen the restriction that states that v * only introduces agent arguments its not entirely clear that this would help the inherent Case proposal if a single head can assign multiple dierent theta-roles 1 , 2 , . . ., inherent Case associated with 1 should not extend to arguments receiving 2 (and if it does, its not quite clear what remains of the notion inherent) II. Concede that erg Case in Basque is not triggered thematically then how is it triggered. . . ?

Before addressing this question, let us consider another relatively well-known instance of
non-agent arguments receiving erg Case in Basque: (13) Rojorekin minduta d-a-go-ela ematen a. Jokalariren bat player one(abs) Rojo.with hurt 3.abs-cop-sg.abs-comp seem d--u-. 3.abs-sg.abs-have-3sg.erg seem > :  It seems that some player is upset with Rojo. > seem:  b. Jokalariren bat-ek Rojorekin minduta d-a-go-ela ematen player one-erg Rojo.with hurt 3.abs-cop-sg.abs-comp seem d--u-. 3.abs-sg.abs-have-3sg.erg seem > :  Some player seems upset with Rojo. > seem: 
[Artiagoitia 2001:(42ab)]

LSA Winter Meeting

Baltimore, MD January, 2010

Artiagoitia (2001): this construction involves raising of the argument (jokalariren bat player one) from the embedded clause in particular, the availability of the seem > interpretation in (13b) rules out several prominent alternatives to raising, as an analysis of the pattern in (13ab) e.g., copy-raising, (backwards-)control Artiagoitia also demonstrates that the upstairs subject positionwhen ematen (seem) takes a nite clausal complementis a non-thematic position by showing it cannot enter into control relations, contra the quasi-argumental subject of weather predicates

What this data points to, is that erg in Basque is triggered by a particular position (rather
than, say, a particular theta-role)

5. The Problem, and a Sketch of a Solution

problem :
How can erg Case be lexically-conditioned (as demonstrated in 2), and at the same time be triggered by a specic position (as demonstrated in 4)?

the beginnings of a solution :

Suppose that erg is assigned to a DP in [Spec,vP]but crucially, a DP can nd itself in [Spec,vP] in one of two ways:1 (i) the DP is base-generated in [Spec,vP] As per usual, base-generation of a DP results in assignment of a theta-role to that DP in this case, agent (since this is the role associated with the [Spec,vP] position) (ii) the DP moves into [Spec,vP] If thematic roles are only discharged at the tail of a chaincontra Hornstein (2001), and related workthis DP will have whatever theta role it got in its base-position Thus, verbs that are selected by an agent-introducing v0 will always have an erg-marked agent but verbs that are selected by a v0 that does not introduce an agentincluding, for example, Levin and Rappaport-Hovavs (1995) verbs of emissioncould also have an erg-marked argument by virtue of their theme argument moving to/through [Spec,vP] Admittedly, this relegates the problem to explaining why verbs of emission have this kind of movement-triggering (perhaps, EPP-bearing) v0 , while one-place unaccusatives do not but we at least have a theory of erg Case that is compatible with the data!
1 If we maintain that vP is a strong phase (Chomsky 2001, et seq.), then movement operations that move an element out of vP must also pass through [Spec,vP]; clearly, it would be undesirable to predict that any such movement through [Spec,vP] would also result in erg marking on the moving element. This could be handled in one of two ways: the position relevant for erg-marking could be specically restricted to the rst specier of vP. Alternatively, one could take the verb-level phasal category to be distinct from vPe.g.,

10

Basque unergatives, Case-competition, and ergative as inherent Case

Omer Preminger (omerp@mit.edu)

6. Conclusion
In this talk, I have. . . Presented three arguments (two of them new) for the lack of an implicit object in unergatives in Basque (contra Hale and Keyser 1993) unergatives whose putative implicit object does not exist as a nominal in the language

unergatives where the putative implicit object is enclosed in an adpositional shell agreement with low absolutives in LDA-unergatives
dealing a blow to theories of erg Case that rely on case-competition (e.g., Marantz 1991) Presented existing evidence against viewing erg Case as inherent: the existence of non-agent arguments that bear erg Case raising-to-erg constructions (Artiagoitia 2001) Suggested that this tensionbetween erg being lexically-conditioned (2), and erg being positionally-triggered (4)can be resolved: by assuming that erg is invariably discharged in [Spec,vP] whether the DP found itself there by way of base-generation (i.e., lexical conditioning) or by way of movement (positional conditioning)

References
Alexiadou, Artemis, Elena Anagnostopoulou, and Florian Schafer. 2006. The properties of anticausatives crosslinguistically. In Phases of Interpretation, ed. Mara Frascarelli, 187212. Berlin: Mouton de Gruyter. Arregi, Karlos. 2004. The have/be Alternation in Basque. Ms., University of Illinois at UrbanaChampaign. Artiagoitia, Xabier. 2001. Seemingly Ergative and Ergatively Seeming. In Features and Interfaces in Romance: Essays in honor of Heles Contreras, eds. Julia Herschensohn, Enrique Mallen, and Karen Zagona, 122. Amsterdam: John Benjamins. Baltin, Mark R. 2007. Deletion Versus Pro-Forms: A False Dichotomy?, Ms., New York, NY. url: <http : //www.nyu.edu/gsas/dept/lingu/people/faculty/baltin/papers/ baltin-deletion-vs-pro-forms.pdf>. Bowers, John. 1993. The syntax of predication. Linguistic Inquiry, 24:591656. Chomsky, Noam. 1995. The Minimalist Program. Cambridge, MA: MIT Press. Chomsky, Noam. 2001. Derivation by phase. In Ken Hale: A life in language, ed. Michael Kenstowicz, 152. Cambridge, MA: MIT Press. Collins, Chris. 1997. Local economy. Cambridge, MA: MIT Press. Eguren, Luis. 1995. Syntax and morphology in Basque verbal inection. Paper presented at the XXI Incontro di Grammatica Generativa, Istituto S. Raaele, Milan.

Alexiadou, Anagnostopoulou, and Schafers (2006) and Harleys (2007) VoiceP (as proposed by Baltin 2007, for example).

11

LSA Winter Meeting

Baltimore, MD January, 2010

Etxepare, Ricardo. 2003. Valency and Argument Structure in the Basque Verb. In A Grammar of Basque, eds. Jose Ignacio Hualde and Jon Ortiz de Urbina, 363426. Berlin: Mouton de Gruyter. Etxepare, Ricardo. 2006. Number Long Distance Agreement in (Substandard) Basque. In Studies in Basque and Historical Linguistics in Memory of Robert L. Trask, eds. Joseba A. Lakarra and Jose Ignacio Hualde, vol. XL, Supplements of the Anuario del Seminario de Filologia Vasca Julio de Urquijo 12, 303350. Hale, Kenneth and Samuel Jay Keyser. 1993. On argument structure and the lexical expression of syntactic relations. In The View from Building 20: Essays in linguistics in honor of Sylvain Bromberger, eds. Kenneth Hale and Samuel Jay Keyser, 53109. Cambridge, MA: MIT Press. Harley, Heidi. 2007. External arguments: On the independence of Voice0 and v0 . Paper presented at GLOW 30, Tromso. url: <http://glow.uit.no/GLOWXXX/abstracts/Harley.pdf>. Holgun, Justin. 2007. The Status of Ergative Case in Basque: A Minimalist Approach. Honors thesis, Reed College, Portland, OR. url: <http://ling.auf.net/lingBuzz/000449>. Hornstein, Norbert. 2001. Move! A minimalist theory of construal. Oxford: Blackwell Publishers. Kratzer, Angelika. 1996. Severing the external argument from its verb. In Phrase Structure and the Lexicon, eds. Johan Rooryck and Laurie Zaring, 109137. Dordrecht: Kluwer Academic Publishers. Laka, Itziar. 1996. A Brief Grammar of Euskara, the Basque Language (ISBN: 84-8373-850-3). Ms., Vitoria-Gasteiz: Euskal Herriko Unibertsitatea (University of the Basque Country). url: <http : //www.ei.ehu.es/p289 - content/eu/contenidos/informacion/grammar_ euskara/en_doc/index.html>. Laka, Itziar. 2006. On the Nature of Case in Basque: Structural or Inherent?, in Organizing Grammar, eds. Hans Broekhuis et al., 374382. Berlin: Mouton de Gruyter. Legate, Julie Anne. 2008. Morphological and Abstract Case. Linguistic Inquiry, 39:55101. Levin, Beth and Malka Rappaport-Hovav. 1995. Unaccusativity. Cambridge, MA: MIT Press. Marantz, Alec. 1991. Case and Licensing. In Proceedings of the 8th Eastern States Conference on Linguistics (ESCOL 8), eds. German Westphal, Benjamin Ao, and Hee-Rahk Chae, Reprinted as Marantz (2000), Ithaca, NY: CLC Publications, 234253. Marantz, Alec. 2000. Case and Licensing. In Arguments and Case: Explaining Burzios Generalization, ed. Eric Reuland, 1130. Amsterdam: John Benjamins. Marantz, Alec. 2001. Words. Ms., Cambridge, MA: MIT. Oyharabal, Bernard. 1993. Verb agreement with non-arguments: On allocutive agreement. In Generative Studies in Basque Linguistics, eds. Jose Ignacio Hualde and Jon Ortiz de Urbina, 89114. Amsterdam: John Benjamins. Preminger, Omer. 2009. Breaking Agreements: Distinguishing Agreement and CliticDoubling by Their Failures. Linguistic Inquiry, 40:619666. Rezac, Milan. 2007. Escaping the Person Case Constraint: Reference-set computation in the -System. In Linguistic Variation Yearbook, eds. Pierre Pica, Johan Rooryck, and Jeroen van Craenenbroeck, vol. 6, 97138. Amsterdam: John Benjamins. Thrinsson, Hskuldur. 2007. The Syntax of Icelandic. Cambridge: Cambridge University Press.

12

Basque unergatives, Case-competition, and ergative as inherent Case

Omer Preminger (omerp@mit.edu)

Woolford, Ellen. 1997. Four-Way Case Systems: Ergative, Nominative, Objective and Accusative. Natural Language and Linguistic Theory, 15:181227. Woolford, Ellen. 2006. Lexical Case, Inherent Case, and Argument Structure. Linguistic Inquiry, 37:111130. This is svn-revision 2318.

13