Semantic case, as I have discussed above, is associated with the semantic inter- pretation of a noun phrase or sentence based on the morphological case. The instrumental case that can be licensed by the verbs described in this section is an instance of semantic case: The difference in case results in a difference in inter- pretation of the sentence based on the relationship between the noun phrase and the rest of the clause. Furthermore, instrumental case is associated with a few specific meanings when it is used as a semantic case (e.g., in adverbial noun phrases, such as valgyti šaukstu ‘to eat with a spoon:ins’). The interpretation of the instrumental arguments in the examples above is that of a true instrument or a means of performing the action. The choice of case is not based on a particular lexical item (as is true with lexical case), but based on meaning, so it must be a semantic case.
Finally, there is the question as to why not all verbs participate in this alter- nation. One possible solution is that these verbs are unique in the theta role that they assign to their internal arguments. Parsons (1990) suggests that certain argu- ments can have two theta roles, such as instrument-theme or agent-theme (p. 81).
This, however, means that the case licensing is not solely related to either the theta role or the semantic relationship between the verb and the object, but relies on both. Thus, the case alternation is limited to the specific classes of verbs where both the multiple theta-role assignment occurs, and the difference in prominence is possible.
These accusative-instrumental case alternations reflect the event structure and argument structure: The morphological case is a reflection of the argument’s role in the clause and in determining the type of event. As seen with the verbs of moving a body part, for instance, the addition of a directional resultative prepo- sitional phrase eliminates the possibility of an alternation. Similarly, Levin and Rappaport Hovav (1997) discuss resultatives as a test for ergativity and unaccusa- tivity and determining event type. Therefore, the accusative-instrumental alter- nation may be limited to verbs whose event structure is flexible enough to accom- modate these differences in interpretation.
Finally, these case alternations show that a difference in case can reflect a difference in event structure. Here, the morphological case is not necessarily con- nected to a particular lexical item, although the theta role assigned by the verb may still play an important role. Case is also an important indicator of argument structure alternations, as seen with verbs of dressing in particular.
6 Conclusion
There are many instances of non-canonical morphological case in Lithuanian that appear to violate the tenets of case theory that hold that case must either be based on the structural position or is tied to a particular lexical item or theta role in some way. In this chapter, I explored three constructions with internal argu- ments marked with a case other than accusative: instrumental-accusative alter- nations on internal arguments, dative and genitive objects in purpose clauses, as well as the passivization of oblique case-licensing verbs with agreeing passive participles and nominative subjects.
These constructions are exceptional because each instance has elements of structural case and semantic case licensing. I have argued that there are multi- ple types of non-structural case and that the traditional view of inherent case is too narrow to account for the variety of case licensing found in Lithuanian.
Rather, there are three kinds of case in addition to structural: lexical, inherent, and semantic. Lexical case is a strong requirement on a particular verb or prepo- sition and can be overridden in passivization in Lithuanian. Inherent case is, as with the traditional definition, associated with a particular theta role. Following Woolford (2006), dative recipients and goals are examples of inherent case.
Semantic case is found when the morphological case contributes significantly to the overall interpretation of the sentence. This sometimes corresponds with a particular theta role that may be assigned instead of the usual patient role for direct objects. By examining a language with rich morphological case like Lithu- anian, we can gain a better understanding of how case can interact with other
elements in the clause, as well as how case can reflect the argument structure and event structure of a verb.
Acknowledgments
Thanks to Leonard Babby, James Lavine, and Edwin Williams for their useful comments and to the participants of the Contemporary Approaches to Baltic Linguistics conference and the audience at FASL 20 and the Syntax Supper at the CUNY Graduate Center. Thanks also go to the reviewers of a previous version of this article and the editors of this volume for their useful comments. I am also very grateful to the native speakers who helped with the Lithuanian data:
Artūras Judžentis, Kristina Lenartaitė, Rolandas Mikulskas, Žydrūnė Mladineo, Elvyra Petrašiūnienė, Giedrius Subačius, Martynas Vasiliauskas, and Virginija Vasiliauskienė.
Abbreviations
1 first person 2 second person 3 third person acc accusative aux auxiliary cnt continuative cnv converb dat dative def definite
f feminine
gen genitive imp imperative ins instrumental m masculine
neg negation nom nominative pa active participle part partitive
pp passive participle pl plural
prep preposition prs present prf (verbal) prefix prt particle pst past
q yes-no question marker refl reflexive
sg singular
References
Aissen, Judith. 2003. Differential Object Marking: Iconicity vs. Economy. Natural Language and Linguistic Theory 21(3): 435–483.
Ambrazas, Vytautas (ed.). 2006. Lithuanian grammar. Vilnius: Baltos lankos.
Anderson, Cori. 2009. Oblique passivization: Evidence from Lithuanian and Slavic. In Issledovanija po Slavjanskim Jazykam [Studies in Slavic Languages] 14: 137–154.
Anderson, Cori. 2011. Case theory and case alternations: Evidence from Lithuanian. Baltic Linguistics 2: 9–35.
Anderson, Cori. 2013a. Passivization and argument structure in Russian and Lithuanian.
Presentation at the annual meeting of American Association of Teachers of Slavic and Eastern European Languages, Boston, MA.
Anderson, Cori. 2013b. Case and event structure in Russian and Lithuanian. Princeton University doctoral dissertation.
Anderson, Cori. 2013c. Case alternation and event structure: Evidence from Lithuanian and Russian. In Formal Approaches to Slavic Linguistics 20, 1–16. Ann Arbor, MI: Michigan Slavic Publications.
Arkadiev, Peter. 2014. Case and word order in Lithuanian revisited. In Axel Holvoet & Nicole Nau (eds.) Grammatical relations and their non-canonical encoding in Baltic, 43–95.
Amsterdam, Philadelphia: John Benjamins.
Armoškaitė, Solveiga. 2006. Accomplishment VPs: Construction of telicity. A case study of Lithuanian. In Claire Gurski & Milica Radisic (eds.) Proceedings of the 2006 annual conference of the Canadian Linguistic Association. http://westernlinguistics.ca/
Publications/CLA2006/Armoskaite.pdf.
Babby, Leonard H. 1986. The locus of case assignment and the direction of percolation: Case theory and Russian. In Richard Brecht & James Levine (eds.) Case in Slavic, 170–219.
Columbus, OH: Slavica.
Babby, Leonard H. 1994. Case theory. In Carlos P. Otero (ed.) Noam Chomsky: Critical assessments (Vol I: Linguistics: Tome II), 630-652. London, New York: Routledge.
Borer, Hagit. 2005. Structuring Sense. Vol. II: The normal course of events. Oxford: Oxford University Press.
Chomsky, Noam. 1981. Lectures on Government and Binding: The Pisa Lectures. Holland: Foris Publications.
Chomsky, Noam. 1986. Knowledge of language: Its nature, origin, and use. New York: Praeger.
Chomsky, Noam. 1995. The minimalist program. Cambridge, MA: MIT Press.
Cuervo, Maria Cristina. 2003. Datives at large. Doctoral dissertation, MIT.
Demjjanow, Assinya & Anatoli Strigin. 2000. Case assignment to conceptual structures:
The Russian instrumental. In Markus Kracht & Anatoli Strigin (eds.) Papers on the Interpretation of Case, 75–107. University of Potsdam.
Dowty, David. 1991. Thematic proto-roles and argument selection. Language 67(3): 547–619.
Emonds, Joseph. 2006. Adjectival passives. In Martin Everaert & Henk C. van Riemsdijk (eds.) The Blackwell Companion to Syntax, Vol 1, 16–60. Oxford: Blackwell.
Fillmore, Charles. 1968. The case for case. In Emmon Bach & Robert Harms (eds.) Universals in linguistic theory, 1–88. New York: Holt, Rinehart, and Winston.
Fowler, George. 1996. Oblique passivization in Russian. Slavic and East European Journal 40(3): 519–545.
Freidin, Robert. 1992. Foundations of generative syntax. Cambridge, MA: MIT Press.
Franks, Steven & James E. Lavine. 2006. Case and word order in Lithuanian. Journal of Linguistics 42(2): 239–288.
Geniušienė, Emma. 1987. The typology of reflexives. Berlin, Mouton de Gruyter.
Geniušienė, Emma. 2006. Passives in Lithuanian (with comparison to Russian). In Werner Abraham & Larisa Leisiử (eds.) Passivization and typology: Form and function, 29–61.
Amsterdam, Philadelphia: John Benjamins.
Holvoet, Axel. 1991. Transitivity and clause structure in Polish (a study in case marking).
Warsaw: Slawistyczny Ośrodek Wydawniczy.
Holvoet, Axel & Loreta Semėnienė. 2004. Linksnio teorijos pagrindai [Basics of case theory].
In Axel Holvoet & Loreta Semėnienė (red.) Gramatinių kategorijų tyrimai [Research on grammatical categories], 11–33. Vilnius: Lithuanian Language Institute.
Holvoet, A. & A. Judžentis, 2004: Tranzityvumo samprata [Notion of Transitivity]. In: Holvoet &
Semėnienė, eds., Gramatinių kategorijų tyrimai [Research on Grammatical Categories], Lietuvių kalbos institutas, 61–76.
Jablonskis, Jonas. [1922] 1997. Lietuvių kalbos gramatika [Grammar of the Lithuanian language]. Vilnius: Mokslo ir enciklopedijų leidykla.
Kiparsky, Paul. 1998. Partitive case and aspect. In Miriam Butt & Wilhelm Geuder (eds.) The projection of arguments, 265–308. Stanford: CSLI.
Lavine, James. 2006. “Is There a Passive Evidential Strategy in Lithuanian?” In Papers from the 42nd Meeting of the Chicago Linguistic Society, eds. Jacqueline Bunting et al., 41–55.
Lenartaitė, Kristina. 2009. Tiesioginiai objektai ir ditranzityvinių konstrukcijų klausimas lietuvių kalboje [Direct objects and questions of ditransitive constructions in Lithuanian].
In Axel Holvoet & Rolandas Mikulskas (red.) Gramatinių funkcijų prigimtis ir raiška [The nature and expression of grammatical functions], 69–98. Vilnius: Vilniaus universitetas &
Asociacija „Academia Salensis“.
Lenartaitė, Kristina. 2010. Argumentų raiškos alternavimas lietuvių kalboje [Alternations of argument expression in Lithuanian]. Vilnius University doctoral dissertation.
Letučij, Aleksandr. 2007. Tvoritel’nyj padež pri russkix perexodnyx glagolax: konstrukcii s ob’’ektami-častiami tela [Instrumental case with Russian transitive verbs: Constructions with body parts as objects]. In Russkaja Filologija 18. Sbornik naučnyx rabot molodyx filologov [Russian Philology 18. Collection of works by young philologists], 50–55. Tartu University Press.
Levin, Beth & Malka Rappaport Hovav. 1997. Lexical semantics and syntactic structure. In Shalom Lappin (ed.) The handbook of contemporary semantic theory, 487–507. Oxford:
Blackwell.
Matushansky, Ora. 2008. A case study of predication. In Frank Marušič and Rok Žaucer (eds.) Studies in formal Slavic linguistics. Contributions from formal description of Slavic Languages 6.5, 213–239. Frankfurt am Main: Peter Lang.
Matushansky, Ora. 2010. Russian predicate case, encore. In Gerhild Zybatow, Philip Dudchuk, Serge Minor, & Ekaterina Pshehotskaya (eds.) Formal studies in Slavic linguistics, Proceedings of FDSL 7.5, 117–135. Frankfurt am Main: Peter Lang.
McFadden, Thomas. 2004. The position of morphological case in the derivation: A study on the syntax-morphology interface. University of Pennsylvania doctoral dissertation.
Parsons, Terence. 1990. Events in the Semantics of English. Cambridge: MA: MIT Press.
Ramchand, Gillian. 2008. A first-phase syntax. Cambridge: Cambridge University Press.
Richardson, Kylie. 2008. Case and aspect in Slavic. Oxford: Oxford University Press.
Rouveret, Alain & Jean Roger Vergnaud. 1980. Specifying reference to the subject: French causatives and conditions on representations. Linguistic Inquiry 11: 97–202.
Šukys, Jonas. 2005. Veiksmo atlikimo priemonės raiška [Expressing means of performing an action]. Kalbos kultūra 78: 133–143.
Svenonius, Peter. 2006. Case alternations and the Icelandic passive and middle. Ms., lingBuzz/000124 ([v2 in 2005]).
Tenny, Carol. 1994. Aspectual roles and the syntax-semantics interface. Dordrecht: Kluwer.
Wiemer, Bjửrn. 2006. Relations between actor-demoting devices in Lithuanian. In Werner Abraham & Larisa Leisiử (eds.) Passivization and typology. Form and function, 274–309.
Amsterdam, Philadelphia: John Benjamins.
Wierzbicka, Anna. 1980. The case for surface case. Ann Arbor, MI: Karoma Publishers.
Woolford, Ellen. 2006. Lexical case, inherent case, and argument structure. Linguistic Inquiry 37(1): 111–130.
Zaenen, Annie & Joan Maling. 1990. Unaccusative, passive, and quirky case. In Joan Maling &
Annie Zaenen (eds.) Modern Icelandic syntax, 137–152. New York: Academic Press.
An obliqueness-based approach
1 The problem stated
The article deals with the interpretation of grammatical relations in Latvian constructions like the ones shown in examples (1)–(3).
(1) Man patīk šī grāmat-a.
1sg.dat please.prs.3 this.nom.sg.f book-nom.sg
‘I like this book.’
(2) Man vajag tav-u palīdzīb-u.
1sg.dat be.needed.prs.3 1sg.poss-acc.sg help-acc.sg
‘I need your help.’
(3) Man jā-lasa šī grāmat-a.
1sg.dat deb-read this.nom.sg.f book-nom.sg
‘I have to read this book.’
These three constructions have one thing in common: They contain dative NPs that normally occur clause-initially, constitute the unmarked topic of the sen- tence, and might be regarded as being semantically, in some sense, the subject of the clause. In this sense, and perhaps also in other respects, they could raise claims to the status of “non-canonically marked subjects”. Other things differ:
(1) has a nominative marked NP that might raise rival claims to subjecthood;
(2) contains no nominative and would traditionally be described as impersonal, whereas (3) is superficially similar to (1) but for the fact that the debitive jālasa
‘one must read’ is an inflectional form of the verb lasīt ‘read’ and one could be tempted to derive evidence from this fact, say, to the effect that grāmata ‘book.
nom’ must be the object of jālasa just as grāmatu ‘book.acc’ is an object in lasu grāmatu ‘I am reading a book’ – a type of evidence that is not available, e.g., for (1), where the interpretation of grāmata as clause subject is at least an option.
But if one accepts the interpretation of grāmata in (3) as an object, one could extend this interpretation to grāmata in (1). One could even adduce evidence in support of this parallel interpretation. The datives with patikt ‘please’ and with
the debitive may both control reflexivization, which would plead in favor of their interpretation as subjects.1
(4) Tev patīk sav-i klasesbiedr-i?
2sg.dat please.prs.3 refl.poss-nom.pl.m classmate-nom.pl
‘Do you like your classmates?’
http://www.formspring.me/r/tev-pat-k-savi-klasesbiedri/253602291771249079 (accessed July 2012)
(5) Tik-uš-i rīko-t-i internacionāl-ie
aux-ppa-nom.pl.m organize-ppp-nom.pl.m international-nom.pl.m.def
vakar-i, kad katr-am bij-is jā-atnes
evening-nom.pl when each-dat.sg.m be-ppa-nom.sg.m deb-bring
sav-s nacionāl-ais ēdien-s
refl.poss-nom.sg.m national-nom.sg.m.def dish-nom.sg http://www.icelo.lv/lat/stasti-un-galerijas/ttstasti/32795/ (accessed July 2012) ‘(It is told that) international evenings were held at which everybody had to bring their national dish with him.’
The reflexivization criterion is not a strong one, because it is not clear what exactly enables control of reflexive pronouns in general. The conditions that have been invoked in the literature on reflexives include: c-command (the standard view in generative grammar), a thematic hierarchy (as proposed by Jackendoff 1972), the obliqueness hierarchy as proposed in HPSG (Pollard & Sag 1994: 238–280), etc.
Examples discussed in this article point to topichood as an important condition.
But, as we will see below, for many putative “non-canonically marked subjects”, control of reflexivization is the only syntactic subjecthood test that works well, and I will therefore use it as a pis aller: At least if there are two candidates to quasi-subjecthood, the one that can control reflexivization is probably a better candidate than the one that cannot, although control of reflexivization is not by itself a sufficient criterion of subjecthood.
1 Here and in what follows, I will be using reflexivization tests involving the reflexive possessive pronoun savs ‘one’s (own)’ rather than the reflexive pronoun proper, the reason for this being that constructions with the reflexive possessive pronoun are of wider application. The reflexive pronoun sevis lacks a nominative, so that with a verb like patikt (which occurs with dative and nominative NPs), the ability of the dative NP to control reflexivity can only be revealed by tests in- volving the reflexive possessive pronoun, which can occur in nominative NPs. The use of tests in- volving reflexive possessive pronouns is not quite unproblematic: for Russian, Rappaport (1986) claims that the reflexive possessive may have arbitary antecedents. These cases, however, seem to involve emphatic (“of one’s own”) rather than properly reflexive uses, and the antecedents are always generic. No uses of this type occur in the material dealt with here. Where tests with reflexives proper and reflexive possessives yield different results, this will be noted.
In the case of (1), the reflexivization criterion leads to difficulties: It is also possible to find examples where the stimulus argument with patikt controls refle- xivization, although only when certain additional conditions, to be discussed below, are satisfied:
(6) Pirmdzimtais var būt krietns lielais brālis vai laba lielā māsa.
Tād-s viņ-š patīk sav-iem
such-nom.sg.m 3-nom.sg.m please.prs.3 refl.poss-dat.pl vecāk-iem.
parent-dat.pl
‘The firstborn can be a decent elder brother or a good elder sister. That’s how his parents like him.’
http://wow7.blogs.lv/2010/11/10/cela-cirtejs-diplomats-vai-dumpinieks-1/
(accessed July 2012)
We will obviously need some criteria to establish which of the noun phrases in (1) should be interpreted as subject. These criteria should apply to all three types of constructions illustrated here, and they should, of course, be consistent with criteria that could be formulated for other languages with “non-canonical sub- jects”. The aim of the article will be, then, to test explanations that have been formulated elsewhere on Latvian, and to characterize the evidence Latvian can contribute to a satisfactory account of datival quasi-subjects.
The structure of the article is as follows. First, I will clarify the basic notions I will be invoking, especially that of obliqueness as understood in the article.
Next, I will discuss grammatical relations in the three clause types illustrated by examples (1)–(3). I will show that they display no clear subject properties in any argument, that they are also intransitive, and that they can be described in terms of a demoted intransitive subject occurring alongside a less-oblique datival quasi-subject. A comparison of the three types discussed in the article is given in the final part, and some generalizations are attempted.
2 The framework
The account I will propose for a number of Latvian constructions without nomi- native subjects relies heavily on the notion of obliqueness. Obliqueness is a hie- rarchy of noun phrases in a clause that, but for the fear of causing confusion with levels of structure in phrasal syntax, we could formulate as based on “relative depth of syntactic embedding”. Rather than to putative levels of phrase structure, obliqueness refers to the empirically verifiable differences in “ accessibility”
(Keenan & Comrie 1977): Less oblique NPs are more easily accessible to various syntactic and morphosyntactic operations than more oblique ones. Obliqueness
is a composite notion comprising at least the following elements: (i) relative position in the topic-comment structure, this structure being conceived not as dichotomous but as a hierarchy of primary and secondary topics (for a recent study highlighting the importance of secondary topics, cf. Dalrymple & Nikolaeva 2011); (ii) semantic role (theta role), a level of semantics whose hierarchical struc- ture is widely recognized; (iii) inherent categorial features related to animacy and individuation; (iv) empathy (in the case of symmetrical predicates like John met Mary, where the hierarchy reduces to differences of perspective or vantage point).
The obliqueness hierarchy has been invoked to explain not only accessibi- lity differences, but also universal regularities of word order (Pullum 1977) and binding (Pollard & Sag 2004). The notion may therefore be considered to be well- established in the literature. In what follows, however, I will slightly modify it.
The obliqueness hierarchy is usually formulated as a hierarchy of grammati- cal relations (actually Keenan and Comrie’s accessibility hierarchy is formulated in this way), but I suggest there is only a default correspondence between gram- matical relations and positions in the obliqueness hierarchy. By default, the least oblique NP becomes clause subject, but it is possible for the relation of subject not to be conferred at all (or, to be more precise, whether and to what extent it must be conferred will be a matter of cross-linguistic variation). Therefore, we will say that the obliqueness hierarchy constitutes the foundation of the hierar- chy of grammatical relations and grammatical relations are its principal manifes- tation, but the two should not be completely identified. Alongside least-oblique arguments that are subjects, we will also have least-oblique arguments that are subject-like but lack one or more features of fully fledged subjecthood (especially coding properties), thereby justifying terms like “quasi-subject”.
I assume that the obliqueness hierarchy is valid cross-linguistically, inde- pendently of language-specific morphosyntactic marking. For example, the same obliqueness pattern is realized morphosyntactically in the English construction I like this book and underlies linguists’ intuition that the datival argument in Latvian example (1) is also some kind of subject – an intuition without which the whole discussion on non-canonically marked subjects would probably never have started.
As will be clear from what was said above, I do not regard obliqueness as a primitive notion: It can certainly be decomposed into more elementary prin- ciples. The default correspondences between the different hierarchies may be overridden, e.g., an NP that would be least oblique in virtue of being animate and an agent or experiencer may be outranked in topicality when a shift in topic- comment structure occurs in a specific type of communicative situation (we will see examples of this further on); still, it is the default assignment of topichood, not that associated with a particular utterance, that will be grammaticalized.