Variation and Typology: New trends in Syntactic Research

Helsinki, August 25.-27.2011

i

Table of Contents

PLENARY PRESENTATIONS .............................................................................................................................................1

BICKEL, BALTHASAR .........................................................................................................................................................1 Putting Variation Center Stage: Beyond ‘Language’ (Or ‘Dialect’) as the Basic Data Unit in Typology (and Elsewhere)

BRESNAN, JOAN .............................................................................................................................................................2 The Development of Syntactic Variation in the Individual: Are There Implications for Typology?

HELASVUO, MARJA-LIISA ..................................................................................................................................................3 Subject Expression in Finnish: Interactional Perspectives

WORKSHOP ON FINNISH DIALECT SYNTAX ....................................................................................................................4

PRESENTATIONS ............................................................................................................................................................5

ANDERSON, GREGORY D. S. ..............................................................................................................................................5 Doubled, Split and Split/Doubled Inflectional Patterns in Auxiliary Verb Constructions

ADLI, ARIA ....................................................................................................................................................................7 Variation and Information Structure

AKHLAGHI, FARYAR; ASADPOUR, HIWA ................................................................................................................................9 A Functional-Typological Study of Complementation in Contemporary Persian and Two Dialects of Kurdish: A Comparative Study

BEHNKE, LARS ..............................................................................................................................................................12 Variation as a Stabilizing Factor? Evidence from Contact-Influenced Syntactic Variation in Eastern Polish Dialects

CREISSELS, DENIS ..........................................................................................................................................................13 Syntactic Contrasts between Closely Related Dialects: Mandinka Antipassive and Causative

FORKER, DIANA; KHALILOVA, ZAIRA ..................................................................................................................................15 Microtypology and the Tsezic Languages: A Case Study of Syntactic Properties of Converbal Clauses

FORSBERG, HANNELE .....................................................................................................................................................17 The structure of Focused yes-no Questions in Finnish Dialects

FORSBERG, HANNELE; SIIROINEN, MARI; VILKUNA, MARIA ....................................................................................................18 Impersonal Readings of Third Person Plural in Context

ERSCHLER, DAVID .........................................................................................................................................................20 Cross-Dialectal Variation in Person Case Constraints: Evidence from Ossetic

GRAFMILLER, JASON; SHIH, STEPHANIE ..............................................................................................................................22 New Approaches to End Weight

HACK, FRANZISKA MARIA ...............................................................................................................................................24 Between Dialect Syntax and Typology – Variation in Dolomitic Ladin Question Formation

HAIG, GEOFFREY...........................................................................................................................................................26 Word-Order Variation and Change in Real Time: Areal Influence and Universal Factors

HALLA-AHO, HILLA ........................................................................................................................................................28 Why Did the Romans Misuse Their Accusatives?

HAMUNEN, MARKUS .....................................................................................................................................................30 Non-Finites of MANNER/MEANS in Finnish Cross-Linguistically

HIIETAM, KATRIN ..........................................................................................................................................................31 The Estonians and the Estonian Language in Northern England

HILL, VIRGINIA .............................................................................................................................................................33 Variation in Verb Movement in Old Romanian

ii

HILLBERG, SANNA .........................................................................................................................................................35 Zero Relativization in Educated Written Scottish English

HUANG, WANTIN STACY ................................................................................................................................................36 Accusative or Ergative or Neither: Justifying for Symmetrical Voice System in Tao

HÅKANSSON, DAVID ......................................................................................................................................................38 A Variationist Approach to Syntactic Change

KARTTUNEN, MIIA ........................................................................................................................................................39 Syntactical Observations on the Future Time Reference in a Language without a Future Tense

JENDRASCHEK, GERD .....................................................................................................................................................40 Core versus Periphery in Basque Morphosyntax

LINDSTEDT, JOUKO ........................................................................................................................................................42 Dialect Syntax or Bad Translation? Interpreting Old Dialect Texts Written in a Multilingual Setting

LANGE, CLAUDIA; BERNAISCH, TOBIAS ...............................................................................................................................43 The Typology of Focus Marking in South Asian Englishes

LEINO, JAAKKO .............................................................................................................................................................44 Variation, Diffusion, and Typological Implications: Finnish Spatial Particles as Typological Data

LINDSTRÖM, LIINA ........................................................................................................................................................45 Patient-like Experiencer Construction in Estonian

MANKILA, ANJA ...........................................................................................................................................................46 Factors in the Choice of the Case of Nouns after the Pronoun čto (what) in Russian

NEDJALKOV, IGOR .........................................................................................................................................................47 Variation of Recipient Marking in Evenki (in comparison with Other Tungus-Manchu Languages)

NOEL, PATRIZIA ............................................................................................................................................................49 The Emergence of V1 Polar Questions in Germanic

ODRIA TUDANCA, ANE; BERRO URRIZELKI, ANE ...................................................................................................................50 Subject and Object Marking Alternation in Bivalent Predicates: Dialectal Evidence from Basque

PEKKARINEN, HELI.........................................................................................................................................................52 The grammaticalization from a Pronoun se 'it' to an Expletive in Finnish Necessive Constructions. Evidence from Dialect Data.

PELTOLA, REA ..............................................................................................................................................................54 Finnish Jussive Mood in Dialects and in Standard Language

PEŠKOVÁ, ANDREA ........................................................................................................................................................56 Overt and Null Subject Pronouns in Experimental Data: The Case of Porteño Spanish

RIIONHEIMO, HELKA ......................................................................................................................................................58 Colliding Passives: A Morphosyntactic Perspective on Finnic Language Contacts

ROSENKVIST, HENRIK .....................................................................................................................................................59 Diachronic Typology and Synchronic Variation – Referential Null Subjects in Old and Modern Germanic

SCHIKOWSKI, ROBERT ....................................................................................................................................................61 Obligatory Coreference in Chintang

SINNEMÄKI, KAIUS ........................................................................................................................................................63 Where to Find Language Universals?

SEILER, GUIDO .............................................................................................................................................................64 Non-Standard Average European: How Robust Is the European Sprachbund beyond the Codified Standard Varieties?

SIDORKEVICH, DARIA .....................................................................................................................................................66 Some Notes on Ingrian Finnish in Western Siberia

SIIROINEN, MARI ..........................................................................................................................................................67 Constructions of the Type ”X becomes Y” in Finnish Dialects

TOKIZAKI, HISAO ...........................................................................................................................................................68 Correlation between Word Order and Phonology: Variation in Dialects and Languages

iii

USSERY, CHERLON ........................................................................................................................................................70 Variation in Icelandic Agreement: Multiple Agree, Multiple Grammars

VERHOEVEN, ELISABETH .................................................................................................................................................72 Word Order Variation with Non-Canonical Arguments: A Parallel Corpus Study

VIERROS, MARJA ..........................................................................................................................................................73 Contact-Induced Variation in Hellenistic Egypt: Pronominal Suffixes and Other Constructions

VILKUNA, MARIA ..........................................................................................................................................................74 Between Dislocation and Expletive

WEBER, TOBIAS............................................................................................................................................................76 Variation in Agent Marking Conditioned by Clausal Properties

1

Plenary Presentations

Bickel, Balthasar University of Zürich

Putting Variation Center Stage: Beyond ‘Language’ (Or ‘Dialect’) as the Basic Data Unit in Typology (and Elsewhere)

It is commonly assumed in typology (and elsewhere) that the right level for collecting and aggregating data is defined by units like ‘language’ (or ‘dialect’): most linguistic analyses are said to hold of a language (or dialect), and the basic data unit of typological databases is most often languages (or dialects). In the face of ubiquitous variation within languages and dialects, I propose to move away from this assumption and instead take it as an empirical (statistical) question to what extent a unit such as a specific language shows a bias towards a modal value or type and does not reflect a uniform distribution of many values/types in a given linguistic variable. If there is a bias in a unit (e.g. towards accusative case alignment across the variants created by split systems in a language; or towards a certain degree of using overt NPs across speakers of a language), this suggests that a preference has developed in that unit and that the unit has in fact shaped the distribution.

Languages (and dialects) are not the only units that shape the distribution of linguistic variables. In fact, two other kinds of units tend to be of greater interest than languages and dialects:

1) The distribution of variables can be shaped by larger genealogical units than languages, such as entire families. Biases in such units (e.g. a general bias towards accusative alignment in Indo-European) reflect diachronic preferences within the units: retention if the proto-language already showed the preferred structure; or innovation if the the proto-language was different. If the same bias holds across many families, or across most families under certain conditions (e.g. certain word order patterns), this reflects a diachronic universal.

2) The distribution of variables can be shaped by other (linguistic or nonlinguistic) structures, without languages developing norms or families developing diachronic preferences. This happens when exposure to certain structures leads speakers to adopt systematic habits, such as habits in NP usage that directly reflect speakers' specific social environment and the processing of certain agreement constructions. If the same effects are present around the world, this reflects a synchronic universal.

I will illustrate these phenomena by recent case studies, showing that rather than being an

obstacle for typological generalizations, research on variation provides a unique opportunity to better understand such generalizations.

2

Bresnan, Joan Stanford University

The Development of Syntactic Variation in the Individual: Are There Implications for Typology?

TBA

3

Helasvuo, Marja-Liisa University of Turku

Subject Expression in Finnish: Interactional Perspectives

This paper explores subject expression in Finnish. In Finnish, the predicate verb agrees with the subject in number (singular vs. plural) and person (1st, 2nd and 3rd; see e.g. Sulkala & Karjalainen 1992). In standard Finnish, (non-3rd-person) subjects are generally not expressed, and verbs are marked for per¬son and number of the subject. When subjects are overtly expressed, however, they serve some spe¬cial discourse function (e.g. contrast). Interestingly enough, the colloquial varieties differ from the written standard in the patterns for subject expression, as it is more common for both the subject pronoun and verbal person marking to be overtly expressed. This paper focuses on patterns subject expression in conversational Finnish.

Both double-marking (pronominal subject + verbal person marking) and single-marking (verbal person marking only) are grammatically possible in conversational Finnish, but previous research (see Duvallon 2006, He¬lasvuo 2009) suggests that the expression vs. ellipsis of subject is contextually constrained. On the basis of a syntactically coded database, I will discuss the variation in subject marking in conversational Finnish from the point of view of the grammatical and interactional context. I have analyzed the contexts for subject expression both in terms of interactional organization (sequential structure, social action) and grammar (word order, verb type, presence of object).

In the literature, several principles have been proposed which guide the choice of referential form in person reference. Sacks and Schegloff (1979) maintain that there is a general preference for minimization in reference to persons. They also note a general preference for recipient design: speakers should use reference forms which allow the recipient to identify (“recognize” in Sacks and Schegloff’s terminology) the referent (see also Hacohen & Schegloff 2006). Similarly, Levinson (2007) sets forth several principles which according to him guide the indexing of person, most importantly, economy, recognition and circumspection. I will discuss these principles and show that they are clearly problematic in the light of conversational Finnish.

4

Workshop on Finnish dialect syntax

The workshop brings together researchers working with questions relating to syntactic phenomena in Finnish dialects and presents new empirical results from FinDiaSyn, a four-year project on Finnish dialect syntax started in 2008. The Fennistic research tradition is like that of a number of other languages in that while much weight has been given to dialectology, the work has concentrated on data collection and phonological, morphological, and lexicological questions. As a result, a respectable amount of corpus material of Finnish dialects has been collected, and the morphophonological history behind Finnish dialects has been widely discussed. There is also a rich body of results on present-day sociolinguistic variation in morphophonological features. Much less has been said and done with regard to syntactic phenomena in Finnish dialects.

Accordingly, the prevailing view of Finnish is largely based on historical morphology, on the one hand, and standard language syntax, on the other. While this view is certainly correct with regard to standard written Finnish, much existing variation is lost by overlooking the variation underlying the standard variety. This is especially regrettable given that standard varieties of different languages tend to show similarities which may not have a basis in the vernacular varieties but, rather, in the European literary tradition and a common tradition of standardization.

The papers presented in the workshop aim at questioning the existing standard view of the Finnish language: To what extent can we accept the received typological picture of Finnish at face value? And to what extent should we alter the standard view?

Dialect syntactic research has been very lively throughout Europe in recent years (see e.g., http://www.dialectsyntax.org/, http://uit.no/scandiasyn/). The research objectives have largely arisen out of the Generative/Minimalist paradigm, and much useful experience has been gathered of careful questionnaire-based methods. While dealing with similar questions, the papers in the present workshop are more usage-based in orientation, operating in a constructionist-functional-cognitive framework and drawing from existing databases of natural discourse. These databases are rich as such but fairly narrow in terms of genre or situational variables, which has necessitated considering new data gathering methods. Questions of data and methodology have therefore a prominent place in the workshop.

5

Presentations

Anderson, Gregory D. S. Living Tongues Institute for Endangered Languages and National Geographic Society

Doubled, Split and Split/Doubled Inflectional Patterns in Auxiliary Verb Constructions

Auxiliary Verb Constructions [AVCs]–verbal constructions consisting of a lexical verbal element and a functional verbal element (the auxiliary)–are widespread among the languages of the world. When viewed cross-linguistically, the functional categories expressed by AVCs cover the full range of tense, aspect, Aktionsart, and mood categories as well as less common categories such as negation, orientation/grammaticalized deixis, version/grammaticalized discourse salience, and even adverbial notions of various sorts. In terms of linear syntax, both AUX V and V AUX orders are widespread. Frequently (but not exclusively) AVCs pattern in a manner analogous to the relative order between a transitive verb and its associated subcategorized object(s), with the auxiliary verb standing in the ‘syntactic head’ position, e.g., in OV languages one often finds V AUX structure while in VO languages the order is commonly AUX V–a distribution known already at least since Greenberg (1963/1966). The ‘semantic head’ of an AVC is the lexical verb, while the auxiliary adds a functional specification to the overall meaning of the construction. Thus, the English sentence she has eaten an apple predicates of an event of ‘eating’, not of an event of ‘having’, etc.

Thus, AVCs appear to be relatively straightforward in terms of accounting for the types of variation seen in their linear syntactic and semantic structure. Two areas where this is not true is in the encoding of inflectional categories within AVCs, and in the types of source constructions that get grammaticalized or functionally specialized as AVCs. In terms of the source constructions for AVCs, one finds not only the verb+complement structures that typify most European languages and better known Asian languages like Hindi, Korean, or Japanese, but also serial verb constructions (Givón 2009) and clausechaining formations as well (Anderson 2006). The distribution of patterns of inflection in the morphosyntax of AVCs is even more varied, showing at least five common macropatterns, three of which are presented in this paper. The familiar structure–in which the auxiliary is the ‘inflectional head’ (encoding the categories obligatory to make the construction grammatical) while the lexical verb appears in a construction-determined, dependent or non-finite form (1)–is but one of these patterns, called the ‘AUX-headed pattern’ in the most recent comprehensive study of AVCs cross-linguistically (Anderson 2006). As this is a well-known pattern and well accounted for in various formal approaches to the syntax of AVCs, I focus here on three other common inflectional patterns seen in AVCs. In the doubled inflectional pattern, obligatory inflectional categories necessary to render the construction grammatical are doubled, appearing on both the lexical verb and the auxiliary verb (2). In the split pattern, there are two subtypes of inflectional markers, one of which only occurs with the lexical verb and one only with the auxiliary verb (3). Finally, in the split/doubled inflectional pattern, some categories are limited to either the lexical verb or the auxiliary verb, while others appear doubly marked (4). In all instances, lexical verbs can appear in a given AVC in a construction-determined non-finite form so thus appear to be formally ‘semi-finite’. In this study, I present these three patterns in some detail. There are clear correlations between the source constructions and the inflectional pattern seen in the grammaticalized AVC, which helps explains in part the variation attested.

6

Bibliography Anderson, G. D. S. 2006. Auxiliary Verb Constructions. Oxford: Oxford University Press. Chung, C-H. and K-J. Chung. 1996. Kuot grammar essentials. In J. Clifton (ed.) Two Non-

Austronesian Grammars from the Islands, 1-75. Ukarumpa: SIL. Givón, T. 2009. Multiple routes to clause union: The diachrony of complex verb phrases. In Givón,

T. and M. Shibatani (eds.) Syntactic Complexity, 81–118. Amsterdam: John Benjamins. Greenberg, J. H. 1963/1966. Some universals of grammar with particular reference to the order of meaningful elements. In J. H. Greenberg (ed.) Universals of grammar, 2nd edition, 73-113.

Cambridge, Mass: MIT Press. Heine, B. 1993. Auxiliaries. Cognitive Forces and Grammaticalization. New York: Oxford

University Press. Keegan, J. 1997. A Reference Grammar of Mbay. Munich: Lincom Europa. Leyew, Z. 2003. The Kemantney Language. A Sociolinguistic and Grammatical Study of Language Replacement. Cushitic Language Studies 20. Köln: Rüdiger Köppe Verlag. Lukas, J. 1971. Die Personalia und das primare Verb im Bolanci (Nordnigerien). Afrika und

Übersee 55: 114-139. Lawes, W. G. 1896. Grammar and Vocabulary of the Motu Tribe. Sydney: Charles Potter. Nurse D. 2008. Tense and aspect in Bantu. Oxford: Oxford University Press. Rassadin, V. I. 1978. Morfologija tofalarskogo jazyka v sravnitel’nom osveshchenii [The

morphology of Tofa in a comparative light]. Moscow: Nauka. Sharma, D. D. 1988. A Descriptive Grammar of Kinnauri. Studies in Tibeto-Himalayan Languages

1. Delhi: Mittal. Tucker, A. N. and M. A. Bryan 1966. Linguistic Analyses: The Non-Bantu Languages of North-

Eastern Africa. London: Oxford University Press.

7

Adli, Aria Universität Freiburg

Variation and Information Structure

The large question of this paper is: What does variation tell us about the relation between grammar and usage?

The working hypothesis is that grammar and usage are clearly distinct (Newmeyer 2003), but that both are essential elements of human syntax with information structure being the bridge between them.

This issue is investigated by means of a quantitative study on optional subject pronoun realization in Spanish, complemented by a comparable pilot study on Persian. Both are consistent null subject languages. I will analyze three dimensions of variation and discuss their mutual relations:

The first dimension is typological: Spanish is a subject-prominent and Persian a topic-prominent language (Karimi 2005). This dimension refers to a cross-linguistic difference in the relation between subject and topic (Li and Thompson 1976; Maslova and Bernini 2006).

The second dimension of variation is language-internal: Here, I concentrate on information structure, namely the difference between familiar topics (signaling continuity of reference) and shift topics (signaling new topic referents). This distinction has already been relied on by Keenan (1976) and Givón (1976). I take up Frascarelli’s (2007) reformulation of the classic Avoid Pronoun rule, which essentially states that the contrast between overt vs. null subject pronouns correlates with the contrast between shift vs. familiar topics.

Finally, the third dimension is social. I will discuss systematic correlations between different social and demographic variables with the overt pronoun rate (see e.g. Otheguy et al. 2007).

The data is extracted from the sgs database (Adli 2011). It consists of 54 interviews with native speakers recorded in Spain and 20 interviews with native speakers recorded in Iran. The results for each language show that the language-internal dimension is significant. However, the patterns of variation differ between Spanish and Persian. This cross-linguistic difference will be discussed in the light of the typological contrast between subject-prominence and topic-prominence. I will argue that the binary nomenclature itself (subject- vs. topic-prominence) is problematic.

The social dimension is studied with the larger Spanish sample. The results show a significant interaction between economic power in terms of income and the type of topic.

I will propose that the complex picture of language-internal and social variation can be modeled, if we assume that the same grammar of one language or dialect can be used differently, depending on the speaker. Furthermore, we explain the idea of different uses of the same grammar by systematic differences in the speaker’s strategies of information packaging, reflected in the interaction between the social dimension and topic type. In other words: Information structure is the bridge between grammar and usage.

8

Bibliography Adli, Aria. 2011. Gradient Acceptability and Frequency Effects in Information Structure: a

quantitative study on Spanish, Catalan, and Persian. Habilitationsschrift, Universität Freiburg.

Frascarelli, Mara. 2007. Subjects, topics and the interpretation of referential pro: An interface approach to the linking of (null) pronouns. Natural Language and Linguistic Theory 25(4). 691–734.

Givón, Talmy. 1976. Topic, Pronoun, and Grammatical Agreement. In Charles N. Li (ed.), Subject and Topic, 149–188. New York, San Francisco & London: Academic Press, Inc.

Karimi, Simin. 2005. A Minimalist Approach to Scrambling: Evidence from Persian. Berlin: Mouton de Gruyter.

Keenan, Edward L. 1976. Towards a Universal Definition of "Subject". In Charles N. Li (ed.), Subject and Topic, 303–334. New York, San Francisco & London: Academic Press, Inc.

Li, Charles N. & Sandra A. Thompson. 1976. Subject and Topic: A New Typology of Language. In Charles N. Li (ed.), Subject and Topic, 457–490. New York, San Francisco & London: Academic Press, Inc.

Maslova, Elena & Giuliano Bernini. 2006. Sentence topics in the languages of Europe and beyond. In Giuliano Bernini & Marcia L. Schwartz (eds.), Pragmatic Organization of Discourse in the Languages of Europe, 67–119. Berlin & New York: Mouton de Gruyter.

Newmeyer, Frederick J. 2003. Grammar is Grammar and Usage is Usage. Language 79(4). 682–707.

Otheguy, Ricardo, Ana Celia Zentella & David Livert. 2007. Language and Dialect Contact in Spanish in New York: Toward the Formation of a Speech Community. Language 83(4). 770–802.

9

Akhlaghi, Faryar; Asadpour, Hiwa Research Institute to the Iranian Cultural Heritage, Handicrafts and Tourism Organization, Tehran, Iran. Linguistic Department, Science and Research branch, Islamic Azad University, Tehran, Iran.

A Functional-Typological Study of Complementation in Contemporary Persian and Two Dialects of Kurdish: A Comparative Study

Complementation refers to a situation that a verb in place of one of its obligatory arguments takes a clause instead of a nominal group, and the presence of that clause is essential for wellformedness of the sentence. In the present study, a functional-typological approach has been applied, in order to study and compare complementation in contemporary Persian, Sorani and Kurmanji Kurdish in Northwest of Iran. These variations have been selected due to some features like accusativity and ergativity. The variations in North Western part of Iran still preserved some original features like case, ergativity etc. while in some other regions like Iraqi Sorani Kurdish or Sorani in Kurdistan region of Iran because of cross linguistic phenomenon some changes appeared and many original characteristics have been lost.

In functional-typological approach the main underlying assumption in investigating complementation is that the semantic features of the predicate of the main clause has a major roll in determining the structure of the complement clause. Henceforth, complement relations will be identified on the basis of semantic classes of the predicate coding the main clause. Further, in each semantic class the constructions used to code dependent events are classified with respect to one basic parameter, whether or not they differ in structure from an independent declarative clause taken in isolation.

Derivation from the independent declarative clause standard is measured with respect to two parameters: verb form and participant coding. Verbs coding dependent events may differ in structure, for example they might not display all of the categorical distinctions (tense, aspect, mood, and person) which are allowed to verbs coding independent events. If so, they are called ‘deranked’ and if they have the same structure as the independent clause verb they are called ‘balanced’. Regarding participant coding it should be noted that the participants of the dependent event may not be expressed in the same way as independent events, for example they might be not expressed overtly.

Results show that in case of contemporary Persian complementation obeys the subordination Deranking Hierarchy and it supports the trace of typological findings. The research also shows that in all of its complement relations the prevailing tendency is to omit the subject of the complement clause, and if it is coreferent with the subject of the main clause usually it has to be omitted. The contemporary Kurdish esp. case of Sorani exhibits both ergative and nonergative patterns in its grammatical structures. Split ergativity in Sorani, based on the overlapping syntactic categories of aspect and clause type, gives rise to a theoretical structural dilemma that is resolved in distinct way by North and South Kurdish due to language contact. In some cases it obeys the same structure as in Persian and in some other cases vice versa. Above all, although Kurmanji is another dialect of Kurdish and in Sprachbund with Sorani and some other neighboring languages, the structures in this ergative dialect is quite in opposition with those of Persian and even Sorani Kurdish. The findings show that in Persian in semantic classes of modals, manipulatives with positive implication and manipulative without implication two subjects cannot be coreferent and in other classes they can and as for two other languages under investigation, Sorani beside differences with Persian it has some similarities as well while in Kurmanji we have a different phenomena with those of Sorani and Persian.

10

Bibliography Aarts, B. 2006. ‘‘Subordination’’, Encyclopedia of Language and Linguistics, Edited by: Keith

Brown, Boston: Elsevier, 248-254. Browning, M. and E. Karimi. 1990. “Scrambling in Persian”, Paper presented at the Tilburg Scrambling Conference. Bachlh, K., 1992. TiirkGe izahh Kiirtqe grameri (Kiirman~~a lehqesi). Istanbul: MED. [first

published 19651 Bamas, U. and J. Salzer, 1994. Lehrbuch der kurdischen Sprache. Berlin: Wegan$n instituya Kurdi.

Bedir Khan, E.L. and R. Lescot, 1986. Kurdische Grammatik Kurmanci-Dialekt. Bonn: Verlag fiir Kultur und Wissenschaft. [German translation of the French original (Paris, 1970)].

Bynon, T., 1980. ‘‘From passive to active in Kurdish via the ergative construction’’. In: E. Traugott, R. Labrum and S. Shepherd (eds.), Papers from the 4th International Conference on

Historical Linguistics, 151-163. Amsterdam: Benjamins. Cristofaro, S. 2005. Subordination. Oxford: Oxford University Press. Croft, W. 2003a. Typology and Universals. Cambridge: Cambridge University Press. --------. 2003b. ‘‘Typology’’, The Handbook of Linguistics, Edited by: M. Aronoff and J. Rees-

Miller, UK: Blackwell, 337-369. Dik, S. C. 1989. The Theory of Functional Grammar: Part 1: The Structure of the Clause.

Dordrecht: Foris. -------- 1997a. The Theory of Functional Grammar: Part 1: The Structure of the Clause. (2nd revised

edition) ed. Kees Hengeveld. Berlin and New York: Mouton de Gruyter. -------- 1997b. The Theory of Functional Grammar: Part 2: Complex and Derived Constructions.

(2nd revised edition) ed. Kees Hengeveld. Berlin and New York: Mouton de Gruyter. Dixon, R. M. W. 2006. ‘‘Complementation’’, in . Complementation, Edited by R.M.W. Dixon and A. Y. Aikhenvald, Oxford: Oxford University Press, 1-50. Dixon, R. M. W. and A. Y. Aikhenvald. 2006. Complementation. Oxford: Oxford University Press. Dorleijn, M., 1996. The decay of ergativity in Kurdish. Language

internal or contact induced? Tilburg: Tilburg University Press. Edmonds, C.J. 1955. Prepositions and personal affixes in Southern Kurdish. Bulletin of the School

of Oriental and African Studies XVII:490–502. Fattah, Isma¨ıl Kamandˆar. 2000. Les dialectes kurdes m´eridionaux. Etude linguistique et

dialectologique. Leuven: Peeters. Fattah, Muhammad Maruf. 1997. A generative grammar of Kurdish. University of Amsterdam: Unpublished PhD Thesis.

Foley, W. A. and R. D. Van Valin, Jr. 1984. Functional Syntax and Universal Grammar. Friend, Robyn. 1985a. “Ergativity in Suleimaniye Kurdish”. Unpublished paper presented at the

Middle East, Studies Association Annual General Meeting, November 1985 (copy available at http//www.home.earthlink.net/˜rcfriend/esk.htm) .

Friend, Robyn. 1985b. Some syntactic and morphological features of Suleimaniye Kurdish. Unpublished PhD thesis: UCLA. Cambridge: Cambridge University Press.

Givon, T. 1980. “The binding hierarchy and the typology of complements”, Studies in Language 4, 333-77.

-------- 1990. Syntax: A Functional-Typological Introduction Vol.2. Amsterdam and Philadelphia: John Benjamins.

-------- 2001. Syntax. 2 Vols. Amsterdam and Philadelphia: John Benjamins. Green, G. M. 1976. “Main clause phenomena in subordinate clauses”, Language 53, 382- 97. Haig, Geoffrey. 1998. ‘‘On the interaction of morphological and syntactic ergativity: Lessons from

Kurdish’’. Lingua 105:149–173. Haig, Geoffrey. 2001. ‘‘Linguistic diffusion in present-day East Anatolia: From top to Bottom’’. In

Areal diffusion and genetic inheritance: problems in comparative linguistics, ed. Alexandra

11

Aikhenvald and R.M.W. Dixon, 195–224. Oxford: Oxford University Press. Haiman J. and S. A. Thompson 1984. ‘ “Subordination” in universal grammar’, in

Proceedings of the Tenth Annual Meeting of Berkeley Linguistic Society, pp. 510-23. Haiman, J. 1985. Natural Syntax. Cambridge: Cambridge University Press. Hajati, A.K. 1977. “ke construction in Persian: descriptive and theoretical aspects”, Unpublished

PhD dissertation, University of Illinois, Urbana. Halliday, M.A.K.2004. An Introduction To Functional Grammar (Third Edition). London: Arnold. Huddleston, R. and G. K. Pullum. 2002. The Cambridge Grammar of the English Language.

Cambridge: Cambridge University Press. Karimi, S. 1989. Aspects of Persian syntax, specify, and the theory of grammar, Ph.D. Dissertation,

University of Washington. Lambrecht, K. 1994. Information Structure and Sentence Form. Cambridge: Cambridge University

Press.Lambton, A.K.S. 1976.Persian Grammar. Cambridge: Cambridge University Press.

Langacker, R. W. 1987. Foundations of Cognitive Grammar. Vol. I: Theoretical Prerequisites. Stanford: Stanford University Press.

-------- 1991. Foundations of Cognitive Grammar. Vol. II: Descriptive Applications. Stanford: Stanford University Press. Lazar, G. 1957. Grammaire du Persian Contemporian. Paris: Librairie C. Kincksieck. Lehmann, C. 1988. “Towards a typology of clause linkage”, in J. Haiman and S. A. Thompson

(eds), Clause Combining in Grammar and Discourse, pp. 181-225. Amsterdam and Philadelphia: John Benjamins.

Miller, J.2006. “Subordination in spoken discourse”, Encyclopedia of Language and Linguistics. Edited by: K. Brown 14 vols. Boston: Elsevier.

Noonan, M. 1985. “Complementation”, Languge Typology and Syntactic Description, volume 2: Complex Constructions, Edited by: T. Shopen, pp. 42-140 Cambridge: Cambridge University Press.

-------- 2007. “Complementation”, Languge Typology and Syntactic Description, volume 2: Complex

Constructions, Edited by: T. Shopen, second Edition. pp. 52-151 Cambridge: Cambridge University Press.

Palmer, F. R. 1979. Modality and the English Modals. London: Longman. -------- 1994. “Mood and Modality”, Encyclopedia of Language and Linguistics, 10

vols: 2533- 2540. -------- 2001. Mood and Modality. Cambridge: Cambridge University Press. Quirk, R, Greenbaum, S, Leech, G and Svartvik, J. 1985. A Comprehensive Grammar of English

Language. London: Longman. Searle, J. R. 1983. Intentionality. Cambridge: Cambridge University Press. Shibatani, M. 1976. “The grammar of causative constructions: a conspectus”, in Mhibatani (ed.),

The Grammar of Causative Constructions, pp. 1-40. NewYork: Academic Press. Van Valin, Jr., R.D. and R. J. LAPollA. 1997. Syntax. Cambridge: Cambridge. Von Wright, E. H. 1951. An Essay in Modal Logic. Amsterdam: North Holland. Windfuhr, G.L.1979. Persian Grammar: History and State of Study, Trends in Linguistics, State-of-

the-art Report 12. The Hague: Mouton Publisher.

12

Behnke, Lars University of Oldenburg, Institute of Slavonic Studies

Variation as a Stabilizing Factor? Evidence from Contact-Influenced Syntactic Variation in Eastern Polish Dialects

In structural linguistics “free variation” is an inherent part of the language system (e.g. Trubetzkoy 1939). Dialectology or contact linguistics, on the other hand, focus on instances of variation that can be treated as symptomatic of their different genetic heritage, i.e. of different systems. Different degrees of fusion between two language varieties are measured against the optionality of forms in identical linguistic contexts.

This study focuses on internal variation in dialectal contact varieties and presents data from a corpus analysis of morphosyntactic variation between dative and prepositional constructions with dla ‘for’ plus genitive in Eastern Polish dialects (EPD) along the present Eastern Polish border. As a common feature all these varieties share a contact situation with varieties, genetically belonging to other languages, i.e. Lithuanian, Byelorussian, Ukrainian and Russian. In functional terms two forms that refer to the same thing (i.e. synchronic variation) are usually treated as a transitional stage within a diachronic differentiation process of the two markers. This study asks, however, if variation between two markers can have a stabilizing effect and provide a model in language contact situations.

While Standard Polish allows variation in the marking of the “third participant” in benefactive constructions, e.g. Kupiłem jejdat / dla niejprep kwiaty ‘I bought her flowers’, an analysis of 3200 dialect EPD-utterances reveals that the use of these two markers can be extended to other dative types here, e.g. the external possessor (Ścięli jejdat / dla niejprep głowę ‘They cut her head off’) or the recipient (Dali jejdat / dla niejprep książkę ‘They gave her the book.’). This phenomenon seems to have an areal rather than genetic extension, since it affects not only Polish, but also dialect varieties of all contact languages within the area, thus marking them off from all roofing standard languages.

As a marker of the external possessor the dative is in variation with the adessive preposition u ‘at’ in East Slavonic languages. The study shows that EPD dla can substitute contact-induced u whenever the latter is in variation with the dative. In other cases contact-induced u remains intact in EPD. The use of dla in u-contexts could be further motivated by the Byelorussian adessive preposition lja ‘at’, that shows some functional equivalence to u, but a formal one to dla. The structural model PREPOSITION-DAT for the external possessor provided by Eastern Slavonic languages together with the formal identification of lja and dla may be the basis for the extension of the dative-dla-variation model known to all contact varieties from other dative types (e.g. beneficiary, experiencer) to the external possessor type, where Standard Polish knows the dative only. On the basis of a model provided by Haspelmath (1999) it is argued that this model is further extended to the recipient type in EPD where the dative is the only option in all contacting languages.

The system-internal factors that govern the choice between two markers thus become relevant in a broader areal pespective which raises the question of their universal significance.

Bibliography Haspelmath, M. (1999): “External possession in a European areal perspective.” In: Barshi,

Immanuel; Payne, Doris L. (eds.), External Possession. Amsterdam, Philadelphia Trubetzkoy, N.S. (1939): Grundzüge der Phonologie. Prague.

13

Creissels, Denis University of Lyon

Syntactic Contrasts between Closely Related Dialects: Mandinka Antipassive and Causative

Mandinka, spoken by approximately 1.5 million speakers in The Gambia, Middle Casamance (Senegal), and Guinea Bissau, is the westernmost member of the Manding dialect cluster, included in the Western branch of the Mande language family. In general, dialectal variation in the syntactic structure of Manding is limited to minor details that are not relevant in a typological perspective. However, in some aspects of valency grammar, Mandinka sharply contrasts with all other Manding varieties, including those spoken in Eastern Senegal, which have a particularly close affinity with Mandinka. Mandinka constitutes the only Manding variety showing clear evidence of the antipassive origin of a suffix described in other varieties as a nominalization marker, and of a possible connection between causativization of transitive constructions and antipassivization.

In Mandinka, the use of most transitive verbs in their non-derived form (either as verbs or action nouns) is bound to strict conditions on the expression of the P argument, and a form marked by a suffix -ri is found exclusively in contexts in which these conditions are not met. -ri can therefore be identified as an antipassive marker, although it differs from typical antipassive markers in that, with just one exception, the ri-form is used exclusively as an action noun. The exception is dómo ‘eat’, whose antipassive form dómó-ri has a verbal use – Ex. (2).

The other Manding varieties have a suffix -li cognate with Mandinka -ri, but also found with intransitive verbs, and compatible with the expression of the second argument of transitive verbs (see a.o. Dumestre 2003: 74-5 on Bambara). Moreover, li-forms are never used as verbs.

The hypothesis that Mandinka represents a more ancient stage in the evolution of this suffix than other Manding varieties (and not the other way round) is supported by comparison with Sooninke (another language of the Western branch of the Mande family). As illustrated by Ex. (3) and (4), Sooninke has a canonical antipassive marker -ndì (Creissels 1991), probably cognate with Manding -ri ~ -li.

Mandinka causativization provides additional evidence of the antipassive origin of -ri ~ -li. Mandinka has two causative suffixes: -ndi, used mainly to causativize intransitive

constructions, and -rindi, used exclusively to causativize transitive constructions, as in Ex. (5). The other Manding varieties have a causative prefix la- functionally equivalent to Mandinka -ndi, but have nothing comparable to Mandinka -rindi.

In a synchronic perspective, the segmentation of fóoríndi ‘make tell’ as fóo-rí-ndi [tell-ANTIP-CAUS] is problematic, since -ndi can attach only to verbal stems, and fóo-ri [tell-ANTIP] can be used only as an action noun. Such a decomposition can however be reconstructed as the origin of the suffix -rindi, as suggested by dómóríndi ‘make eat’ – Ex.(6). In this particular case, the segmentation of -rindi as -ri-ndi [ANTIP-CAUS] is fully consistent with the syntactic properties of dómóri ‘eat (intr.)’ and dómóríndi ‘make eat’: -ri encodes the demotion of the object, making it possible for the initial subject to move to object position when a causer is introduced in subject position.

14

Examples

(1) Díndíŋ-o yé mbuur-ôo dómo. (Mandinka) child-DEF PF.POS bread-DEF eat ‘The child ate the bread.’ (2) Díndíŋ-o dómó-rí-ta. (Mandinka) child-DEF eat-ANTIP-PF.POS ‘The child ate.’ (3) Fàatú dà kómpè ŋ céllà. (Sooninke) Fatou PF.POS room DEF sweep ‘Fatou swept the room.’ (4) Fàatú séllá-ndì. (Sooninke) Fatou sweep-ANTIP ‘Fatou did the sweeping.’ (5) Kew-ó ye díndíŋ-o fóo-ríndi tooñáa la. (Mandinka) man-DEF PF.POS child-DEF tell-CAUS truth.DEF OBL ‘The man made the child tell the truth.’ (6) Kew-ó ye díndíŋ-o dómó-rí-ndi mbuur-ôo la. (Mandinka) man-DEF PF.POS child-DEF eat-ANTIP-CAUStruth.DEF OBL ‘The man made the child eat the bread.’

Abbreviations CAUS: causative, DEF: definite, OBL: postposition in oblique marker function, PF: perfective, POS: positive. Bibliography Creissels, Denis. 1991. ‘La voix en soninké’. Mandenkan 23. 1-24. Dumestre, G. 2003. Grammaire fondamentale du bambara. Paris: Karthala.

15

Forker, Diana; Khalilova, Zaira Max Planck Institute for Evolutionary Anthropology Max Planck Institute for Evolutionary Anthropology

Microtypology and the Tsezic Languages: A Case Study of Syntactic Properties of Converbal Clauses

This paper presents a case study of syntactic variation focusing on clause linkage in the Tsezic languages. The Tsezic languages are five closely related languages that form a subbranch of the Nakh-Daghestanian language family. They can be divided into East Tsezic, comprising Hunzib and Bezhta, and West Tsezic, comprising Khwarshi, Tsez, and Hinuq.

In all Tsezic languages converbal clauses are the main way of expressing adverbial relations between a dependent clause and a main clause. In addition, converbal clauses with ‘contextual’ (i.e. semantically vague) converbs are the main translation equivalents of clauses linked by coordination in most European languages.

In this paper we want to explore the syntactic properties of Tsezic converbal clauses. We focus on three types of syntactic properties:

(i) reference and control properties (i.e. subject coreference, zero and overt arguments in

converbal clauses) (ii) scope properties such as tense, evidentiality, illocutionary force, and focus (iii) word order and extraction We will show that the Tsezic languages behave alike with respect to many of the properties

listed in (i) - (iii), e.g. in all Tsezic languages adverbial clauses express at most relative temporal reference. The absolute temporal reference and the evidentiality value of the main clause also extend to the adverbial clause. But the Tsezic languages also display interesting variations in the syntax of clause linking. One domain of variation is control of pronominal cataphors. For example, although adverbial clauses usually precede their main clauses, in Bezhta and Tsez they may also follow them and contain an NP that controls a preceding pronoun (1a). Similar sentences in Khwarshi and Hinuq are ungrammatical if coreference is intended (1b).

In addition to the investigation of the microvariation within the Tsezic languages we will embed our findings into a broader typological perspective comparing it with the data presented in Bickel (2010).

Examples (1) Bezhta

a. hogoi/j y-uɣo-s [Žamilatii äč’enayig-na iłna ƛi she II-die-PRS Zhamilat.ERG ninety-and six year ömrö-nä b-oh-na]

life(III)-and III-do-PFVCVB ‘After Zhamilati lived for 96 years, shei died.’

Khwarshi

b. žuj/*i qʷaqʷaƛ-še Ø-eč-i [Naziri Ø-ot’q’-aƛa] he laugh-IPFVCVB I-be-WPST Nazir(I) I-come-ANT

‘Hej/*i was laughing when Naziri came.’

Bibliography

16

Bickel, Balthasar. 2010. Capturing particulars and universals in clause linkage: a multivariate analysis. Bril, I. (ed.) Clause-hierarchy and clause-linking: the syntax and pragmatics interface. Amsterdam: Benjamins, 51-101.

17

Forsberg, Hannele University of Eastern Finland

The structure of Focused yes-no Questions in Finnish Dialects

In standard Finnish yes-no (polar) questions are formed by attaching the question clitic -ko/-kö (CLT) to the first constituent of the sentence. When the truth of the whole proposition is being questioned, the first constituent is the finite verb (so the word order is inverted), as in (1), and when just a certain part is questioned, the first constituent is NP as in (2).

(1) Tuleeko se tänään? “Will it come today?” come-3SG-CLT it today

(2) Tänäänkö se tulee? “Is it today when it comes?”

today-CLT it come-SG

Empirical studies (e.g. Forsberg 2006 and sources therein) suggest variation in the formation of yes-no questions in regional dialects. In eastern and northern dialects, the system seems to resemble that of standard written Finnish, whereas south-western dialects exhibit notable differences. Instead of the type in (2), these dialects tend to use cleft structures, prosody, or the clause-final particle vai (originally “or”), as in (3, a constructed example) to mark the constituent that is in the focus of the question.

(3) Tuleek se tänään vai? “Is it today when it comes?” come-ind-3SG-CLT it today PRT

The present paper discusses some results from an empirical study concerning the formation of focused yes-no questions in spoken Finnish varieties. The only question type represented to some extent in the available dialect material (that is, recorded interviews consisting of narrative monologues) is a question that functions as a repair initiation. It appears when the interviewee considers a part of interviewer’s question unclear and makes an offer of interpretation before answering to the question.

The analysis of the data show regional variation in the formation of these questions. In all dialects a bare NP (e.g. Maria? ”You mean Maria?”) can be interpreted as a question in the basis of its sequential placement and the lexical content, but usually an additional element is added. In eastern and northern dialects this element is question clitic (-ko/-kö ~ -k ~ks) as in standard language (e.g. Mariako?), whereas in southern and western dialects the particle vai is used in the same position (Maria vai?). In addition, the data show that there are certain utterance-initial particles that function very similarly as question markers, and these particles seem to be regionally restricted. The utterance-final particle jaa (originally from Swedish ja “yes”) is used especially in the south-western and in the Häme area (Jaa Maria?), and particle niin (“so, yes”) especially in the Ostrobothnian dialects (Niin Maria?).

Bibliography Forsberg, Hannele 2006: Montako-kysymystyypin synty ja kehitys. Virittäjä: 365-396. (Abstract:

Origin and development of the interrogative form montako? “how many?”).

18

Forsberg, Hannele; Siiroinen, Mari; Vilkuna, Maria University of Eastern Finland University of Helsinki Research Institute for the Languages of Finland

Impersonal Readings of Third Person Plural in Context

Third person (3PL) forms are a cross-linguistically widespread resource for the expression of impersonal (arbitrary) reference (Siewierska 2004; Cabredo Hofherr (2006), Siewierska (2008), Siewierska and Papastathi 2011). This use is common in large areas of Eastern Europe, e.g., Polish, Russian, Hungarian and several other Uralic languages, in which it is typically confined to “pro-drop” instances such as Hungarian (1), but also occurs in Western Europe, e.g., in English, where an overt pronominal subject is used.

(1) Ellopták a biciklimet. steal-PST-3Pl DEF bicycle-3SGPOSS-ACC

‘My bike was stolen ~ They have stolen my bike.’

In Finnish, the same function is typically served by the impersonal passive (imp-passive), as in (2a) and (2b), but 3PL impersonals such as (2c) are found as well, particularly in Eastern dialects. The example is from an interview discussing traditional customs.

(2) a. Q: Kuivattiinko kalloo dry-PST.PASS-Q fish-PART ’Was there fish-drying?’ b. A: no 'kuivattaan/ "lahnankalat 'kuivattaa suuret

ja 'haovit./ PTL dry-PASS bream-fish-PL dry-PASS big-PL

and pike-PL ‘Yes there is, bream is dried, big ones, and pike.’ c. 'joka 'talossa 'teälläek kuivovvaat. each farm-INE here-TOO dry-3PL ‘They dry [them] at every farm here too.’ (Liperi, North Carelian)

The present paper discusses the status of 3PL impersonals as an Eastern Finnish phenomenon and methodological issues that arise in its interpretation. 3PL impersonals tend to preserve part of the meaning of their personal use, viz. exclusion of the speaker from the referential range. This accounts for the choice between 3PL and imp-passive; in (2), line (b) is about general customs in the speaker’s home region, while (c) conveys that the speaker positions herself as not belonging to the households in question. The exclusive/inclusive parameter may also help to explain why impersonal 3PL is more prominent in Eastern than Western Finnish. Imp-passive is used as the1st person plural form with a pronominal subject (‘we’) throughout colloquial Finnish. Since this development has started from Eastern dialects, the speaker-inclusive use of imp-passive may be more firmly entrenched in the east and may in turn have strengthened the use of a more distancing impersonal, the 3PL.

19

Although the existence of 3PL impersonals has been observed by Finnish linguists, the phenomenon has attracted little attention. This is presumably due to the existence and versatility of the imp-passive, an impersonal par excellence. In other words, while the impersonal passive (without an 1PL subject pronoun) always leaves open the interpretation of the agent, 3PL forms in principle always have both impersonal and anaphoric-referential readings. This makes it hard to assess the precise circumstances of its use, both when interpreting natural data such as (2) and when conducting data elicitation. In our paper, we contrast past-oriented interviews with elderly people with data from more varied situations such as service encounters and free conversation to see how impersonal interpretations are built.

Bibliography Cabredo Hofherr, Patricia 2006: ‘Arbitrary’ pro and the theory of pro-drop. In Ackema, Peter et al

(eds) Agreement and Arguments, pp. 230–246. Oxford: Oxford University Press. Siewierska, Anna, 2004: Person. Cambridge: Cambridge University Press. Siewierska, Anna 2008: Ways of impersonalizing: Pronominal vs verbal strategies. In María De Los

Ángeles Gómez Gonzáles & al. (eds.) Current trends in contrastive linguistics: functional and cognitive perspectives, pp. 3–26. Amsterdam: John Benjamins.

Siewierska, Anna & Papastathi, Maria 2011: Toward a typology of third person impersonals. Linguistics 43(2): 575–610.

20

Erschler, David Max Planck Institute for Developmental Biology

Cross-Dialectal Variation in Person Case Constraints: Evidence from Ossetic

I use evidence from two closely related East Iranian languages, Iron and Digor Ossetic1

to explore the variation in Person Case Constraint for cases other than the accusative and dative. ,

Unlike full pronouns, clitics are not freely combinable with each other. Constraints of this type have received the name of “PERSON CASE CONSTRAINTS” (PCC). (1) illustrates this for French:

(1) a. DO is the 1st or 2nd person. *Il me/te= lui= envoya he I/you.ACC he.DAT sent ‘He sent me/you to him.’ (Nicol 2005)

b. DO is the 3rd person. Il me/te= l’= envoya he I/you.DAT he.ACC sent ‘He sent him to me/you.’ (Ibid.)

PCC constraints have been intensely scrutinized in the literature (see, a.o., Bonet 1991, Haspelmath 2004, Anagnostopoulou 2005). It is well known that native speakers show a certain variation in their judgments (Bonet 1991). The extant studies deal mostly with combinations of the accusative with the dative, although the question makes sense for combinations of other cases too.

Digor and Iron clitic inventories are identical: enclitics exist for the accusative, ablative, allative, superessive, and inessive. For the “classical” combination of cases, i.e. DAT and ACC, the PCC constraint has the form2

(2) A clitic standing in case A should be higher or identical to the clitic standing in case B in thestandard hierarchy of persons (1>2>3).

This is valid for both idioms, and all consultants. For other cases, the situation is more variegated.

In Digor, constraint (2) is operative for the combinations of DAT with any other case (i.e. the accusative, ablative, inessive, allative, and superessive), and for the combination of ALL & ABL. Furthermore, ALL and SUP may not co-occur in a cluster at all. No robust restrictions are imposed on admissible values of person for the rest of case pairs. To illustrate the operation of the constraint (2) in Digor:

(1) a. ɐma=min=dɐbɐl či rcɐwuʒɐj? and=DAT.1SG=SUP.2SG what will.happen ‘What (should I expect) that will happen to you?’

b.*ɐma=din=mɐbɐl či rcɐwuʒɐj? and=DAT.2SG=SUP.1SG what will. happen ‘What (should you expect) that will happen to me?’ In Iron, constraint (2) is also operative for the combinations

1 Digor and Iron are often considered dialects. The data were obtained from native speakers (5 for Digor and 2 for Iron) and written texts. This type of constraint 2 This type of constraint is called “superstrong” PCC in the literature.

21

In Iron, constraint (2) is also operative for the combinations of DAT with any other case, and ALL cannot co-occur with Sup either. However, there exists an additional constraint: for ALL and ACC, ALL.3SG is incompatible with the ACC.1SG/PL and ACC.2SG/PL. The same constraint is operative for ABL & SUP. (It corresponds to what is called “Weak PCC” for the combination of DAT and ACC).

Ossetic data indicate that, first, the dative plays a pivotal role in PCCs, something which is not evident when the combination of only ACC and DAT is considered. Second, only already attested types of PCC recur for “non-canonical” cases. Third, robust constraints for rarely occurring combinations of cases testify against treating PCC as the grammaticalization of most frequent patterns (Haspelmath 2004).

22

Grafmiller, Jason; Shih, Stephanie Stanford University Stanford University; University of California Berkeley

New Approaches to End Weight

The Principle of End Weight maintains that constituents will occur in order of increasing weight (Behagel 1909; Quirk et al. 1985), but the precise definition of “weight” has been heavily debated. Previous proposals have defined weight as syntactic complexity (syntactic nodes or words), processing (dependencies), phonological complexity (lexical stresses), or phonological length (syllables). This paper presents a systematic investigation of these measures on constituent ordering in two constructions in spoken English. Our results show that the relative influences of phonological and syntactic weight measures vary across syntactic domains, and that weight effects cannot be reduced to a single dimension.

As approximations of syntactic complexity, counts of orthographic words or syntactic nodes have frequently been shown to be reliable predictors of constituent ordering (Wasow 2002; Szmrecsányi 2004; Bresnan and Ford 2010; a.o.). Phonological weight measures have also been suggested to affect constituent ordering (Benor and Levy 2006; McDonald et al. 1993; Anttila et al. 2010; Selkirk 1984; Zec and Inkelas 1990). However, these studies neither controlled for the influences of non-phonological predictors (see esp. Bresnan et al. 2007), nor did they account for the high degree of correlation between weight measures—e.g. the number of stresses increases as word count increases—which potentially masks the independent effects of individual variables.

In this study we examined the influences of various weight measures in genitive and dative construction choice (the car’s wheel ~ the wheel of the car; give the dog the bone ~ give the bone to the dog) in spoken American English using two relatively novel statistical techniques: information-theoretic model averaging (Burnham and Anderson 2004) and conditional random forests analysis (RFA) (Strobl et al. 2009). RFA is robust to collinear data, and is capable of detecting the independent importance of each variable. On the other hand, model averaging rejects the assumption of a single “best” model, and instead allows us to make inferences about evidence from multiple candidate models with varying sets of parameters.

We studied five measures of weight, including the number of syntactic nodes, words, lexical stresses, syllables, and discourse-new referents in the constituents of genitive and dative constructions, while controlling for other known predictors of construction choice (Shih et al. 2009; a.o.). In both constructions, we found syntactic node count to be a highly predictive weight measure, but the reliability of other weight measures varied across the two constructions. Primary stresses were more reliable in predicting the genitive construction, while word and referent count were more reliable in the datives. We propose that this variation is evidence of differing constraints on organization within the NP and VP domains. More importantly, our study reveals that each of the measures investigated makes a distinct contribution to the choice of construction, thus challenging the common practice of reducing weight effects to a single syntactic or phonological variable. Finally, we note that while the two analytical methods used differ in important ways, there is a striking consistency between them, and we advocate this complementary approach to the study of linguistic variables.

23

Bibliography Anttila, Arto, Matthew Adams, and Michael Speriosu. 2010. The role of prosody in the

English dative alternation.” Language and Cognitive Processes. 25(7/8/9): 946-981. Behagel, O. 1909. “Beziehungen zwischen Umfang und Reihenfolge von Satzgliedern.

Indogermanische Forschungen. 25: 110-42. Benor, Sarah Bunin and Roger Levy. 2006. The Chicken or the Egg? A Probabilistic

Analysis of English Binomials. Language. 82(2): 233-278. Bresnan, Joan, Anna Cueni, Tatiana Nikitina, and R. Harald Baayen. 2007. Predicting

the Dative Alternation. In Bouma, G., I. Kraemer, and J. Zwarts (ed). Cognitive Foundations of Interpretation. Royal Netherlands Academy of Science. 69-94.

Bresnan, Joan and Marilyn Ford. 2010. Predicting syntax: Processing dative constructions in American and Australian varieties of English. Language. 86(1): 168-213.

Burnham, Kenneth P. and David R. Anderson. Multimodel inference. Sociological Methods and Research. 33(2): 261-304.

Comrie, Bernard. 2003. On explaining language universals. In Tomasello, M. (ed). The New Psychology of Language: Cognitive and Functional Approaches to Language Structure. Mahwah, NJ: Erlbaum. 195-210.

Gibson, Edward. 1998. Linguistic Complexity: locality of syntactic dependencies. Cognition. 68: 1-76.

Gibson, Edward. 2000. The dependency locality theory: A distance-based theory of linguistic complexity. In Miyashita, Y., A. Marantz, and W. O’Neil (ed). Image, Language, Brain. Cambridge, MA: MIT Press. 95-126.

Hawkins, John A. 1994. A Performance Theory of Order and Constituency. Cambridge: Cambridge University Press.

McDonald, Janet L., Kathryn Bock, and Michael H. Kelly. 1993. Word and World Order: Semantic, Phonological, and Metrical Determinants of Serial Position. Cognitive Psychology. 25: 188-230.

Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London and New York: Longman.

Selkirk, Elisabeth O. 1984. Phonology and Syntax: the Relation between Sound and Structure. Cambridge, MA: MIT Press.

Shih, Stephanie, Jason Grafmiller, Richard Futrell, and Joan Bresnan. 2009. Rhythm’s role in genitive and dative construction choice in spoken English. Paper presented at the 31st annual meeting of the Linguistics Association of Germany (DGfS). University of Osnabrück, Germany. March 4, 2009.

Strobl, Carolin, James Malley, and Gerard Tutz. 2009. An introduction to recursive partitioning: Rationale, application, and characteristics of classification and regression trees, bagging and random forests. Psychological Methods. 14(4): 323-348.

Szmrecsányi, Benedikt. 2004. On Operationalizing Syntactic Complexity. Journées internationals d’Analyse statistique des Données Textualles. 7: 1031-38.

Wasow, Tom. 2002. Postverbal Behavior. Stanford, CA: CSLI Publications. Zec, Draga and Sharon Inkelas. 1990. Prosodically Constrained Syntax. In Inkelas, S. and D. Zec

(ed). The Phonology-Syntax Connection. Stanford, CA: Center for the Study of Language and Information.

24

Hack, Franziska Maria University of Oxford

Between Dialect Syntax and Typology – Variation in Dolomitic Ladin Question Formation

In this paper, I address the relation between dialectsyntax and typology on the basis of syntactic variation in Dolomitic Ladin question formation. The Dolomitic Ladin varieties Gherdëina, Badiot and Fascian constitute a particularly promising research area as they are closely related but differ in (morpho)syntactic properties such as the V2-property or the use of subject clitics. In the first part of the talk, I show in what ways these properties correlate and point out the loci and limits of variation in Dolomitic Ladin question formation.

(1) gives a non-exhaustive overview of question formation strategies in the three subvarieties of Fascian (Cazet, Brach, Moenat) collected in a recent fieldwork study of my own.

(1) ‘Where are you going?’ question formation strategy: a. Olà vas=te? wh-S(ubject)V(erb)I(nversion) where go-2SG=SCL b. Olà vas=te pa3

where go-2SG=SCL PA ? wh-SVI-pa

c. Olà pa tu vas? wh-pa-SVO where PA you go-2SG d. Ola po vas=to? wh-pa-SVI where PA go- SG=SCL e. Olà è=lo che tu vas? wh-cleft-che-SVO where is=it that you go f. Olà che tu vas? wh-che-SVO where that you go-2SG

In the second part of the talk, I will be concerned with the consequences of this case study for the relation between dialectsyntax and typology. Typology can guide dialectologists regarding phenomena to investigate. On the other hand, findings from microsyntactic research can crucially complete or even rectify the insights gained from typology. I illustrate this point by arguing that wh-question formation in some varieties of Dolomitic Ladin represents a counterexample to Cheng’s (1991) cross-linguistic Clausal Typing Hypothesis.

The third part of the talk will then deal with various problems microsyntactic variation poses for typology.

3 pa is a question particle. Depending on the subvariety of Fascian, the particle is realized as [pa] or [pɔ], rendered in

orthography as pa and po, respectively. In Fascian, as opposed to other Dolomitic Ladin varieties, the use of the particle is not obligatory but conventionalized.

25

a) Typological data selection: Grammars and other work on dialects may be biased, e.g. for normative reasons. A case in point is question formation strategy (1f) which entered into Fascian during the 20th century (cf. Chiocchetti 1992) and involves the complementizer che followed by direct word order SVO. It was feared that this novel structure would supersede the traditional wh-SVI and in this way detach Fascian from the other Dolomitic Ladin varieties which maintain and exclusively use SVI in questions. As a consequence, the wh-che-construction was banished systematically from schoolbooks and grammars, leaving no evidence of it for typologists. However, recently collected microsyntactic data show that these normative measures have not been successful; on the contrary, wh-che-SVO has even gained ground being now among the younger speakers the preferred strategy in question formation. This shows that cross-linguistic typological work crucially depends on the accuracy of grammars or dialect studies and suggests that reference grammars should represent all variants observed with their conditions of use so as to enable typologists to draw cross-linguistic conclusions.

b) Borderline cases between dialectal and typological variation: These arise in Dolomitic Ladin in particular due to its debated status as a language’ vs. ʻdialect group’ and due to properties that some varieties of Dolomitic Ladin share with neighbouring varieties of a different language family (Germanic).

In conclusion, both the microsyntactic and the typological perspectives are necessary to determine universal (or common) properties of languages and to shed light on the architecture of the language faculty. I therefore appeal for considering the relation between dialectsyntax and typology as a division of labour. It is the task of dialectsyntacticians to undertake a thorough investigation of microsyntactic variation and pass their findings on to typologists who then process them on the cross-linguistic level.

Bibliography Cheng, L.1991. On the typology of wh-questions. Doctoral dissertation. Cambridge, MA: MIT. Chiocchetti, F. 1992. Evoluzioni sintattiche dell’interrogativa nel fassano. Mondo Ladino

26(3-4):199-219. Kortmann, B. (ed.) 2004. Dialectology meets Typology. Dialect Grammar from a Cross-Linguistic

Perspective. Berlin: Mouton de Gruyter.

26

Haig, Geoffrey University of Bamberg

Word-Order Variation and Change in Real Time: Areal Influence and Universal Factors

Several cases of fundamental word-order change from OV to VO or vice versa are attested in the literature (e.g. Amharic, Harris & Campbell 1995; Germanic, Kiparsky 1996, VO to OV in dialects of Northeast Neoaramaic, Khan 2008). Such changes do not occur in isolation, but involve the complex orchestration of interlinked subchanges, which take many centuries to work through the entirety of the grammar, and the varieties of the languages concerned. Even for comparatively well-documented cases, such as Germanic, the causes and relative sequencing of these processes remain poorly understood, and hotly debated.

The VO / OV parameter correlates in various ways with other word order parameters, for example pre- vs. postpositional ordering (Dryer 2008). A number of theories have been advanced to account for such correlations, e.g. Dryer’s (1992) Branching Direction Theory, Hawkins’ (1994) Principle of Early Immediate Constituents, or Holmberg’s (2000) Final over Final Constraint (FoFC)). While each of these models make certain predictions regarding the sequence of possible changes, none provides a really satisfactory explanation of how change is triggered in the first place.

This paper investigates variation in word order across distinct areal varieties of a highly dispersed language community, Northern Kurdish (also known as Kurmanji; North West Iranian, Iranian, Indo-European). Northern Kurdish straddles a sub-continental buffer zone located at the periphery of two larger-scale macro-areas, the central Asian OV area and the Western European / Mediterranean VO area (Stilo 2006, Wälchli 2009: 91). It has long-standing language contacts with Neo-Aramaic, Arabic, Armenian, Turkish/Azeri, as well as related Iranian languages (Haig 2001). While Kurdish, like the rest of Iranian, is typically characterized as OV, the ordering of other objects (here: Goals (G), including Recipients and Addressees) relative to the verb shows a high degree of language-internal variation, with both GV and VG well attested in texts of all provenience.

I will present results from a systematic quantitative investigation of word order patterns in natural spoken discourse (traditional narratives) from three data points across the Kurdish speech zone: North Iraq, Mush/Erzurum, and Armenia. The results are statistically analysed with a view to assessing the relative importance of universalist claims on word order correlations, and areal influence.

Bibliography Dryer, M. 1992. The Greenbergian word order correlations. Language 68: 81-138. Dryer, Matthew S. (with Gensler, Orin D.). 2008. Order of Object, Oblique, and Verb. In:

Haspelmath, Martin & Dryer, Matthew S. & Gil, David & Comrie, Bernard (eds.). The World Atlas of Language Structures Online. Munich: Max Planck Digital Library, chapter 84. Available online at http://wals.info/feature/84. Accessed on 2011-01-10.

Haig, G. 2001. Linguistic diffusion in present-day East Anatolia: From top to bottom. In: Aikhenvald, A. and Dixon, R.M.W. (eds.). Areal diffusion and genetic inheritance. Problems in comparative linguistics. Oxford: OUP, 195-224.

Harris, A. & Campbell, L. 1995. Historical syntax in cross-linguistic perspective. Cambridge: CUP. Hawkins, J. 1994. A performance theory of order and constituency. CUP. Holmberg, A. 2000. Deriving OV order in Finnish. In: P. Svenonius (ed.). The Derivation of VO

and OV. New York: Oxford University Press, 123–152. Khan, G. 2008. The Jewish Neo-Aramaic Dialect of Urmi. Gorgias Press.

27

Kiparsky, P. 1996. The shift to head-initial VP in Germanic. In: Thrainson, H., Epstein, D. & Peters, S. (eds.). Studies in Comparative Germanic Syntax, Vol. II. Dordrecht: Kluwer, 140-179.

Stilo, D. 2006. Circumpositions as an areal response: The case study of the Iranian zone. In: Johanson / Bulut (eds.). Turkic-Iranian contact areas. Historical and linguistic aspects. Wiesbaden: Harrassowitz, 310-333

Wälchli, B. 2009. Motion events in parallel texts. A study in primary-data typology. Unpublished Habilitationsschrift, University of Bern.

28

Halla-aho, Hilla University of Helsinki, Department of World Culture

Why Did the Romans Misuse Their Accusatives?

This topic relates to the role of the standard variety in historical linguistic research. Classical Latin has a notoriously ‘logical’ syntax where all parts in a sentence are not only carefully organized in a sequence, but also inflected according to their function. An essential part of the inflectional system is that grammatical subjects of transitive verbs are in the nominative, and grammatical direct objects of such verbs are in the accusative. The nominative-accusative distinction is deep-rooted in the Indo-European prehistory of Latin. In the course of the development towards late Latin, things changed to the extent that the Latin accusative in most cases is the source of the new Romance base form.

But in fact the whole picture within ancient Latin is more complicated than that described above, and various nonstandard uses of the accusative case are in evidence from the earliest written evidence onwards, in texts that are not part of the Classical canon. Most commonly, these non-standard usages are found at the beginning of the sentence.

These nonstandard usages include, for example, the following constructions: (1) The proleptic accusative where the subject of the subordinate clause is syntactically the object of the governing verb (sometimes called ‘raising’):

pernam, callum, glandium, sumen facito ham:F.ACC.SG skin:N.ACC.SG glandule:N.ACC.SG udder:N.ACC.SG make:IMP.2SG in aqua iaceant in water:F.ABL.SG lie:PRES.SBJ.3PL “The ham, the skin, the glandule, the udder (acc.) make that [they] lie in water” (Plautus, Pseudolus 166)

(2) The thematic accusative where a thematic element in the accusative precedes the predication (the predication usually has a co-referential element in the accusative):

columellam ferream quae in miliario stat pivot:F.ACC.SG of-iron: F.ACC.SG that:F.NOM.SG in mill:M.ABL.SG stand:PRES.3SG eam rectam stare oportet in medio ad perpendiculum it: F.ACC.SG straight: F.ACC.SG stand:PRES.INF must:PRES.3SG in middle:M.ABL.SG towards plumb- line:N.ACC.SG “The central pivot of iron (acc.) which stands in the mill, it (acc.) should stand straight in the middle perpendicularly” (Cato, De agricultura 20, 1)

(3) Attraction, i.e. the so-called attractio inversa where the head of the relative clause is attracted into the case of the relative pronoun (usually into the accusative):

hunc chlamydatum, quem vides, ei that:M.ACC.SG chlamys-wearing:M.ACC.SG who:M.ACC.SG see:PRES.2SG he:M.DAT.SG Mars iratust Mars:M.NOM.SG be angry:PRES.3SG “That man wearing a chlamys (acc.) whom you see, to him Mars is angry” (Plautus, Poenulus 644)

29

In the present paper I analyse these deviant usages, the degree of their marginality in the Latin system, and the implications of their existence for the later development of the language. I suggest that a partial reassessment concerning the potential of the accusative case is in order and that such a reassessment can help in understanding the later development from Latin to Romance. Bibliography Álvarez Huerta, O. 2005: Attraction régressive et corrélation en latin. Structures parallèles et

corrélatives en grec et en latin (Actes du Colloque de Bordeaux, Septembre 2002), Paulo de Carvalho & Frédéric Lambert (eds), Publications de l’Université de Saint-Étienne. 181-197.

Rovai, F. 2005: L’estensione dell’accusativo in latino tardo e medioevale. Archivio glottologico italiano 90: 54-89.

30

Hamunen, Markus University of Helsinki

Non-Finites of MANNER/MEANS in Finnish Cross-Linguistically

Finnish (and more broadly Finno-Ugric) has been generally considered a language containing plenty of non-finite verb forms, those of infinitives, participles and (deverbal) action nominals, to put them with more language-specific category labels. In this paper, I will focus on two Finnish non-finites which have traditionally in Finnish grammars been interpreted as defining (loosely) the way how something is done by a predicate verb or a clause. The one is MALLA-infinitive (1), and the other is TEN-infinitive (2).

(1) puhu-malla-ha asija paranoo speak-MALLA-CL matter get.better ‘by speaking things get better, you know’

(2) kakara-t läht-i parkuj-en tiä-he-nsä children take.off-3SG.PRT cry.out-TEN way-ILL-3PX4

‘children made off crying ~ and cried’

Syntactically both MALLA and TEN are free adverbials or adjuncts. In Finnish grammar tradition they have been considered as sentence equivalents (lauseenvastike) and as such thought to be parallels to adverbial subordinate clauses. Then, semantically they, nevertheless without usually having any clear inherent regular meaning, tie an interpretative relation (circumstantial, sequential or else) between finite verb or main clause and a non-finite in question. For example in (1) MALLA form denotes ‘means’ of getting things to better stance and, respectively, in (2) TEN form denotes ‘concomitance’ or ‘attendant circumstance’ of the action denoted by the finite verb. MALLA and TEN have some other functions as well.

Typologically TEN and MALLA (and few other Finnish non-finites too, see Ylikoski 2003: 203–205) have been considered converbs (e.g. König 1995, van der Auwera 1998). Now, after having shown different ambiguous interpretations of these converbs with empirical dialectal data in Finnish, I will raise some methodological implications when combining (‘horizontal’) typology and (‘vertical’) language-specific dialectology. Particularly, I will discuss the typological character of Finnish as converbal language in respect to MALLA and TEN. More broadly speaking, I will reflect ‘natural’-oriented language description based on empirical data on general methods being used in typology.

Bibliography Van der Auwera, Johan 1998. Defining Converbs. Typology of Verbal Categories. Papers

presented to Vladimir Nedjalkov on the occasion of his 70th birthday, Leinid Kulikov & Heinz Vater (eds.), Tübingen: Max Niemeyer Verlag, 273–282.

König, Ekkehard 1995. The meaning of converb constructions. Converbs in cross-linguistic perspective. Structure and meaning of adverbial verb forms – adverbial participles, gerunds –, Martin Haspelmath & Ekkehard König (eds.), Empirical Approaches to Language Typology 13, Berlin: Mouton de Gryuter, 57–95.

Ylikoski, Jussi 2003. Defining Non-finites: Action Nominals, Converbs and Infinitives. SKY Journal of Linguistics 16: 185–23.

4 Abbreviations: CL = clitic, 3SG = 3. person singular, PRT = preteriti, MALLA = mA-infinitive adessive, TEN = (T)e-infinitive instructive, ILL = illative, 3PX = 3. person possessive suffix

31

Hiietam, Katrin Unaffiliated

The Estonians and the Estonian Language in Northern England

This paper gives an overview of the Estonian language spoken by the Estonian community in Northern England. The migration has taken place in five distinct waves (see also Väravas 2004). The first wave of migration took place at the beginning of the 20th century and the fifth one was induced by Estonia gaining independence in 1991 and especially by Estonia joining the EU in 2004. The exact size of the Estonian community in the UK is difficult to establish but an estimate is around 3000 members (Hiietam 2010).

The data used in the study was collected by interviewing the members of the community by using a set questionnaire. We could say that Estonian in the UK is currently used by three generations and I have divided the informants into three groups as follows: 1) the first generation aged 69–85, the second generation 48–50 and the third 12–23.

Based on the material collected, we can claim that the Estonian used in Great Britain has noticeable influences, across all the three generations, from the majority language – English. The language of the first generation is best preserved; people use Estonian grammar and syntax in relative conformity to standard common Estonian. Direct borrowings and foreign influences are few. With the first generation only, one could notice the influence of Estonian phonology on their English. The most common characteristic of the language used by the first generation was lexical code switching, i.e. English terms were used in an otherwise Estonian conversation (e.g. semi, tšätti (chatty), õupen aus (open house)). Also, mixed forms were common, where the English terms had been adapted to the Estonian morphology (e.g. monstrid (monsters), tressitud (dressed)). To a certain extent, the language of the first generation also presents with the syntactic code switching, e.g. ja mina saan inglise keelt kirjutada ka (ingl I can write English as well compare with the Standard Estonian: ’ma oskan ka inglise keeles kirjutada’)

The language of the second generation is not a homogeneous phenomenon, however, in most cases, it is characterised by language attrition at an individual level which results from fewer possibilities of using the language. A fairly common phenomenon is code mixing, both morphological and syntactic, where a conversation that was started in Estonian is finished in English, especially if the other party is a member of the third generation. It is common for the speakers to apply the s. c. English subject requirement to their Estonian and lexical borrowings do not conform to the norms of the Standard Estonian in terms of their morphology. Some informants from the second and third generation had difficulties with pronouncing the Estonian õ regardless of the native Estonian dialect of their parents.

Among the third generation there are only a few who speak Estonian. Those who have been taught Estonian at home are able to express themselves on topics used at home. When speaking about science, technology or social life, it is common for the informants to make use of English as they lack the necessary vocabulary in Estonian. With the third generation, code mixing is common and in syntactic code switching there are recognizable English sentence patterns.

Not surprisingly, the Estonian language used in the UK shows signs of moving towards a more Northern-Germanic language pattern with erosion of morphological case endings, copying of the English sentence structure and a compulsory but semantically empty subject in the sentence.

Bibliography Hiietam, Katrin. (2010). Eestlasted ja eesti keel Põhja-Inglismaal. (Estonians and Estonian language

in Northern-England). A book chapter in a collective volume, Ed by Viikberg, “Eesti keel välismaal“ (Estonian Language abroad). EKSA (Eesti Keele Sihtasutus), Tallinn

32

Väravas, Tiina-Maria (2004). Identiteedi ja kohanemise aspekte: Inglismaal elavate eestlaste näitel. (Aspects of identity and aclimatization: based on the Estonians living in England). Bakalaureusetöö (BA thesis). Tallinn. [Käsikiri TLÜ Eesti Humanitaarinstituudis.]

33

Hill, Virginia University of New Brunswick - Saint John

Variation in Verb Movement in Old Romanian

This is a case study of syntactic variation: the level of verb (V) movement in Old Romanian (OR) declarative clauses, where the location of the verb is either high (1) or low (2) in relation to clitics. The texts indicate free alternation between (1) or (2).

Background. Historical linguistics assigns the word order in (1) to a phonological requirement on second position clitics (Wackernagel’s law – W-l); this typological group includes Romance ((Meyer-Lübke 1890) and Slavic languages (Slawski 1946). Romanian is a Romance language under language contact with South Slavic, so the variation in (1) vs (2) is assumed to show a transitional loss of W-l (Frâncu 2009). Generative grammar assigns a particular configuration to W-l: Rivero (1993) argues that W-l triggers Long Head Movement (LHM), i.e., the non-finite verb stem raises across clitics (pronouns and/or auxiliary), targeting the highest head position (Xº) in the clause (1a). This operation occurs in complementary distribution with the fronting of phrasal constituents (XP). Thus, W-l triggers either [Xº > Clitic] or [XP > Clitic] orders.

Objective. We check on the application of W-l in OR from Rivero’s perspective. The main data source: the Moldavian chronicles (17th-18th c.). Framework: the clause cartography in Rizzi (1997), where LHM means V movement to Force in the hierarchy [Force > Topic > Focus > Finite > IP] (Fischer 2003).

Analysis. First, OR displays both finite (1b) and non-finite (1a) verb stems above the clitics; hence, we have V fronting, not only LHM. Second, V fronting/LHM co-occur with phrasal constituents fronted to Topic (3). This invalidates W-l as the trigger: the enclitics occur in 3rd, 4th position. Furthermore, the first written texts, translated from Slavonic (where W-l is strong) display “mistakes” in the placement of clitics (4): we see LHM plus proclitics. Although double spell-out of clitics is common at this time (5), it involves low V movement, not LHM/V fronting. On the other hand, there is evidence that the V movement alternation in (1) and (2) is discourse related: (i) LHM/V fronting disallow constituents with contrastive focus reading; (ii) LHM/V fronting is cancelled by clause coordinators that double as “narrative fillers” (‘and’, ‘but’) (6a); higher placed connectors (‘therefore’, ‘so’) show no effects on it (6b); (iii) option for (1) or (2) depends on the writing style. Thus, the analysis exploits the discourse-syntax interface: high V movement targets Focus (vs Force) to check a discourse operator, when such operator exists (it depends on discourse type) and when no other items (e.g. “narrative fillers”) do the checking. The enclisis mimics W-l but follows from different triggers probing V from a different position. W-l is not active in 17-18th c. OR.

Conclusion. By explaining the variation in V movement within OR, the syntactic approach uncovers the need to revisit the typological classification of OR, which had been done on the basis of phonology only. The syntactic analysis brings diachronic clarifications for the typological classification w.r.t. the second position clitics variable.

Bibliograhy Fischer, Olga. 2003. Rethinking the Tobler-Mussafia Law. Diachronica 20 (2): 259-288. Frâncu, Constantin. 2009. Gramatica limbii române vechi (1521-1780). Iaşi: Demiurg. Meyer-Lübke, Wilhelm. 1890/1902. Grammatik der romanischen Sprachen. Strasburg. Rivero, Maria Luisa. 1993. LHM vs V2 and null subjects in Old Romance. Lingua 89: 217-245. Rizzi, Luigi. 1997. The fine structure of the left periphery. In Haegeman, L. (ed.). Elements of

grammar: A handbook of generative syntax. 281-337. Dordrecht: Kluwer.

34

Slawski, Franciszek. 1946. Miejsce enklityki odmiennej w dziejach jrzyka bulgarskiego (The Place of the Inflected Enclitic in the History of Bulgarian), Krakow: Polska Akademia Umiejetnosci.

Examples:

1a) Rămasu-i- au pomană în ţară mănăstirea ... [- finite] V form left- of.him-has memory in country monastery-the V>Clitic>Aux ‘In the country a monastery has been left to his memory….’ (Costin apud Panaitescu 1979: 33) 1b) Cunoaşti-se că au fost neaşezaţi… [+ finite] V form tell- refl that have been non.settled V > Clitic ‘One can tell that they were not settled’ (Ureche apud Panaitescu 1958: 73) 2a) -au tras cuvintul in Tara Leseasca pan astadzi,… Clitic>Aux>V refl-has stayed word.the in Country.the Polish until today… ‘The rumor has stayed in Poland until today,…’ (Costin apud Panaitescu 1979: 161) 2c) Să rădicase în dzilele aceştii domnii…. Clitic > V refl arose in days.the this.Gen government ‘During the time of this government, there arose….’ (Costin apud Panaitescu 1979: 51) 3) Acest domn…dupa doi ani.. rădicatu-s- au de la Tara this king-TOP after 2 years -TOP risen- refl-has of from Country-the Muntenească cu multă …oaste.. (Ureche apud Panaitescu 1958: 19 v p 90) Munteneasca with much army ‘After two years, this king alighted from Vallachia with a big army.’ 4) părinţii noştri….i- ai mîntuitu-i-ai TopP>Clit>AUX> LHM parents.the ours them-have blessed-them-have (hypercorrection) ‘you blessed our parents’ (Densuşianu 1997: 707 – PH.xxi, 5 – 16th c.) 5) De la Focseni le-au dusu-li cu căruţi… Clitici > AUX > V > Clitici from at Focsani them have transported them with carts ‘they transported them from Focsani by carts’ (Neculce apud Iordan 1955: 193) 6a) …. Niteia bătutu-o-au şi o au dobîndit. +LHM > ‘and’ > -LHM Niteia beat-it-has and it has conquered ‘Niteia, he has beat it and conquered it.’ (Ureche apud Panaitescu 1958: 71) 6b) Dece intrat-au tătarii in ţară,… ‘so’ > +LHM so invaded-have Tartars in country ‘So the Tartars invaded the country,…’ (Neculce apud Iordan 1955: 143)

35

Hillberg, Sanna University of Eastern Finland

Zero Relativization in Educated Written Scottish English

Zero relativization, i.e. the omission of a relativizer in a relative clause, is a feature which is rarely addressed in studies concerning relativization in standard varieties of English. This paper focuses on zero relativization in the written variety of Educated Standard Scottish English (ESSE), and in two of its subvarieties: Highland and Island Scottish English (HIScE) and Lowland Scottish English (LScE), while Standard British English (BrE) and Irish English (IrE) are used as points of comparison. Comparisons between BrE and IrE widen the perspective of the study, since there are no previous comparative studies on relativization in educated written language between these British English varieties. The database consist of the Corpus of Scottish English On-line Press News (CSEOPN), which is a corpus I have compiled of Scottish on-line newspaper articles, and the written news reportage sections of two components of the International Corpus of English (ICE), viz. ICE-Great Britain Release 2 and ICE-Ireland.

Previous studies on Scottish English relativization have concentrated on spoken LScE, while relativization strategies of ESSE and HIScE have been largely ignored. HIScE and LScE have been influenced by different language contacts, which are still evident in their modern use. The Scottish Highlands and Islands were previously Gaelic-speaking, whereas in the Lowland areas Scots and English, sharing OE as their origin, have existed side by side for centuries.

The purpose of this paper is to provide a detailed report on the zero relativizer use in ESSE. The main research question concerns the possible differences and similarities in the zero relativizer use between the Scottish varieties as well as between the different varieties of the British Isles. Earlier studies have shown that the zero relativizer is a prevalent feature of spoken LScE. However, no comparable data is available for HIScE. The preliminary findings of the current study indicate that differences in the zero relativizer use between the ESSE varieties are minor. The zero relativizer functions similarly in both varieties. Interestingly, prepositional complementation with the zero relativizer is rather common in ESSE, whereas almost nonexistent in Standard English. The current data also shows that the systems operating in ScE written and spoken zero relativizer use differ from each other considerably, just as they do in other varieties of English, e.g., in the uses of subject zero relativizer.

Bibliography Aitken, A.J. 1984a. ‘Scottish accents and dialects’. In Trudgill, P. (ed.), Language in the British

Isles. Cambridge: Cambridge University Press, 94–114. Biber D., S. Johansson, G. Leech, S. Conrad and E. Finegan. 1999. Longman Grammar of Spoken

and Written English. London: Longman. Herrmann, T. 2005. ‘Relative clauses in English dialects of the British Isles’. In Kortmann B., T.

Herrmann, L. Pietsch and S. Wagner, A Comparative Grammar of British English Dialects. Berlin, New York: Mouton de Gruyter, 21–123.

Miller, J. 1993. ‘The Grammar of Scottish English’. In Milroy, J. and L. Milroy. (eds), Real English: The Grammar of English Dialects in the British Isles. London: Longman, 99–138.

Tagliamonte, S., J. Smith and H. Lawrence. 2005. ‘No taming the vernacular! Insights from the relatives in northern Britain’. Language Variation and Change, 17: 75–112.

36

Huang, Wantin Stacy Student at Graduate Institute of Linguistics at National Cheng Chi University (Taiwan)

Accusative or Ergative or Neither: Justifying for Symmetrical Voice System in Tao

Tao is an Austronesian language, spoken by about 3,000 speakers resided in the Orchid Island located near the southeastern coast of Taiwan. Unlike other Austronesian languages spoken in Taiwan, it shares innovations with Philippine languages and is classified as one of Malayo-Polynesian languages, a primary branch of Austronesian language family. Philippine languages exhibit a unique agreement between verb and its nominative NP, in that the verbal affix consistently agrees with the thematic role of the nominative NP. This has been referred to as focus. Later on, researchers have suggested to refer to this agreeing phenomena as voice even it does not share a lot in common with the well-known English active-passive voice system (Ross and Tang 2005).

Earlier studies on Philippine languages have based on the assumption that the basic transitive active sentence is the one in (1b) and classify these languages as nominative-accusative languages (Bloomfield 1917, Blake 1925). Later on, researchers take the construction in (1c) as the basic active transitive sentence, and have claimed that these languages should be classified as ergative languages (Payne 1982). Researchers based on different assumptions have made different claims regarding case system and verbal voice of Austronesian languages. The diverse perspectives are based on different assumptions on the basic active transitive sentence, which remains controversial.

A detail examination of Tao voice system is done and possible evidence for considering it as accusative and ergative language are examined. The language data have shown that the consideration of either (1b) or (1c) as the basic transitive sentence in Tao is not adequate. The indefiniteness of the patient nominal in agent voice construction (AV), as in (1b), and the definiteness of patient nominal in patient voice construction (PV), as in (1c), has lead researchers to consider (1c) as the basic transitive sentence based on Hopper and Thompson’s transitivity scale (1980). Neither AV nor PV exhibits zero verbal marking for voice, as English active voice sentences. Antipassive affix is not found with AV verbs in Tao. Tao verbal markings do not manifest it as either accusative or ergative language. Furthermore, nonnominative NP in AV and PV construction has similar patterning. While researchers has considered the definite agent nominal in PV construction as a core argument, it is not adequate to consider the indefinite patient nominal in AV as a demoted object and treats the AV construction as an antipassive construction. Hence, it is more suffice to suggest that Tao exhibits a symmetrical voice system -- “[n]o one NP type is preferred for pivot choice; regardless of which choice is made, all are signaled by some overt verbal voice morpheme.” (Foley 2008: 42). The symmetrical voice system might not be an economical system that require more verbal and case marking than asymmetrical voice system, but it reflects the language facts. It also supports that Polynesian languages have undergo change from either ergative-to-accusative or accusative-to-ergative proposal, and Tao is a product of it (Ball 2007).

1) Agent Voice Construction a. ya mi-anoanood sira AUX AV-sing 3.P.NOM5

‘They ae singing.’ (Ho 1990: 116)

b. ya ni-man-rakat o kanakan so kois AUX PERF-AV-kill NOM child OBL pig ‘The child has killed a pig/pigs.’

Patient Voice Construction c. ya na ni-rakat no kanakan o kois

5 Interlinear glosses in the examples are modified for the present study.

37

AUX 3.S.GEN PERF-kill GEN child NOM pig ‘The child has killed the pig(s).’ (Ho 1990: 64)

Bibliography Ball, Douglas. 2007. On Ergativity and Accusativity in Proto-Polynesian and Proto-

Central Pacific. Oceanic Linguistics, 46: 128-153. Blake, F.R. 1925. A Grammar of Tagalog Language. New Haven: American Oriental

Society. American Oriental Series, 1. Bloomfield, Leonard. 1917. Tagalog texts with grammatical analysis. University of

Illinois Studies in Language and Literature 3:2-4. Urbana: University of Illinois. Chang, Claire Hsun-huei. 1997. On thematic structure of Yami verbs. NSC report 85-2418- H-004-

001. Chang, Claire Husin-huei. 2000. Yami reference grammar. Taipei: Yuen Liu. Deng, Dun-Hong. 2005. Argument-Function Linking in Yami: An Optimality-Theoretic Account.

MA Thesis. National Cheng-Chi University. Dixon, R. M. W. 1994. Ergativity. Cambridge: Cambridge University Press. Foley, William, A. 1998. Symmetrical voice systems and precategoriality in Philippine languages.

University of Sydney. Foley, William, A. 2008. The Place of Philippine Languages in a Typology of Voice Systems. In

Peter K. Austin and Simon Musgrave, ed., Voice and Grammatical Relations in Austronesian Languages, 22-44. Standford: CSLI Publications.

Ho, Arlene Yue-ling. 1990. Yami Structure: A Descriptive Study of the Yami Language. M.A. Thesis, National Tsing Hua University.

Hopper, Paul J. and Sandra A. Thompson. 1980. Transitivity in Grammar and Discourse. Language 56: 251-299.

Payne, Thomas. 1982. “Role and reference related subject properties and ergativity in Yup’ik Eskimo and Tagalog.” Studies in Language, 6:75-106.

Ross, Malcolm and Stacy Fang-ching Teng. 2005. Formosan Languages and Linguistic Typology. Language and Linguistics 6.4:739-781.

Walton, Charles. 1986. Sama Verbal Semantics: Classification, Derivation, and Inflection. Manila: Linguistic Society of the Philippines.

Rau, D. Victoria. And Ma-nuy Tung. 2006. Tao Dictionary: Reference Grammar, Texts, and Dictionary. Language and Linguistics Monograph. Taipei: Academia Sinica.

Van Valin, Robert D., Jr. 1981. Grammatical Relations in Ergative Languages. Studies in Language 5.3. 361-394.

38

Håkansson, David Uppsala University

A Variationist Approach to Syntactic Change

Much of the research on syntactic change has been carried out within a generative framework, more specifically within the principles-and-parameter theory of universal grammar. One of the strengths of this theory is that it makes it possible to relate a number of apparently unrelated linguistic phenomena to one single parameter. In diachronic studies, the simultaneous change of a number of syntactic phenomena can therefore be explained as the result of a new parameter setting (Lightfoot 1991).

Recently, however, the principles-and-parameter theory has been heavily criticized. This criticism concerns the theoretical status of the parameters (Holmberg 2010) as well as the scope of the typological generalizations (Newmeyer 2004). Taking this criticism seriously, we seem to be left without any valid explanation of why a number of syntactic features changes practically simultaneously, as shown for example in Swedish (Platzack 1987; Falk 1993).

Drawing on the distinction between one the one hand innovations and on the other hand changes (Milroy 2003) in my talk I will argue that a theory of universal grammar can only be used to explain innovations – i.e. the origin of variation in the speech community – but never to explain changes – i.e. the origin of a new norm in the speech community. In order to adequately account for changes, we therefore have to look beyond theories of universal grammar towards sociolinguistic approaches to linguistic alteration.

The empirical basis for my talk consists of investigations of a number of word order phenomena in Early Modern Swedish (1526–1732). Building on the results of these investigations, I will show that a sociolinguistic approach to syntactic change can shed new light on word order change and make a contribution to the discussion of why several apparently unrelated word order patterns change simultaneously during the Early Modern Swedish period.

Bibliography Falk, Cecilia (1993): Non-Referential Subjects in the History of Swedish. Ph.D. Diss. Lund:

Department of Scandinavian Languages. Holmberg, Anders (2010): Parameters in minimalist theory: The case of Scandinavian. Theoretical

Linguistics 36.1: 1–48. Lightfoot, David (1991): How to Set Parameters: Arguments from Language Change.

Cambridge/London: MIT Press. Milroy, Jim (2003): When is a sound change? D. Britain & J. Cheshire (eds.): Social Dialectology.

In Honour of Peter Trudgill, 219–231. Amsterdam: Benjamins. Newmeyer, Frederick (2004): Against a parameter-setting approach to typological variation. P. Pica

(ed.): Linguistic Variation Yearbook 4, 181–234. Amsterdam: Benjamins. Platzack, Christer (1987): The Scandinavian Languages and the Null-Subject Parameter. Natural

Languages and Linguistic Theory 5.3: 377–401.

39

Karttunen, Miia Kymenlaakso University of Applied Sciences University on Eastern Finlandsss

Syntactical Observations on the Future Time Reference in a Language without a Future Tense

The Finnish language tense system has the opposition past/non-past (see e.g. Comrie 1985: 43–45, 49). Past time reference involves three tenses, while non-past time reference only one, present tense. It refers not only to present events but also to future events. Finnish is claimed to lack a grammatical, i.e. inflectional future tense. According to WALS, the inflectional and non-inflectional futures are equally common in the languages of the world (Dahl & Velupillai 2008).

In Finnish, there are no means to express future by verbal inflection, but instead, the future time reference in Finnish can take various forms, e.g. present tense verb forms both alone and combined with temporal adverbials, constructions with certain infinitive forms or modal verbs (see e.g. Metslang 1994; Hakulinen et al. 2004 § 1542;). The present tense forms are construed as future time reference depending on logical, situational and contextual factors.

In addition, there are different syntactical constructions but either their role in the tense system is controversial. These structures are not solely used for future reference or their usage is restricted to certain contexts (e.g. the tulla ’to come’ construction, antaa ‘to let’ construction, the progressive construction, the be-about-to construction). The variety is wide, particularly, when colloquial language comes into question, but the constructions tend to be quite marginal but still recognizable across languages. (Xrakovskij 2001.) The goal of this paper is to focus on the syntactical structures with future time reference.

The paper will discuss the future time reference system from the syntactical viewpoint and focus on spoken Finnish varieties. The study is based on the empirical data which come two different corpora of conversational language recorded in Eastern Finland (Kainuu dialects and Eastern Savo dialects) in 1990 and 2000 and. Irrespective of the dialectal basis, I believe that the data of the study will illuminate the Finnish system of future as a whole. The data come from both interviews and service encounters. Because of the nature of the data, the syntactical structures will be analysed in respect the contextual and the interactional factors.

In my presentation, I would like to provide more evidence and detailed analysis on the syntactical structures. Although I deal with the data from Finnish dialects, I will also discuss my data-based observations from the cross-linguistic point of view. Bibliography Comrie, Bernard. 1985. Tense. Cambridge Textbooks in Linguistics. Press Syndicate of the Univer-

sity of Cambridge. Dahl, Östen and Velupillai, Viveka. 2008: The Future Tense. In World Atlas of Language

Structures Online. Munich: Max Planck Digital Library. Available on line at http:// http://wals.info/feature/67.

Hakulinen, Auli, Vilkuna, Maria, Korhonen, Riitta, Koivisto, Vesa, Heinonen, Tarja Riitta and Alho, Irja. 2004. Iso suomen kielioppi. Available on line: http://scripta.kotus.fi/visk.

Metslang, Helle. 1994. Temporal relations in the predicate and the grammatical system of Estonian and Finnish. Oulu: Oulun yliopiston suomen ja saamen kielen laitoksen tutkimus-raportteja 39.

Xrakovskij, Viktor (ed.). 2001. Typology of Imperative Constructions. LINCOM Studies in Theoretical Linguistics. Lincom Europa. Muenchen

40

Jendraschek, Gerd University of Regensburg

Core versus Periphery in Basque Morphosyntax

This talk deals with the morphological reflection of the interaction between argument relations and TAM exponents in Basque, as well as the diachronic, sociolinguistic, and diatopic variation observed in this area. This variation can be accounted for in terms of core vs. peripheral morphosyntax.

As an example of how the distinction between core and periphery manifests itself consider (1) to (5). The direct object of transitive constructions can be freely exchanged in real mood (see (1) and (2)), but not in potential mood ((3) vs. (4)). (4) represents marked values for both person and mood. It is (still) used by some speakers (as indicated by ‘%’), but many never use this form. The more a form represents marked values, the higher its proneness to obsolescence. This explains why an auxiliary form like zintzakedan ‘I could V you’ in (5) is virtually non-existent in corpora (and classified as ‘very rare’ in Hualde 2003: 229).

Contrasting (3) and (4), morphemic obliteration and routinization led to the reanalysis of a once polymorphemic sequence to a single morpheme deza-; cf. Bybee (2003) and Bybee & Hopper (2001: 16) on morphologically complex forms being stored as ‘single processing units’. As for the sequence underlying zaitza-, it was much less frequent than deza-, and therefore never entered the morpheme inventory as a fused morpheme. As a result, forms like zaitzaket ‘I V you’ were marginalized and pushed to the periphery of the linguistic system.

The discrepancy between the combinatorial potential of verbal morphology on the one hand, and the frequency of forms in real-language corpora has important consequences for an adequate description of the language as well as its development. The possible number of forms is immense – this is why we find large numbers of tables with elaborate paradigms in most grammatical descriptions of Basque. However, these descriptions produce a false impression, as present-day spoken Basque uses only a small subset of all possible forms. The two approaches – list all the possible forms vs. list frequently attested forms – can be unified though, if we locate possible forms along a cline leading from the core of the grammatical system to its periphery (a similar approach can be found in King 1993: 190 ; see also Haase 1992: 159-174). Whereas the core contains the most frequent and stable forms, the outer periphery contains forms which exist only in theory, with intermediate forms being subject to a great deal of intra-speaker, inter-speaker, and interdialectal variation. And while a peripheral form does not violate any grammatical rule, (most) speakers will avoid it, as their primary aim is not to exploit the morphological potential of the system, but to make themselves understood.

The concepts of core vs. periphery can be applied to other grammatical systems where an inflation of morphological forms leads to a hierarchical organization of the system. As for linguistic typology, the Basque scenario is an illustration of what is likely to happen when a language is simultaneously high on the indices of synthesis and fusion (cf. Comrie 1989: 46-51).

1) Ikus-ten du -t. see-IPFV PRS.O.3SG -A.1SG

‘I see him.’ 2) Ikus-ten zaitu -t. see-IPFV PRS.O.2SG -A.1SG

‘I see you.’ 3) Ikus deza -ke -t. see PRS.SUBJ.O.3SG -POT -A.1SG

‘I can see him.’

41

4) %Ikus zaitza -ke -t. see PRS.SUBJ.O.2SG -POT -A.1SG

‘I can see you.’ 5) %Ikus zintza> -ke -da<n. see <PST.SUBJ.O.2SG> -POT -A.1SG ‘I could see you.’

Bibliography Bybee, Joan 2003. Mechanisms of change in grammaticization: the role of frequency. Handbook of

historical linguistics, Brian Joseph & Richard D. Janda (eds.), Oxford: Blackwell, 602-623.

Bybee, Joan & Paul Hopper (eds.) 2001: Frequency and the emergence of linguistic structure. Amsterdam/Philadelphia: John Benjamins (Typological studies in language, 45).

Comrie, Bernard 1989: Language Universals and Linguistic Typology. Second edition. Oxford: Blackwell.

Haase, Martin 1992: Sprachkontakt und Sprachwandel im Baskenland. Die Einflüsse des Gaskognischen und Französischen auf das Baskische. Hamburg: Buske.

Hualde, José Ignacio 2003. Finite forms. A Grammar of Basque, José Ignacio Hualde & Jon Ortiz de Urbina (eds.), Berlin/New York: Mouton de Gruyter, 205–246.

King, Alan R. 1993. Communicative grammar of the Basque verb (selected aspects). Ph.D. diss. London University.

42

Lindstedt, Jouko University of Helsinki

Dialect Syntax or Bad Translation? Interpreting Old Dialect Texts Written in a Multilingual Setting

The Slavic dialects of Northern Greece are seriously endangered and strongly influenced by the Greek language. Unique material for the study of earlier stages of these dialects can be found in some Slavic Gospel manuscripts that were written in Greek letters in the late 18th and in the 19th century, notably the Kulakia Gospel (Mazon & Vaillant 1938) and the Konikovo Gospel (Lindstedt & Spasov & Nuorluoto 2008). These vernacular Slavic Gospel translations were produced in multilingual Macedonia, a part of the Ottoman Empire where different varieties of Greek, Bulgarian, Macedonian, Albanian, and Vlach (Aromanian) were all written in Greek letters, Greek being the language of high culture among the Christian population.

Syntactic features typical of these old texts include differential object marking (the marking of certain direct objects with a preposition adopted from the indirect object), and the levelling of certain distinctions in the Slavic pronominal system, especially the loss of some reflexive pronouns. Differential object marking is a well-known areal feature of the languages spoken in Macedonia (Adamou 2009); the loss of pronominal distinctions appears to derive from mechanical copying of the Greek original by the Gospel translators. However, differential object marking can also be analysed as confusion due to the accusative/dative syncretism in the northern dialects of Greek, whereas the substitutes for the reflexive pronouns do not seem to mechanically follow the Greek original.

The methodological question is how to distinguish occasional translation interference in an old text from features which were a genuine part of the dialect syntax, albeit originally contact-induced. Internal evidence can be sought by looking at the consistency of the syntax in the text. External evidence may be found in the results of the earliest scientific dialect field work, which in this case was done as late as in the 1930s (Małecki 1934, 1936). But given the extensive multilingualism in Ottoman Macedonia, it can even be asked whether the distinction between occasional transfers and borrowed dialect features is applicable to such texts at all.

Bibliography Adamou, Evangelia. 2009. Le marquage differential de l’object en nashta et en pomaque (slave,

Grèce). Bulletin de la société de linguistique de Paris 105, 1. 383–410. Lindstedt, Jouko & Ljudmil Spasov & Juhani Nuorluoto (eds.). 2008. The Konikovo Gospel –

Кониковско евангелие: Bibl. Patr. Alex. 268 (Commentationes Humanarum Litterarum 125). Helsinki: Societas Scientiarum Fennica.

Małecki, Mieczysław. 1934. Dwie gwary macedonskie : Suche i Wysoka w Solunskiem . Cz. 1: Teksty (Bibljoteka "Ludu słowian skiego", A 2). Kraków: Gebethner i Wolff.

Małecki, Mieczysław. 1936. Dwie gwary macedonskie : Suche i Wysoka w Solunskiem . Cz. 2: Słownik (Bibljoteka "Ludu słowian skiego", A 3). Kraków: Gebethner i Wolff.

Mazon, André & André Vaillant. 1938. L’Évangéliaire de Kulakia: Un parler slave du Bas-Vardar (Bibliotheque d’études balkaniques 6). Paris: Librairie Droz.

43

Lange, Claudia; Bernaisch, Tobias Justus Liebig University Giessen Justus Liebig University Giessen

The Typology of Focus Marking in South Asian Englishes

The emergence of grammatical norms in postcolonial varieties of English has been argued to manifest itself in quantitative preferences rather than in categorical distinctions (cf. Schneider 2007: 46). Several studies on Indian English (IndE), however, have shown that this South Asian variety has developed innovative uses for the additive focus marker also and the restrictive focus markers only and itself as presentational focus markers (Bhatt 2000, Lange 2007, Balasubramanian 2009), e.g.

Since 7 am itself, schoolchildren started to reach the venue smartly dressed and armed with their queries and waited patiently for more than two hours for the programme to begin. <SAVE-TOI-38032>

The grammaticalization of the focus marker itself to an invariant particle is indicated by examples where the form occurs with a plural antecedent, e.g.

He said the temporary peace achieved by leaders of the country was a victory for the Sri Lankan Security Forces itself as it was gained by the Security Forces at the expense of their lives. <SAVE-DN-2004-07-02>

The present study is concerned with variation and convergence in the use of focus marking with itself in South Asian Englishes, specifically in IndE and in Sri Lankan English (SLE). Although both are independent varieties of English in their own right, and even though there is a tendency for speakers of SLE to maximize the "abstand" (Kloss 1967: 29) to IndE, both varieties show a striking parallelism in adapting innovative patterns of focus marking.

On the basis of the South Asian varieties of English (SAVE) corpus, an 18-million word web-based newspaper corpus featuring acrolectal language use of the varieties under scrutiny (cf. Bernaisch et al. 2011), we will report on the pervasiveness of (presentational) focus marking with itself. Although the novel usage of itself as illustrated above certainly represents a feature of South Asian English, there is a clear pattern characterised by unity and diversity with regard to the individual varieties of English in South Asia.

Bibliography Balasubramanian, Chandrika. 2009: Circumstance adverbials in registers of Indian English. World

Englishes 28(4): 485–508. Bernaisch, Tobias, Christopher Koch, Joybrato Mukherjee & Marco Schilk. 2011: Manual to the

SAVE Corpora: Compilation, Cleanup Process, and Details on the Individual Corpora. Giessen: Justus Liebig University.

Bhatt, Rakesh M. 2000: Optimal Expressions in Indian English. English Language and Linguistics 4(1): 69-95.

Kloss, Heinz. 1967: Abstand-Languages and Ausbau-Languages. Anthropological Linguistics 9(7): 29-41.

Lange, Claudia. 2007: Focus marking in Indian English. English World-Wide 28(1): 89-118. Schneider, Edgar W. 2007: Postcolonial English: Varieties around the World. Cambridge:

Cambridge University Press.

44

Leino, Jaakko University of Helsinki

Variation, Diffusion, and Typological Implications: Finnish Spatial Particles as Typological Data

Finnish is typically cited as having more postpositions than prepositions. The aim of my paper is to point out some of the complexities hidden behind this innocent-looking typologically oriented piece of trivia.

More well-informed sources cite Finnish as using, rather than having, more postpositions, reflecting the fact that most Finnish adpositions may be used either as prepositions or as postpositions, even though the latter usage tends to be more frequent.

1) Talo o liki Munttiste kylä. house-NOM be-3SG liki Munttinen-GEN village-PTV ‘The house is near the village of Munttinen.’

2) Kraatari asus kaupa liki. tailor-NOM live-PST-3SG store-GEN liki ‘The tailor lived near the store.’

Furthermore, many (if not most) Finnish “adpositions” may also be used as adverbs.

3) siinä oli liki niitä riihiä. there be-3SG liki they-PTV drying.house-PTV ‘There were some of those drying houses nearby.’

Categorizing each individual word as either adposition or adverb proves both construction-dependent and arbitrary. To further complicate matters, the usage of each of the words differs dramatically between data sets. For example, halki (‘through’, ‘across’) is nearly always a postposition in modern newspaper texts, an adverb in old written data, and either preposition or adverb in dialects. This variation is not systematic, however: each word behaves differently in each data set, and no overall trend (e.g. a shift from adverb through postposition to preposition) can be shown to exist.

Given all this, we need to reconsider the validity not only of the claim that “Finnish has more postpositions than prepositions” but of claims of this type in general. To what extent do we operate with oversimplified data? How could we take into account the variation which inevitably underlies such seemingly simple statements in reference grammars? Do we (always) need to care? And, indeed: to what extent do traditional linguistic categories like preposition or adverb direct our observations of language, and are there alternative categories or approaches which make it easier to take into account the variable nature of language data?

The paper is divided in two parts. The first part illustrates the aforementioned issues by presenting an overall view of (a group of) Finnish spatial particles and their different usages in different kinds of data. The second part makes a methodological point on the basis of this example case and discusses methodological questions related to the interplay of typological research, usefully simple descriptions, and underlying variation. The focus of the second part will be on questions rather than answers.

45

Lindström, Liina University of Tartu

Patient-like Experiencer Construction in Estonian

In Finnic languages, the Patient-like experiencers (PE) typically occur with predicates which are called also causative emotional verbs (cf. Siiroinen 2001). In most of the Finnic languages, the PE construction has an experiencer argument in the partitive (=object), and the stimulus in the nominative (=subject; ex. 1). The stimulus may be missing in most of the Finnic languages: it is not an obligatory part of the construction (ex. 2).

1) Estonian Min-d huvita-vad kriminaalromaani-d. I-PRT interest-PL3 detective_novel-PL Exp Stimulus Obj Subj

’ I am interested in detective stories.’ 1) Finnish

Matti-a pelo-tta-a. (Pörn 2008: 202) Matti-PRT be.afraid-CAUS-SG3 ‘Matti is scared.'

Estonian – and evidently also Livonian – differs from other Finnic languages by the obligatoriness of the stimulus argument (see ex. 1). This seems to be an influence of the surrounding Indo-European languages. As the stimulus is fixed to the construction the overall construction is structurally close to the typical transitive clause, but it still has some properties by which it deviates from typical transitive clauses.

In this paper, the aim is to investigate the properties which make the construction different from typical transitive clause. Special attention is payd on the subject properties of the object argument (Experiencer) in Estonian in comparision to other Finnic (Finnish, Veps, Votic) and Indo-European languages, mainly Latvian and Russian. Additionally, some properties of the construction connected to the agentivity of the stimulus will be investigated: the use of the construction in impersonal and passive, restrictions in the use of imperative, and changes in word order.

Bibliography Siiroinen, Mari 2001: Kuka pelkää, ketä pelottaa. Nykysuomen tunneverbien kielioppia ja

semantiikkaa. Suomalaisen Kirjallisuuden Seura: Helsinki. Pörn, Michaela 2008: Psychophysical and Physical Causative Emotion Verbs in Finnish: The

Temporal Structure of Causative Emotion Verb + Infinitive 1 -Constructions within Conceptual Semantics. - SKY Journal of Linguistics 21, 201–218.

46

Mankila, Anja University of Helsinki

Factors in the Choice of the Case of Nouns after the Pronoun čto (what) in Russian

In the contemporary Russian language, variation between the genitive case and the nominative case takes place within the NP “čto (what) + a noun”, the noun being most commonly a substantive adjective. My point of departure is that there exist two separate NPs: one with agreement and the other one with government, which can be demonstrated as follows. If, on the one hand, the noun stands in the genitive case, the pronoun čto acts as the head of the NP and the noun is governed by it. On the other hand, if the noun stands in the nominative case, it acts as the head of the NP and the pronoun čto is acts as an attribute agreeing with it. An example is given by the following pair:

čto novogo (what new GEN ) – čto novoe (what new NOM)

The function of the pronoun čto can be interrogative, relative or indefinite, or a combination

of them. My basic hypothesis is that the share of the nominative case increases when moving from the interrogative function toward the indefinite function. Also the referential properties of the noun and the syntactic position of the whole NP are assumed to be factors affecting the choice of the case. In addition, contextual factors, such as the presence of adverbials and the distance between the constituents of the NP are taken into consideration. The construction and testing of hypotheses is being done stepwise in light of materials, collected from Integrum, a huge Russian database with more than 50 billion running words.

In Integrum, several examples are found of duplicated sentences with the changed case form as the only difference:

Vse, čto est’ horošego/ horošee – eto ty.

(All what good GEN / good NOM there is – it is you.) In this presentation, some factors identified so far are discussed, starting with the relative

function of the pronoun čto. When the pronoun čto is used in the basic relative function, both the absence of adverbials and the use of the superlative form of the substantive adjective are shown to factors, leading to a highly significant increase of the share of the nominative case. The underlying reason for the first factor is that in the absence of adverbials, the focus is on the NP and thus the variant with agreement is more frequently required. As for the second factor, the use of the superlative form includes a comparison to a reference group, and thus the NP is more definite than the one with the positive form.

The transition from the basic relative function toward the indefinite function, accompanied by the conditional conjunction esli (if) seems to lead to a considerable increase of the share of the nominative case. An experiment with the conditional-concessive construction, introduced by the compound conjunction esli i (and if) shows that the adjacent position of the constituents of the NP leads to a highly significant increase of the share of the nominative case.

To conclude, some remarks are made on the role of these factors in a wider context, as well as on the construction of new hypotheses to be tested.

47

Nedjalkov, Igor St-Petersburg State University

Variation of Recipient Marking in Evenki (in comparison with Other Tungus-Manchu Languages)

The paper discusses the main case alternations of three-place Evenki verbs with the meanings GIVE, FEED / GIVE TO EAT, GIVE TO DRINK, TELL, SHOW, TEACH, EXPLAIN and SEND, i.e. verbs whose case frames include Agent, Theme (object or information given / transferred to Recipient) and Recipient (R). In the first part of the paper sets of argument marking patterns for each Evenki verb with above mentioned meanings are discussed, in the second part of the paper I will discuss each Evenki pattern in comparison with other Tungus-Manchu languages (TML). R is understood in the paper as the semantic actant which receives some object as a result of such ‘canonical three-participant events like ‘give’ and ‘send’ (cf. Kittilä 2005: 270). So, Rs are present in the semantic frames of such Evenki verbs expressing either physical or mental transfer as alagu- ‘teach’, anii- ‘give as a gift’, buu- ‘give’, guun- ‘tell’, nuŋni- ‘show’, ulgučen- ‘tell’, ulii- ‘feed’, uŋ- ‘send’, uniye- ‘sell’, and also derived causative verb forms devu-vken- ‘feed / give to eat’, umi-vkan- ‘give to drink, iče-vken- ‘show’, tyli-vken- ‘explain’.

Four cases can flag Rs in Evenki (i.e. their markers are added to the Recipient-(pro)nouns): Dative, Allative, Locative, and Accusative. One case (Designative) can index the R being added to the Theme-noun (see (5)):

1) Etyrken sulaki-du imuren-me buu-re-n. old.man-NOM fox-DAT fat –ACC give-NONFUT-3SG ‘The-old-man gave the fat to the fox’; 2) Girki-tki-vi tara-ve guu-kel. friend-ALL-REFL.POSS that-ACC tell-2SG.IMP ‘Tell it to your friend’; 3) Eni sin-dule ukumni-ve uŋ-če-n. mother you.SG-LOC milk-ACC send-PAST-3SG ‘Mother sent you (the) milk’; 4) Atyrkan beyetken-me ulle-t dev-uvken-deče-n. old.woman boy-ACC meat-INSTR eat-CAUS-IMPF-3SG ‘The old woman fed the boy with meat’; 5) Purta-ja-v buu-kel. knife-DES-1SG give-2SG.IMP ‘Give me a/the knife’ (lit. ‘Knife-for-me give’).

Other patterns of R-marking or indexing also discussed in the paper include reflexive possession markers added to R-nouns (see (6)), and personal possession markers added to purposive converbs on which R-s are indexed (see (7)):

6) Asi tare edy-vi guun-e-n woman that husband-REFL.POSS tell-NONFUT-3SG savoda-l-vi d’av-du nee-t-če-de-n. thing-PL-REFL.POSS boat-DAT put-DISTR-IMPERF-PURP.CONV-3SG ‘That woman told her husband to put their belongings into the boat.’ 7) Buu-kel tan-daa-v. give-2SG.IMP smoke-PURP.CONV-1SG.POSS ‘Give me [a pipe, tobacco, etc.] to smoke’ (lit. ‘Give so-that-I-smoked // in-order-to-smoke-for-me’).

48

The major construction pattern with the verbs meaning ‘GIVE’ in TML, as in other Altaic languages, is indirective – Dative flagging of Rs and Accusative suffix marking the Theme (see example (1)). Recipients with GIVE verbs in almost all TML – Even, Negidal, Solon, Nanai, Ulcha, Udihe, Oroch, and Manchu (with the only exception of Uilta/Orok) are flagged by the Dative case suffix -du. In Uilta the Allative case (but not the Dative) is obligatorily used with the verb buu- ‘give’ (see (8)). In Udihe alongside with the predominant use of the Dative case marking Rs there exists also the peripheral pattern with the Allative case marker –tigi (see (9)):

8) Uilta (Ozolinya 2001: 45) Bi mapa-tai ulise bu-hem-bi. I old.man-ALL meat give-PAST-1SG ‘I gave the meat to the-old-man’; 9) Udihe (Kyalundzyuga & Simonov 1998: 178) Si niŋka eže-tigi-ni baŋčau-ve you Chinese emperor-ALL-3SG ginseng-ACC bu-li, digana-ja… give-CONV tell-IMP ‘When you give this ginseng to the Chinese emperor, tell him…’

Verbs of speech are characterized by the highest degree of R-marking variation in Evenki. For instance, the verb guun- ‘tell’ may take Rs partly depending of the particular Evenki dialect in the Dative, Allative, or the Accusative case form, or with the reflexive possession marker (see (6)). Southern Tungusic languages (e.g. Nanai and Udihe; see: Avrorin 1981; Nikolaeva & Tolskaya 2001) are characterized by the developed syntactic type of ditransitive constructions with two objects expressing both the Theme and the Recipient in the Accusative case forms (Double Accusative construction type is very rare in Evenki in general), cf.

10) Nanai. Morim-ba mue-ve omi-vam-bori. horse-ACC water-ACC drink-CAUS-IMPERS.PRES.PART ‘It is necessary (it is time) to give water to the horse to drink’.

The choice of different case or possession markers (and their alternations) depends on a number of factors which are at play in different TMLs: 1) the lexical meaning of a concrete three-participant verb, 2) the particular dialect, 3) morphological or lexical causative verb form, 4) causative vs. permissive meaning of the causative verb (e.g. in Even and Udihe), 5) degree of activeness of the causer and passiveness of the causee in causative constructions (correlating with the feature of contact vs. distant causation), 6) aspectual differences of the causative situation expressed (single and concrete situation vs. iterative situations).

Bibliography Аvrorin, Valentin. 1981. Sintaksicheskiye issledovaniya po nanajskomu jazyku [Syntactic research of the Nanai language]. Leningrad. Kittilä, Seppo. 2005. Recipient-prominence vs. beneficiary-prominence. Linguistic Typology 9 (2): 269-297. Kyalundzyuga, Valentina, Simonov, Mikhail. An Udehe Language Dictionary. Khor river region dialect. Volume 1. Stęszev. IIEOS. 1998. Ozolinya Larisa. Oroksko-russkij slovar’ [The Orok-Russian Dictionary]. Novosibirsk. 2001. Nikolaeva, Irina & Tolskaya, Maria. 2001. A Grammar of Udihe. Berlin: Mouton de Gruyter.

49

Noel, Patrizia University of Bamberg

The Emergence of V1 Polar Questions in Germanic

In the modern Germanic languages, polar questions are indicated by word order. This way of constructing polar questions is very rare from a typological point of view. According to the World Atlas of Language Structures (Dryer 2005: 470-473), interrogative word order as an only marker of polar questions occurs only 12 times in a sample of 842 languages. Among these 12 languages are all analysed Germanic languages, i.e. Danish, Dutch, English, Frisian, German, Norwegian and Swedish. In contrast, languages with question particles form the most common class, with 532 languages in the corpus. What is striking from a historical perspective is that the question particle pattern, which is so common in the languages of the world, was given up in the history of the Germanic languages in favour of a marked syntactic pattern.

Clitic question particles are both a feature of Indo-European syntax and attested in Gothic (Eythórsson 1996), which means that their loss in the Germanic languages is not trivial. (1) gives the Latin construction, (2) the Gothic question particle in Wackernagel position after the finite verb, and (3) the Gothic question particle in Wackernagel position after the verbal prefix.

1) Lat. venisti-ne domum ad tuos penates (Catull, Carmen 9.3) ‘Have you returned home to your household gods?’ 2) Goth. skuld-u ist kaisaragild giban? (Mc 12, 14; cf. Braune 1912: § 216) ‘Is it lawful to give tribute to Caesar?’ 3) Goth. ga-u-laubjats þatei magjau þata taujan? (Mt 9, 28; cf. Braune 1912: § 216) ‘Believe ye that I am able to do this?’

In the paper, the loss of the old question particle is motivated by means of a language-specific interaction of Germanic syntax, information structure, and prosody. A hierarchy is proposed for the interacting linguistic subsystems on the basis of OV syntax, accent-based quantity, and the informational status of sentence mood markers. I offer a solution for the establishment of the rare verb-first only pattern following language-specific variation in phonological strengthening and syntactic reorganisation within the Germanic dialects of Gothic and Old-High German. In doing so, I also demonstrate why Germanic wh-questions do not show verb-first syntax.

Bibliography Braune, Wilhelm. 1912. Gotische Grammatik: Mit Lesestückchen und Wörterverzeichnis. 8th ed.

Halle: Niemeyer. Catullus, C. Valerius. Sämtliche Gedichte. Translated and edited by Michael von Albrecht. Stuttgart

2001: Philipp Reclam. Dryer, Matthew S. 2005. Polar questions. In: Martin Haspelmath, Matthew S. Dryer, David Gil und

Bernard Comrie (eds.), The world atlas of language structures. Oxford: Oxford University Press, 470–473.

Eythórsson, Thórhallur. 1996. Functional categories, cliticization, and verb movement in the Early Germanic languages. In: Höskuldur Thráinsson, Samuel David Epstein and Steve Peter (eds.), Studies in Comparative Germanic Syntax II. Dordrecht: Kluwer, 109–39.

50

Odria Tudanca, Ane; Berro Urrizelki, Ane University of Basque Country University of Basque Country

Subject and Object Marking Alternation in Bivalent Predicates: Dialectal Evidence from Basque

Goal: In bivalent predicates of Basque, we find a two way alternation: (i) absolutive/dative case variation in the object, (ii) ergative/absolutive variation in the subject. In this work, we describe and analyse the distribution of these alternations in Basque dialects, and we compare it with several other ergative and nominative languages.

Split-ergativity in Basque: Basque is an ergative language in which the subject of a transitive predicate is assigned ergative case (1) and the object takes absolutive case (1). The same case, absolutive zero case, is assigned to the subject of intransitive predicates (2). However, in intransitive predicates, Basque shows two different alignment patterns; unaccusative (2) –subjects with absolutive case– and unergative (4) –subjects with ergative case–. This division in intransitive subject marking is known as split-ergativity, and such case systems have been termed extended ergative systems (Dixon 1979, 1994). Languages showing this pattern have also been called active (Levin 1983, Rebuschi 1985).

Some of these intransitive predicates can also take a dative object, so that two possible constructions can occur: (i) bivalent unaccusative (absolutive and dative) –common in north-eastern varieties– and (ii) bivalent unergative (ergative and dative) –preferred in south-western varieties (Fernández and Ortiz de Urbina 2010).

III. Object marking alternation: bivalent predicates which take a dative object can be: verbs without alternation (unergative verbs with an ergative DP and a dative DP) and alternating verbs in which the object can be either (canonical) absolutive (5) or (quirky) dative (6) (Etxepare 2003, Mounole 2008). We call quirky dative object to the object of the alternating verb marked by dative case (Ortiz de Urbina 1995, Fernández & Ortiz de Urbina 2010, Fernández & Landa 2009). Due to this different marking, these verbs can occur in two different constructions: transitive –with an absolutive object– (5) and bivalent unergative –with a dative object– (6). Alternation is three variable dependent (Fernández & Ortiz de Urbina 2010): (i) dialectal like in verbs as deitu ‘call’ or entzun ‘hear’; semantic like in laga ‘let, leave’ or jarraitu ‘follow’ –this type of alternation is also found in Djaru (Tsunoda 1994) and Chepang (DeLancey 1981); and (iii) plain alternation like in barkatu ‘forgive’ or bultzatu ‘push’. It is worth noting that this marking alternation in the object is also found in nominative languages such as Spanish (Fernández Ordóñez 1999), Catalan, Occitan, Italian (Pinneda, 2011), Galician (Rivas 2005) and Icelandic (Gísli Jònsson, 2010) and other ergative languages such as Chechen (Nichols 1984) or Daghestanian languages (Kibrik 1987).

IV. Subject marking alternation: Ergative/absolutive alternation in bivalent constructions has been reported in (i) verbs with aspectual datives, (ii) verbs having quirky dative objects and (iii) other bivalent verbs. When the object is marked with dative case, an absolutive subject is preferred in north-eastern dialects (7), contrary to south-western ones where ergative subjects are used (6). Change from ergative subject to absolutive with the presence of a dative object has also been attested in other ergative languages such as Kalkutungu (Blake 1982) and Yalarnnga (Mallison and Blake 1981).

51

V. Examples

Mirenek liburua erosi du Miren.ERG book.ABS buy.ASP (3ABS).have.3sgERG ‘Miren has bought the book’ Miren etorri da Miren.ABS come.ASP (3sgABS).be ‘Miren has come’ Mirenek alabari liburua erosi dio Miren.ERG daughter.DAT book.ABS buy (3ABS).(have).DF.3sgDAT ‘Miren has bought the/a book to her daughter’ Mirenek oso ondo dantzatzen du. Miren.ERG very well dance-ASP expl.have.3sgERG ‘Miren dances very well’ Mirenek Peio jarraitzen du Ane.ERG Jon.ABS follow-ASP (3ABS).have.3sgERG ‘Miren follows Pello’ Mirenek Pellori jarraitzen dio Miren.ERG Pello.DAT follow-ASP expl.(have).DF.3sgDAT.3sgERG ‘Miren follows Pello’ Miren Pellori jarraikitzen zaio Miren.ABS Pello.DAT follow-IMP expl.be.DF.3sgDAT ‘Miren follows Pello’

Bibliography DeLancey 1981. “An interpretation of split ergativity and related patterns”, Language 57. 626-657. Dixon. 1979. “Ergativity”, In Language 55/1. 59-138. ____.1994. Ergativity. Cambridge University Press. Etxepare, R.. 2003. “Valency and argument structure in the Verb” In Jose Ignacio Hualde & Jon

Ortiz de Urbina (eds.). A grammar of Basque. Berlin: Mouton de Gruyter. Etxepare, R. & B. Oyharçabal.2008a. ‘Locational datives and dative agreement in the eastern

dialects of Basque’. In European Dialect Syntax III, Venice, September 2008. ____. 2008b. ‘Bi datibo egitura ifar ekialdeko zenbait hizkeretan’, In Aldaketak, aldaerak,

bariazioak euskaran eta euskal testugintzan’, Baiona, December 2008. ____. 2009. ‘Datiboa ifar ekialdeko zenbait hizkeretan’, IKER, April 2009. Fernández, B. & J. Landa. 2009. “Datibo komunztadura beti zaindu, inoiz zaindu ez eta batzuetan

baino zaintzen ez denean. Hiru ahoko aldagaia, datu iturri bi, eta erreminta bat: Corsintax”. To appear in: Lapurdum.

____. 2010. Datiboa hiztegian. Bilbo: UPV/EHUko Argitalpen Zerbitzua. Fernández-Ordoñez, I. 1999. “Leismo, laismo y loismo”. In Ignacio Bosque & Violeta Demonte

(eds.). Gramática descriptiva de la lengua española I. Madrid, Espasa Calpe. 1317-1394.

52

Pekkarinen, Heli Research Institute for the Languages of Finland

The grammaticalization from a Pronoun se 'it' to an Expletive in Finnish Necessive Constructions. Evidence from Dialect Data.

Although Finnish is usually considered as a language lacking expletive subjects, they are common in subjectless constructions in colloquial Finnish. However, they are not obligatory and they often lend the expression an affective tone. This paper aims to show that the grammaticalization from a pronoun into an expletive in Finnish takes place within modal or otherwise subjective constructions. Since a pronoun frequently occurs in a construction that itself has subjective discourse functions, it gradually receives an affective reading and looses its referentiality.

I will concentrate on the grammaticalization of the pronoun se 'it' to an expletive in a construction that includes of a modal verb or compound (‘be’ + passive present participle) expressing obligation or recommendation. This necessive construction, as it is called here, forms a syntactically specific construction in Finnish. It takes a genitive subject instead of a nominative one and hence does not show person agreement; the verb always appears in the 3rdperson singular form. In actual language data, however, the oblique subject is very often missing, creating a open reference. (In more detail, see Helasvuo & Vilkuna 2008.)

In subjectless necessive constructions, the preverbal position typical of the subject in Finnish is often occupied by se 'it'. Based on empirical data obtained from large collections of transcribed material from Finnish dialects, I aim to show that there is a continuum between se as a referential pronoun exemplified in (1) and expletive se as in (2) rather than a strict dichotomy since the pronoun se may involve a reference to a situation in general or some aspects of the situation. I suggest that even expletive se is not totally devoid of referential meaning but seems to retain some aspects of the general reference as it gradually turns into an expletive-like element (cf. Bolinger 1973).

1) nyt piika pitää tyhyjentä-ä now maid must.3SG.PRES empty.INF kaffipannu ja keittä-ä uut-ta ja coffee pot and boil.INF new.PART and se pitää pestä hyväks. (Veteli) it must.3SG.PRES wash.INF good.TRA ”The maid must empty the coffee pot and make new coffee and it (the coffee pot) must be washed properly”

2) jos saeraala-an meno tuli nii se if hospital-ILL going come.3SG.PST so it täöty sitte men-näk kunnantoemisto-oj must.3SG.PST then go.INF municipal office.ILL ja siellä sitte päättää että mihinkä (Kestilä) and there then decide.INF that where “If you needed hospital care, you had to go to the municipial office where they decide in which hospital.”

This paper supports the view that the grammaticalization is takes place within specific constructions (eg. Bybee 2010: 106). Consequently, it is not arbitrary that it is precisely the nominative form the pronoun 'it' which is gradually turning into an expletive in the necessive construction since the form is determined by the properties of the construction.

53

Bibliography Bolinger Dwight, 1973, Ambient it is Meaningful Too. Journal of Linguistics. 1973;9:261-270. Bybee, Joan 2010: Language, Usage and Cognition. Cambridge University Press, Cambridge. Helasvuo Marja-Liisa & Maria Vilkuna, 2008, "Impersonal is Personal: Finnish perspectives",

Transactions of the Philological Society106, s. 216-245

54

Peltola, Rea Institut national des langues et civilisations orientales (INALCO), Paris, University of Helsinki

Finnish Jussive Mood in Dialects and in Standard Language

The aim of this presentation is to examine the syntactic and semantic factors contributing to the interpretation of Finnish jussive mood, that is the third person and passive forms of imperative, in dialects and in standard language. Jussive displays a wider variety of uses in the Eastern dialects than in the Western ones or in standard Finnish, which is a particularity jussive shares with other imperative forms (see Forsberg, forthcoming).

While in standard Finnish jussive is formed with the affix -kO and the personal endings -On and -Ot, usually considered as coding the difference between singular and plural (e.g. osta-ko-on ‘buy-JUSS-JUSS.SG’ > ‘let him/her buy’; osta-ko-ot ‘buy-JUSS-JUSS.PL’ > ‘let them buy’), in non-standard variants it is marked by a larger series of affixes and often without the final consonant of the personal ending. Unlike second person imperatives, jussive can occur with an explicit subject, the word order following the same principles as that of an indicative clause. Furthermore, jussive diplays passive conjugation and a compound form. As the intention coded by jussive is addressed to a third person, or to a non-specific person, instead of the interlocutor, a jussive clause tends to be interpreted rather as an expression of permission than as a command. (See Hakulinen et al. 2004: § 1666–1667.)

Permission can entail active authorization or passive non-opposition (see Lyons 1977: 836–837; J. Leino 2003: 137). I argue that the permissive reading of jussive is highly dependent, not only on the semantics of the verb and on other modalities present in the context, but also on the syntactic position the jussive clause holds. First, in main clauses, jussive tends to express non-opposition to a necessity that in standard language uses is typically imposed by an intentional third party. This produces a reading of active authorization. Note that the source of intention is not necessarily human: in (1), the referent of the NP vanhuus (‘old age’) is presented as intentional, as indicated by the jussive form tulkoon and the infinite construction tullakseen (see Penttilä 2002 [1963]: 492–493; P. Leino 2005).

(1) Oleminen käy raskaaksi, ellemme hyväksy omaa ikäämme. Tulkoon vanhuus sitten, kun on tullakseen. (FTC, Turun Sanomat 1999.) ’It’s hard to live, if we don’t accept our own age. Let the old age come, when it is to come.’

In Eastern dialects, permission can also manifest itself as not opposing to a necessity imposed by non-intentional circumstances, as in (2) and (3). This type of use is typical of contexts where jussive clause codes a consequence. In the examples presented here, the cause – or rather the condition – is expressed by a conditional clause, in (2), and by an autonomous relative clause, in (3).

(2) jos ol semmone köyhä kaaso ni antoha ne sillekkii t... siit jottai mut jos ol rikkaamp ni / kävelköö ilma. (LaX, Nuijamaa.) ’if it was a poor bridesmaid then they surely gave her as well… some of it but if it was a richer one then / she would have to go without.’

(3) ka mitäpä siinä / vietettii kell- ol' viinaa se joi viinaa ja mitäs siinä ol'i kell- ei ollu ni sitte olkoo juomatta. (LaX,

Sortavalan mlk.)

55

’well nothin special / one celebrated he who had liquor he drank liquor and that’s it he who didn’t have well then he would have to be without’

Second, in both standard language and in dialects, the meaning of not opposing can in complex constructions occur as an interclausal concessive relation, as in example (4). In standard Finnish, it is the jussive form that marks the first clause as concessive – and subordinate, in functional terms (cf. p. ex. Cristofaro 2003), whereas in dialectal data, jussive clause can be introduced by the concessive conjunction vaik(ka) (‘though’), as in (5).

(4) Olkoon [hiihto]putken sijoituspaikka Kuopio, Nilsiä tai Leppävirta, se on saatava. (FTC, Aamulehti 1999.) ’Be it situated in Kuopio, Nilsiä or Leppävirta, we need to have this ski track.’

(5) ja ko se pakotti oikei julmast tuost tuota selkää / ni miäkii ku sanon' / sisarellein- et ku pakottoa selkeä ja on / ni se sano miule ni- et / eis se märkeä tieh vaik sitä pakottakkoo. (LaX, Taipalsaari.) ’and as my back was aching really bad there / I said / to my sister that my back is aching and it’s / so she said to me that / it won’t fester no matter if it aches.

In the light of this large scale of permissive readings, connected to the syntactic position of the clause, Finnish jussive can be considered as an example of a form that covers different layers of modal meanings, from agent-oriented to subordinate, as suggested by the typological model of Bybee, Perkins & Pagliuca (1994).

Bibliography Bybee, Joan, Revere Perkins & William Pagliuca 1994: The Evolution of Grammar.

Chicago/London: The University of Chicago Press. Cristofaro, Sonia 2003: Subordination. Oxford Studies in Typology and Linguistic Theory. Oxford

University Press. Forsberg, Hannele, forthcoming: Imperatiivi kannanottona – varoituksesta paheksuntaan. Article

manuscript. Hakulinen, Auli, Maria Vilkuna, Riitta Korhonen, Vesa Koivisto, Tarja Riitta Heinonen & Irja Alho

(eds.) 2004: Iso suomen kielioppi. Helsinki: Finnish Literature Society. Leino, Jaakko 2003: Antaa sen muuttua. Suomen kielen permissiivirakenne ja sen kehitys. Helsinki:

Finnish Literature Society. Leino, Pentti 2005: Kun tarkoitus häviää – finaalisesta infinitiivistä temporaaliseksi. Elävä

kielioppi. Suomen infiniittisten rakenteiden dynamiikkaa, I. Herlin, Ilona & L. Visapää (eds.), Helsinki: Finnish Literature Society, 194-230.

Lyons, John 1977: Semantics. Volume 2. Cambridge: Cambridge University Press. Penttilä, Aarni 2002 [1963]: Suomen kielioppi.

56

Pešková, Andrea University of Hamburg

Overt and Null Subject Pronouns in Experimental Data: The Case of Porteño Spanish

This paper investigates the realization of pronominal subjects in semi-spontaneous speech of Porteño, the variety of Spanish spoken in Buenos Aires. Based on data from an elicited production task it is shown that various pragmatic and semantic factors play an important role in the overt realization of subject pronouns.

Spanish is a pro-drop or null-subject language, which thus permits the omission of pronominal subjects. The question arises as to what triggers their overt realization. Whereas the Hispanic traditional grammar argues for reasons such as ambiguity resolution, contrast, and emphasis (e.g. Alarcos Llorach 1994), studies by Barrenechea/Alonso 1977 and Silva-Corvalán 1994 among others show that it is important to also take into account various linguistic or even extra-linguistic factors. My study concentrates on the following factors: (1) grammatical person, (2) verb semantics (epistemic vs. perceptive verbs), (3) type of sentence (declaratives, interrogatives), and (4) type of clause (simple, main, subordinate).

In order to determine the relevance of these factors I conducted a production experiment with 13 monolingual speakers of Porteño-Spanish in Buenos Aires. A questionnaire with 126 everyday (non-contrastive and non-ambiguous) situations was created to obtain target-sentences, in which the speaker had the possibility to either realize a subject pronoun or not (e.g. a situation like “Ask your father what he thinks about Argentina”, presented by the researcher can evoke a target-question from the participant such as ¿Qué opinás (vos) de Argentina? ‘What do [you] think of Argentina?’). Of a total of 1638 sentences only 12 had to be discarded due to speaker misperformance.

The results show that the speakers realize the subject pronoun in 48% of all contexts. As for factor (1), singular persons are more often overtly realized than plural persons (Table 1). Moreover, there is a correlation between the overt pronoun rate and the grammatical persons (e.g. vos vs. usted), which differ in percentages although they have been controlled for by the same contextual conditions (Table 2). Regarding verb semantics, subjects are more often overtly realized with epistemic verbs than with perceptive verbs (Table 3). Concerning factor (3), the realization of subject pronouns is higher in interrogatives than in declaratives (Table 4). And finally (factor 4), subject pronouns are more often used in simple or main clauses than in subordinate clauses (Table 5).

The data confirm previous findings and furthermore support the hypothesis that the overt realization of pronominal subjects not only depends on ambiguity resolution or contrastiveness, but also on factors such as grammatical person. Furthermore, dialectal differences can be relevant as well: Speakers of Buenos Aires Spanish seem to realize pronominal subjects more frequently than speakers of the peninsular variety, but less frequently than speakers of Caribbean dialects (cf. Hochberg 1986, Soares da Silva 2006). Moreover, the experimental technique used in this study can be applied to future cross-dialectal or even cross-linguistic research, since it is possible to obtain controllable and comparable sets of data, which are difficult to obtain with other methods.

Bibliography Alarcos Llorach, E. 1994. Gramática de la lengua española. Madrid: Espasa-Calpe. Barrenechea, A. M./ Alonso, A. 1977. Los pronombres personales sujetos en el español hablado en

Buenos Aires. Estudios sobre el español hablado en las principales ciudades de América, Lope Blanch, J. M. (ed.), México: Universidad Autónoma de México, 333-349.

57

Hochberg, J. G. 1986. Functional Compensation for /s/ Deletion in Puerto Rican Spanish. Language 62(3): 609-621.

Silva-Corvalán, C. 1994. Language contact and change: Spanish in Los Angeles. Oxford: Oxford University Press.

Soares da Silva, H. 2006. O parâmetro do sujeito nulo: Confronto entre o português o e espanhol. MA thesis. Universidade Federal do Rio de Janeiro.

Person Overt pronoun (of total sentences) In %

Singular 477 (932) 51% Plural 304 (694) 44% TOTAL 781 (1626) 48%

Table 1: Use of pronominal subject according the number.

Person Overt pronoun (of total sentences) In % vos (‘you’, familiar Sg.) 76 (232) 33% usted (‘you’, formal Sg.) 164 (234) 70% ustedes (‘you’, familiar Pl.) 110 (232) 47% él/ella (‘he/she’) 128 (232) 55% ellos/ellas (‘they’) 109 (229) 48% nosotros (‘we’) 85 (233) 36% yo (‘I’) 109 (234) 47%

Table 2: Use of pronominal subject according the grammatical person.

Type of Verb Overt pronoun (of total sentences) In % Epistemic verbs 462 (815) 57% Perceptive verbs 319 (811) 39%

Table 3: Use of pronominal subject according the verbal semantic.

Type of Sentence Overt pronoun (of total sentences) In % Wh-questions 124 (236) 53% Yes/no questions 277 (537) 52% Declaratives 380 (853) 45%

Table 4: Use of pronominal subject according the type of sentence.

Type of Clause Overt pronoun (of total sentences) In % Simple 359 (720) 50% Main 299 (553) 54% Subordinate 123 (353) 35%

Table 5: Use of pronominal subject according the type of clause.

58

Riionheimo, Helka University of Eastern Finland

Colliding Passives: A Morphosyntactic Perspective on Finnic Language Contacts

Contacts among cognate languages involve the juxtaposition of partly overlapping grammatical sys-tems and structures. Due to their shared history, many of the structures have a common origin but have later developed into different directions. In contact, the similarities form a hole through which the differences may flow from one language to the other leading to cross-linguistic interference. This presentation examines a contact of multifaceted morphosyntactic structures: the use of so-called passive forms in a contact between two Finnic languages, Finnish and Estonian. The topic invokes themes related to contact linguistics and studies on dialect syntax (see e.g. Kortman 2004). The Finnic passive is an impersonal verbal category that leaves the agent/subject argument unspeci-fied, implies the involvement of a human agent/subject in the situation, and allows both (semanti-cally) plural and singular and speaker/hearer exclusive and inclusive interpretations of the agent/subject (see e.g. Shore 1988, Helasvuo & Vilkuna 2008). In modern Finnish dialects, the pas-sive form is also routinely used as the PL1 verb form with a subject pronoun (myö männään ulos we go-PASS out ‘we are going out’). The Estonian passive forms have a fairly similar basic function: the passive structures are subjectless and refer to an unspecified human agent. Estonian differs from Finnish in that the passive forms are not used inclusively with a PL1 pronoun. Furthermore, Esto-nian has several passive structures that are not known in Finnish (see e.g. Erelt 2000, Tragel & Lindström 2007). Previous research (Riionheimo 2007, Kokko 2007) has shown that the contact of complex overlapping features may result in widespread variation. In this presentation, the variable use of passive forms will be analyzed on the basis of authentic, audio-recorded data.

Bibliography Erelt, M. (ed.) 2003. Estonian language. Linguistica Uralica supplementary series / volume 1. Tal-

linn: Estonian Academy of Sciences. Helasvuo, M. L. & Vilkuna, M. 2008. Impersonal is personal: Finnish perspectives. Transactions of

the Philological Society. Volume 106:2, 216–245. Kokko, O. 2007. Inkerinsuomen pirstaleisuus. Eräiden sijojen kehitys murteen yksilöllistymisen

kuvastajana [Scattered Ingrian Finnish.The development of selected cases as reflectors of a dialect]. Joensuu: University of Joensuu.

Kortmann, B. (ed.) 2004. Dialectology Meets Typology. Dialect Grammar from a Cross-linguistic Perspective. Trends in Linguistics, Studies and Monopraphs, vol 153. Berlin: Mouton de Gruyter.

Riionheimo, H. 2007. Muutoksen monet juuret. Oman ja vieraan risteytyminen Viron inkerinsuoma-laisten imperfektinmuodostuksessa [The multiple roots of change. Mixing native and borrowed influence in the past tense formation by Ingrian Finns in Estonia]. Helsinki: Finnish Literature Society.

Shore, S. 1988. ‘On the so-called Finnish passive’. Word 39, 151–176. Tragel, I. & Lindström, L. 2007: Eesti keele impersonaali ja seisundipassiivi vahekorrast

adessiiargumendi kasutamise põhjal [Estonian Impersonal and Stative Passive, and the Use of the Adessive Argument]. – Keel ja kirjandus 7/2007.

59

Rosenkvist, Henrik Centre for Languages and Literature, Lund University

Diachronic Typology and Synchronic Variation – Referential Null Subjects in Old and Modern Germanic

Referential null subjects (RNS; cf. Rixxi 1986, Jaeggli & Safir 1989, Barbosa 1995, Ackema et al 2006, Frascarelli 2007, Barbosa 2009, Cole 2009, Biberauer et al 2010 etc) occur in both Old and Modern Germanic languages (Rosenkvist 2009).6

On the other hand, the RNS in Old Germanic languages (Old High German, Old English,

However, although it has been claimed that the null subjects that can be found in e.g. Zürich German are a direct continuation from Old High German (Axel & Weiß to appear), Fuß (2005) show that RNS may develop through grammaticalization (as in e.g. Bavarian), and Rosenkvist (2009, 2010) argue that the RNS that are present in the modern Germanic vernaculars are typologically different from the ones found in Old Germanic. There are several indications that point to this conclusion, such as differences regarding distribution in main/embedded clauses, the relation to distinct verb agreement, person reference and frequency.

7 Old Icelandic, Old Swedish) are remarkably similar, from a syntactic perspective. This suggests that these language varieties could be regarded as dialects (cf. Davis 2006). Similarly, the RNS that can be found in Modern Germanic vernaculars (Bavarian, Lower Bavarian, Zürich German, Frisian, Schwabian, Yiddish8

The picture that emerges is thus that within the respective synchronic stages (Old v. Modern Germanic), there are merely dialectal differences when it comes to the principles for RNS, while there is a typological difference separating all Old Germanic languages from all Modern Germanic vernaculars. This observation raises several questions: why are e.g. Old Icelandic and Old High German so similar, and why are RNS in Övdalian, which is spoken in central Sweden (cf. Rosenkvist 2009, 2010, Garbacz 2010), restricted in the same manner as RNS are restricted in e.g. Bavarian and Frisian? Given that some of these language varieties are spoken far away from each other it seems unreasonable to assume that language contact has caused the similarities, but in that case, how can the attested similarities be explained?

and Övdalian) follow remarkably similar syntactic conditions in all of these language varieties, and again the properties of the RNS indicate that in this particular aspect, we are dealing with dialectal variation within Modern Germanic.

In this talk, I will first demonstrate that RNS appear in all of the language varieties that have been mentioned above, and show that the RNS can be divided in two separate groups (Old v. Modern Germanic) with the aid of syntactic criteria. Then I will discuss the possible causes for the observed situation, arguing that there are language internal mechanisms that regulate the distribution and reference of RNS, which are part of the typological core. Hence, it seems that there has been a general Germanic typological shift (Old > Modern Germanic) which also has affected the possibilities for identification of omitted referential subjects (cf. Kiss 1995, Holmberg & Nikanne 2002, Faarlund 2003, Håkansson 2008).

6 Jaeggli & Safir (1989) and Rohrbacher (1999) claim that RNS are universally incompatible with V2-word order; this

has however been shown to be incorrect. 7 Hulk & van Kemenade (1995:245) claim that ”The phenomenon of referential pro-drop does not exist in Old English”,

but see van Gelderen (2000:137) and Rosenkvist (2009). 8 Speas (2006:60) states that “Yiddish does not allow null referential pronouns”, but see Prince (1998), Geller (1999),

Jacobs (2005) and Rosenkvist (2009).

60

Bibliography Ackema, P, Brandt, P, Schoorlemmer, M, & Weerman, F. 2006. The Role of Agreement in the Expression of Arguments. In: Ackema, Peter, Patrick Brandt, Maaike Schoorlemmer and Fred Weerman (eds.). Arguments and Agreement, 1-34. Oxford: OUP. Axel, K. & Weiß, H. To appear. Pro-drop in the History of German. From Old High German to the modern dialects. Empty pronouns. Ed. P. Gallman and M. Wratil. Berlin: De Gruyter. Barbosa, Pilar. 1995. Null Subjects. Cambridge, Mass. The MIT Press. Barbosa. M. 2009. Two kinds of subject pro. Studia Linguistica 63,1:2–58. Biberauer, Theresa & Anders Holmberg, Ian Roberts & Michelle Sheehan (eds.). 2010.

Parametric Variation: Null Subjects in Minimalist Theory. CUP, Cambridge. Cole, M. 2009. Null Subjects: a Reanalysis of the Data. Linguistics 47.3:559–587. Davis, G. 2006. Comparative Syntax of Old English and Old Icelandic. Peter Lang, Bern . Frascarelli, M. 2007. Subjects, Topics and the Interpretation of Referential pro. Natural Language and Linguistic Theory 25:691-734. Fuß, E. 2005. The Rise of Agreement. John Benjamins, Amsterdam/Philadelphia. Garbacz, P. 2010. Word Order in Övdalian. Ph. D. thesis. The Centre for Languages and

Literature, Lund University. van Gelderen, Elly. 2000. A History of English Reflexive Pronouns. John Benjamins,

Amsterdam/Philadelphia. Geller, Ewa. 1999. Hidden Slavic Structure in Modern Yiddish. Jiddische Philologie. Festschrift für Erika Timm. Max Niemeyer Verlag, Tübingen. Hulk, A. & van Kemenade, A. 1995. Verb Second, Pro-drop, Functional Projections and Language

Change. Clause Structure and Language Change, 227–256. Ed. Adrian Battye and Ian Roberts. OUP, Oxford. Jacobs, N. 2005. Yiddish. A Linguistic Introduction. CUP, Cambridge. Jaeggli, O. & Safir, K. 1989. The null-subject parameter and parametric theory. The Null Subject Parameter, 1-44. Ed. Osvaldo Jaeggli & Ken Safir. Dordrecht, Kluwer. Prince, E. 1998. Subject-prodrop in Yiddish. Focus: linguistic, cognitive and computational perspectives:82-104. Ed. P. Bosch & R. van der Sandt. Cambridge, CUP. Rizzi, L. 1986. Null objects in Italian and the theory of pro. Linguistic Inquiry 17:501-557. Rohrbacher, B. W. 1999. Morphology-Driven Syntax: A Theory of V to I Raising and Pro-

Drop. John Benjamins, Amsterdam/Philadelphia. Rosenkvist, H. 2009. Null Referential Subjects in Germanic – an Overview. WPSS 84:151– 180. Rosenkvist, H. 2010. Null Referential Subjects in Övdalian. Nordic Journal of Linguistics 33.3:231–267. Speas, M. 2006. Economy, Agreement, and the Representation of Null Arguments. In: Ackema,

Peter, Patrick Brandt, Maaike Schoorlemmer and Fred Weerman (eds.). Arguments and Agreement, 35–75. Oxford: OUP.

61

Schikowski, Robert University of Leipzig

Obligatory Coreference in Chintang

Chintang (Tibeto-Burman > Kiranti, spoken in Eastern Nepal) offers a wide range of mechanisms for combining clauses (Bickel 2008; Paudyāl et al. 2010). The various constructions it uses are extremely heterogeneous with respect to several factors. This talk focusses on coreferentiality as one of them. Consider the following examples:

1) I-cop-si u-ti-a-ŋs-e naŋ. 2sPOSS-look.at-PURP 3[ns]S-come-PST-PRF-IND.PST but ‘But they came to have a look at you.’ [CLLDCh1R02S03a.108]

2) Darkem ma-ce sa u-sed-a=go Darkem mother-ns meat 3[ns]S-kill-PST=NMLZ1 na-khutt-i=kha? 3>2-bring-2pO=NMLZ2 ‘Did Darkem’s mother bring you some of the meat they killed?’ [CLDLCh2R02S02.241]

3) Bhale-ŋa thok-ma na-kon-no goneı. cock-ERG peck-INF 3>2[s]-search-IND.NPST ATTN ‘Watch out, the cock is trying to peck you.’ [CLLDCh1R07S02 847]

4) Cekt-a-kh-a, anne-ŋa na-khem-lok! speak-IMP-CON-IMP[.2sS] elder.sister-ERG 3>2[s]-hear-[SUBJ.]SIM ‘Speak so that the girl can hear you!’ [Ch4R12S08.444]

Each of these sentences contains a subordination marker: the purposive -si in (1), the nominaliser =go in (2), the infinitive -ma in (3), and the simultaneous converb -lok in (4). The corresponding constructions do not only differ with respect to factors such as finiteness and the function of the embedded clause in the matrix but also with respect to coreferentiality constraints.

• Syntactic function: What type of syntactic entities does the constraint make reference to (e.g. case, grammatical relations, semantic roles, information structure...)?

• Coreferent values: Which values of this function are concerned (e.g. nominative, subject, S, topic...)? • Symmetry: Do these values have to be the same in all involved clauses? • Overtness: In which of the involved clauses can the shared referent be expressed overtly? • Number of shared referents: How many referents must be shared (one or several)? • Sharing space: Is the shared referent located in a single or in several mental spaces (Fauconnier 1984)

which are linked? • Form of referent: If the shared referent is expressed overtly, is it restricted to a certain part of speech (e.g.

pronouns) or must it be marked in some special way (e.g. with a special case - cf. Bickel & Yadava 2000 on Hindi, Nepali, and Maithili)?

• Form of predicate: Do the combined predicates have to be marked in some special way (e.g. with respect to agreement or diathesis)?

The talk explores these component variables in greater detail and shows how they can be used to distinguish between the various relevant constructions of Chintang. The same variables may be easily applied to other languages as well, thus providing a means of rendering the description of coreferentiality in general more precise.

Bibliography Bickel, Balthasar. 2008. Grammatical relations in Chintang. Paper presented at the Max Planck

Institute for Evolutionary Anthropology, Leipzig, October 7, 2008.

62

Bickel, Balthasar. 2010. Capturing particulars and universals in clause linkage: a multivariate analysis. In Isabelle Bril (ed.), Clause-hierarchy and clause-linking: the syntax and pragmatics interface, Amsterdam: John Benjamins.

Bickel, Balthasar & Yogendra Yadava. 2000. A fresh look at grammatical relations in Indo-Aryan. Lingua 110. 343--373.

Fauconnier, Gilles. 1984. Espaces mentaux. Aspects de la construction du sens dans les langues naturelles. Paris: Les éditions de Minuit.

Paudyāl, Netra Prasād, Balthasar Bickel, Robert Schikowski, Sabine Stoll, Elena Lieven, Gomā Banjade, Iccha Pūrṇa

Rāī, Manoj Rāī, Martin Gaenszle, Noval Kiśor Rāī & Toyā Nāth Bhaṭṭa. 2010. Non-finite adverbial subordination in Chintang. Nepalese Linguistics 25. 121--132.

63

Sinnemäki, Kaius University of Helsinki

Where to Find Language Universals?

It has been a general trend in language typology to code typological variation at a rather abstract (or reductionist) level. For example, in the World atlas of language structures (Haspelmath et al. 2005), 8% of the grammatical variables are binomial (with values “feature is present” vs. “feature is absent”) and almost half of the variables are classified in no more than four types. However, many typologists have recently called for more analytic and detailed analysis of typological variables, paying attention to the full variation in the data (e.g. Bickel 2007; Dixon 2009: 257-263). This approach takes typology and the study of language universals closer to other variationist subdisciplines, such as sociolinguistics. When the level of abstraction is lowered in this way, possible language universals can be scrutinizing at many different levels of abstraction, from the most abstract to the most detailed coding of variables.

Although this move towards more variationist methods in typology is certainly welcome, it remains to be seen what its consequences are for the study of language universals. In this presentation I would like to preliminarily suggest that the more fully typological variation is coded, the more likely the data will reflect genealogical and areal patterns (or no patterns at all) instead of reflecting language universals. I also hypothesize that the more often than not language universals will cluster to the more abstract or reductionist levels of coding of typological variables (as in the WALS).

To evaluate these hypotheses, I discuss examples of the variation of language complexity in different domains. The focus is on core argument marking in general but differential object marking is also discussed. Complexity is measured by contrasting different complexity metrics as well as paying attention to different types of complexity (in the spirit of Rescher 1999). Preliminary results support the hypotheses: significant patterns occur mostly at a rather abstract level of coding and the lowering of the level of abstraction does not reveal new global patterns.

But why should language universals cluster towards the more abstract levels of coding? One possible answer is that linguistic diversity is just so overwhelmingly great that once it is taken into account, we no longer find global patterns but rather local or totally random ones. This is not bad news for typology, because local linguistic patterns (or even patterns that at first seem random) may be possible to connect with other local patterns, such as social structure.

Bibliography Bickel, Balthasar 2007. Typology in the 21st century: Major current developments. Linguistic

Typology 11(1): 239-251. Dixon, R. M. W. 2009. Basic Linguistics Theory Volume 1: Methodology. Oxford: Oxford

University Press. Haspelmath, Martin, Matthew S. Dryer, David Gil & Bernard Comrie (eds.) 2005. The world atlas

of language structures. Oxford: Oxford University Press. Rescher, Nicholas 1998. Complexity: A philosophical overview. New Brunswick: Transaction.

64

Seiler, Guido University of Freiburg

Non-Standard Average European: How Robust Is the European Sprachbund beyond the

Codified Standard Varieties? (Western) European languages display a number of morphosyntactic convergences (Standard

Average European features, henceforth SAE, Haspelmath 2001). It is a well-established fact that these languages form a linguistic area, yet the evidence for SAE is mainly taken from the written standard languages. So far, it is an open question how pervasive SAE features are at the level of vernaculars (but see Kortmann 2009 for varieties of English, Murelli 2008 for a comparative investigation of nonstandard relativization). In this programmatic paper we argue that including non-standard varieties into typological comparison is not just a matter of granularity, but it may create a qualitatively different picture of Europe's position in typological space. We hypothesize that SAE features are less widespread in vernaculars than in standard varieties.

A list of SAE features (as presented by Haspelmath 2001) will be discussed from the perspective of German dialects, in particular Upper German (based on questionnaire studies, transcripts of spontaneous conversations, and grammatical descriptions). Whereas for some features no substantial difference between standard German and German dialects is found, others are indeed indicative of a less SAE structure of vernaculars. Crucially, none of the SAE features is more articulate in vernaculars than in the standard language. Less "European" traits of vernaculars are found in the following morphosyntactic domains:

1. Relativization (Fleischer 2004, Murelli 2008)

2. Negation (Weiß 2002)

3. (Non)Obligatoriness of subject pronouns (Cooper 1994)

4. (Non)Identity of intensifiers and reflexive pronouns

5. Alienable/inalienable opposition in adnominal possession

6. Particles marking discourse pragmatic notions

7. Specialized markers for depictive secondary predication (Bucheli Berger 2005)

65

Given that SAE is a gradual rather than a categorical entity, our preliminary conclusion is that the varieties under inspection are less prototypical representatives of the SAE type than standard German.

The consequences of our study are threefold. First, our findings shed new light on the historical emergence of SAE. If it is right that the European linguistic area is a matter of written standard languages in the first place, SAE must be the result of common pathways of standardization rather than the result of areal convergence as such. Second, Chambers (2004) introduces the notion of vernacular universals in order to explain reoccurring structural differences between standard languages and vernaculars. If SAE features are crosslinguistically rare and cluster in standard varieties in the first place, it is unnecessary to assume a specialized set of vernacular universals (vernacular universals are simply language universals). Third, the present paper is also intended to be an invitation. The documentation of vernacular morphosyntax has made significant progresses in recent years. It is the right time to reconsider Europe's position in typological space, taking into account all varieties without any privileges for the codified standard varieties. Only through a joint effort of typologically minded variationists with different language specializations it will become clear how far our preliminary conclusions can be generalized.

Bibliography Langacker, Ronald W. 2009: A dynamic view of usage and language acquisition. Cognitive

Linguistics 20(3): 627-640. Langacker, Ronald W. 2009: Investigations in Cognitive Grammar. Berlin, New York: Mouton de

Gruyter. Divjak, D. 2006. Ways of intending: Delineating and structuring near-synonyms. Corpora in

cognitive linguistics, St. Th. Gries & A. Stefanowitsch (eds), Berlin: Mouton, 19-56. Arppe, A. 2008. Univariate, bivariate and multivariate methods in corpus-based lexicography - a

study of synonymy. Dept. of General Linguistics, University of Helsinki, http://urn.fi/URN:ISBN:978-952-10-5175-3.

Bucheli Berger, C. 2005: Depictive agreement and the development of a depictive marker in Swiss German dialects. In: Secondary predication and adverbial modification. Crosslinguistic explorations in the syntax and semantics of depictives, N. P. Himmelmann & E. Schultze-Berndt (eds), Oxford: Oxford University Press, 141-171.

Chambers, Jack K. 2004: Dynamic typology and vernacular universals. Dialectology meets typology: dialect grammar from a cross-linguistic perspective, B. Kortmann (ed), Berlin / New York: de Gruyter, 127-145 .

Fleischer, Jürg 2004: A typology of relative clauses in German dialects. Dialectology meets typology: dialect grammar from a cross-linguistic perspective, B. Kortmann (ed), Berlin / New York: de Gruyter. 211-243.

Kortmann, Bernd 2009: Die Rolle von (Nicht-Standard-)Varietäten in der europäischen (Areal-) Typologie. Eurolinguistik: Entwicklung und Perspektiven, U. Hinrichs, N. Reiter & S. Tornow (eds), Wiesbaden: Harrassowitz, 165-187.

Murelli, Adriano 2008: Relative constructions in European non-standard varieties. An integrative approach. Diss. University of Pavia.

Weiß, Helmut. 2002: Three types of negation: a case study in Bavarian. Syntactic Microvariation, S. Barbiers et al. (eds), Amsterdam: Meertens-Instituut, 305-332. www.meertens.knaw.nl/books/synmic/

66

Sidorkevich, Daria Institute for Linguistic Studies, Russian Academy of Sciences

Some Notes on Ingrian Finnish in Western Siberia

The presentation tells about a Finnish language variety spoken by a small group of Ingrians of mixed Finnish-Izhorian descent residing in Ryzhkovo village in Western Siberia. The research is based upon field data collected in Ryzhkovo settlement in 2008, 2009, 2010 and 2011. Present-day Ryzhkovo is a multiethnic settlement where Ingrians, Estonians, Latvians and Russians live together (Korb 2007, 2003). Although Ryzhkovo is not the only place in Siberia where Ingrians were historically present (Juntunen 1983, Zlobina 1972), today it is the last place where they still regard themselves as a group and still use their native language regularly.

The first Ingrians came to Western Siberia in 1804 (Juntunen 1982). Forty years later their first settlement, Ryzhkovo, became an official Lutheran colony. Through the last decades of the 19th century Lutheran colonies in Siberia were under strong influence of Finnish Lutheran clergy and, as a result, under considerable influence of Standard Finnish. The situation changed rapidly after the Russian Revolution signifying the end of cultural frame within which Ryzhkovo Ingrians were regarded as Finns. In the process of collectivization the record of their ethnic origin was changed to Estonian, although the population census of 1920 and 1926 had demonstrated that before 1930’s Ryzhkovo Ingrians regarded themselves as Finns. The shift in their official status took place most likely due to the ethnic construction processes accompanying the forming of the so-called national kolkhozes. For a short period of approximately 10 years Ryzhkovo Ingrians had Estonian as the language of schooling. Later on, especially after the World War II, contacts with Russian-speaking community became an inevitable part of Ryzhkovo social life. By the end of the 20th century a rather rapid language shift has taken place. Ethnic languages, such as Estonian, Latvian and Finnish are now spoken in Ryzhkovo only by elder generations.

The language of Ryzhkovo Ingrians is considered ‘mixed one’ by its speakers (Korb 2007). Indeed, multiple traces of various language contacts can be attested in it, the contact with Russian being the most influential. Surprisingly, the degree of Estonian influence here is rather modest. The main dialectological traits of Ryzhkovo Finnish still make it possible to correlate it with the Lower Luga Ingrian Finnish variety presently spoken in several villages in the western part of Ingermanland (Nirvi 1972).My presentation deals with some peculiar examples of contact-induced change that can be regarded as cases of syntactic transfer from Russian to Ryzhkovo Ingrian Finnish, in particular, with new domains of the so-called adessive-allative case.

Bibliography Juntunen A. 1983: Suomalaisten karkottaminen Siperiaan autonomian aikana ja karkotetut

Siperiassa. Helsinki. Juntunen A. 1982: Länsi-Siperian inkeriläiset siirtolat. Turun Historiallinen Arkiso 38: 350-367. Korb A. 2003: Virulased, a Multiethnic and Multicultural Communitiy in Ryzhkovo Village, West-

Siberia. Pro-Ethnologia 15: 29-47. Korb A. 2007: Rõžkovo virulased pärimuskultuuri kandjaina. Tartu. Nirvi R.E. 1972: Siperian inkeriläisten murteesta ja alkuperästä. Kotiseutu 2/3: 92-95. Zlobina V. 1972: Mitä alkujuurta Siperian suomalaiset ja korlakat ovat. Kotiseutu 2/3: 86-92.

67

Siiroinen, Mari University of Helsinki

Constructions of the Type ”X becomes Y” in Finnish Dialects

In standard Finnish there are three slightly different constructions expressing the meaning ‘X becomes Y’. The verb used in these constructions is tulla (also meaning ‘to come’). All the sentences 1–3 have the meaning ‘the porridge became thick’.

1. Puuro tuli sakeaa. NOM PART

2. Puurosta tuli sakeaa. ELA PART

3. Puuro tuli sakeaksi. NOM TRANSL

The difference between these constructions is in the cases of the nominal constituents. The first constituent, whose change is at issue, is either in the nominative (1 and 3) or elative case (2), and the predicative complement, which expresses the new state or property, in either in the partitive9

Of these three constructions, type 1 is special because it is very infrequent in modern standard Finnish and often regarded by native speakers as odd and ungrammatical. Still it has been one of the standard examples mentioned in the traditional reference grammars, and in old dialectal data this construction is actually quite frequent.

(1 and 2) or translative case (3). All these constructions are also found in Finnish dialects.

In my presentation I will discuss the dialectal distribution of type 1 and compare it with types 2 and 3. I will also try to pin down the subtle differences in meaning between the constructions. This will be done in the gist of the so-called Principle of Contrast: “If two grammatical structures occur in the same language to describe the ‘same’ experience, they will differ in their conceptualization of that experience in accordance with the difference in the two structures.” (Croft 2001: 111.)

In today’s standard language type 1 seems to have virtually vanished. I try to find out what might have caused its demise: has its conceptualization changed, and if so, for what reason?

Bibliography Croft, William 2001: Radical Construction Grammar. Syntactic Theory in Typological Perspective.

Oxford:Oxford University Press.

9 Or nominative, if the first argument is countable.

68

Tokizaki, Hisao Sapporo University

Correlation between Word Order and Phonology: Variation in Dialects and Languages

In this paper, I show that the correlation between word orders and phonology can be seen in dialects, related languages and languages of the world. It is argued that variation within and across languages is governed by a single parameter about the prominence location in a word.

Donegan and Stampe (1983, 2004) and Donegan (1996) investigate two types of Austroasiatic languages, which have initial or final accent in phrases and in words. They argue that these accent patterns (falling or rising rhythms in their terms) apply to both phonology and syntax, e.g., stress and word orders such as OV vs. VO, as shown in (1).

Hashimoto (1981) argues that languages in East Asia including Manchu, Beijing Chinese (Mandarin), Guangzhou Chinese (Yue dialect) and Thai have a gradational variation of tone complexity and coda inventory. I show that these languages and dialects also have a gradational variation of syntactic head-complement orders, as shown in (2). Assuming that verb, adposition and noun is the head and adjective is its complement, southern languages or dialects have more head-complement orders (V-O, P-NP and N-A) than northern ones.

In addition to these regional variations, I argue that the gradational variation of syntactic head-complement orders and word-stress location can be found in languages from Africa to East Asia. I show an analysis of the word-order data in Dryer (2005) and the stress-location data in Goedemans and van der Hulst (2005). Typical languages in each region show the shift of word-stress location from the right-edge to the left-edge of a word, which corresponds to the increase of complement-head orders (Stem-Affix, Genitive-N, O-V, NP-Postposition and Clause-Adverbial Subordinator), as shown in (3). I also show an analysis of the correlation between word-stress location and head-complement orders in languages of the world.

The result of this study shows that the same covariation between syntax and phonology is found in dialects and languages of the world. It is noteworthy that variation in syntactic word orders always goes with phonological variation such as stress, rhythm, the complexity of tone and syllable. This correlation between syntax and phonology opens up the possibility to find one parameter in syntax or phonology that determines the whole range of parameters in both components of grammar. This line of research gives an answer to the question of parameter settings in language acquisition. I argue that a phonological parameter such as location of stress or prominence in a word determines the head-complement orders in a dialect or a language. The change of prominence location triggers word-order changes. This research sheds light on the study of variation and typology in syntax and phonology and their interactions.

Bibliography Donegan, Patricia Jane. 1996. Rhythm and vocalic drift in Munda and Mon-Khmer. Linguistics of

the Tibeto-Burman Area 16:1, 1-43. Donegan, Patricia Jane and David Stampe. 1983. Rhythm and the holistic organization of linguistic

structure, Papers from the Parasession on the Interplay of Phonology, Morphology, and Syntax, J. Richardson et al. (eds.), Chicago: Chicago Linguistic Society. 337–353.

Donegan, Patricia Jane and David Stampe. 2004. Rhythm and the synthetic drift of Munda, The Yearbook of South Asian Languages and Linguistics 2004, Rajendra Singh (ed.), 3-36.

Dryer, Matthew S. 2005. Prefixing vs. suffixing in inflectional morphology/Order of object and verb/Order of adposition and noun phrase/Order of genitive and noun/Order of adverbial subordinator and clause, Haspelmath et al. (eds.), 110-113/338-341/346-349/350-353/382-385.

69

Goedemans, Rob and Harry van der Hulst. 2005. Fixed stress locations/Weight-sensitive stress, Haspelmath et al. (eds.), 62-65/66-69.

Hashimoto, Mantaro. 1981. Gendai Hakugengaku (Modern linguistics). Tokyo: Taishukan. Haspelmath, Martin, Matthew S. Dryer, David Gil and Bernard Comrie. 2005. The World Atlas of

Language Structures. Oxford: Oxford University Press. Tables (1) Languages in Austroasiatic: Munda a. Grammar: Synthetic Analytic

Mon-Khmer

b. Word Order: Head-last: OV, Postposition Head-first: VO, Preposition c. Phrases: Falling (initial) Rising (final) d. Words: Falling (trochaic) Rising (iambic/monosyllabic) e. Affixation: Pre/infixing, Suffixing Pre/infixing or Isolating f. Timing: Isosyllabic/Isomoric Isoaccentual g. Fusion: Agglutinative Fusional h. Syllables: (C)V(C) (C(ə) + (C)V(ː)/V)(C) i. Consonants: Stable/Assimilative Shifting/Dissimilative j. Tonality: Level (rare) Contour (common) k. Vowels: Harmonizing/Stable Reducing/Diphthongizing (2) Languages in East Asia from north to south: Languages tone coda a. Manchu 0 n/ŋ OV NP-Postposition Adjective-Noun

head-complement orders

b. Beijing 4 n/ŋ VO Preposition-NP Adjective-Noun c. Guangzhou 8(9) m/n/ŋ/p/t/k VO Preposition-NP Adj-Noun/Noun-Adj d. Thai 8 m/n/ŋ/p/t/k VO Preposition-NP Noun-Adjective (3) Languages from Africa to East Asia Language/Genus Bantu Romance English German Uralic Jap/Kor Word stress Penult R-edge R-orient R-orient Initial Initial

a. Stem-Affix +– + + + + + b. Word (C)-Word (H) – – + + + + c. Modifier-Noun – – +– +– + + d. Object-Verb – – – +– + + e. Object-Adposition – – – – + + f. Clause-Subordinator – – – – – +

70

Ussery, Cherlon Carleton College

Variation in Icelandic Agreement: Multiple Agree, Multiple Grammars

This paper provides an analysis of agreement with post-verbal third person Nominatives in Icelandic. While several analyses have attempted to account for the person restriction which bans first and second person Nominative objects, (e.g., Alexiadou 2003, Anagnostopoulou 2005, Sigurðsson and Holmberg 2008, Taraldsen 1995), less attention has been paid to the optionality of agreement with third person Nominative objects, shown in (1).

(1) Henni leiddist/leiddust þeir. her.Dat bored.3sg/3pl them. ‘She found

Nom.pl them

One notable analysis of constructions such as (1) is provided by Sigurðsson and Holmberg (2008) (henceforth S&H). S&H categorize Icelandic speakers into three dialect groups – those who prefer agreement, those who disallow agreement, and those who optionally allow agreement. Based on the results of a survey of sixty-one native Icelandic speakers that I conducted at the University of Iceland, I show that agreement with 3rd person post-verbal Nominatives depends not only on the speaker, but also on the type of construction. I argue that as the number of interveners between T and the Nominative increases, the likelihood of T probing the Nominative decreases.

boring.’

A summary of the survey results is shown in (2); the percentage reflects the rate of agreement for items of this type.

(2) a. Intransitive 100% agreement Það slógust/*slóst fjórir nemendur á ballinu

there fought.3pl/*3sg four student.Nom.pl at dance.the ‘Four students fought at the dance.’

b. Transitive 47% agreement Sumum gömlum mönnum líkar/líka pípuhattar. some.Dat.pl old.Dat.pl men.Dat.pl like.3sg/3pl top hats.Nom.pl ‘Some old men like top hats.’

c. Transitive Expletive 36% agreement Það líkar/líka sumum gömlum mönnum pípuhattar. expl like.3sg/3pl some.Dat.pl old.Dat.pl men.Dat.pl top hats.Nom.pl ‘Some old men like top hats.’

d. Biclausal Transitive 36% agreement Einum dómara sýndist/sýndust þessar athugasemdir vera óréttlátar. one.Dat.sg judge.Dat.sg understood.3sg/3pl these comments.Nom.pl be unfair

‘One judge understood these comments to be unfair.’ e. Biclausal Transitive Expletive 18% agreement

Það sýndist/sýndust einum dómara þessar athugasemdir vera óréttlátar. expl understood.3sg/3pl one.Dat.sg judge.Dat.sg these comments.Nom.pl be unfair ‘One judge understood these comments to be unfair.’

71

I argue that while Agree is obligatory, Multiple Agree (Hiraiwa 2001) is optional. Agreement always obtains in (2a) because T probes the expletive, which is in Spec,TP. Since the Nominative is the associate, the expletive values the Number feature on T. However, agreement is marginal in (2b) because more than one Agree relation is required for T to probe the object. Since the Dative is closer, T necessarily probes the Dative. Datives cannot value features on T, so T the number feature on T remains unvalued. Crucially, T only optionally continues probing until it finds the Nominative. Any derivation in which T fails to probe the Nominative results in the default form of the verb. The analysis in (3) delivers the agreeing form of the verb for the constructions in (2).

(3) a. [EXPL1 T Nom1] =(2a) b. [T Dat Nom] =(2b) c. [EXPL1 T Dat1 Nom] =(2c) d. [T+finite Dat [TP T-finite Nom] ] =(2d) e. [EXPL1 T+finite Dat1 [TP T-finite Nom] ] =(2e)

While in (3b), two Agree relations are necessary for T to probe the Nominative, in (3c) three Agree relations are required; T probes the expletive just as it does in (3a). In both (3d) and (3e), T must probe the Dative and the complement clause before reaching the Nominative. However, agreement is more degraded in (3e) because T must also probe the expletive.

The analysis above accounts for variation throughout a population, and I also provide an analysis of the grammars of individuals. Approximately 8% of participants prefer non-agreement in all constructions. These speakers never selected the agreeing form of the verb, suggesting that their grammars do not allow Multiple Agree. Approximately 12% allow agreement in Dative-verb-Nominative constructions only, e.g., (2/3b). Even though these speakers allow Multiple Agree, T can only probe two goals. Approximately 28% allow agreement in biclausal expletive constructions, e.g. (2/3e), suggesting that Multiple Agree can apply freely for these speakers. Crucially, only two speakers in this category consistently prefer agreement in this type of construction, underscoring the optionality of Multiple Agree. Even speakers who freely allow Multiple Agree do so only optionally.

Bibliography Alexiadou, Artemis. 2003. On nominative case features and split agreement. In E. Brandner and

H.Zinsmeister, eds., New Perspectives on Case Theory, CSLI Publications: Stanford, 23–52.

Anagnostopoulou, Elena. 2005. Strong and weak person restrictions: a feature checking analysis. In L. Heggie and F. Ordonez, eds., Clitics and Affixation, John Benjamins:

Amsterdam, 199–238. Hiraiwa, Ken. 2001. Multiple agree and the defective intervention constraint in Japanese. In Ora

Matsushansky et al., eds., The Proceedings of the MIT-Harvard Joint Conference (HUMIT 2000) Cambridge, MA: MITWPL. 40:67-80.

Sigurðsson, Halldór Ármann and Anders Holmberg. 2008. Icelandic dative intervention: person and number are separate probes. In Robertda D’Alessandro, ed. Agreement Restrictions. Berlin:Mouton de Gruyter, 251-280.

Taraldsen, Knut Tarala. 1995. On agreement and nominative objects in Icelandic, in H. Haider, S. Olsen, and S. Vikner, eds., Studies in Comparative Germanic Syntax, Kluwer:

Dordrecht, 307–327.

72

Verhoeven, Elisabeth Humboldt University Berlin

Word Order Variation with Non-Canonical Arguments: A Parallel Corpus Study

It has been shown for many languages that object constituents that bear the semantic role of experiencer exhibit a number of semanto-syntactic features that are subject-like or at least non-object-like (see e.g. Aikhenvald et al. (eds.) 2001, Bhaskararao & Subbarao (eds.) 2004, Landau 2010). Among these features is the property that experiencer-objects tend to occur in an earlier position than the stimulus subject even in ‘S before O’ languages. The present study investigates this observation in a number of genealogically and typologically mostly unrelated languages, i.e., German, Chinese, Turkish, and Modern Greek. All of these languages display ‘S before O’ in their unmarked order and allow for alternative orders including object preposing. This paper presents original evidence from a parallel corpus study that has been conducted in the mentioned languages and allow for a detailed comparison of the crosslinguistic variation.

The parallel corpus study was designed in the following way: 10 representative verbs of three verb groups (canonical transitive verbs, e.g., equivalences of order, criticize, etc.; accusative experiencer-object verbs, e.g., equivalences of amuse, concern, etc.; dative experiencer-object verbs, e.g. equivalences of appeal to, seem to, etc.) were identified in each language. A randomly selected sample of 400 occurences of the selected verbs with two noun phrases was extracted from a large corpus for each language and annotated for (a) word order and (b) animacy properties of the NPs. On the one hand, the data reveal universally explainable tendencies: In all languages, experiencer-object preposing is more frequent than canonical object preposing. This can be interpreted as an experiencer-first effect. At the same time, animacy plays a role in triggering object preposing in the following way: in all languages experiencer-object preposing is more frequent when the stimulus subject is inanimate than when it is animate (animate-first effect). On the other hand, typologically conditioned variation between the languages exists as regards the frequency of experiencerobject preposing: experiencer-object preposing is very frequent in German and Modern Greek (German: 24% in the prefield, 66% in the middle field; Greek: 27%) but less frequent in Chinese and Turkish (Chinese: 3%, Turkish: 7%). This result is in line with previous studies that analyze experiencer-object verbs in German and Greek as non-canonical transitive verbs while the respective verbs in Chinese and Turkish display the semanto-syntactic properties of canonical transitive verbs.

The crucial observation is that this cross-linguistic empirical study reveals a difference between languages at a global level (if we abstract away from the properties of the individual items (i.e., different verbs). However, within each language we also observe substantial variation between different items belonging to one verb class. For instance, in German the experiencer-object verb interessieren ‘interest’ has a robust difference in the frequency of object preposing from the experiencer-object verb enttäuschen ‘disappoint’. This talk will demonstrate that lexically-conditioned variation is observed in all languages and will discuss the question, how we can establish a typological difference at the global level and consider the lexically conditioned variation at the same time.

Bibliography Aikhenvald et al. (eds.) 2001, Non-canonical Marking of Subjects and Objects. Bhaskararao & Subbarao (eds.) 2004, Nonnominative subjects. Landau 2010, The locative syntax of experiencers.

73

Vierros, Marja University of Helsinki

Contact-Induced Variation in Hellenistic Egypt: Pronominal Suffixes and Other Constructions

Bilingual notaries whose L1 was Egyptian produced private contracts in Greek in Upper Egypt in the second and first centuries BCE. In this paper, I present a few examples of how the notaries struggled with Greek syntax producing structures that show contact-induced variation. I examine what is the typological distance between Egyptian and Greek in the relevant structures and how the results fit into common views on what type of features are likely to be transferred.

This paper is based on my doctoral dissertation where a structural qualitative analysis of ca. 150 Greek notarial contracts was performed. The notaries' Greek presents good administrative koine Greek, but, on the other hand, contact-induced variation was a fact in the idiolects of a few notaries. Variation mostly existed at the syntactic level, as can be expected in a situation of substratum/adstratum influence. For example, the notaries usually produced the grammatical cases correctly, but their use was more problematic. A strategy of phrase initial inflection was used in instances where a group of proper nouns should have been in agreement. Only the phrase initial element (proper name, pronoun, article or participle) was inflected in the required case and the rest of the NP was usually in the nominative. The predominance of the nominative case is also visible in certain recurrent phrases, resulting in, for example, an interesting hybrid of nominative+accusative with infinitive.

Egyptian as an Afro-Asiatic language had no morphological cases, but operated with word order, prepositions and pronominal suffixes. The structural pattern of demotic Egyptian relative conversion, for example, transferred to the Greek relative clause of some notaries. In the Egyptian structure, the subject of the relative clause was marked with a pronominal suffix attached to the relative marker. Following that logic, the notaries inflected the Greek relative pronoun in agreement with the subject of the relative clause instead of the subject of the main clause, the correct antecedent of the relative pronoun in Greek. The Egyptian word order, especially the placement and function of the pronominal suffixes, seems to have been behind some other confusions as well. Moreover, Egyptian marked possession with word order and a prepositional marker, which was weak and could be left out. This, together with the pronominal suffix indicating subject/object, had the consequence that the notaries had a mixed pattern of use for the Greek cases indicating the object (accusative) or possession (genitive), especially in certain phrases. However, sometimes we cannot take the L1 as the only explanation. Internal Greek variation and working methods of the notaries may have influenced simultaneously.

Bibliography Vierros, Marja 2011: Bilingual Notaries in Hellenistic Egypt. A Study of Language Use. Diss.

Helsinki.

74

Vilkuna, Maria Research Institute for the Languages of Finland

Between Dislocation and Expletive

In Finnish, an overt syntactic subject is not a necessary part of sentence structure, and the subject/topic position may be freely occupied by another constituent. Thus, Finnish is generally portrayed as not needing and not having expletives. Still, non-standard varieties display expletive-like elements (usually based on the pronoun se ‘it’) whose function seems to be to mark the topic position by their presence. On the basis of a survey of such patterns, the use of expletives can be said to provide (i) a preferred clause structure starting with a light but non-verbal element, (ii) a disambiguated information structure when the pre-topic position is filled, or (iii) a host for particles that favor the post-topic position. The starting point in the present paper is that expletives are best regarded as essential parts of specific clausal constructions (for a generative treatment, see Holmberg and Nikanne 2002).

Many Finnish expletive constructions are transparently similar to Germanic ones (e.g , infinitival extraposition like It is stupid to pay too much). While Germanic models cannot be ruled out, there are other potential sources for expletive constructions. One of these is the so-called right dislocation (RD) or doubling (Holmberg and Nikanne 2008), which is common in all spoken Finnish. In RD, an argument’s canonical place is filled by a pronoun and its lexical content is provided later, with a great degree of freedom of placement. A typical RD variant consists of a pronoun and a lexical part whose referent is given and marked as such by a pronominal determiner as in (1).

1) kyl se oli "kurjan "näkönen se "mökki

sure pron was miserable looking pron cottage

’The cottage sure looked miserable.’

As shown by Karhu (1994), however, RD is not restricted to given referents. Karhu also sketches a grammaticalization path from RD to an expletive construction. In RD, there is a coreference relation between a pronoun and the lexical NP. An expletive construction results when coreference is not relevant; many occurrences are structurally ambiguous in this respect. Based on a survey of existing dialectal data, the present paper discusses a specific point on the dislocation–expletive continuum, viz. sentences like (2) where, unlike in (1), the lexical part of a potential RD consists of a bare noun with no determiner.

2) kyllä se 'om "maailma 'kääntynyt "tykkenään 'toisellaiseks

sure PRON has world turned altoghether different

’The world has sure turned into a different kind of place [during my life]’

It is cases like this that often give rise to an expletive reading, particularly if the NP is an indefinite element with low referential continuity like in (2). However, it seems that even constructions like (2) with a clearly definite lexical part have a conventional function in certain dialects. These phenomena seem to be especially prominent in Eastern and Northern dialects, which display a great variety of expletive structures.

75

Bibliography Holmberg, Anders & Nikanne, Urpo 2002: Expletives, Subjects and Topics in Finnish. In Peter

Svenonius (ed.), Subjects, expletives, and the EPP. Oxford: Oxford University Press, pp. 71–105.

Holmberg, Anders & Nikanne, Urpo 2008: Subject doubling in Finnish: the role of deficient pronouns. In: Barbiers, S., Koeneman, O. & Lekakou, M. (ed.), Microvariations in Syntactic Doubling. Bingley: Emerald, pp. 325–349.

Karhu, Anna 1994: Suomen itämurteiden oikealle lohkeavat rakenteet ja niiden lähirakenteet: Dislokaatioita, korviketopiikkeja, diskurssinmerkitsimiä ja pragmaattisia partikkeleita. Unpublished Phil.lic. thesis, University of Joensuu, Department of Finnish language, Literature and Culture.

76

Weber, Tobias University of Zürich

Variation in Agent Marking Conditioned by Clausal Properties

In this paper I explore the variation in the case marking of the agent argument of a transitive construction conditioned by clausal properties across languages. The following properties can be responsible for the variation:

1) Tense, aspect and mood 2) Polarity 3) Morphological form of the predicate (e.g. verb stem alternations) 4) Clause type (main vs. subordinate clause) 5) Word order 6) Scenario (i.e. the variation in agent marking is conditioned by properties of another

argument) Furthermore, what adds to the degree of complexity is that in some languages several of these

properties are responsible for the variation in agent marking, and additionally, optional case marking of the agent argument (i.e. presence vs. absence of a case marker in one specific lexico-grammatical environment) can also come into play: in Lhasa Tibetan, for instance, we find a split according to aspect: in the perfective aspect, the ergative marking is obligatory on all agents, but in the imperfective aspect, the ergative marking occurs only on agents with certain discourse-pragmatic and semantic properties (cf. Zeisler 2004: 258).

The present paper aims at focusing on the following questions: 1) What clausal conditions trigger a variation in agent marking in a specific language? 2) What case markers are involved in the agent marking? 3) What are the functions of these case markers? 4) What interaction patterns can be found? Some methodological prerequisites should also be mentioned here: 1) Case marking is defined here in fairly broad terms, including any element of

dependent marking on the clause level irrespective of their morphological nature (affixes, clitics and separate words), since the properties (and definitions) of words vary widely across languages (see Dixon and Aikhenvald 2002).

2) Arguments (and valence) are defined in purely semantic terms since the application of syntactic criteria of argumenthood poses problems for the crosslinguistic investigation of arguments (cf. Witzlack-Makarevich 2010: 41-47).

Examples are drawn from a worldwide sample of languages that have some sort of variation in agent marking conditioned by clausal properties.

Bibliography Dixon, R. M. W. and Alexandra Y. Aikhenvald. 2002. Word: a typological framework. In R. M. W.

Dixon and Alexandra Y. Aikhenvald (eds.), Word: a Cross-Linguistic Typology. Cambridge: Cambridge University Press, 1-41.

McGregor, William B. 2010. Optional ergative case marking systems in a typological-semiotic perspective. Lingua 120: 1616-1636.

Witzlack-Makarevich, Alena. 2010. Typological Variation in Grammatical Relations. Phd thesis, Universität Leipzig.

Zeisler, Bettina. 2004. Relative Tense and Aspectual Values in Tibetan Languages: A Comparative Study. Berlin: Mouton de Gruyter.