Abstracts

Henrik Jørgensen (Aarhus)
Pronominal Case Re-Opened -
Some notes on pronominal inflection and object shift in Mainland Scandinavian
      The mutual relation between 'object shift' (with a conventional, though somewhat misleading term) and the case inflection on Mainland Scandinavian pronouns is an interesting issue. The situation in Modern Danish, where the pronominal inflection is suspended when the pronouns retain the stress and remain in situ (which they do when they contribute deictically to the identification of a referent in the discourse), might be interpreted in such a way that the inflection is in some way connected with the off-site positions achieved through 'object shift'). However if this should be true, we would expect Mainland Scandinavian dialects with sparse pronominal inflection to lack object shift as well, and Mainland Scandinavian dialects without Object Shift to have sparse inflections. Both of these expectations seem not to hold, as far as I informed. Thus, we still need an explanation why only absolutely anaphoric pronouns in Modern Danish inflect and the deictic ones do not.

Eva Engels and Sten Vikner (Aarhus)
Object Shift, Scrambling and Optimality Theory
      Although they both place an object to the left of a sentential adverbial, Scandinavian Object Shift (OS) and continental West Germanic Scrambling (SCR) are normally treated as two different phenomena since they do not take place under exactly the same circumstances. The main difference between OS and SCR is that the former presupposes movement of the main verb whereas the latter does not.
      The optimality-theoretic account to be suggested will not distinguish between (Scandinavian) OS and (continental West Germanic) SCR as such. The differences will simply follow from more general constraints on object movement (in particular a constraint on Order Preservation), given the linear differences between the languages (VO in Scandinavian vs. OV in the continental West Germanic languages). A distinction will be made concerning the complexity of the moved element, weak pronouns vs. complex phrases. Cross-linguistic contrasts in the availability of the movement operations and in their restrictions will be derived from differences in the language-specific rankings of constraints.

Eva Skafte Jensen (Roskilde)
Case in Early Middle Danish - a Double Content System
      The somewhat surprising distribution of casemarkings in certain manuscripts of Early Middle Danish (appr. 1100-1350) has been regarded as scribal errors. A closer look, however, reveals a pattern in the casemarkings that suggests that formally marked nominative and accusative in part appear in accordance with a system alternative to the traditional case system. The two cases accusative and nominative thus seem to serve two functions, one has to do with well known case functions of syntactic and semantic nature. The other function concerns pragmatic distinctions such as foreground vs. background information (/focus vs. non-focus).
      Apparently the nominative has developed a secondary function as a marker of foreground information in nouns, adjectives and to some extent pronouns, at first only in words working as subjects and subject complements, later also by other syntactic functions.
      In my paper I'll show how this alternative function manifests itself in manuscripts of Early Middle Danish, primarily in excerpts from the Scanic Law (ms. Stockholm B 69, appr. 1350).

Gunlög Josefsson (Lund)
Object Shift and Grammaticality Judgments: How Grammatical is Grammatical?
      In the literature the following four claims are often assumed to hold for Object shift (OS) in Swedish:

      (i) OS is obligatory for weak pronouns
      (ii) The reversed order of objects – direct object preceding indirect object – is grammatical, if both objects are shifted
      (iii) Long object shift (LOS) (weak object pronouns preceding subjects) is grammatical in Swedish
      (iv) Only weak pronouns can shift

In my talk I will discuss these four claims, showing that the data situation is more complex than what is usually assumed. A consequence is that the notion of grammaticality in this area is not straightforward or simple. Concepts such as relative weight in the clause, syllable structure of the pronoun, and lexical properties have to be taken into account when describing and analysing the phenomenon. There also seems to be a large amount of idiosyncratic, and maybe also dialectal, variation, among speakers.

Jonas Blom (SDU, Odense)
Blocking
      I will discuss the relation between empty and filled slots in Lars Heltoft’s and Ole Togeby’s revised sentence schemes. Normally, the so-called empty slots are all treated in the same way despite the fact that there are important differences at work. In order to explain these different types of emptiness, I will make a threefold distinction of slots as either filled, empty, or blocked. This will indicate that something is present in the shadows of blocked slots, traces of something.  I will claim, though, that the traces are not to be found within the slots, on the contrary the blocked slots are the traces themselves topologically governed by the predicate frame, the illocutionary frame and various sorts of constituent ‘movement’, primarily object shifts. 

Hans-Martin Gärtner (ZAS, Berlin)
On Object-Shift in Icelandic and Partial Iconicity
      In this talk, based on a paper published in Lingua (2004), I am going to discuss what role, if any, OT-systems accord to the functionalist notion of "iconicity." As a first step, this will involve a somewhat idiosyncratic reformulation of Vikner's (1997; 2001) OT-analysis of Icelandic object shift in terms of a constraint called "UNAMBIGUOUS ENCODING" (UE). (The obvious shortcomings of) UE will then serve as a backdrop against which the following issues are explored.
      (i) Icelandic object-shift partially exhibits a disambiguation pattern known equally as "Horn's division of pragmatic labor" or "iconicity," which consists in (un-)marked forms associating with (un-)marked meanings.
      (ii) OT-syntactic approaches to iconicity rely on (more or less "shallow") input/output constraints. The ease with which these can be recast so as to license "anti-iconicity" raises the question as to how arbitrary the relation between such contraints and iconicity patterns may be. Put differently, we have to ask whether such constraints can be grounded in language use.
      (iii) The symmetric bidirectional OT-framework proposed by Blutner (2000) induces iconicity through its optimization procedure and thus provides a deeper entrenchment of iconicity, bidirectionality being at the heart of OT-learning theory and OT-pragmatics (cf. Blutner & Zeevat, to appear; Smolensky, 1996). However, this brand of bidirectional OT - as opposed to OT syntax - lacks a satisfactory way for treating partial iconicity. Finally,
      (iv) a treatment of Icelandic object-shift in terms of Aissen (2000), built on "harmonic alignment" and "local conjunction," raises the the same questions as other OT syntactic approaches. This is in spite of the aim of modeling iconicity, which underlies this approach. However, as shown by Zeevat & Jäger (2002) and Jäger (to appear), there exist ways of grounding Aissen's theory in terms of principles of language use and OT-learning theory.

Tavs Bjerre (Aarhus)
The Position of Nonfinite Verbs in a Topological Sentence Model for Danish
      There is no abstract.

Sten Vikner and Eva Engels (Aarhus)
Object Shift, Remnant VP-topicalisation and Optimality Theory
      Holmberg (1997, 1999) assumes that Holmberg's generalisation is derivational, prohibiting Object Shift across an intervening non-adverbial element at any point in the derivation. Consequently, placement of the non-finite verb in clause-initial position as in (1), from Holmberg (1997:205), could not possibly be derived by Object Shift prior to remnant VP-topicalisation, but would have to involve V°-topicalisation with subsequent Object Shift.

(1)

Sw.

Kysst  kissed

har  have

jag  I

henne  her

inte  not

___

___

(  (

bara  only

hållit  held

henne  her

i  by

handen  hand-the

). )

Counterexamples to Holmberg's hypothesis are given in Fox & Pesetsky (2005) which show that remnant VP-topicalisations may leave behind an object in Object Shift position in Scandinavian as long as the VP-internal order relations are maintained, as illustrated by the asymmetry between stranding of a direct object and stranding of an indirect object in double object constructions (examples from Fox & Pesetsky 2005:25):

(2)

Sw.

a.

?

[VP

Gett  given

henne  her

___

]

har  have

jag  I

den  it

inte.  not

b.

*

[VP

Gett  given

___

den  it

]

har  have

jag  I

henne  her

inte.  not

Extending the empirical basis concerning remnant VP-topicalisations, we argue that Holmberg's generalisation and the restrictions on object stranding result from one and the same, more general condition on order preservation.
      References:
      Fox, Danny & Pesetsky, David. 2005. Cyclic Linearization of Syntactic Structure. Theoretical Linguistics 31, 1-45.
      Holmberg, Anders. 1999. Remarks on Holmberg's Generalization. Studia Linguistica 53, 1-39.

Johannes Kizach (Aarhus)
Complexity and Word Order in Russian
      John Hawkins claims that certain orders are more efficient (for the processor/parser) than others. Efficiency is calculated using his complexity metric and his concept of construction domains. When efficiencies of alternative orders are calculated, the result is a ranking which is predicted to mirror the relative frequencies of these orders in performance.
      In Russian transitive main clauses, all six possible orderings of S, V and O are possible, and I have attempted to test whether Hawkins´ predictions are borne out by analyzing 600 examples from a Russian corpus.

Maria Melchiors (Tübingen)
The what for-Construction and the Analysis of Dative in German
      This talk deals with the what for-construction, which is illustrated in (1):

(1)	Ge.			Was  what	hast  have	du  you	damals  at-that-time	für  for	Bücher  books	gelesen  read	?
				What kind of books did you read at that time?

One can find very different judgements of German dative splits in the literature. In order to throw light on the limitations of this construction I have applied the magnitude estimation technique to elicit judgements of different examples. I will report the results of three different experiments. The following questions were investigated in the experiments:
      • Are dative splits grammatical?
      • Does animacy matter?
      • Do all splits of dative verbs with two arguments behave in the same way?
Traditionally, what for constructions are analysed as extractions (Corver 1991 and Pafel 1996). It will be shown that such analyses have certain drawbacks. Finally, I discuss what these experiments can tell us about the analysis of dative objects in German.
      References:
      Corver, Norbert (1991): The Internal Syntax and Movement Behavior of the Dutch 'wat voor'-Construction. Linguistische Berichte 133, 190-228.
      Pafel, Jürgen (1996): Die syntaktische und semantische Struktur von was für-Phrasen. Linguistische Berichte 161, 37-67.

Andrej Malchukov (MPI, Leipzig)
Asymmetries in Differential Case Marking and Case Marking Strategies
      The present talk addresses asymmetries between differential object marking (DOM) and differential subject marking (DSM) and provides an explanation for the attested patterns in terms of two competing case-marking strategies. It has been observed that DSM is less consistent cross-linguistically as compared to DOM, which invariably singles out a more prominent (e.g. animate) object for overt marking (cf. Bossong 1985; Aissen 2003). This asymmetry is attributed to the fact that in case of DOM the two basic motivations for case marking, indexing roles and disambiguating between arguments, conspire in producing the same pattern, while in the domain of DSM these motivations are in conflict. It is argued that complexity of animacy effects in differential case marking is due to the fact that some of these patterns result from the differentiating function of case marking, while some other result from the indexing function. Definiteness effects in differential case marking parallel animacy effects and may be provided a similar explanation in terms of the two basic case-marking strategies. Finally examination of DOM patterns in ditransitive construction lends further support to the advocated approach.

Øystein Nilsen (Utrecht/Tromsø)
Object shift by Remnant Movement
      I argue that the standard analysis of Object Shift (OS) in terms of independent leftward movement of (pronominal) arguments must be rejected because of what has come to be known as `Bobaljik Paradoxes' (Bobaljik 1999, Nilsen 2003). I argue that a proper solution to these can be found if we view the left-peripheral position of the finite verb in V2 sentences and the leftward shifting of pronominal objects as results of one (remnant) movement operation. Several, otherwise astonishing properties of OS will be shown allow for a natural characterization under such an analysis, and, perhaps most importantly, Holmberg's Generalization can be seen to follow. Time permitting, I will address the currently popular treatment of Holmberg's Generalization in terms of 'Order Preservation' (Williams 2003; Fox and Pesetsky 2004), which essentially amount to attempts to derive the effects of Holmberg's generalization by taking it as a primitive. I will show that such treatments, at least as currently stated, cannot derive the properties of Object Shift, and that the approach in terms of Remnant Movement fares better.

Participants

In addition to the twelve speakers listed above, the following linguists participated in the workshop: