Registration

The course was open to Ph.D.-students and also to other interested persons. There was no participation fee for the course, but the food and lodgings at Sandbjerg cost DKK 3 000 (EUR 400) per person for the whole period (3 nights and 10 meals in total). We did not have to charge all participants the full amount, however, thanks to financial support from the Aarhus University Research Foundation.

Abstracts (triple lectures)

Christer Platzack
Argument structure
      Cross-linguistically, the number of DP-arguments (i.e., arguments not expressed by the help of adpositions) is restricted: in the Germanic and Romance languages, and presumably more generally, the number of DP arguments varies between zero (det dansades till sent på kvällen = it was danced until late at night) and three (Johan gav Eva en bok = John gave Eva a book). In my lectures I will discuss a possible way to account for this fact within present-day generative grammar (the minimalist framework). With this goal in mind, my lessons will survey the theoretical took-box, presenting and discussing concepts like UTAH, the vP/VP distinction, valued and unvalued features, the operation Agree and EPP. Descriptively, we will take a closer look at impersonal passives, unergative and unaccusative verbs, transitive verbs of different kinds, and ditransitive verbs, investigating how much of the argument structure that is a consequence of the computational system mediating between meaning and form (the faculty of language in narrow sense).

Peter Sells
Hierarchical and linear constraints on structure - constituent order in the Mainland Scandinavian languages
      Based on my work on Swedish (Sells 2001), in these lectures I will present a view of constituent order which makes use of two types of constraints: those that refer to hierarchical structure (e.g., every XP has a head X) and those that refer directly to linear precedence, stated in terms of alignment (e.g., TOPIC-Left). I will discuss how to account for major word order patterns, and variations on them, presenting a view based on Lexical-Functional Grammar (LFG), which allows direct reference to grammatical functions (e.g., TOPIC, OBJ).
      In the first lecture I will present an overview of the facts relevant for constituent ordering in Mainland Scandinavian, from the perspective of a non-transformational approach such as LFG. In the second lecture I will present an analysis in terms of ranked constraints, Optimality-Theory-style, focussing on the particular analysis of the data from the first lecture. In the third lecture, I will consider how to extend the account to cover a wider range of facts within Scandinavian.

Ole Togeby
The meaning of syntactic position or movement in modern Danish syntax
      The terms of ‘syntactic position’ and ‘syntactic movement’ are metaphors of something which must be described as mind processes of prediction and recall during the language user’s interpretation of a text. These processes are triggered by lexical items with meaning, and so word order in it self carries meaning too. In my talks I’ll discuss the meaning of the positions in Diderichsen's sentence scheme, and of the different movements in modern generative syntax. I suggest that the requirement to these explanations is that they are not incompatible with what we know about how the brain works - which in fact both ‘position’ and ‘movement’ are.
      Examples to be discussed are objects in front position in Danish, e.g.: Ham kender jeg (Him know I), so-called light objects before the negation, e.g.: Jeg kender ham ikke (I know him not), and normal objects in the final position: Jeg kender ikke ham der (I know not him there).

Eva Engels & Sten Vikner:
Object shift and scrambling
      We will give an overview over object movement devices in the Scandinavian languages (object shift) and the continental West Germanic languages (scrambling), covering the following:
­   •    Although both place an object to the left of a sentential adverbial, OS and SCR are normally treated as two different phenomena since they do not take place under exactly the same circumstances.
­   •    We want to investigate their properties and show that they are quite similar in what moves and which position movement can target provided one considers the entire range of OS and SCR languages.
   •    The main difference between OS and SCR is that the former presupposes movement of the main verb whereas the latter does not.
­   •    This property might be related to the contrast in basic verb placement, VO in Scandinavian vs. OV in the continental West Germanic languages.
   •    Moreover, some specific characteristics of OS and SCR will be discussed (contrast between OS of arguments and OS of adverbials, differences in the information-structural contexts that facilitate OS and SCR, V- vs. VP-topicalisation).

Abstracts (double lectures)

Tavs Bjerre:
Formalizing Diderichsen's sentence model
      These two lectures will be devoted to a formalization of a revised version of Diderichsen's sentence model within HPSG. Because of its nonderivational nature with parallel representation of the various levels of linguistic information (syntactic, semantic, contextual etc.) HPSG seems well suited for the task. The version of HPSG to be employed, Linearization-based HPSG, holds the view also shared by Danish functionalists that linear order and constituency are two different levels of description, which to a considerable extent are independent of each other.

Henrik Jørgensen:
From Diderichsen's sentence model to a functional approach
      In these two lectures, I want to deal with both the formal and the functional aspects of Paul Diderichsen’s ‘sentence model’. I am interested in its double nature, being at the same time a vehicle of pragmatic competence and a formalizable system (cf. the double lecture by Tavs Bjerre). The focus of my investigation is the object positions, but in this context I want to emphasize the overall structure of the theoretical framework.
      The two lectures will be devoted to a discussion of the object as a part of the functional sentence perspective. I shall raise the discussion to which degree the object (in the traditional understanding: the second transitive valency phrase of the sentence) has any inherent connection with the focus of the sentence and continue with the problems raised by the actual positions occupied by Danish objects in their different functional aspects.

Abstracts (single lectures)

Werner Abraham
Reinforcing the event typological concept of verb valence - and terminally demolishing it for prepositional ­objects
      According to Talmy (1985), the typologically parameterized semantic decomposition of predicates of motion leads to multiple semantic categories (such as path, place of departure, goal, R-direction) which turn out to be overt or covert across languages. However, they need to be accounted for in a strictly structural representation whose PF-deriva­tion is subject to all calibrated and highly constrained syntactic mechanisms valid in Min­im­alism. This will be executed and illustrated by means of a number of German prepositions accompanying verbs of motion. The wealth of subatomic se­man­tic components, their strictly and universally constrained configuration, and the insight that pre­positional objects need be merged outside of VP (and act there as probes) lead to the conclusion that the traditional claim that the valence of verbs determines which prepositions are governed is wrong and that the valence relation should be inverted: i.e., that verbs need to be assigned to specific semantic spaces defined by the semantics of prepositions.

Eckhard Bick
Corpus tools for a quantitative study of object positions in written language
      In my talk I will present a corpus-based approach to the study of object positions. In order to empirically exemplify and quantify a complex syntactic topic like object positions, it is desirable to be able to build concordances from grammatically searchable corpora. In the context of the SDU-based VISL project, Constraint Grammar parsers are used to add form and function tags to word tokens in running text, which can then be searched for patterns such as ACC-OBJECT VFIN SUBJECT, ACC-OBJECT PRP-OBJECT, SUBJECT VFIN ACC-OBJECT VINF etc. In the internet-based CorpusEye search interface, function patterns like the above are accessible in a menu based way, allowing multi-token constituents and quantifiers, as well as sorting and quantification. Currently, Corpuseye offers grammatically annotated corpora for 8 Germanic and Romance languages (about 1 billion words), mostly from the written language domain. Finally, the lecture will discuss the concept of manually revised syntactic treebanks, built to allow a refinement of structural syntactic description in either constituent or dependency grammar format. Samples from the Danish Arboretum treebank will be used to demonstrate object topology in Danish.

Per Bærentzen
Word order patterns in German
      In my lecture I will briefly present the general model for German sentences. In more detail I will show that, in contrast to English and Danish, German has two different principles for the placing of embedded non-finite verb phrases, one of which leads to the ”famous” German word order represented in (ich will) ihn sie Geige spielen hören lassen (literally: 'I will him her violin play hear let', meaning 'I will let him hear her play the violin') with the group of objects assembled before the group of verbs. Finally I will try to formulate word-ordering rules in a way that makes them equally applicable to German and to other languages. Such rules enable us to make contrastive analyses which bring out the similarities and differences between German and other languages.

Ken Ramshøj Christensen
Object positions in the brain
      Studies of language deficits as well as neuroimaging studies have shown that Broca’s area in the brain plays a crucial role in the processing of syntactic structure, and that it seems to be particularly involved in the processing of syntactic movement. I shall present data from a neuroimaging study on object movement in Danish and compare it with the results of a number of other neuroimaging studies on syntax and the brain. I shall argue that a number of apparent counterexamples in the literature to the correlation between activation in Broca’s area and word order variation can be accounted for by taking linguistic theory seriously, and that the brain activations reported in the neurolinguistic literature reflect the interfacing between syntactic structure building and other cognitive systems, including information structure and pragmatics.

Ralf Vogel
Prosodic constraints on NP/PP placement
      I will study restrictions on the positioning of noun phrases and prepositional phrases that come from prosodic factors. Starting with a discussion of Scandinavian object shift, I will show that the same underlying prosodic regularities are responsible for placement restrictions for weak pronouns in Scandinavian as well as in English, German and Dutch. I will present an optimality theoretic model of prosody-syntax interaction offering a uniform account of weak pronoun shift. We will then see how this model can be extended to deal with further phenomena, in particular, exceptions to the wh-island constraint, observed in English and a couple of other languages, and a class of prosodic restrictions on extraposition, observed in German.