Linguistik
Refine
Year of publication
Document Type
- Article (1206)
- Part of a Book (784)
- Working Paper (254)
- Review (178)
- Conference Proceeding (166)
- Preprint (122)
- Book (108)
- Part of Periodical (60)
- Report (58)
- Doctoral Thesis (23)
Language
- English (1397)
- German (1057)
- Croatian (298)
- Portuguese (118)
- Turkish (39)
- French (21)
- Multiple languages (21)
- mis (16)
- Spanish (7)
- Polish (4)
Is part of the Bibliography
- no (2987) (remove)
Keywords
- Deutsch (435)
- Syntax (151)
- Linguistik (124)
- Englisch (123)
- Semantik (112)
- Spracherwerb (96)
- Phonologie (85)
- Rezension (77)
- Kroatisch (68)
- Fremdsprachenlernen (67)
Institute
- Extern (438)
- Institut für Deutsche Sprache (IDS) Mannheim (113)
- Neuere Philologien (43)
- Sprachwissenschaften (43)
- Universitätsbibliothek (4)
- Sprach- und Kulturwissenschaften (3)
- Gesellschaftswissenschaften (2)
- Medizin (2)
- Präsidium (2)
- SFB 268 (2)
We adopt Markert and Nissim (2005)’s approach of using the World Wide Web to resolve cases of coreferent bridging for German and discuss the strength and weaknesses of this approach. As the general approach of using surface patterns to get information on ontological relations between lexical items has only been tried on English, it is also interesting to see whether the approach works for German as well as it does for English and what differences between these languages need to be accounted for. We also present a novel approach for combining several patterns that yields an ensemble that outperforms the best-performing single patterns in terms of both precision and recall.
Tagging kausaler Relationen
(2005)
In dieser Diplomarbeit geht es um kausale Beziehungen zwischen Ereignissen und Erklärungsbeziehungen zwischen Ereignissen, bei denen kausale Relationen eine wichtige Rolle spielen. Nachdem zeitliche Relationen einerseits ihrer einfacheren Formalisierbarkeit und andererseits ihrer gut sichtbaren Rolle in der Grammatik (Tempus und Aspekt, zeitliche Konjunktionen) wegen in jüngerer Zeit stärker im Mittelpunkt des Interesses standen, soll hier argumentiert werden, dass kausale Beziehungen und die Erklärungen, die sie ermöglichen, eine wichtigere Rolle im Kohärenzgefüge des Textes spielen. Im Gegensatz zu “tiefen” Verfahren, die auf einer detaillierten semantischen Repr¨asentation des Textes aufsetzen und infolgedessen für unrestringierten Text m. E. nicht geeignet sind, wird hier untersucht, wie man dieses Ziel erreichen kann, ohne sich auf eine aufwändig konstruierte Wissensbasis verlassen zu müssen.
When a statistical parser is trained on one treebank, one usually tests it on another portion of the same treebank, partly due to the fact that a comparable annotation format is needed for testing. But the user of a parser may not be interested in parsing sentences from the same newspaper all over, or even wants syntactic annotations for a slightly different text type. Gildea (2001) for instance found that a parser trained on the WSJ portion of the Penn Treebank performs less well on the Brown corpus (the subset that is available in the PTB bracketing format) than a parser that has been trained only on the Brown corpus, although the latter one has only half as many sentences as the former. Additionally, a parser trained on both the WSJ and Brown corpora performs less well on the Brown corpus than on the WSJ one. This leads us to the following questions that we would like to address in this paper: - Is there a difference in usefulness of techniques that are used to improve parser performance between the same-corpus and the different-corpus case? - Are different types of parsers (rule-based and statistical) equally sensitive to corpus variation? To achieve this, we compared the quality of the parses of a hand-crafted constraint-based parser and a statistical PCFG-based parser that was trained on a treebank of German newspaper text.
In the past, a divide could be seen between ’deep’ parsers on the one hand, which construct a semantic representation out of their input, but usually have significant coverage problems, and more robust parsers on the other hand, which are usually based on a (statistical) model derived from a treebank and have larger coverage, but leave the problem of semantic interpretation to the user. More recently, approaches have emerged that combine the robustness of datadriven (statistical) models with more detailed linguistic interpretation such that the output could be used for deeper semantic analysis. Cahill et al. (2002) use a PCFG-based parsing model in combination with a set of principles and heuristics to derive functional (f-)structures of Lexical-Functional Grammar (LFG). They show that the derived functional structures have a better quality than those generated by a parser based on a state-of-the-art hand-crafted LFG grammar. Advocates of Dependency Grammar usually point out that dependencies already are a semantically meaningful representation (cf. Menzel, 2003). However, parsers based on dependency grammar normally create underspecified representations with respect to certain phenomena such as coordination, apposition and control structures. In these areas they are too "shallow" to be directly used for semantic interpretation. In this paper, we adopt a similar approach to Cahill et al. (2002) using a dependency-based analysis to derive functional structure, and demonstrate the feasibility of this approach using German data. A major focus of our discussion is on the treatment of coordination and other potentially underspecified structures of the dependency data input. F-structure is one of the two core levels of syntactic representation in LFG (Bresnan, 2001). Independently of surface order, it encodes abstract syntactic functions that constitute predicate argument structure and other dependency relations such as subject, predicate, adjunct, but also further semantic information such as the semantic type of an adjunct (e.g. directional). Normally f-structure is captured as a recursive attribute value matrix, which is isomorphic to a directed graph representation. Figure 5 depicts an example target f-structure. As mentioned earlier, these deeper-level dependency relations can be used to construct logical forms as in the approaches of van Genabith and Crouch (1996), who construct underspecified discourse representations (UDRSs), and Spreyer and Frank (2005), who have robust minimal recursion semantics (RMRS) as their target representation. We therefore think that f-structures are a suitable target representation for automatic syntactic analysis in a larger pipeline of mapping text to interpretation. In this paper, we report on the conversion from dependency structures to fstructure. Firstly, we evaluate the f-structure conversion in isolation, starting from hand-corrected dependencies based on the TüBa-D/Z treebank and Versley (2005)´s conversion. Secondly, we start from tokenized text to evaluate the combined process of automatic parsing (using Foth and Menzel (2006)´s parser) and f-structure conversion. As a test set, we randomly selected 100 sentences from TüBa-D/Z which we annotated using a scheme very close to that of the TiGer Dependency Bank (Forst et al., 2004). In the next section, we sketch dependency analysis, the underlying theory of our input representations, and introduce four different representations of coordination. We also describe Weighted Constraint Dependency Grammar (WCDG), the dependency parsing formalism that we use in our experiments. Section 3 characterises the conversion of dependencies to f-structures. Our evaluation is presented in section 4, and finally, section 5 summarises our results and gives an overview of problems remaining to be solved.
This paper is concerned with the tagging of spatial expressions in German newspaper articles, assigning a meaning to the expression and classifying the usages of the spatial expression and linking the derived referent to an event description. In our system, we implemented the activation of concepts in a very simple fashion, a concept is activated once (with a cost depending on the item that activated it) and is left activated thereafter. As an example, a city also activates the nodes for the region and the country it is part of, so that cities from one country are chosen over cities from different countries. A test corpus of 12 German newspaper articles was tested regarding several disambiguation strategies. Disambiguation was carried out via a beam search to find an approximately cost-optimal solution for the conflict set of potential grounding candidates for the tagged spatial expression. Test showed that the disambiguation strategies improved accuracy significantly.
Using a qualitative analysis of disagreements from a referentially annotated newspaper corpus, we show that, in coreference annotation, vague referents are prone to greater disagreement. We show how potentially problematic cases can be dealt with in a way that is practical even for larger-scale annotation, considering a real-world example from newspaper text.
We investigate methods to improve the recall in coreference resolution by also trying to resolve those definite descriptions where no earlier mention of the referent shares the same lexical head (coreferent bridging). The problem, which is notably harder than identifying coreference relations among mentions which have the same lexical head, has been tackled with several rather different approaches, and we attempt to provide a meaningful classification along with a quantitative comparison. Based on the different merits of the methods, we discuss possibilities to improve them and show how they can be effectively combined.
In this paper, we investigate the usefulness of a wide range of features for their usefulness in the resolution of nominal coreference, both as hard constraints (i.e. completely removing elements from the list of possible candidates) as well as soft constraints (where a cumulation of violations of soft constraints will make it less likely that a candidate is chosen as the antecedent). We present a state of the art system based on such constraints and weights estimated with a maximum entropy model, using lexical information to resolve cases of coreferent bridging.
Im Fokus dieser Magisterarbeit stehen Präpositionalphrasen (PP), deren Komplement eine unikale Komponente ist. Es handelt sich bei diesen Komplementen um Nomen, die außerhalb einer PP nicht vorkommen bzw. in anderen Umgebungen nicht die selbe Bedeutung haben. Um dieses Phänomen zu beschreiben wird eine Analyse innerhalb der Head-Driven Phrase Structure Grammar (HPSG) entwickelt. Grundkenntnisse über Struktur und Begrifflichkeiten der HPSG werden in dieser Arbeit vorausgesetzt, als Referenz siehe [PS94]. Die Gliederung gestaltet sich wie folgt: Zunächst werden die zu untersuchenden Daten im Detail dargestellt. Anschließend werden verschiedene Analysemöglichkeiten innerhalb der Theorie der HPSG in Betracht gezogen, nämlich Selektion, Konstruktion und Kollokation. Dabei muss festgestellt werden, dass die existierenden Ansätze den Daten nicht oder nur unbefriedigend gerecht werden können. Der Ansatz, der letztendlich verfolgt wird, besteht darin, den bereits existierenden Selektionsmechanismus über SPEC zu generalisieren. Dieses Vorgehen erlaubt dann der unikalen NP, die Präposition, mit welcher sie einhergeht, zu selegieren. Hierzu werden einige, jedoch vertretbare Änderungen in der HPSG-Architektur vorgenommen und es wird gezeigt, wie mit dem generalisierten Mechanismus die Daten behandelt werden können. Daran anschließend folgt eine Erweiterung des Phänomenbereichs auf Paarformeln. Ferner wird ein Einwand im Zusammenhang mit der Analyse des Komplements als NP bzw. DP diskutiert und zur weiteren Motivation des Ansatzes wird noch ein weiteres lokales Phänomen,die Distribution der Spur, mit der hier vorgestellten Herangehensweise modelliert. Darüberhinaus wird die Frage untersucht, ob man nicht auch PPs mit festen Verben geschickt analysieren kann. Dazu wird ein Weg, Lexeme zu selegieren, eingeführt und der entwickelte Mechanismus erweitert. Diese Erweiterung findet Anwendung bei der Modellierung der lokalen Distribution einer Partikel. Eine Zusammenfassung, sowie ein Ausblick auf weiterführende Fragestellungen schließen die Arbeit ab.
Freeze (1992) argued on the basis of data from several different languages that there is a close relationship between existential sentences (stating the existence of an entity) and locative sentences (stating the location of an entity). Freeze (1992) proposes that they are both derived from the same base structure and that the surface differences are rather due to the distinct information structures.This paper argues against this position with the data from Serbian existentials, which show clear syntactic differences from the locatives. Thus, the close relationship between existential and locative sentences that Freeze (1992) observes is conceptual, but not (necessarily) part of the syntax of the language. In order to account for the data, we propose that existential sentences originate from a different syntactic predication structure than the locative ones. The existential meaning arises, as we will show, from the interaction of this predication structure with the structure and meaning of the noun phrase.
In this paper I will discuss the formation of different types of yes/no questions in Serbian (examples in (1)), focusing on the syntactically and semantically puzzling example (1d), which involves the negative auxiliary inversion. Although there is a negative marker on the fronted auxiliary, the construction does not involve sentential negation. This coincides with the fact that the negative quantifying NPIs cannot be licensed. The question formation and sentential negation have similar syntactic effects cross-linguistically. This has led to various attempts to formulate a unifying syntactic account of the phenomena (ever since Klima 1964). One striking fact about the two syntactic contexts is that both license weak NPIs (Negative Polarity Items). It has been suggested (cf. Laka 1990, Culicover 1991) that the derivation of both interrogatives and negatives involves the same type of functional projection PolP (polarity phrase). One such account of the formation of negative interrogatives in Serbo- Croatian is offered by Progovac (2005). She proposes that there are two PolPs optionally cooccurring in the same clause, in which both positive and negative polarity items check their positive or negative features (following Haegeman and Zanuttini (1991) feature-checking account of negative structures, and the insights of Brown(1999) on the negation in Russian). On her account, the negative auxiliary question in (1d), is the case when both polarity phrases are present. The higher has [-pos +neg] features, and the lower one (below TP) is [-pos -neg]. Although her account correctly predicts the ungrammaticality of (2a) in contrast with (1c), it wrongly predicts the (2b) to be grammatical. I will argue that Progovac’s theory regarding the nature of the PolP is wrong. It employs both the binary feature valuation on the polarity head and the hierarchical ordering of the two polarity phrases, which eventually leads to overgeneration. On the account presented here the nature of the question marker (li vs zar) is highly relevant. Notice that (1b) and (1d) express presuppositions regarding the truth value of the propositions. In this way they contrast with (1a) and (1c). In addition, the type (1b) (with the question particle zar) can introduce both the positive and negative presupposition as shown in (3), which, semantically, makes this construction compatible with negative auxiliary questions in English (4a). The polarity items licensed in the relevant structures are also of the same type in both languages. The fronted-negative-auxiliary questions (1d) in Serbian are only possible with the particle li. In this case the presupposition is exclusively positive. The peculiar question/focus marking function of li (in Bulgarian and Russian) is well known. However, it is always assumed that its focus marking role is not relevant for the formation of yes/no questions. This I believe is not correct. The syntactic explanation of the interpretational facts points to the following: A) The possibility of the separate lexical encoding (particle zar) of the ‘rhetorical’ yes/no questions in Serbian allows the embedding of both positive and negated sentences, in which case the (weak) NPIs can remain in local relation with the negated verb. B) Recall that Serbian is an NC language, which requires local/c-command relation between the verbal negative marker and the NPI. With the negative inverted auxiliary questions this condition is not met, and the licensing of an n-word is not possible. C) The impossibility of licensing a weak NPI (i-words in the examples below) is due to the nature of the question marker li. (1) a. Da li je Vera videla ikoga / nekoga / *nikoga? DA Q aux Vera see.part.F.Sg anyone someone noone “Did Vera see anyone/someone/noone?” b. Zar je Vera videla ikoga / nekoga / *nikoga? ZAR aux Vera see.part.F.Sg anyone someone noone “Is it really the fact that Vera saw anyone/someone?” c. Je li Vera videla ikoga / nekoga /*nikoga? aux Q Vera see.part.F.Sg anyone someone noone “Did Vera see anyone/someone/noone?” d. Nije li Vera videla *ikoga / nekoga / *nikoga? neg+aux Q Vera see.part.F.Sg anyone someone noone “Didn’t Vera see someone?”/ “Vera saw someone, didn’t she?” (2) a. *Nije li Vera videla nikoga? neg+aux Q Vera see.part.F.Sg noone b. *Nije li Vera videla ikoga? neg+aux Q Vera see.part.F.Sg anyone (3) a. Zar je Vera videla nekoga / ikoga? ZAR aux Vera see.part.F.Sg someone/anyone b. Zar Vera nije videla nekoga/nikoga? ZAR Vera neg+aux see.part.F.Sg someone/anyone (4) a. Didn’t Vera (NOT) see someone/anyone? b. Vera saw someone, didn’t she?
The renowned Grimm Dictionary (1854-1961) makes the statement that the German copula sein (to be) is “the most general and colourless of all verbal concepts” (der allgemeinste und farbloseste aller verbalbegriffe). A more concise summary of the linguistic issues surrounding the copula is hardly possible. These two properties (and the latent tension between them!) make copulas a particularly interesting and vexing subject of linguistic research. Copulas appear to be almost colourless, i.e., devoid of any concrete meaning, thus leading to the question of why such expressions exist at all, not only in German but in the majority of the world’s languages. And at the same time copulas presumably provide the best window into the core of verbal concepts thereby telling us what it actually means to be a verb – at least in a language like German or English. While there is a rather rich body of research on copulas in philosophical and formal semantics including several in-depth studies on the copular systems of individual languages, copulas have received comparably little attention from a typological perspective. The monograph of Regina Pustet sets out to fill this gap. She presents an extensive cross-linguistic study of copula usage based on a sample of 154 languages drawn from the language families of the world. The analysis is embedded in the theoretical framework of functional typology. The study aims at uncovering universal principles that govern the distribution of copulas in nominal, adjectival, and verbal predications. Its major objective is the development of a “semantically-based model of copula distribution” (p.62) by means of which the presence vs. absence of copulas can be motivated through the inherent meaning of the lexical items they potentially combine with. Drawing mainly on the work by Givón (1979, 1984) and Croft (1991, 2001), who provide a functional foundation of the traditional parts of speech, Pustet identifies four semantic parameters which, if taken together, are claimed to support substantial generalisations on copula distribution – within a given language as well as crosslinguistically. These parameters are DYNAMICITY, TRANSIENCE, TRANSITIVITY, and DEPENDENCY. Pustet goes on to argue – and this is in fact the driving force behind the overall monograph – that the distributional behaviour of copulas, in turn, yields a useful methodology for developing a general approach to lexical categorization. Thus, in the long run Pustet aims at contributing to a better understanding of the traditional parts of speech, noun, adjective, and verb by defining them in terms of “semantic feature bundles, which can be arranged in [a] coherent semantic similarity space” (p.193).
Since Donald Davidson’s seminal work “The Logical Form of Action Sentences” (1967) event arguments have become an integral component of virtually every semantic theory. Over the past years Davidson´s proposal has been continuously extended such that nowadays event(uality) arguments are generally associated not only with action verbs but with predicates of all sorts. The reasons for such an extension are seldom explicitly justified. Most problematical in this respect is the case of stative expressions. By taking a closer look at copula sentences the present study assesses the legitimacy of stretching the Davidsonian notion of events and discusses its consequences. A careful application of some standard eventuality diagnostics (perception reports, combination with locative modifiers and manner adverbials) as well as some new diagnostics (behavior of certain degree adverbials) reveals that copular expressions do not behave as expected under a Davidsonian perspective: they fail all eventuality tests, regardless of whether they represent stage-level or individual-level predicates. In this respect, copular expressions pattern with stative verbs like know, hate, and resemble, which in turn differ sharply from state verbs like stand, sit, and sleep. The latter pass all of the eventuality tests and therefore qualify as true “Davidsonian state” expressions. On the basis of these empirical observations and taking up ideas of Kim (1969, 1976) and Asher (1993, 2000), an alternative account of copular expressions (and stative verbs) is provided, according to which the copula introduces a referential argument for a temporally bound property exemplification (= “Kimian state”). Considerations on some logical properties, viz. closure conditions and the latent infinite regress of eventualities, suggest that supplementing Davidsonian eventualities with Kimian states may yield not only a more adequate analysis of copula sentences but also a better understanding of eventualities in general.
Davidsonian event semantics has an impressive track record as a framework for natural language analysis. In recent years it has become popular to assume that not only action verbs but predicates of all sorts have an additional event argument. Yet, this hypothesis is not without controversy in particular wrt the particularly challenging case of statives. Maienborn (2003a, 2004) argues that there is a need for distinguishing two kinds of states. While verbs such as sit, stand, sleep refer to eventualities in the sense of Davidson (= Davidsonian states), the states denoted by such stative verbs like know, weigh,and own, as well as any combination of copula plus predicate are of a different ontological type (= Kimian states). Against this background, the present study assesses the two main arguments that have been raised in favour of a Davidsonian approach for statives. These are the combination with certain manner adverbials and Parsons (2000) so-called time travel argument. It will be argued that the manner data which, at first sight, seem to provide evidence for a Davidsonian approach to statives are better analysed as non-compositional reinterpretations triggered by the lack of a regular Davidsonian event argument. As for Parsons´s time travel argument, it turns out that the original version does not supply the kind of support for the Davidsonian approach that Parsons supposed. However, properly adapted, the time travel argument may provide additional evidence for the need of reifying the denotatum of statives, as suggested by the assumption of Kimian states.
One aspect of the progress being made is that the focus of attention has widened. Adverbials, though still the heart of the matter, now form part of a much larger set of constituent types subsumed under the general syntactic label of adjunct; while modifier has become the semantic counterpart on the same level of generality. So one of the readings of Modifying Adjuncts stands for the focus on this intersection. Moreover, recent years have seen a number of studies which attest an increasing interest in adjunct issues. There is an impressive number of monographs, e.g. Alexiadou (1997), Laenzlinger (1998), Cinque (1999), Pittner (1999), Ernst (2002), which, by presenting in-depth analyses of the syntax of adjuncts, have sharpened the debate on syntactic theorizing. Serious attempts to gain a broader view on adjuncts are witnessed by several collections, see Alexiadou and Svenonius (2000), Austin, Engelberg and Rauh (in progress); of particular importance are the contributions to vol. 12.1 of the Italian Journal of Linguistics (2000), a special issue on adverbs, the Introductions to which by Corver and Delfitto (2000) and Delfitto (2000) may be seen as the best state-of-the-art article on adverbs and adverbial modification currently on the market. To try and test a fresh view on adjuncts was the leitmotif of the Oslo Conference “Approaching the Grammar of Adjuncts” (Sept 22–25, 1999), which provided the initial forum for the papers contained in this volume and initiated a period of discussion and continuing interaction among the contributors, from which the versions published here have greatly profited. The aim of the Oslo conference, and hence the focus of the present volume, was to encourage syntacticians and semanticists to open their minds to a more integrative approach to adjuncts, thereby paying attention to, and attempting to account for, the various interfaces that the grammar of adjuncts crucially embodies. From this perspective, the present volume is to be conceived of as an interim balance of current trends in modifying the views on adjuncts. In introducing the papers, we will refrain from rephrasing the abstracts, but will instead offer a guided tour through the major problem areas they are tackling. Assessed by thematic convergence and mutual reference, the contributions form four groups, which led us to arrange them into subparts of the book. Our commenting on these is intended (i) to provide a first glance at the contents, (ii) to reveal some of the reasons why adjuncts indeed are, and certainly will remain, a challenging issue, and thereby (iii) to show some facets of what we consider novel and promising approaches.
“Comments are very welcome!” This basic attitude and the many ways of implementing it contribute immensely to the fascination of engaging in scientific research. I am grateful to Theoretical Linguistics for providing a public platform for this kind of scholarly exchange and I thank all commentators for their thoughtful, stimulating, and often challenging contributions to my target article. My response will address two main issues that are raised by the commentaries. The first issue is shaped by a cluster of questions relating to ontology. The second issue concerns questions of methodology pertaining in particular to the problem of judging data.
The article offers evidence that there are two variants of adverbial modification that differ with respect to the way in which a modifier is linked to the verbs eventuality argument. So-called event-external modifiers relate to the full eventuality, whereas event-internal modifiers relate to some integral part of it. The choice between external and internal modification is shown to be dependent on the modifiers syntactic base position. Event-external modifiers are base-generated at the VP periphery, whereas event-internal modifiers are base-generated at the V periphery. These observations are accounted for by a refined version of the standard Davidsonian approach to adverbial modification according to which modification is mediated by a free variable. In the case of external modification, the grammar takes responsibility for identifying the free variable with the verbs eventuality argument, whereas in the case of internal modification, a value for the free variable is determined by the conceptual system on the basis of contextually salient world knowledge. For the intriguing problem that certain locative modifiers occasionally seem to have nonlocative (instrumental, positional, or manner) readings, the advocated approach can provide a rather simple solution.
Das Zustandspassiv : grammatische Einordnung – Bildungsbeschränkungen – Interpretationsspielraum
(2005)
Semantic research over the past three decades has provided impressive confirmation of Donald Davidsons famous claim that “there is a lot of language we can make systematic sense of if we suppose events exist” (Davidson 1980:137). Nowadays, Davidsonian event arguments are no longer reserved only for action verbs (as Davidson originally proposed) or even only for the category of verbs, but instead are widely assumed to be associated with any kind of predicate (e.g. Higginbotham 2000, Parsons 2000).1 The following quotation from Higginbotham and Ramchand (1997) illustrates the reasoning that motivates this move: "Once we assume that predicates (or their verbal, etc. heads) have a position for events, taking the many consequences that stem therefrom, as outlined in publications originating with Donald Davidson (1967), and further applied in Higginbotham (1985, 1989), and Terence Parsons (1990), we are not in a position to deny an event-position to any predicate; for the evidence for, and applications of, the assumption are the same for all predicates. (Higginbotham and Ramchand 1997:54)" In fact, since Davidson’s original proposal the burden of proof for postulating event arguments seems to have shifted completely, leading Raposo and Uriagereka (1995), for example, to the following verdict: "it is unclear what it means for a predicate not to have a Davidsonian argument (Raposo and Uriagereka 1995:182)" That is, Davidsonian eventuality arguments apparently have become something like a trademark for predicates in general. The goal of the present paper is to subject this view of the relationship between predicates and events to real scrutiny. By taking a closer look at the simplest independent predicational structure – viz. copula sentences – I will argue that current Davidsonian approaches tend to stretch the notion of events too far, thereby giving up much of its linguistic and ontological usefulness. More specifically, the paper will tackle the following three questions: 1. Do copula sentences support the current view of the inherent event-relatedness of predicates? 2. If not, what is a possible alternative to an event-based analysis of copula sentences? 3. What does this tell us about Davidsonian events? The paper is organized as follows: Section 2 first reviews current event-based analyses of copula sentences and then gives a brief summary of the Davidsonian notion of events. Section 3 examines the behavior of copula sentences with respect to some standard (as well as some new) eventuality diagnostics. Copula expressions will turn out to fail all eventuality tests. They differ sharply from state verbs like stand, sit, sleep in this respect. (The latter pass all eventuality tests and therefore qualify as true “Davidsonian state” expressions.) On the basis of these observations, section 4 provides an alternative account of copula sentences that combines Kim’s (1969, 1976) notion of property exemplifications with Ashers (1993, 2000) conception of abstract objects. Specifically, I will argue that the copula introduces a referential argument for a temporally bound property exemplification (= “Kimian state”). The proposal is implemented within a DRT framework. Finally, section 5 offers some concluding remarks and suggests that supplementing Davidsonian eventualities by Kimian states not only yields a more adequate analysis for copula expressions and the like but may also improve our treatment of events.
A pragmatic explanation of the stage level/individual level contrast in combination with locatives
(2004)
One important difference between stage level predicates (SLPs) and individual level predicates (ILPs) is their behavior with respect to locative modifiers. It is commonly assumed that SLPs but not ILPs combine with locatives. The present study argues against a semantic account for this behavior (as advanced by e.g. Kratzer 1995, Chierchia 1995) and proposes a genuinely pragmatic explanation of the observed stage level/individual level contrast instead. The proposal is spelled out using Blutners (1998, 2000) optimality theoretic version of the Gricean maxims. Building on the observation that the respective locatives are not event-related but frame-setting modifiers, the preference for main predicates that express temporary properties is explained as a side-effect of “synchronizing” the main predicate with the locative frame in the course of finding an optimal interpretation. By emphasizing the division of labor between grammar and pragmatics, the proposed solution takes a considerable load off of semantics.
The study offers a discourse-based account of the Spanish copula forms ser and estar, which are generally considered to be lexical exponents of the stage-level/individual-level contrast. It argues against the popular view that the distinction between SLPs and ILPs rests on a fundamental cognitive division of the world that is reflected in the grammar. As it happens, conceptual oppositions like “temporary vs. permanent” or “arbitrary vs. essential“ provide only a preference for the interpretation of estar and ser. In addition, the evidence for an SLP/ILP impact on the grammar turns out to be far less conclusive than is currently assumed. The study argues against event-based accounts of the ser/estar contrast in particular, showing that ser and estar pattern alike in failing all of the standard eventuality tests. The discourse-based account proposed instead assumes that ser and estar both display the same lexical semantics (which is identical to the semantics of English be, German sein, etc.); estar differs from ser only in presupposing a relation to a specific discourse situation. By using estar a speaker restricts his or her claim to a specific discourse situation, whereas by using ser, the speaker makes no such restriction. The preference for interpreting estar predications as denoting temporary properties and ser predications as denoting permanent properties follows from economy principles driving the pragmatic legitimation of estars discourse dependence. The analysis proposed in this paper can also account for the observation that ser predications do not give rise to thetic judgements. The proposal is couched in terms of the framework of DRT.
In recent years, research in parsing has extended in several new directions. One of these directions is concerned with parsing languages other than English. Treebanks have become available for many European languages, but also for Arabic, Chinese, or Japanese. However, it was shown that parsing results on these treebanks depend on the types of treebank annotations used. Another direction in parsing research is the development of dependency parsers. Dependency parsing profits from the non-hierarchical nature of dependency relations, thus lexical information can be included in the parsing process in a much more natural way. Especially machine learning based approaches are very successful (cf. e.g.). The results achieved by these dependency parsers are very competitive although comparisons are difficult because of the differences in annotation. For English, the Penn Treebank has been converted to dependencies. For this version, Nivre et al. report an accuracy rate of 86.3%, as compared to an F-score of 92.1 for Charniaks parser. The Penn Chinese Treebank is also available in a constituent and a dependency representations. The best results reported for parsing experiments with this treebank give an F-score of 81.8 for the constituent version and 79.8% accuracy for the dependency version. The general trend in comparisons between constituent and dependency parsers is that the dependency parser performs slightly worse than the constituent parser. The only exception occurs for German, where F-scores for constituent plus grammatical function parses range between 51.4 and 75.3, depending on the treebank, NEGRA or TüBa-D/Z. The dependency parser based on a converted version of Tüba-D/Z, in contrast, reached an accuracy of 83.4%, i.e. 12 percent points better than the best constituent analysis including grammatical functions.
This paper profiles significant differences in syntactic distribution and differences in word class frequencies for two treebanks of spoken and written German: the TüBa-D/S, a treebank of transliterated spontaneous dialogues, and the TüBa-D/Z treebank of newspaper articles published in the German daily newspaper die tageszeitung´(taz). The approach can be used more generally as a means of distinguishing and classifying language corpora of different genres.
This paper profiles significant differences in syntactic distribution and differences in word class frequencies for two treebanks of spoken and written German: the TüBa-D/S, a treebank of transliterated spontaneous dialogs, and the TüBa-D/Z treebank of newspaper articles published in the German daily newspaper ´die tageszeitung´(taz). The approach can be used more generally as a means of distinguishing and classifying language corpora of different genres.
This paper presents an approach to the question whether it is possible to construct a parser based on ideas from case-based reasoning. Such a parser would employ a partial analysis of the input sentence to select a (nearly) complete syntax tree and then adapt this tree to the input sentence. The experiments performed on German data from the Tüba-D/Z treebank and the KaRoPars partial parser show that a wide range of levels of generality can be reached, depending on which types of information are used to determine the similarity between input sentence and training sentences. The results are such that it is possible to construct a case-based parser. The optimal setting out of those presented here need to be determined empirically.
Quantitative evaluation of parsers has traditionally centered around the PARSEVAL measures of crossing brackets, (labeled) precision, and (labeled) recall. However, it is well known that these measures do not give an accurate picture of the quality of the parsers output. Furthermore, we will show that they are especially unsuited for partial parsers. In recent years, research has concentrated on dependencybased evaluation measures. We will show in this paper that such a dependency-based evaluation scheme is particularly suitable for partial parsers. TüBa-D, the treebank used here for evaluation, contains all the necessary dependency information so that the conversion of trees into a dependency structure does not have to rely on heuristics. Therefore, the dependency representations are not only reliable, they are also linguistically motivated and can be used for linguistic purposes.
The purpose of this paper is to describe the TüBa-D/Z treebank of written German and to compare it to the independently developed TIGER treebank (Brants et al., 2002). Both treebanks, TIGER and TüBa-D/Z, use an annotation framework that is based on phrase structure grammar and that is enhanced by a level of predicate-argument structure. The comparison between the annotation schemes of the two treebanks focuses on the different treatments of free word order and discontinuous constituents in German as well as on differences in phrase-internal annotation.
The earliest known extensive texts in Gullah (and perhaps African American Vernacular English as well) to appear in print were published in The Riverside Magazine for Young People in November, 1868, under the title "Negro Fables" (p. 505-507). These are four animal stories, which the editor of the magazine, Horace Elisha Scudder, described in his column only as having been "taken down from the lips of an old negro, in the vicinity of Charleston" (see Appendix for the editor´s comments and the full text of the stories).2 The Story-Teller was evidently a genuine "man of words" (Abrahams, 1983), a true raconteur who could artistically embellish a simple traditional account (perhaps further embellished by the transcriber) in a variety of ways. That he commanded a certain range of Gullah is evident from particular signature features in the texts, but the absence of other typical Gullah features and the presence of shared Gullah/African American Vernacular English usages, together with the periodic appearance of standard English forms, demonstrate that these texts provide perhaps the earliest actual documentation (apart from early tertiary comments, cited e.g. in Feagin, 1997, p. 128-129) of register variation or style/code-switching among Gullah speakers. ...
The ACL 2008 Workshop on Parsing German features a shared task on parsing German. The goal of the shared task was to find reasons for the radically different behavior of parsers on the different treebanks and between constituent and dependency representations. In this paper, we describe the task and the data sets. In addition, we provide an overview of the test results and a first analysis.
The Conference on Computational Natural Language Learning features a shared task, in which participants train and test their learning systems on the same data sets. In 2007, as in 2006, the shared task has been devoted to dependency parsing, this year with both a multilingual track and a domain adaptation track. In this paper, we define the tasks of the different tracks and describe how the data sets were created from existing treebanks for ten languages. In addition, we characterize the different approaches of the participating systems, report the test results, and provide a first analysis of these results.
Recent approaches to Word Sense Disambiguation (WSD) generally fall into two classes: (1) information-intensive approaches and (2) information-poor approaches. Our hypothesis is that for memory-based learning (MBL), a reduced amount of data is more beneficial than the full range of features used in the past. Our experiments show that MBL combined with a restricted set of features and a feature selection method that minimizes the feature set leads to competitive results, outperforming all systems that participated in the SENSEVAL-3 competition on the Romanian data. Thus, with this specific method, a tightly controlled feature set improves the accuracy of the classifier, reaching 74.0% in the fine-grained and 78.7% in the coarse-grained evaluation.
Das Chunkparsing bietet einen besonders vielversprechenden Ansatz zum robusten, partiellen Parsing mit dem Ziel einer breiten Datenabdeckung. Ziel beim Chunkparsing ist eine partielle, nicht-rekursive syntaktische Struktur. Dieser extrem effiziente Parsing-Ansatz läßt sich als Kaskade endlicher Transducer realisieren. In diesem Beitrag wird TüSBL vorgestellt, ein System, bei dem die Eingabe aus spontaner, gesprochener Spache besteht, die dem Parser in Form eines Worthypothesengraphen aus einem Spracherkenner zur Verfügung gestellt wird. Chunkparsing ist für eine solche Anwendung besonders geeignet, da es fragmentarische oder nicht wohlgeformte Äußerungen robust behandeln kann. Des weiteren wird eine Baumkonstruktionskomponente vorgestellt, die die partiellen Chunkstrukturen zu vollständigen Bäumen mit grammatischen Funktionen erweitert. Das System wird anhand manuell überprüfter Systemeingaben evaluiert, da sich die üblichen Evaluationsparameter hierfür nicht eignen.
The purpose of this paper is to describe recent developments in the morphological, syntactic, and semantic annotation of the TüBa-D/Z treebank of German. The TüBa-D/Z annotation scheme is derived from the Verbmobil treebank of spoken German [4, 10], but has been extended along various dimensions to accommodate the characteristics of written texts. TüBa-D/Z uses as its data source the "die tageszeitung" (taz) newspaper corpus. The Verbmobil treebank annotation scheme distinguishes four levels of syntactic constituency: the lexical level, the phrasal level, the level of topological fields, and the clausal level. The primary ordering principle of a clause is the inventory of topological fields, which characterize the word order regularities among different clause types of German, and which are widely accepted among descriptive linguists of German [3, 6]. The TüBa-D/Z annotation relies on a context-free backbone (i.e. proper trees without crossing branches) of phrase structure combined with edge labels that specify the grammatical function of the phrase in question. The syntactic annotation scheme of the TüBa-D/Z is described in more detail in [12, 11]. TüBa-D/Z currently comprises approximately 15 000 sentences, with approximately 7 000 sentences being in the correction phase. The latter will be released along with an updated version of the existing treebank before the end of this year. The treebank is available in an XML format, in the NEGRA export format [1] and in the Penn treebank bracketing format. The XML format contains all types of information as described above, the NEGRA export format contains all sentenceinternal information while the Penn treebank format includes only those layers of information that can be expressed as pure tree structures. Over the course of the last year, more fine grained linguistic annotations have been added along the following dimensions: 1. the basic Stuttgart-Tübingen tagset, STTS, [9] labels have been enriched by relevant features of inflectional morphology, 2. named entity information has been encoded as part of the syntactic annotation, and 3. a set of anaphoric and coreference relations has been added to link referentially dependent noun phrases. In the following sections, we will describe each of these innovations in turn and will demonstrate how the additional annotations can be incorporated into one comprehensive annotation scheme.
Part-of-Speech tagging is generally performed by Markov models, based on bigram or trigram models. While Markov models have a strong concentration on the left context of a word, many languages require the inclusion of right context for correct disambiguation. We show for German that the best results are reached by a combination of left and right context. If only left context is available, then changing the direction of analysis and going from right to left improves the results. In a version of MBT (Daelemans et al., 1996) with default parameter settings, the inclusion of the right context improved POS tagging accuracy from 94.00% to 96.08%, thus corroborating our hypothesis. The version with optimized parameters reaches 96.73%.
The definition of similarity between sentences is formulated on the levels of words, POS tags, and chunks (Abney 91; Abney 96). The evaluation of this approach shows that while precision and recall based on the PARSEVAL measures (Black et al. 91) do not reach state of the art Parsers yet (F1=87.19 on syntactic constituents, F1=77.78 including functionargument structure), the parser shows a very reliable performance where function-argument structure is concerned (F1=96.52). The lower F-scores are very often due to unattached constituents.
The problem of vocalization, or diacritization, is essential to many tasks in Arabic NLP. Arabic is generally written without the short vowels, which leads to one written form having several pronunciations with each pronunciation carrying its own meaning(s). In the experiments reported here, we define vocalization as a classification problem in which we decide for each character in the unvocalized word whether it is followed by a short vowel. We investigate the importance of different types of context. Our results show that the combination of using memory-based learning with only a word internal context leads to a word error rate of 6.64%. If a lexical context is added, the results deteriorate slowly.
In dieser Arbeit soll erst ein kurzer Überblick über die Gebiete der Wortklassifizierung und des maschinellen Lernens gegeben werden (Kap. 1). Dann wird der Ansatz der transformationsbasierten fehlergesteuerten Wortklassifizierung (Transformation-Based Error-Driven Tagging) von Brill (1992, 1993, 1994) vorgestellt und für die Verwendung für deutschsprachige Korpora angepaßt (Kap. 2). Hierbei handelt es sich um ein regelbasiertes System, bei dem die Regeln im Gegensatz zu den bisher vorhandenen Systemen nicht manuell erarbeitet und dem System vorgegeben werden; das System erwirbt die Regeln vielmehr selbst anhand von wenigen Regelschemata aus einem kleinen bereits getaggten Lernkorpus. In Kapitel 3 werden die Ergebnisse aus der Anwendung des Systems auf Teile eines deutschsprachigen Korpus dargestellt. In Kapitel 4 schließlich werden andere Taggingsysteme vorgestellt und mit dem System von Brill (1993) anhand von acht Kriterien verglichen.
In syntax, the trend nowadays is towards lexicalized grammar formalisms. It is now widely accepted that dividing words into wordclasses may serve as a laborsaving mechanism - but at the same time, it discards all detailed information on the idiosyncratic behavior of words. And that is exactly the type of information that may be necessary in order to parse a sentence. For learning approaches, however, lexicalized grammars represent a challenge for the very reason that they include so much detailed and specific information, which is difficult to learn. This paper will present an algorithm for learning a link grammar of German. The problem of data sparseness is tackled by using all the available information from partial parses as well as from an existing grammar fragment and a tagger. This is a report about work in progress so there are no representative results available yet.
This paper presents a comparative study of probabilistic treebank parsing of German, using the Negra and TüBa-D/Z treebanks. Experiments with the Stanford parser, which uses a factored PCFG and dependency model, show that, contrary to previous claims for other parsers, lexicalization of PCFG models boosts parsing performance for both treebanks. The experiments also show that there is a big difference in parsing performance, when trained on the Negra and on the TüBa-D/Z treebanks. Parser performance for the models trained on TüBa-D/Z are comparable to parsing results for English with the Stanford parser, when trained on the Penn treebank. This comparison at least suggests that German is not harder to parse than its West-Germanic neighbor language English.
How to compare treebanks
(2008)
Recent years have seen an increasing interest in developing standards for linguistic annotation, with a focus on the interoperability of the resources. This effort, however, requires a profound knowledge of the advantages and disadvantages of linguistic annotation schemes in order to avoid importing the flaws and weaknesses of existing encoding schemes into the new standards. This paper addresses the question how to compare syntactically annotated corpora and gain insights into the usefulness of specific design decisions. We present an exhaustive evaluation of two German treebanks with crucially different encoding schemes. We evaluate three different parsers trained on the two treebanks and compare results using EVALB, the Leaf-Ancestor metric, and a dependency-based evaluation. Furthermore, we present TePaCoC, a new testsuite for the evaluation of parsers on complex German grammatical constructions. The testsuite provides a well thought-out error classification, which enables us to compare parser output for parsers trained on treebanks with different encoding schemes and provides interesting insights into the impact of treebank annotation schemes on specific constructions like PP attachment or non-constituent coordination.
In the last decade, the Penn treebank has become the standard data set for evaluating parsers. The fact that most parsers are solely evaluated on this specific data set leaves the question unanswered how much these results depend on the annotation scheme of the treebank. In this paper, we will investigate the influence which different decisions in the annotation schemes of treebanks have on parsing. The investigation uses the comparison of similar treebanks of German, NEGRA and TüBa-D/Z, which are subsequently modified to allow a comparison of the differences. The results show that deleted unary nodes and a flat phrase structure have a negative influence on parsing quality while a flat clause structure has a positive influence.
Transforming constituent-based annotation into dependency-based annotation has been shown to work for different treebanks and annotation schemes (e.g. Lin (1995) has transformed the Penn treebank, and Kübler and Telljohann (2002) the Tübinger Baumbank des Deutschen (TüBa-D/Z)). These ventures are usually triggered by the conflict between theory-neutral annotation, that targets most needs of a wider audience, and theory-specific annotation, that provides more fine-grained information for a smaller audience. As a compromise, it has been pointed out that treebanks can be designed to support more than one theory from the start (Nivre, 2003). We argue that information can also be added to an existing annotation scheme so that it supports additional theory-specific annotations. We also argue that such a transformation is useful for improving and extending the original annotation scheme with respect to both ambiguous annotation and annotation errors. We show this by analysing problems that arise when generating dependency information from the constituent-based TüBa-D/Z.
Chunk parsing has focused on the recognition of partial constituent structures at the level of individual chunks. Little attention has been paid to the question of how such partial analyses can be combined into larger structures for complete utterances. Such larger structures are not only desirable for a deeper syntactic analysis. They also constitute a necessary prerequisite for assigning function-argument structure. The present paper offers a similaritybased algorithm for assigning functional labels such as subject, object, head, complement, etc. to complete syntactic structures on the basis of prechunked input. The evaluation of the algorithm has concentrated on measuring the quality of functional labels. It was performed on a German and an English treebank using two different annotation schemes at the level of function argument structure. The results of 89.73% correct functional labels for German and 90.40%for English validate the general approach.
In this paper, we investigate the role of sub-optimality in training data for part-of-speech tagging. In particular, we examine to what extent the size of the training corpus and certain types of errors in it affect the performance of the tagger. We distinguish four types of errors: If a word is assigned a wrong tag, this tag can belong to the ambiguity class of the word (i.e. to the set of possible tags for that word) or not; furthermore, the major syntactic category (e.g. "N" or "V") can be correctly assigned (e.g. if a finite verb is classified as an infinitive) or not (e.g. if a verb is classified as a noun). We empirically explore the decrease of performance that each of these error types causes for different sizes of the training set. Our results show that those types of errors that are easier to eliminate have a particularly negative effect on the performance. Thus, it is worthwhile concentrating on the elimination of these types of errors, especially if the training corpus is large.
Prepositional phrase (PP) attachment is one of the major sources for errors in traditional statistical parsers. The reason for that lies in the type of information necessary for resolving structural ambiguities. For parsing, it is assumed that distributional information of parts-of-speech and phrases is sufficient for disambiguation. For PP attachment, in contrast, lexical information is needed. The problem of PP attachment has sparked much interest ever since Hindle and Rooth (1993) formulated the problem in a way that can be easily handled by machine learning approaches: In their approach, PP attachment is reduced to the decision between noun and verb attachment; and the relevant information is reduced to the two possible attachment sites (the noun and the verb) and the preposition of the PP. Brill and Resnik (1994) extended the feature set to the now standard 4-tupel also containing the noun inside the PP. Among many publications on the problem of PP attachment, Volk (2001; 2002) describes the only system for German. He uses a combination of supervised and unsupervised methods. The supervised method is based on the back-off model by Collins and Brooks (1995), the unsupervised part consists of heuristics such as ”If there is a support verb construction present, choose verb attachment”. Volk trains his back-off model on the Negra treebank (Skut et al., 1998) and extracts frequencies for the heuristics from the ”Computerzeitung”. The latter also serves as test data set. Consequently, it is difficult to compare Volk’s results to other results for German, including the results presented here, since not only he uses a combination of supervised and unsupervised learning, but he also performs domain adaptation. Most of the researchers working on PP attachment seem to be satisfied with a PP attachment system; we have found hardly any work on integrating the results of such approaches into actual parsers. The only exceptions are Mehl et al. (1998) and Foth and Menzel (2006), both working with German data. Mehl et al. report a slight improvement of PP attachment from 475 correct PPs out of 681 PPs for the original parser to 481 PPs. Foth and Menzel report an improvement of overall accuracy from 90.7% to 92.2%. Both integrate statistical attachment preferences into a parser. First, we will investigate whether dependency parsing, which generally uses lexical information, shows the same performance on PP attachment as an independent PP attachment classifier does. Then we will investigate an approach that allows the integration of PP attachment information into the output of a parser without having to modify the parser: The results of an independent PP attachment classifier are integrated into the parse of a dependency parser for German in a postprocessing step.
Maschinelles Lernen wird häufig zur effzienten Annotation großer Datenmengen eingesetzt. Die Forschung zu maschinellen Lernverfahren beschränkt sich i.a. darauf unterschiedliche Lernverfahren zu vergelichen oder die optimale größe der Trainingsdaten zu bestimmen. Bisher wurde jedoch nicht untersucht, in wie weit sich linguistisches Wissen bei der Aufgabendefinition positiv auswirken kann. Dies soll hier anhand des Lernens von Base-Nominalphrasen mit drei unterschiedlichen Definitionen untersucht werden. Die Definitionen unterscheiden sich im Grad der linguistisch motivierten Erweiterungen, die zu einer eher praktisch motivierten ersten Definition hinzu kamen. Die Untersuchungen ergaben, dass sich die Anzahl der falsch klasssifizierten Wörter um ein Drittel reduzieren lässt.
This report explores the question of compatibility between annotation projects including translating annotation formalisms to each other or to common forms. Compatibility issues are crucial for systems that use the results of multiple annotation projects. We hope that this report will begin a concerted effort in the field to track the compatibility of annotation schemes for part of speech tagging, time annotation, treebanking, role labeling and other phenomena.
This paper reports on the SYN-RA (SYNtax-based Reference Annotation) project, an on-going project of annotating German newspaper texts with referential relations. The project has developed an inventory of anaphoric and coreference relations for German in the context of a unified, XML-based annotation scheme for combining morphological, syntactic, semantic, and anaphoric information. The paper discusses how this unified annotation scheme relates to other formats currently discussed in the literature, in particular the annotation graph model of Bird and Liberman (2001) and the pie-in-thesky scheme for semantic annotation.