Thomas Herbst

The Status of Generalizations: Valency and Argument Structure Constructions

This article outlines why valency theory and Goldbergs theory of argument structure constructions should be combined. In the light of empirical valency research it is argued that the amount of item-specificity to be observed in language must be accounted for in a theory of complementation and that generalizations concerning parallels between semantic properties of verbs and their occurrence in particular valency patterns must be treated with a certain degree of caution. It is suggested that one way of combining the two approaches is to make use of specific formal categories and to introduce a valency realisation principle in a theory of argument structure constructions.

Collo-Creativity and Blending: Recognizing Creativity Requires Lexical Storage in Constructional Slots

Abstract This article explores to what extent linguistic creativity can be accounted for by investigating the collocational and collostructional properties of different constructions. A distinction is made between intended creativity and creativity caused by despair. It is argued that while unexpected combinations of constructions (in the sense of lexical items and syntactic constructions) are not the only type of linguistic creativity, they are of particular relevance to linguistic theory because they can only be appropriately accounted for in terms of a model that takes an integrative view of grammar and lexicon and allows for a considerable amount of lexical knowledge concerning the description of constructions.

From collocations and patterns to constructions – an introduction

Linguistic usage patterns are not just coincidental phenomena on the textual surface but constitute a fundamental constructional principle of language. At the same time, however, linguistic patterns are highly idiosyncratic in the sense that they tend to be item-specific and unpredictable, thus defying all attempts at capturing them by general abstract rules. A range of linguistic approaches inspired by surprisingly different background assumptions and aims have acknowledged these insights and tried to come up with ways of emphasizing the importance of linguistic repetitiveness and regularity while doing justice to unpredictability and item-specificity. Their efforts are epitomized in the terms enshrined in the title of the present volume, whose aim is to provide a multifaceted view of Constructions, Collocation and Patterns. What all of these approaches share, in addition to their interest in recurrent patterns, is a strong commitment to the value of usage, be it in the wider sense of usage as an empirical basis for sound linguistic analysis and description or in the narrower sense of usage as constituting the basis for the emergence and consolidation of linguistic knowledge. The first and presumably oldest (though to some perhaps not the most obvious) tradition takes the perspective of foreign language linguistics. Any reflection upon what is important in the learning – and, consequently, also in the teaching – of a foreign language will have to take into account the crucial role of conventionalized but unpredictable collocations. Any attempt by a learner to achieve some kind of near-nativeness will have to include facts of language such as the fact that it is lay or set the table in English, but Tisch decken in German, and mettre la table in French. 1 It is thus not at all surprising that foreign language linguistics has resulted in extensive research on collocations and how they can best be taught and learnt. In fact, the very origin of the term collocation can be traced back to the Second Interim Report on English Collocations by Harold E. Palmer published in 1933 (Cowie 2009: 391-393, Stubbs 2009). Secondly, while the phenomenon of collocation concerns the associations between lexical items, verb complementation or valency patterns present learners with the same kind of difficulty, since a similar element of unpredictability can be observed in this area in that you can say Sie erklärte

Is Language a Collostructicon? A Proposal for Looking at Collocations, Valency, Argument Structure and Other Constructions

This chapter argues in favour of not regarding collocation and valency as strictly discrete categories but rather seeing them as near neighbours in the lexis-grammar continuum. Following Bybee’s (Usage-based theory and exemplar representation of constructions. In Hoffmann T, Trousdale G (eds) The Oxford handbook of construction grammar. Oxford University Press, Oxford, pp. 49–69, 2013) analysis of the drive me crazy construction, a suggestion will be made for presenting both collocational and valency phenomena in terms of constructions. It will be argued that the constructicon representing speakers’ linguistic knowledge contains both item-specific information and generalized information in the form of Goldbergian argument structure constructions (Goldberg 2016) and in particular that the description of valency slots should provide exemplar representations based on the principles of collostructional analysis as developed by Stefanowitsch and Gries (Inter J Coprus Lingusitics 8:209–243, 2003).

Do Constructions make a Difference? Introduction to a Special Issue of ZAA on Aspects of Construction Grammar

Is construction grammar a model of language for “syntaktisch Unbegabte” – people with no talent for syntax? This question was raised (though not answered in the affirmative) by Leiss (2009, 19) and leads to the question of why constructionist approaches have gained in popularity in recent years. It is one of the central tenets of constructionist and cognitive approaches to language to reject the idea of a strict dividing line between syntax and lexis – something they have in common with many traditional and corpus linguistic approaches. This recognition of a lexicogrammatical continuum or a gradient between lexis and grammar may in fact be one of the reasons that make this framework attractive to scholars who may be typical contributors to a journal such as ZAA, namely linguists who tend to analyse a language that is not their mother tongue: dealing with a foreign language raises the awareness of the crucial role of phenomena that do not lend themselves to a treatment in terms of easy generalizations or “rules” such as collocation, valency and idiomaticity, as indeed do many findings of corpus linguistics. Constructionist approaches acknowledge this and consequently do not make a distinction between a core grammar and a periphery. There may be other reasons for the increasing popularity of constructionist and usage-based approaches. One may be that some of the models developed within this overall framework do not make use of explicit formal representations to a great extent, which, however, is by no means true of all constructionist approaches.1 A third reason – and maybe one that applies in particular to European linguistics – could be that basic assumptions of constructionist approaches show

Chunk parsing in corpora

After a brief introduction to the concept of “chunk”, we describe the expectations in (“shallow”) chunk parsing from a syntactic and semantic perspective. For syntactic chunking, the task is separated in two steps, segmentation and prototyping. For segmentation, a sequence of words in a corpus is annotated by “IOB” tags (for inside, outside, and beginning of chunk), where the B tags are augmented by a POS (part of speech) tag. This has to be done either manually or by using an already annotated reference corpus in a semi-automatic fashion. An annotated corpus can be used for training a chunker by well-known techniques from pattern recognition. Setting up transformation-based learning as an iterative process can lead to precision and recall rates of about 94% on unseen corpora. For semantic chunking, additional information about the contents of phrases is required, e.g., by tagging noun phrases with person, location, etc. Experiences with tasks to identify thematic role fillers for verbs as agent, patient, or theme by shallow parsing are still significantly less successful than parsing with “full” syntax. 1 What on Earth is a Chunk? 1.1 Basic Features of Chunks A fundamental analytical task in Natural Language Processing (NLP) is the segmentation and labeling of texts. In a first step, texts are broken up into sentences as sequences of word forms (tokens). “Chunking” in general means to assign a partial structure to a sentence. “Tagging” assigns labels to the tokens which represent word specific and word form specific information such as the word category and morphological features. Chunk parsing regards sequences of tokens and tries to identify structural relations within and between segments. Chunk parsing as conceived by Abney [1] originated from a psycholinguistic motivation: “I begin with an intuition: when I read a sentence, I read it a chunk at a time. For example, the previous sentence breaks up something like this: (1) [I begin] [with an intuition]: [when I read] [a sentence], [I read it] [a chunk] [at a time] These chunks correspond in some way to prosodic patterns. It appears, for instance, that the strongest stresses in the sentence fall

INTRODUCTION: THE MYSTERY OF COLLOCATION

The term collocation presents an almost prototypical example of the phenomenon of polysemy. In what one could call a rather neutral sense it is used to refer to any combination of two words that co-occur in language text. Other uses of the term identify more specific types of combination but differ as to what they consider to be the relevant criterion that makes up specificity. One type, the sandy beaches-type, refers to specificity in statistical terms of co-occurrence in the language, or in a corpus, – where statistical significance is not necessarily determined in terms of absolute frequency of co-occurrence but calculated on the basis of some sort of measure of mutual expectancy. In the second type, the guilty conscience-type, the combination is significant because it is established or institutionalized, to use a term common in word formation, and somehow unpredictable on the grounds of the meanings of the words. It is relatively obvious why the first type, referred to as “quantitative” by Hausmann and Blumenthal (2006, 3) should feature prominently in corpus linguistics, whereas the second, which Hausmann and Blumenthal (2006, 3) characterize as “essentiellement qualitative” (essentially qualitative) and in terms of “coocurrence lexicale restreinte” (restricted lexical co-occurrence), has been a key concept of foreign language linguistics for a very long time. That a foreign speaker of English who knows the noun tea will need to know (or be able to find out from a dictionary) that it can be qualified as weak tea but not as feeble tea or light tea or that the verb that accompanies tea to express the idea of preparation is make and not cook or boil is an insight that is reflected in Makkai’s (1972) classification of such combinations as “encoding idioms” and in Hausmann’s fundamental distinction between Basis (tea) and Kollokator (collocate), which is of particular relevance to foreign language teaching and foreign language lexicography. While the original vagueness of the term as used by Firth can be seen as leading to the emergence of its different uses, there is also considerable overlap