Constructional Approaches to Syntactic Structures in German

Hans C. Boas, Alexander Ziem, Hans C. Boas, Alexander Ziem

• Publisher: De Gruyter Mouton
• Year: 2018
• eBook ISBN: 9783110455168
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

This book provides a state of the art collection of constructional research on syntactic structures in German. The volume is unique in that it offers an easily accessible, yet comprehensive and sophisticated variety of papers. Moreover, various of the papers make explicit connections between grammatical constructions and the concept of valency which has figured quite prominently in Germanic Linguistics over the past half century.

Part I:Grammatical constructions and valency

Stefan Engelberg

The argument structure of psych-verbs: A quantitative corpus study on cognitive entrenchment

Note: I am grateful to Alexander Koplenig for valuable comments on the statistical analyses and for conducting the cluster analysis presented in Section 2.4. The reviewers also provided a number of helpful comments.

Stefan Engelberg, Institut für Deutsche Sprache, R5 6-13, 68161 Mannheim, Germany, [email protected]

1Introduction

1.1Psych-verbs and argument structure alternations

Psych-verbs exhibit a fairly large variation in argument structure patterns, as the following examples from German show. Many of these verbs allow alternations between stimuli and experiencers in subject position (1a vs. 1b), between nominal and clausal realizations of arguments (1a vs. 1c, 1b vs. 1d), between inanimate and agent-like animate stimuli (1a vs. 1e), between simple stimuli and “split-stimuli” that are spread out over two constituents, a subject NP and a PP (1a vs. 1f), and between explicit and implicit argument realization (1b vs. 1g).

Traditionally, psych-verbs have been investigated because the alternation between stimulus and experiencer subjects poses some interesting problems for linking theories (cf., e.g., Belletti and Rizzi 1988; Grimshaw 1990; Dowty 1991). Other argument structure alternations with psych-verbs – of which the examples in (1) only represent a small part – have attracted much less attention. In particular, there has been little discussion of why a particular argument structure is chosen and why the alternations differ so strongly with respect to their frequency of occurrence.

The latter question has of course not been addressed because quantitative investigations in general led a marginal existence in the linguistics of the 1980s and 1990s with its strong tendency to separate competence and performance, system and use. This is currently changing, and the present article adheres to a conception of linguistics that assumes a strong connection between linguistic knowledge and language use.

1.2Frequency in usage-based linguistics

While to a certain degree, frequency data has always been discussed in linguistics with respect to language acquisition, language processing and language change, the assumption that frequency of use is an important factor in the cognitive representation of synchronic grammatical structure is a more recent development. Within usage-based linguistics, language is seen “as fluid and dynamic, changing through the interaction of social usage events with the cognitive processes characteristic of the human brain in general” (Bybee and Beckner 2010: 854).15 This has consequences for the grammatical system: “[…] language structure comes about through the application of a handful of common mechanisms that recur when human beings use language. The domain-general processes of sequential learning, chunking, categorization, and inference-making, along with the effect of partial or complete repetition, lead to the establishment and conventionalization of the categories and structures we find in languages. This bottom-up and emergentist perspective, we argue, may turn out to be indispensable to our understanding of linguistic processes and structure” (Bybee and Beckner 2010: 853).

This conception of linguistics requires weakening or even giving up on the separation between language use and linguistic structure in favor of a dynamic model of grammar (Diessel 2007: 123–124). Thus, recurrence and co-occurrence of linguistic expressions shape our linguistic system. One of the concepts that is closely attached to the recurrence and co-occurrence of expression is cognitive entrenchment. The entrenchment of a linguistic item or pattern into the cognitive system is strongly influenced by its frequency of occurrence and its frequency of co-occurrence with other entities or patterns. This presumes that speakers have linguistic knowledge that is based on a statistical assessment of the input they are confronted with. Ellis (2002) describes this from a psycholinguistic point of view: “[…] psycholinguistic studies of sentence processing show that fluent adults have a vast statistical knowledge about the behavior of the lexical items of their language. They know the strong cues provided by verbs, in English at least, in the interpretation of syntactic ambiguities. Fluent comprehenders know the relative frequencies with which particular verbs appear in different tenses, in active versus passive and in intransitive versus transitive structures, the typical kinds of subjects and objects that a verb takes, and many other such facts. This knowledge has been acquired through experience with input that exhibits these distributional properties and through knowledge of its semantics” (Ellis 2002: 160).

Similarly, corpus linguists who take a cognitive stance towards language argue: “It is common practice in corpus linguistics to assume that the frequency distribution of tokens and types of linguistic phenomena in corpora have – to put it as generally as possible – some kind of significance. Essentially, more frequently occurring structures are believed to hold a more prominent place, not only in actual discourse but also in the linguistic system, than those occurring less often” (Schmid 2010: 101). Thus, cognitively oriented corpus linguists “try to correlate the frequency of occurrence of linguistic phenomena (as observed in corpora) with their salience or entrenchment in the cognitive system. A corollary of this assumption is that patterns of frequency distributions of lexico-grammatical variants of linguistic units correspond to variable degrees of entrenchment of cognitive processes or representations associated with them” (Schmid 2010: 102).

Entrenchment has become quite a popular concept within cognitive and usage-based linguistics. In this paper, entrenchment will be viewed not so much as the absolute strength of the representation of a linguistic item in memory, but rather as a relative notion. I will investigate how strongly a lexical item is associated with a pattern in which it occurs. Thus, the paper addresses the question of whether and how strongly a verb is cognitively entrenched relative to an argument structure pattern and how strongly an argument structure pattern is entrenched with respect to particular verbs. This might be called relative, or associative entrenchment. It is of course undisputed that the relation between quantitative corpus data and cognitive processes as revealed by psycho- and neurolinguistic experiments is not a straightforward one. However, most linguists using the concept of entrenchment agree that there is a strong connection between quantitative corpus data and cognitive processes (cf. Schmid 2010, Blumenthal-Dramé 2012, Gries 2012b). Since I will not be discussing data from experimental studies in this paper, I shall leave the exact nature of the relationship between data from corpus and experimental studies open.

1.3Quantitative corpus studies on argument structure

The present article starts from the observation that each verb shows particular frequencies of occurrence with respect to its argument structure patterns and each argument structure pattern seems to attract some verbs more strongly than others. The basic assumption is that the observed quantitative distribution patterns can be accounted for, on the one hand, by numerous diverse linguistic factors and, on the other, by basic functions of the human memory system, such as the entrenchment structures as a consequence of recurrent use. I will assume that argument structure patterns constitute entities in our linguistic memory system that accumulate traces of use that determine their variation and diachronic dynamics.

Corpus-based quantitative studies on argument structure have continually – albeit rather infrequently – been published since the 1990s. A number of early studies collected frequency data in order to explain certain phenomena in language processing. MacDonald (1994) and MacDonald, Perlmutter, and Seidenberg (1994) demonstrate how the frequency of argument structures with particular verbs serves to resolve syntactic ambiguities in argument ...

Table of contents

• Cover
• Title Page
• Copyright
• Preface
• Contents
• Approaching German syntax from a constructionist perspective
• Part I: Grammatical constructions and valency
• Part II: Comparing constructions in German and English
• Part III: Prepositional constructions in German
• Part IV: Constructional Productivity
• Author index
• Subject index