Date: Thu, 23 Jan 2003 17:19:24 +0100 From: Johannes Reese Subject: Learning the Meaning of Change-of-State Verbs: A Case Study of German Child Language
Wittek, Angelika (2002) Learning the Meaning of Change-of-State Verbs: A Case Study of German Child Language. Mouton de Gruyter, viii+233pp, hardback ISBN 3-11-017304-2, Studies on Language Acquisition 17.
DESCRIPTION OF THE BOOK The book deals with how children learn the meaning of causative change- of-state verbs in German. It consists of seven chapters.
The first chapter covers the literature in the field of change-of-state (Aktionsart) and language acquisition. She bases her thoughts on the aspect/aktionsart literature derived from Vendler (1967), which has been extended by Dowty (1979) -though she does not cite the latter, and used by different other researchers, some of which are cited in her work. Wittek mentions the lexical decomposition scheme invented by Dowty. She also drops a few words on the role arguments play in that circumstance.
More relevant seems the link to Talmy's 1985 work on the typology of motion verbs. The lexical decomposition approach is used as a tool for language acquisition theory in order to explain differences in argument marking and conceptualization deviances among young children like marking the object with 'throw' earlier than with 'read' or mistaking the words buy and sell for give and take; the Semantic Feature Hypothesis is the anchor between language acquisition theory and this theory of the lexicon. In recent theories of grammatical development and lexical structure, the causative change-of-state-verbs, the topic of the book, are considered to be basic, and thus easy to learn. Language acquisition research, on the other hand, has shown that this is not the case. At least certain phases of these verbs are often understood very late. This paradox forms the challenge for Wittek.
Curiously, Wittek presents the structuring of the book and the acknowledgments at the end of her Chapter One, after having described the main theoretical background.
The second chapter deals with the language acquisition research. In the chapter, some non-acquisitional issues are touched as well. E.g., a distinction is made between manner verbs and result verbs.
In the one-word stage, children tend to content themselves with the expression of completions of events. In satellite-framed languages, they use the verbal particles rather than the verbs. They also tend to use preterite forms with completive verbs and present forms with non- completive ones. On the other hand, as another research has shown, children (apparently English-speaking ones, at this point, language specifics are dropped from attention) have the more difficulties to understand the end-state component of a change-of-state verb the younger they are (Manner Bias). Other (newer) research, however, came to different results, they found no or only a slight Manner Bias.
She poses this scheme as the basis for her main idea: testing if the Manner Bias is a typological one; she does so by opposing complex predicates (with both result satellites and manner verbs) to mere result verbs. If children have more problems with the latter than with the former, the Manner Bias would turn to be a typological one, according to her (Transparent Endstate Hypothesis). She uses German as target language instead of English due to the stronger Romance influence on English.
In Chapter 3, this first experiment is presented. The chapter begins with a deeper explanation of what she wants to test. Fully transparent satellite constructions as opposed to semi-transparent or non- compositional ones are for her constructions with a manner verb and a satellite that can by itself serve (with a copula) as the predicate of the end state. Those constructions should be more reluctant to negligence of the end state than mere verbs containing an end state, a kind of a deviation to the satellite frame.
After that, she gives a very detailed description of her experiment settings, which seem quite sophisticated - as do all of the following ones. Her results were that it is possible to reproduce to a limited degree the findings that lead to the Manner Bias Hypothesis, but these are due rather to "learners' problems in interpreting certain form- meaning mappings", e.g. a focal saliency bias, than to a general predisposition to learn manner verbs easier than result verbs.
In Chapter 4, she reconsiders the results of Chapter 3 in more detail; she presents some of the interviews in order to get a deeper insight into why the children acted as they did. Apparently, the children consider the endstates as an entailment of a verb's meaning. It is the starting point for a brilliant and well-informed discussion of Aktionsarten. There is a aspectual gap in the system of the children that makes them treat "perfective" telic verbs as conative. She cites research from other languages (Japanese, Chinese, Tamil), where result verbs do not necessarily entail the result to be achieved. That's the system of the young children, too. They turn out to be a class of verbs that pragmatically favor the achievement of an endstate, but do not necessarily entail them, similar to the weak endstate verb languages. In English and German, there is such a class within manner verbs.
She then discusses several approaches as to the question how children finally learn the correct entailments of change-of-state verbs in Germanic languages. Forgetting about some hints from chapter one (as to different morphological role marking) she states that children don't distinguish between different roles due to different Aktionsart and argues that there wouldn't even be any overt cues for them.
Chapter 5 tries to solve this question. It adds another idea to aspectological theory as well, using modifiers (i.e. adverbs) as a test. She presents the Adverbial Modification Cue Hypothesis, using the adverb 'again', which can, in one reading, refer to restitution of a previous state and shows thus in due case that the end state is connected to the verb's meaning (indefeasible). German children are aware of this reading; therefore Wittek argues that they can defer the relevance of the end state by this adverb. Again, she supposes a Transparency bias, i.e. that verbs with the result as a satellite would be used more easily than simple verbs in combination with 'again', at least in German. That is what corpus studies told her. She then tests this hypothesis empirically. Her prediction proved to be correct.
In Chapter 6, Wittek examines her hypothesis further. She presents an experiment testing if children can learn novel verbs by the clue of the adverb 'again' in restitutive reading. As a result, Wittek managed to teach children the change-of-state component of novel verbs by the cue of the adverb 'again'. They learnt the new verbs as change-of-state verbs more often than in the control groups.
Chapter 7 is summary of the whole book. The settings are presented in more detail as appendices.
EVALUATION The merit of the book is based on two fundaments. At first, the overt structure is very comfortable. Every chapter ends with a summary, and there is an overall summary at the end of the book as Chapter 7. This makes the book very good to read.
At second, the research presented is very sophisticated. It is based on another two fundaments: Aktionsart theory and language acquisition research. In the former, there is the almost one and only critical remark to be made: she tends not to cite the "inventors" of a tool, but works who used it: example: there is no hint to Davidson (1967), but she mentions his event argument. Nevertheless, her reflections don't cover a wide range of Aktionsart literature (cf. e.g. Sasse 2002), but the ones she decided for are came out in a research very fruitful to aspectual and Aktionsart theory. And she uses useful hints from other languages. Her deep (Aktionsart research based) reflections helped her detect mistakes in other language acquisition based works.
Summing up, the book is a step forward not only in language acquisition research, but as well in Aktionsart theory. It is worth reading also for people without any interest in both German and language acquisition. One of the key merits for them is her discussion of pragmatically favored endstate verbs.
I have especially liked the fact (though it is of course only partly or not at all determined by her will) that some of the experiments failed; although it forced her to do more experiments, it was a step (or she made it look like a step) towards finding the points that aspectologists are interested in.
One final critical remark: end notes are a nuisance.
REFERENCES Davidson, Donald (1967): The Logical Form of Action Sentences. In: Rescher, N. (ed.): The Logic of Decision and Action. Pittsburgh: University of Pittsburgh Press. 81-95
Sasse, Hans-Juergen (2002): Recent activity in the theory of aspect: Accomplishments, achievements, or just non-progressive state? Linguistic Typology 6.2(2002), 199-271
ABOUT THE REVIEWER:
ABOUT THE REVIEWER Johannes Reese is a research fellow at the Universitaet Zuerich, Switzerland. He is currently writing his dissertation on aspect in Moroccan Arabic. His main interests include above all fields relevant for facilitating language learning, currently especially aspect, genericity, focus, and formal semantics.