How do children know what is meant by words without a constant general content like he or she, this orthat, some or any, given that the meaning of such words has to be inferred from preceding utterances? How can children learn to use such words to maintain focus on one specific thing or person out of a presumably great number of persons and things mentioned in the previous discourse? The research project is concerned with the basic knowledge types which the child has to acquire in order to keep track of the referents meant by the communication partners. Consider the following examples from German:
a. Der Vater traf gestern seinen Bruder. Er hat sich sehr gefreut.
'The father met his brother yesterday. He was really happy.'
b. Der Vater traf gestern seinen Bruder. Der hat sich sehr gefreut.
'The father met his brother yesterday. This one was really happy.'
German speakers preferably use the pronoun er in (a) to talk about the subject of the clause, the father. The pronoun der in (b) signals that they will go on to talk about the less central referent, the brother.
The single steps and factors involved in the acquisition of such pronoun resolution strategies call for a more profound investigation. This will provide useful information about the process of language acquisition and enlarge the empirical basis of linguistic theory. Contextually referring entities (anaphora) present the theoretical linguist with an opportunity to investigate the interaction of syntactic, semantic and pragmatic factors in the organisation of the language system. Anaphora research gains in importance through the practical problems it has to solve in natural language processing, e.g. machine translation. The correct depiction of referential dependences established in complex texts and dialogues presents a challenge not only in the domain of computer implementation, but needs a deeper understanding also in the field of human language acquisition. Even for the purpose of a simple conversation children have to sort out who or what is being talked about, and what are the appropriate expressions to go on with the topic. The use of anaphoric means becomes complicated by the complex interactions of the numerous factors involved. Language acquisition studies and especially comparative studies can contribute to our knowledge of the basic factors and strategies guiding the anaphoric reference establishment.
Central to the project investigation are comparative experimental studies conducted with German, Russian and Bulgarian children at the age of 3 to 6 years. At the current stage, we concentrate on monolingual children acquiring German, Russian and Bulgarian as a first language. Based on the findings about the mechanisms of anaphoric use in monolingual children, the second stage of the project will target the co-reference establishment strategies of bilingual children.
The theoretical discussion of adult anaphora resolution brings to the fore the following main factors. These are proposed to promote a single referent out of those available from the immediately preceding utterance: the syntactic role of the referent (subject vs. object), its animacy status, its information status (old or new information) or its syntactic or semantic parallels to the anaphoric pronoun. We seek to provide answers to the following questions:
1. Which of these factors are operative in child anaphora resolution? From which age on do these factors constitute a part of the "anaphora" knowledge of the child?
2. Are these factors relevant for the anaphora acquisition process in all three investigated languages or are there language-specific differences? If there are any, how can they be explained?
3. Which general steps can be singled out in the acquisition of the investigated pronominal anaphora? Which factors and strategies are the basic ones?
4. Adults are thought to refer to the most salient referent using the formally simplest (shortest) anaphor but to referents which are not at the focus of attention by more complex means. Do children employ such a reversed mapping between referent's salience and pronominal complexity?
At the second stage of the project, we will focus on two main research topics: detection of differences between monolingual and bilingual anaphora acquisition and verification of techniques and factors found to be relevant in monolingual co-reference establishment. We plan to compare bilingual anaphora acquisition in the language pairings German/ Russian and German/Bulgarian. The basic questions remaining the same, particular effects of factors operative in monolingual anaphora resolution will be investigated in greater depth against the background of bilingual language acquisition.