Implications of Early Word Learning for Theories of Meaning
Meaning is one of the central concepts in the philosophy of language. Despite, or maybe due to this central role, there is little consensus on the nature of meaning itself. Is meaning identical to a certain mental concept, or is it even external to the speaker? If it is a concept, what structure does said concept have? Is it a definition of some kind, or a mental image? A list of typical features or one of the many other alternatives? The discussion on meaning, and more specifically word meaning, contains all of the above options and many different arguments for and against all of them.
The discussion is rich and diverse in content to a point where it is sometimes hard to gain an overview of the relative strength of the different proposals in the race. I aim to move towards that goal of a better overview by confronting as many of the above theories of word meaning as possible with the same body of phenomena. In this way, we can determine how well each theory accounts for this body and thereby gain an impression of the strength of each position, at least a partial one relative to the body of evidence we discuss.
For this project, however, we need a body of empirical research in an area that is both, clearly enough established and relevant to as many theories of word meaning as possible. We find such a body of research in the psychological investigations of early word learning. Studies in this field empirically establish many phenomena in the beginnings of language learning that bear a relation to many theories of word meaning. In my thesis, I present the respective theories and phenomena and map out the implications of the latter for the former, with the goal of achieving a better overview of the different theories' respective ability to account for this body of evidence. This is a first step towards an overall better understanding of the argumentative strength of all the different theories of word meaning discussed in philosophy of language.