This is the frontpage of my Ph.D. dissertation. Here you find short introductions to each of the five chapters, and below
is the Abstract of the thesis. The complete text is available for download.
Please refer to the thesis as:
Winter, Simon. 1998. Expectations and Linguistic Meaning. Ph.D.
thesis. Cognitive Science Department, Lund University. http://www.lucs.lu.se/People/Simon.Winter/Thesis/
Please do not quote from the web pages, but only from the downloaded
What is the relation between the words in language and our everyday actions?
Are linguistic structures dependent on our actions, or does language function
on its own?
This thesis deals with the pragmatic foundations of language and proposes
a model of meaning in language that is based on our expectations about the
world and about other people. On this view, language is seen as composed
of three functional layers of pragmatics, semantics and morpho-syntax, with
each layer having a certain autonomy: semantics captures useful generalizations
from the pragmatic level, and morpho-syntax captures generalizations from
the semantic level.
The thesis consists of an introduction and five
papers that examine different aspects of this overarching model.
- The first paper explores what happens when individual
cognitive structures are shared in language and proposes a model of how
names, nouns and adjectives emerge as different levels of abstraction.
Further, some cognitive prerequisites for referential communication are
- The second paper proposes a model of how breakdowns
in an expert-novice task trigger discourse on the levels of instructions,
coordinations and labels (words). These levels are correlated with a gradual
conventionalization from pragmatics to semantics.
- The third paper shows how the modal verbs can
be modeled in terms of expectations and social power.
- The fourth paper argues that meaning in language
can be traced back to features of the environment of all living creatures
that are inherently meaningful. Examples of such features are food, mates
and shelter. The space of meaningful features can be extended by categorization:
previously meaningless features gain meaning by association with meaningful
- The fifth paper, finally, uses computer simulations
to model the pragmatic process of stabilizing a simple lexicon.