mundane level, we sometimes have dreams in which someone who looks like one person is“really”someone else (I
was talking to Uncle Harold but for some reason he looked like Bill Clinton). In such cases a mind/self/soul has been“dotted”
with a different body.
The“dotted”aspects of speech acts are each connected to the appropriate“dotted”aspects of the speaker: the body
makes the noise, thereby conveying the information“contained in”the mind. Interestingly, social roles, which are
abstract concepts in the mental/social domain, are often embodied in the physical domain by“dotting”the mwith
aspects of physical appearance such as uni-forms, costumes, or haircuts. These physicalaspects thus become symbols
of the social role: a symbol is one object regarded as simultaneously something else, hence yet another sort of dot-
object.
It remains as a challenge for future research to work out the details of dot-objects and their qualia structure, and to
sort out which aspects are innate (surely some of them must be) and which are culturally determined. In particular,
what kinds of ontological entity can be“dotted”with each other and how can they be connected in qualia structure?
Are the possibilities totally open, or is there a delimited set (say several dozen or a hundred) of dot-object types?
However the answer turns out, it raises further questions of principle and of detail for learning and evolution.
11.11 Beyond
I want to touch briefly on one further layer of complexity in lexical semantics before closing this chapter.
One of the classic examples of lexical decomposition isbachelor, analyzed by Katz and Fodor (1963) as‘unmarried
adult human male.’This is often taken as a parade case of analytic decomposition. But consider the putative feature
‘unmarried’: what sort of feature is thisanyway? It is approximately on a par with[HAS-A-BACK] in theanalysisofchair.
It is certainly not primitive, but any attempt to further decompose it requires a whole analysis of the social institution
of marriage. That is, the meaning ofbacheloris inseparable fro mthe understanding of a co mplicated social fra mework
inwhichitis embedded. Lakoff (1987) makes muchofthisfact;heobserves, for instance,thatitis odd tospeak ofthe
Pope as a bachelor, because the Pope falls outside the social institutions in which marriage is an available option. In
addition, of course, it is odd to consider an adult unmarried man living in a long-term relationship with a woman as a
bachelor. Obviously‘unmarried’is not a sufficient condition. The upshot is that the prospects for a simple analytic
feature decompositionforbachelordissolvein favor of somethingwitha great deal more implicitstructure, in particular
overall understanding of larger social frames.