be instantiated in a combinatorial system, and therefore that a simple semantic network account of meaning is not
feasible.He is also properly concerned with issues of concept acquisition. We willcome back later to his proposals for
thecombinatorialstructureofconceptsand for thenatureofacquisition. For themomentI wishtoconcentratesimply
on the term Fodor has chosen for this combinatorial system: the Language of Thought (LoT).
Like“syntax,”“language”has a number of senses. First is the everyday sense: English, French, Chinese, and so forth.
Languages in this sense are systems composed of phonology, syntax, and semantics, plus the relations between them
established by interfaces (including the lexicon). Fodor clearly cannot have this sense in mind when he speaks of the
Language of Thought: LoT has no phonology, and no (narrow) syntax. Nevertheless, the ter mis so meti mes
interpreted in this sense. For example, one sometimes encounters statements to the effect that“Mentalese/The
Language of Thought is like a natural language”(e.g. Barnden 1996). This is misguided; it is comparable to saying,“A
wheel is likea bicycle.”Similarly, sometimes LoT is taken to be composed of“sentences in an inner/private language.”
But a sentence has phonology and (narrow) syntax; a thought doesn't. The same goes for the term“proposition”:ifit
is supposed to be the thought expressed by a sentence (so that one can express the same proposition in different
languages), it cannot be conceived of as sentence-like. The sentences that we use to express propositions have
phonology and (narrow) syntax; the propositions themselves don't. Part of our job here (questions (5) and (6) in the
previous section) is tofigure out the proper for mfor propositions, and how syntax and phonology express the min
different languages.
A second sense of“language”comes from thestudy offormal languages: a language is a setofexpressions and/or the
principles that generate them. For instance, we can speak of the strings in (8a) as expressions in a formal language
generated by the principles in (8b).
(8) a. ab, aabb, aaabbb, aaaabbbb,...
b. S→aSb
S→abThis sense of“language”includes the broad sense of“syntax”in Chomsky's use, but is not restricted to formal
organizations inthemind: formal logicin thissenseis a language, as are programminglanguages and possiblyeventhe
artistic style of Rembrandt. This sense is appropriate to the notion of“thought”as a combinatorial system in Fig. 9.1.
It is not, however, what Fodor has in mind.
Rather, there is a third sense of“language,”in which it denotes a set of expressions in a formal languageplusasetof
mapping principles that“interpret”these expressions into some domain. For instance, the sequences ofasandbsin
(8a)