It is not surprising that the structure of complex words should guide us in their interpretation. What is perhaps surprising is the uniformity of this structure in English: no node ever has more than two branches, and the element on the righthand branch (whether a root, an affix or a word) is usually the head. What is more, the freedom with which complex structures can be embedded in larger complex structures, especially within compounds, provides great scope for the generation of new words; and, since lexical items are typically though not universally words, this freedom facilitates vocabulary expansion too.

Despite the general conformity of meaning with structure, there are occasions where meaning gets the upper hand, so to speak. French history and nuclear physics being institutionalized domains of study, we need terms to denote the people who engage in them; and, since the words historian and physicist exist, French historian and nuclear physicist come readily to hand as labels for the relevant specialists. This seems a good way to make sense of the mismatches. However these examples are to be analyzed structurally, their existence seems to show that, in derivation and compounding as well as in inflection, semantic pressures can sometimes enforce the existence of an expression with a certain meaning, and the expression chosen for that meaning need not be structurally ideal. The language’s acceptance of this expression, nevertheless, shows that, although word-structure guides interpretation, it does not dictate it.