A word and its relatives: 

Relationships between lexemes

We discussed the words perform, performs, performed and performance. I argued that perform, performs and performed were grammatically conditioned variants of one lexeme PERFORM, but performance was not one of these variants. The reason was that, whereas there are grammatical factors that determine the choice between perform, performs and performed (in appropriate contexts), there is no grammatical factor that requires specifically the presence of -ance on performance. To put it another way: there are contexts where, if any verb appears, it must carry the third person singular suffix -s, but there are no contexts where, if a noun appears, it must carry the suffix -ance. The suffix -ance is not one of the small class of suffixes (so-called ‘inflectional’ suffixes) whose use is tightly determined by grammar. What sort of suffix is it, then? A short answer is that, not being inflectional, it must be derivational, since the term ‘derivation’ is used for all aspects of word-structure involving affixation that is not inflectional. The purpose is to put flesh on the bones of this purely negative definition, showing something of how derivation works in English.

Since performance is not a variant of the lexeme PERFORM, it must belong to some other lexeme, which may itself have more than one form. What lexeme could this be? This question is easy to answer when we notice that, alongside performance, there is a plural form performances. Just as cat and cats are the two forms (singular and plural) of the lexeme CAT, it makes sense to regard performance and performances as the two forms of a lexeme PERFORMANCE. This tells us something about the relationship between perform and performance: it is a relationship not between word forms but rather between lexemes. (Strictly, then, in terms of our typographical convention, we should call it a relationship between PERFORM and PERFORMANCE.) Thus derivational morphology is concerned with one kind of relationship between lexemes.

There are many ways in which lexemes can be related. We are not concerned here with relationships solely of meaning (such as the synonymy of AUBERGINE and EGGPLANT) or of sound (such as the homonymy of ROW ‘line of people or things’ and ROW ‘propel with oars’). Rather, we are concerned mainly with relationships involving affixation, and the grammatical and semantic tasks that such affixation can perform. As we will see, both the affixes and their tasks are quite diverse. An encyclopedic coverage of all the English derivational processes would be impossible, but I will attempt to supply a representative selection, so as to equip the reader to notice and to describe, with reasonable confidence, other processes not mentioned here.

I will introduce the term base for the partially complete word form to which an affix is attached so as to create either an inflected word form or a new lexeme. (Equivalently, the base for an affixation process is what remains if the affix is removed.) Some bases are roots, whether bound (e.g. wive-, the base for wives) or free (e.g. cat, the base for cats). Others, however, already contain a root and one or more affixes, such as helpful in its capacity as the base for helpfulness.