Introducing English Linguistics

Why languages change
The development of English
39

---

In linguistics, there is considerable confusion over the use of the term
“evolution” to describe language change. Some linguists use this term
metaphorically, never implying a direct relationship between language and
evolutionary theories of biology but instead using the term to describe the
gradual changes that any language experiences over time. Other linguists,
in contrast, actually maintain that there are direct parallels between how
species and languages evolve. Croft (2000) proposes a “Theory of utterance
selection” that has its basis in Hull’s (1988) “generalized theory of selection,”
a theory that extends the traditional idea of natural selection to “all evolu-
tionary phenomena,” including, Croft (2000: 6) argues, language change.
Croft’s notion of utterance selection is heavily grounded in linguistic
convention, the idea presented in Chapter 1 that an utterance such as the
imperative sentence Leave is based on convention: a linguistic rule that stip-
ulates how this construction is formed as well as a pragmatic principle gov-
erning its use. As speakers of a language communicate, they engage in what
Croft (2000: 7) terms “normal replication” and “altered replication.” The
gradual replacement of whom by who, a process that is ongoing in English,
provides a useful illustration of the two types of replication. Before whom
started disappearing, speakers would have communicated with one anoth-
er with an utterance such as Whom do you trust? Normal replication would
have involved speaker after speaker using this form, passing it down gener-
ation after generation to children in the process of language acquisition. At
some stage, however, some speakers started saying Who do you trust? – an
example of altered replication, a deviation from the so-called “normal” way
of structuring this utterance. Because this change is currently underway,
both forms co-exist, the difference between them being primarily stylistic:
whom is more formal than who. And because both forms are being passed on
simultaneously, we even see instances of what Croft (2000: 121) terms hyper-