This use of the word norms here is inspired by Patrick Hanks’ recent book, Norms and Exploitations. It’s just the latest in a large number of articles, books and theories which have tried to account for the fact that despite an infinite number of possible sentences based on a limited set of innate grammar rules (essentially the Chomskyan account), actual language use tends towards often repeated chunks and phrases that avoid many possible grammatically correct sentences (see Pawley and Syder’s seminal article here). Corpus linguistics has repeatedly shown how there seems to be what Sinclair calls an idiom principle at work where speakers and writers align their usage towards each other resulting in particular phrases and collocation. Michael Hoey has argued quite convincingly that this kind of alignment in lexical choices is seen at a level beyond collocation. A word is typically chosen over a synonym because of the surrounding grammar, because of the position in a text, because of other words in the text and so on.
However, these choices are not absolutes. When Hoey shows his findings on his limited corpora, the results are usually much lower than 100% for one word over another. Indeed, Hoey argues that while we can say there are norms for certain words and phrases, everyone’s English is actually distinct in that we can’t share the exact same experience of language.
This also chimes with the tradition of descriptive grammars in English. Grammars such as Swan’s Practical English Usage or Biber et al’s corpus-based Longman Grammar of Spoken and Written English are explanations based on norms of usage. The “rules” of the grammar are NOT Chomskyan deep grammar rules, but are interpretations of meanings of common patterns in the language which, like words, are difficult to define and open to exploitation within limits (also known as ‘exceptions’).
We might also note that pronunciation is based on norms too, both in terms of individual sounds and what Richard Cauldwell calls the sound shapes of words. The phonetic symbols are not directly connected to one exact sound, but a number of slight variations. Similarly, whenever we hear the words such as ‘that’ or ‘big’ or ‘city’, they are a normalized versions of a limited variety of sound shapes.
Finally, note that here we are dealing with a description of how language is rather than the process of acquisition – something we’ll come back to at another time.
· Try to provide (more) examples of normal probable usage rather than less likely made-up sentences based on grammar rules.
· How far you or your students want to align themselves with the aggregated norms of English as exemplified in a corpus or of a native speaker accent is up to you – and them. However, bear in mind that in the heat of the classroom even accessing your own typical usage of words may be difficult and need planning. (See this post).
· We might use Google searches, corpora-based dictionaries, and other corpora or corpora tools to access or test our ideas of normal usage.
· We might try and imagine conversations and texts in order to access our ‘inner corpora of norms’.
· We might avoid stating absolute rules whether it is grammar or collocations. Saying often or normally may frustrate some students, but in the end that’s how English is.
As well as the links in the article above, here is an article by Richard Cauldwell illustrating variations in sound shapes.
This is a Powerpoint by Michael Hoey – characteristically long – which gives some examples of ways in which he believes word choice is primed.