Hatcher:
Archibald Hill (1902–1992), who organized the symposium, acknowledges with his last comment that ‘broken a law’ is as prescriptive as it gets. Chomsky would say it is not, because he meant a ‘natural’ law – and while today he might not apply the term law to the valency of a particular lexical item, he would still claim to be describing his native-speaker intuition here. But what about Veblen’s intuition? Or Hatcher’s, who finds it ‘unusual’ but not unacceptable? And it’s not about nativeness: Thorstein Veblen (1857–1929) was American born, the son of immigrants who spoke another Germanic language, and so was Chomsky.9
Proposition 3: Prescriptivism Inheres in Use, not Intent
Chomsky’s intent was never prescriptive, but so what? His ungrammatical sentences describe introspective judgements about what is or is not English or Chinese or whatever. Yet the effect is prescriptive, and we generally take the effect of an utterance to matter more than the claimed intent. If I am charged with making a verbal threat and say in my defence that I was only joking, the judge ain’t gonna dismiss the case out of hand.
The classic mantra of descriptivist linguists is that ‘the native speaker cannot err’. This is fine as an axiomatic methodological position, as long as one is prepared to accept variability. Chomsky’s problem in 1958 was his intolerance of grammaticality judgements that did not match his own,10 and not to see, as Hatcher did, how the sentence he was criticizing was evidence of language change. Chomsky might have objected that Veblen’s Theory of the Leisure Class was published in 1899, 29 years before Chomsky himself was born – but that only shows how languages do not change all in one go. That insight was far from unavailable to Chomsky; it was a long-standing tenet of linguistics that had just had an important new updating by John Fischer (1958). Changes in language always set up a choice with implications for speakers who do or do not take up the change – changes that are ‘political’ in the broad sense, since they become indexed for personal value judgements in a way that directly affects the speaker’s rank in the distribution of social capital and power.
Among these are judgements about ‘authenticity’ – who are the real speakers of the language – with all the knock-on effects that idea has for defining the rightful inhabitants of the place. Your intended descriptive analysis may in time serve as information on how people spoke back in 2020 – back when, from the perspective of 2070, say, Edinburgh was really Scots. Your description may get commodified, your data transferred onto tea towels and T-shirts and picked up in the speech of locals who want to perform (pace Chomsky) their localness. That prescriptive function can go on long after you are dead, and you have no control over it. The meaning of prescriptive and descriptive are in the use of your linguistic work – how it is interpreted and applied – rather than in your intention as an analyst, which no one else can know, only infer, in ways that will vary according to their own experience.
Proposition 4: Anti-Prescriptivism is based on an Impoverished View of Language
It implies that language is detached from people, that it is a code for transmitting information, commands, etc. that can be analysed without considering the interpretative freedom exercised by hearers and readers. Both message and speaker are interpreted, the latter indexically. A ‘hermeneiaphobia’ has always characterized linguistics, a fear and loathing of the notion of interpretation at the individual level (see Joseph, 2010). It is accompanied by a desire to contain interpretation, and maybe even control it. A classic example is Chomsky’s (1964: 7) analysis of how hearers interpret a ‘deviant’ sentence such as Colorless green ideas sleep furiously. Their mental grammar assigns a structural description that indicates the manner of its deviation from perfect well-formedness, after which ‘an interpretation can often be imposed by virtue of formal relations to sentences of the generated language’ (Chomsky, 1964: 9). In contrast, with Revolutionary new ideas appear infrequently, the hearer’s mental grammar assigns a structural description indicating that it is perfectly well-formed, and interpretation proceeds automatically. The thick concepts here include deviant obviously, but also well-formed. As for automatically and imposed, which do not have an empirical descriptive basis even on the hearer-response level as well-formed and deviant do, they may qualify as thin evaluative concepts, or at least show that the thick–thin distinction is scalar rather than binary.
Chomsky has revised this in his Minimalist program, where ‘we effectively dispense with the notion of “grammaticality”’ (Ott, 2010: 99), and where, says Chomsky,
‘deviant’ […] is only an informal notion. […E]xpressions that are ‘deviant’ are not only often quite normal but even the best way to express some thought; metaphors, to take a standard example, or such evocative expressions as Veblen’s ‘perform leisure’. […] The only empirical requirement is that SM [Sensorimotor] and C-I [Conceptual-Intentional interface] assign the interpretations that the expression actually has, including many varieties of ‘deviance’. (Chomsky, 2008: 10)11
Fifty years on and deviant is the new normal – but Chomsky still wants to contain it. The scare quotes acknowledge that it is a thick concept, yet that concession acts as a smokescreen, taking away from hearers and readers their freedom to see ‘deviant’ for the thin evaluative concept that it most probably is. He remains no less determined than in 1958 to have the ‘interpretations that the expression actually has’ be assigned rather than individually created. As for what ‘evocative’ may mean to him, I am at a loss even to guess. It appears to be one of those ‘informal’ notions he has just referred to. The Oxford Dictionaries define evocative as ‘Bringing strong images, memories, or feelings to mind’; does perform leisure evoke memories for Chomsky of his long-ago rejection of it as not English?
Here is a passage from probably the most widely used beginning linguistics textbook ever (Fromkin et al., 2014: 424),12 aimed at showing students why ‘prescriptivism’ is counter-natural, and therefore a ludicrous waste of effort.
| CHILD: | Nobody don’t like me. |
| MOTHER: | No, say ‘Nobody likes me.’ |
| CHILD: | Nobody don’t like me. |
| (dialogue repeated eight times) | |
| MOTHER: | Now, listen carefully, say ‘Nobody likes me.’ |
| CHILD: | Oh, nobody don’t likes me. |
The point is that the child will in his own time say what Mother is telling him to, but right now the child’s mental grammar is at a stage where Mother’s utterance could at best be parroted, not genuinely generated. Mother is being as silly as if she expected the kid to play a violin sonata or solve a calculus problem, and she gets her comeuppance when it turns out that her efforts have made the error even worse. The child is Laurel to Mother’s Hardy.
Such examples are powerful because they are part of our own experiences that we forget until they are pointed out to us, whereupon we can try them out on our own and usually get the same basic result. There are, though, aspects of the example which the textbook passes over, starting with Mother’s linguistic behaviour. It is ‘natural’ for mothers, across species, to ‘groom’ their children – a word that has taken on paedophilic overtones in recent years, but a concept that we cannot do without. Grooming is a genuinely universal practice, and inseparable from all the
