Routledge Revivals: The Violence of Language (1990)

Jean-Jacques Lecercle

• Publisher: Taylor & Francis
• Year: 2016
• Print ISBN: 9781138200340
• eBook ISBN: 9781315514673
• Edition: 1
• Topic: Philosophy
• Subtopic: Historical & Comparative Linguistics

First published in 1990, this book argues that any theory of language constructs its 'object' by separating 'relevant' from 'irrelevant' phenomena — excluding the latter. This leaves a 'remainder' which consists of the untidy, creative part of how language is used — the essence of poetry and metaphor. Although this remainder can never be completely formalised, it must be fully recognised by any true account of language and thus this book attempts the first 'theory of the remainder'. As such, whether it is language or the speaker who speaks is dealt with, leading to an analysis of how all speakers are 'violently' constrained in their use of language by social and psychological realties.

1
Linguistics and the Remainder

Something about scouring, or scourging, he can't remember, and a teacher he once had who called his lectures 'lechers'.

(Robert Coover)

A Corpus of Texts

In Somerset Maugham's Cakes and Ale, a Cockney landlady utters the following sentence:

'I'm not so young as I used to was!'¹

Were I to read this in a student's paper, I would underline the last word in petulant red. As a reader of fiction, I enjoy the sentence. But what exactly is it that I enjoy in it? The elaborate narrative setting that enables Maugham to commit a solecism and yet get away with it? The picturesque character of the Cockney dialect at the turn of the century, and the accuracy with which the author of Liza of Lambeth, that notorious slum novel, adds a touch of local colour to enliven his otherwise blameless style? A first answer would be: a little of both.

It is clear that the sentence blatantly breaks an elementary rule of English syntax. Infinitives, not finite verbs, are immediately preceded by 'to', and the phrase 'used to' is therefore followed by an infinitive. The error is all the more enjoyable as it is obvious, a kind of syntactic malapropism. The Cockney landlady allows herself to do what I, as an 'educated' speaker of English, do not dare to do, and I secretly envy her. (Perhaps there is such a thing as sinning against language, which produces the same mixture of guilt and excitement as the common and garden variety.) Like all errors, this solecism requires an explanation - and it is easy enough to find one. You cannot expect an illiterate Cockney to speak like a young gentleman just down from Oxford, or, to put it more kindly, they speak different dialects, of unequal value in social intercourse but of equal interest to the student of language.

This explanation is superficial. It fails to stress what the dialects have in common: the narrator - and the reader - treat it as an English sentence, which is entirely meaningful for all English speakers. It may not be acceptable, but it is understandable. It receives an unambiguous semantic interpretation, and is therefore less problematic than many 'correct' complex sentences. No doubt this is due to the fact that its deviance from the equivalent correct sentence is minimal: merely a matter of substituting 'be' for 'was'. But this raises interesting questions. The flouting of a major syntactic constraint has turned out to be semantically negligible. If we operate within the Chomskyan tradition of the centrality of syntax, this is not what we would expect. For this rogue 'was' does breach a principle of syntax: there is only one verb inflected for tense per clause.

We can try to save the rules of syntax, and avoid the dilemma, by treating the so-called error as rational. The sentence is in fact perfectly acceptable if we consider that here, although not in our usage, 'used to' is not a verb but an adverb. If this is the case, the as-clause has only one finite verb, 'was', and everything is as it should be. We could explain the passage in terms of a portmanteau phrase, produced by the coalescence of 'as I used to be' and 'as I was'. Or we could treat it diachronically, as a case of conversion, where 'used to' has changed, or is changing, its grammatical category. In fact, there is every reason for it to do so. First, from the semantic point of view, suffixes of tense are close to time adverbials, all the more so if the verb that bears them has intrinsic temporal meaning (here the equivalent of a habitual aspect). The sentence might indeed have taken the form:

'I'm not so young as I was before (in the old days).'

It is irrelevant to point out that here the adverbial is placed after, not before, the verb, for it is easy to produce a sentence where it is in the right position:

'He wasn't so chirpy as he generally (usually) was.'

No doubt the existence of 'usually', an adverb phonetically similar to 'used to', has encouraged the conversion.

Second, there are grammatical reasons for the conversion. 'Used to' is a strange verb, which has always caused a good deal of uncertainty in grammarians. It is usually claimed that it behaves like a marginal modal auxiliary, although its meaning is aspectual. Mostly, however, it is a defective verb: there is no present form, even if there is no reason why we should not wish to refer to a habitual event in the present - indeed, this is one of the values of the present tense in English. Because of this semantic gap, the past suffix has tended to be neutralized, and is no longer always perceived as such, or even pronounced: /ju:zd t∂/ has become / ju:st∂/. The uncertain state of the suffix appears in the hesitation over the negative form: 'he used not to smoke* is now often replaced, so my grammar assures me,² by 'he didn't use to smoke' (where the non-existent root form, on which the present tense is based, makes its appearance) and even, more significantly, by 'he didn't used to smoke', in which the same 'error' as the Cockney landlady's seems to have crept into the language. But all is well in the end, since my grammar adds that this last form is 'nonstandard'. It also notes, nevertheless, that this uncertainty over the negative form of 'used to' impels the more astute speaker to avoid the problem by saying 'he never used to smoke'. But we can understand how a marginal and defective verb, which causes problems for everyday users of the language, is likely to turn into a more manageable word, an adverb.

This makes our explanation of this 'error' in terms of a different dialect untenable. This is no longer a case of another dialect of English having its own rationality and its own structures - one is reminded of Labov's description of the English of Harlem, and of his analysis of 'done' (in 'he done gone') as an auxiliary verb.³ Not only do we understand the sentence, we also understand the rationality of the so-called error. This is a case of language changing. Today's terrorist leader is tomorrow's prime minister: today's solecism is tomorrow's rule of grammar. What we first perceived as corruption has turned out to be conversion, a perfectly respectable device, one that every language resorts to. The difference with the more frequent case of nominalization is that this adverbialization is not available to the speaker as an expressive choice, but belongs to the diachronic movement of the system. Who can tell that this sentence will not be standard usage in three generations? Language, in its evolution, is opening up a new path. Maugham, who has the ear of the stylist, is one of the first to venture on it - he is certainly not the last. The reader, even if he perceives the sentence as a malapropism or a Cockneyism, enjoys the experiment. Is not its memorable character the most striking aspect of this sentence? We enjoy sinning against language because the violence we impose on its structure is what makes it alive. A solecism is not so much an aberration from the rules of universal grammar or the grammar of English as a (potential) anticipation of the evolution of structures, the 'universality' of which is strictly historical.

The following sentence belongs to a rare but entirely respectable syntactic pattern - there is no question of solecism:

A poem is a poem is a poem.

The pattern seems to have two uses: emphatic tautology, as is the case here, or equality, as in the sentence 'Crime is money is consideration', where the copula is the linguistic equivalent of the mathematical 'equals' sign.

The sentence is acceptable, and yet it violates a fundamental rule of English (and perhaps universal) syntax, embodied in the first rule of phrase structure in most syntactic theories: S → NP VP. This means not only that a sentence is composed of a subject noun phrase and a predicate verb phrase, but also that it is made up of only one of each. No one would think of writing down a rule of the type: S → NP VP VP. Yet this seems to be the structure of our sentence.

Of course, there are traditional ways out of this quandary. The obvious one is to distinguish the surface from the deep structure of the sentence. Surface structures can have the wildest appearance, but their very wildness can be shown to derive from perfectly simple and regular deep structures. For our sentence, the most obvious candidate is coordination. If the structure 'NP VP VP' is bizarre, the surface structure 'NP VP and VP' is normal. I can paraphrase my second sentence as 'Crime is money; it is also consideration.' The trouble is that this will not do the trick. Our pattern is not a case of coordination. The coordinated sentence is tame and banal, the equality sentence emphatic and striking. This appears even more strongly if I try to paraphrase the first sentence, for 'a poem is a poem and a poem' is totally meaningless. There is no comma or pause between the two verb phrases - it is a case of emphatic repetition, not coordination.

The detour through coordination, however, is perhaps not entirely useless. Coordinated tautologies can produce interesting expressive effects. Suppose I say: 'My butcher is a butcher. And he is also a butcher. But most of all, he is a butcher.' The utterance is no longer meaningless. It cries out for the computation of pragmatic meaning through Grician implicature. Am I not saying that my butcher lives exclusively for his profession, that a more dedicated butcher is hard to find, that for him butchering is no mere avocation but a true calling? Or perhaps the slightest hint of disapproval creeps in, and I am insinuating that he is as narrow- as he is single-minded (if you allow me this German turn of phrase). In other words, if my utterance violates certain pragmatic rules (embodied in the grammar of 'and', 'also', 'but', and 'most of all'), this is only an instance of exploitation that, far from disallowing any interpretation, creates new meaning.

I would like to suggest that the original sentence is an instance of the flouting of syntactic rules, exactly as the utterance about my butcher exploits pragmatic maxims. What we have in 'A poem is a poem is a poem' is not elliptic coordination (a new pattern within the existing rules) but the flouting of basic syntax, a case not of rule-governed, but of rule-breaking creativity - perhaps a case of playing what the French linguist L. J. Calvet calls rhythmic competence against syntactic competence.⁴ Indeed, the simple (and grammatical) tautology 'A poem is a poem' is a strange mixture of symmetry and asymmetry. For the naive user of the language, who follows his ear or his eye, the sentence is symmetrical: (a poem) is (a poem). But for the linguist it is deeply asymmetrical. The pattern NP VP, i.e. NP V NP, whether V is a transitive verb or the copula, has the following structure: (NP (V NP)), not a ternary but a binary, subject-predicate, structure. In the case of our sentence, it would have the form: (a poem (is a poem)). If we add another occurrence of 'is a poem', we only expand the illicit symmetry of our naive reading:

(A poem) [is (a poem)] [is a poem].

This is in the best style of the incremental repetition of Border Ballads. 'Incremental' is the right word, for the repetition threatens to become compulsive. There is no reason why we should not add another VP, as we do in 'a rose is a rose is a rose is a rose'. Only boredom will set a limit to the free-wheeling of this illegal syntax.

The allusion to Border Ballads is relevant in another sense. Repetition not only introduces symmetry — it creates rhythm, a tripodic rhythm (as one talks of the dipodic rhythm of ballads and nursery rhymes): 'A poém is a poém is a poém.' The sentence sounds like a jingle or a slogan and this is undoubtedly one of the sources of its force. Emphatic or playful repetition plays an important part in what Jakobson calls the 'emotive' function of language:⁵: 'No, no, no!', or 'his heart was going pit-a-pat, pit-a-pat'.

With this sentence, therefore, there is no question of an explanation in terms of solecism, local colour, or social dialect. Nor is it a case of language change or individual creative use. Since the pattern is conventional and productive (I can substitute any noun for 'poem'), it is a case of the English language tampering with its own rules. We might be tempted to interpret this in terms of modularity. There is a prosodic module in the grammar, which cooperates (and in this case interferes) with other modules to produce this apparently erratic pattern. The trouble is that in doing this we would stretch the concept of a module to breaking point. The crux is in the passage from 'cooperation' to 'interference'. For the cooperation of modules, rather like the combination of different forces, produces sentences within the acceptable structures favoured by the grammar of a language. Ours is a case of interference: the prosodic component plays with and against the phrase structure. From which we may draw two conclusions: (1) there is a phrase structure component (our pattern, by forgetting that there is one, reminds us of its existence, exactly as jokes about language locally and temporarily break syntactic rules); (2) structures can be violated for expressive purposes without producing gibberish or even solecism. In other words, the syntax of English, if I dare personify this august abstraction, is perfectly capable of treating its own rules as defeasible. Contrary to expectations, rules of syntax can be breached, like pragmatic rules and maxims.

The first sentence of Pride and Prejudice is, quite deservedly, one of the most memorable in the English language:

It is a truth universally acknowledged, that a single man in possession of a good fortune, must be in want of a wife.⁶

This first sentence is also the first paragraph, a motto inscribed on the threshold of the novel, a gnomic pronouncement which, in customary anticipation, spells out the moral of the tale and announces its ending. Except that, as a maxim, there is something wrong with it. At first, the reader is hardly aware of this; it requires the distance of a second reading to realize that there is a certain excess in the sentence, and this is where uneasiness is felt. Certainly, 'a truth universally acknowledged' is not an instance of pleonasm. Not all truths are acknowledged, fewer still universally acknowledged. Yet we may wonder at this insistence, in the first sentence of a novel, not of a treatise of logic or epistemology. This sense of excess is confirmed by the occurrence of the modal 'must', which, it seems, we must take in its acceptation of 'it is logically necessary that'. One might think that the presence of 'must' reinforces the meaning of the first words of the sentence. But such is not the case: this 'must' is not even redundant, it is counterproductive. A sentence reading 'It is a truth universally acknowledged that a man ... is in want of a wife' would have the required meaning of logical necessity. Substituting 'must' for 'is' implies falling into verbal excess, thus inducing the reader to reinterpret 'must' according to another of its values, obligation. The single man in possession of a good fortune is under the moral obligation of marrying a wife.

The reader is now even more uneasy, for this reinterpretation of 'must' makes the sentence incoherent on two counts. It is hard to see how a universally acknowledged truth can be an obligation, and the sentence is grammatically dubious. Most English modal auxiliaries possess two sharply distinguished meanings, often called epistemic and radical - in the case of 'must', logical necessity on the one hand, obligation on the other. In certain contexts, the modal is ambiguous, as in 'your brother must work very hard'. But not here. For there are constraints on the use of radical modals. They do not normally combine with aspects, because they behave like performatives, and perform speech acts (obligation, for instance), which cannot have retroactive effect. You cannot impose an obligation in the past, and 'your brother must hav...

Table of contents

• Cover
• Title
• Copyright
• Original Title
• Original Copyright
• Contents
• Acknowledgements
• INTRODUCTION
• 1 LINGUISTICS AND THE REMAINDER
• 2 THE RAG-BAG
• 3 A THEORY OF THE REMAINDER
• 4 METAPHOR
• 5 CORRUPTION
• 6 THE VIOLENCE OF LANGUAGE
• CONCLUSION
• Index