Journal of Pragmatics 29, 1998, 1-39

Performative speech acts in linguistic theory:

The rationality of Noam Chomsky

Robert de Beaugrande

Maybe everything I’ve done is wrong; I’m not God.

— Noam Chomsky (in Achbar and Wintonick 1992)

Abstract

The discourses of philosophy and science have often revolved around the problems of relating theory with practice, and rationality with materiality. Yet the problems have been evaded more often than solved by ‘speech acts’ which purport to be strictly ‘constative’ (this is the real truth because the universe is like that) but which are mainly ‘performative’ (this is the real truth because we say so and we are the most rational). In the discourse of ‘modern’ linguistics, the question of whether the ‘reality’ of language is mental, material, or social has been evaded by a performative campaign to replace real language with ideal language and to short-circuit mental with material whilst bypassing the social basis of language. A leading idealiser in this evasion has been Noam Chomsky, whose discourse has implied large claims to a hold a superior rationality.

A. ‘Constative’ and ‘performative’ in philosophy and science

1. For analysing the discourses of philosophy and science, ‘speech act theory’ might offer a bridge metaphor. In ‘classical’ agendas, these discourses are officially dominated by the ‘constative’ mode for conveying information and declaring the truth about reality, but unofficially by the ‘performative’ mode for constructing reality and claiming authority. Officially, disagreements occur only because the sole and complete truth is still unattained, or because individual philosophers or scientists are misled by personal interests or relativism (cf. § 84, 103). ‘Non-classical’ or ‘post-classical’ agendas are now seeking to reinterpret philosophy and science as performative enterprises, wherein a ‘theory’, ‘description’, or ‘explanation’ emerges from performative speech acts deriving authority from a superior rationality than your competitors or predecessors (cf. § 4f, 35).

2. Recalling the ancient ‘trivium’, we might say that classical philosophy and science propose to found their rationality upon their logic whilst actually founding it upon their rhetoric. They can exploit the long history of idealising human cognition and communication to be essentially logical and only accidentally (and perhaps regrettably) rhetorical. Within this same history, many projects have sought to reform and rationalise human language and especially grammar, the third member of the ‘trivium’, by merging it with logic (cf. § 5, 10, 36). Despite their inevitable failure, these projects have bequeathed us a curious detritus of prescriptions and proscriptions in schoolbooks, grammar-books, and so on about how we should and should not think or speak; and a deep-seated mistrust that everyday thinking or speaking are not ‘correct’ or ‘grammatical’.

3. The discourses of ‘philosophy’, defined as ‘the rational investigation of the principles of truth, being, or conduct’,¹ propose not merely to proceed rationally but also to establish the very foundations of human rationality. A key question has been whether rationality should be derived from inward introspection or from outward reflection of the material world, whence the venerable gallery of polarities: ideal versus real, internal versus external knowledge, mind versus body, or subject versus object, plus their respective ‘modelling styles’:² idealism versus realism, subjectivity versus objectivity, rationalism versus empiricism, mentalism versus physicalism, and cognitivism versus behaviourism. Philosophers with ambitions to make one side of these polarities definitively triumph over the other enact strenuous performatives purporting to be constatives, as in radical idealism or radical realism:

the various sensations or ideas imprinted on the sense, […] whatever objects they compose, cannot exist otherwise than in a mind perceiving them. (George Berkeley)

of these languages, the physical, or that in which we speak about physical things, in everyday life or physics, is of the greatest importance. […] physical language [is] a universal language comprehending the contents of all other scientific languages. (Rudolf Carnap)

Still, radical idealists and realists coincide in the ambition to establish rationality by deploying their own rationality to determine the ultimate status of materiality and to show that the universe is ordered by rational principles. Either rationality is not encountered in the material world but only in the ‘mind’, e.g., through innate concepts; humans reconstruct the world in the mind and regard experience and observation as agglomerations of accidents. Or, all material is organised according to rational principles humans can discover by extending and refining our experience and observation, e.g., of the architecture of galaxies, stars, and planets.

4. The ambition of philosophy for superior rationality have been inherited and modified by ‘science’, defined to subsume any ‘branch of knowledge or study’ dealing with ‘objectively measurable phenomena’, producing ‘a body of facts or truths systematically arranged, and showing the operation of general laws’.¹ As this definition indicates, the major modification resides in the ambition to reconcile rationality with the materiality of ‘phenomena’ by discovering the ‘facts’, ‘truths’, and ‘laws’ which govern materiality on a plane that is not accessible either to ordinary experience or to philosophical introspection. Science has therefore developed material technologies to enhance its rationality about the universe without having to resolve philosophical disputes over whether the universe itself is rational. Science seeks to maximise its own rationality for approaching the universe from its material side and imposing specialised rational controls upon experience and observation..

5. These divergent ambitions assign differing roles to discourse. In philosophy, most great achievements have been performative discourses whose authority may persist for hundreds or thousands of years. Yet many philosophers have mistrusted whether language and discourse themselves are sufficiently rational, and have sought to explain language itself as a rational system, e.g., by integrating logic with grammar (§ 2, 20, 36).

6. In science, great achievements are rarely performative discourses with long-term authority but rather ‘discoveries’ of ‘laws’ that are universally true, quite independently of the language or discourse used to explain them. Many scientists seem to mistrust language and discourse far more than philosophers do, and try to state their results in alternative modes of representation, such as numbers, charts, tables, formulas, and equations (cf. § 39ff, 77). Philosophers know things only if they can explain them in language; scientists may know things they cannot explain in language, e.g., when ‘superstring theory’ is ‘founded upon elegant mathematical ideas’ and ‘promises to provide a unified description of all forces, all the fundamental particles of matter, and space and time’ (Davies and Brown, 1988: vii).

7. ‘Classical’ science has been an enterprise for observing and explaining reality without having to observe or explain the human operations of observation and explanation, either in ordinary experience or in the academic pursuits of philosophy and science. Thus, ‘the ambition of Newtonian science was to present a vision of nature that would be universal, deterministic, and objective inasmuch as it contains no reference to the observer, and complete inasmuch as it attains a level of description that escapes the clutches of time’ (Prigogine and Stengers, 1984: 213).

8. Perhaps science has drifted apart from philosophy over the last two centuries to evade uncomfortable problems and paradoxes inherent in this ambition and to dissociate itself from the more performative and self-reflecting rationality of philosophical discourse. Whereas classical science accepts the possibility of true knowledge as its fundamental principle from the start, philosophy treats it as an open question to be deliberated from all angles, including, but not being confined by, the disciplined epistemology of science. Scientists who base their knowledge and methods upon refined technological probes of materiality and upon a sparse and sceptical rationality for interpreting the results, can officially highlight external knowledge of the ‘real’ and consign internal knowledge of the ‘ideal’ to the charge of philosophers. A ‘scientific theory’ can purport to be a free-standing construct, independent of the human practices that have produced and advocated it.

9. The ratios between the ‘real’ and the ‘ideal’ and between theory and practice have therefore rarely been an explicit topic in classical science. This topic was eventually addressed by the enterprise called ‘philosophy of science’, whose English name (as compared, say to German ‘Wissenschaftstheorie’) might sound portentous or even paradoxical. This important work has not yet fully addressed the performative discursivity of both philosophy and science. We need more detailed explorations of how ‘new paradigms’ ‘incorporate much of the vocabulary and apparatus, both conceptual and manipulative’, even whilst ‘old terms, concepts, and experiments fall into new relationships’ (Kuhn, 1970: 49).

10. Intriguingly, Kuhn’s book appeared under the auspices of The International Encyclopedia of Unified Science, an enterprise inaugurated long before to ‘unify’ philosophy and science within a stringently ‘classical’ mould, such as ‘physicalism’ and ‘logical positivism’ (cf. Neurath, Bohr, Dewey, Russell, Carnap and Morris, 1938) (cf. § 22). The relation between real and ideal was not so much ‘unified’ as short-circuited: the version of reality projected by this ‘science’ would directly establish which ideas are admissible and what they refer to. Materiality and rationality would thereby converge in a ‘dual reality principle’: physical reality matches and certifies mental reality, at least in the reality addressed by philosophy and science (cf. § 19, 23, 26, 45). This certification would be relayed in ‘physical language’, where Carnap (cited in § 3) could exploit a strategic pun between general language talking about the physical world ‘in everyday life’ versus the specialised scientific language used ‘in physics’. The imposing prospect arises, though I haven’t seen it stated in these terms, of the physical universe itself being a language or discourse which gets most successfully translated or recoded into human knowledge and human language by science, especially by physics (cf. § 22f, 75).

11. ‘Unified science’ renewed the enterprise of constructing models of human cognition and communication upon idealised analogies to the activities of scientists and philosophers, who thus become the ideal prototypes of human beings, though they might well hesitate to say so (cf. § 65). These models project a world of idealised agents possessing rational, full, and correct knowledge of a material reality that determines just which statements can and should be uttered and just what they mean. With a profound irony, idealisations can assume the guise of ultimate realities to the point where actually observed realities appear deficient and deviant (§ 63, 67), even though the observation of reality is the official cornerstone of classical science (§ 7).

12. A ‘non-classical’ or ‘post-classical’ programme could investigate how the cognition and communication of humans in general and of scientists and philosophers in particular construct, negotiate, and advocate concurrent realities. ‘Scientific knowledge’ and related concepts like ‘theory’ and ‘explanation’ can be directly probed as predominantly discursive activities for navigating and mediating between real versus ideal, external versus internal, material versus rational, universal versus specialised, deterministic versus non-deterministic, and objective versus subjective (Beaugrande, 1997).

13. Returning to speech act theory as a bridging metaphor (§ 1), the discourse of classical scientists and philosophers can be said to highlight locutionary acts conveying content and reference and holding the constative status of true reports about reality. These acts officially underwrite illocutionary acts like building or refuting an argument and perlocutionary acts like convincing a community of scientists to adopt a new theory. The constative is projected to dominate and authorise the performative, whilst the physical reality of the material world, having been relayed into the mental reality of scientists’ knowledge, is rationally relayed onward into the truth of scientific discourse.

14. Yet viewed from a non-classical perspective, the discourse of science continually merges locutionary, illocutionary, and perlocutionary, and adapts the constative to suit the performative rather than vice-versa. The discourse is not a relay from reality into truth, but rather a complex construction of adaptive approximations whose authority must be discursively established and maintained. The reconciliation between rationality and materiality remains problematic in principle; the discourse seeks to address the material in ways that count as rational within the current ‘paradigm’ of the science. Performative strategies for displaying rationality imply a shared neutrality for judging the alternatives by their own merits, even whilst your own discourse is constructing and adjusting those merits to your own advantage.

15. Insofar as scientific discourse has to bootstrap its rationality while appearing to take it as given, irrational implications portend, and classical scientists (or ‘normal scientists’) live” at least subconsciously, with paradoxes rather than work out the implications of multiple realities and multiple approximations. The sole truth will be gradually but reliably attained through steady increases in rationality; and yet everyone else but you or your group has been missing the truth so far. The truth is destined by science and rationality to triumph over untruth and irrationality; yet this triumph had to await your own explanation. Paradoxes.

B. The reality of language in the science of language

16. We can now explore how the ‘science of language’ known as ‘modern linguistics’ has deployed discourse to seek accreditation and authority (detailed survey in Beaugrande, 1991). The immense scope and pervasive presence of language in human activities has animated linguists to regard their science as a central and exemplary one: ‘among all the human sciences, linguistics has been the one science whose scientificity is given as an example with a zealous and insistent unanimity’ (Pennycook, 1994: 23f, citing Derrida, 1974: 281). Already, Saussure’s inaugural Course in General Linguistics  declared: ‘language, the most complex and universal of all systems of expression, is also the most characteristic; in this sense linguistics can become the master-pattern for all branches of semiology’ (1966 [1916]: 68). Hjelmslev (1969 [1943]: 78) even contemplated ‘regarding all science as centred around linguistics’.

17. Yet ironically, the performative discourse of linguists from Saussure onward has often sought rationality by reducing the scope and disconnecting language from human activities. Saussure himself observed that ‘in the lives of individuals and societies, speech is more important than anything else’, but he proclaimed that ‘the true and unique object of linguistics is language [French, ‘langue’] studied in and for itself’, such that ‘language will stand apart from everything else’ ‘as a well-defined object in the heterogeneous mass of speech facts’ (§ 25, 72); ‘language is not a function of the speaker’ and ‘exists perfectly only within a collectivity’ (1966: 13f) (cf. § 40). The irony was intensified when Hjelmslev’s (1969 [1943]: 3, 18) Prolegomena declared that ‘language is inseparable from man and follows him in all his works’, yet advised ‘linguistic theory’ to ‘foresee all conceivable possibilities’, including ‘languages that have perhaps never been realised’; such a ‘theory cannot be verified (confirmed or invalidated) by reference to any existing texts and languages’ (cf. § 21, 59f). Citing the physicalist philosopher Carnap (§ 3) for a ‘sign theory’ wherein ‘any semiotic is considered as a mere expression system without regard for its content’, Hjelmslev counselled that ‘a linguistic theory which searches for the specific structure of language through an exclusively formal system’ ‘must seek a constancy, which is not anchored in some “reality” outside language’ (1969: 110f, 8, h.e. [= his emphasis]).

18. These performatives signal how linguistic theorists aspired to construct a reality inside language. They might have drawn upon philosophy (e.g. Humboldt, Vossler), had they felt less self-conscious about maintaining a rational science. Saussure’s review of the ‘stages’ through which ‘the science that has been developed around the facts³ of language passed’ ‘before finding its true and unique object’ (1966: 3), didn’t even mention philosophy, though ‘grammar’ was censured for ‘lacking a scientific approach’ and being ‘based on logic’ and ‘far removed from actual observation’. Soon after, Sapir (1921: 154) portrayed ‘philosophers’ seeking ‘a theory constructed on “general principles”’, which ‘lies beyond the demonstrable’ and ‘is of no real interest’ ‘to linguistic science’. Bloomfield (1933: 3, 270) pointedly contrasted ‘philosophical reasoning’ against ‘scientific language study’ and opined that ‘a good deal of what passes for “logic” or “metaphysics” is merely an incompetent restating of the chief categories of the philosopher's language’. For Hjelmslev (1969 [1943]: 6f), ‘the theory of language must not be confused with the philosophy of language’ lest ‘attempts to form a linguistic theory’ be ‘discredited’ ‘as empty philosophising and dilettantism, characterised by apriorism’. Firth in turn chided the ‘philosophically pretentious’ quality of ‘traditional grammatical categories’ — ‘a nuisance’ ‘out of harmony with the general scientific theory and practice of today’ — and briskly foresaw ‘general linguistics supplanting a great deal of philosophy’ ‘during the next fifty years’ (1964 [1937]: 87; 1959 [1949]: 168) (but cf. § 36).

19. Yet philosophy seems to resemble history in George Santayana’s memorable aphorism: those who forget or ignore it are doomed to repeat it. Linguistics had to start over on the old problems of ideal versus real and of rationality versus materiality. The ‘mentalism’ that predominated in early stages engendered its own ‘dual reality principle’ (cf. § 10, 23, 26, 45): language has a primary mental reality and a secondary material reality that linguistic theory can invoke without having to specify. Vowing that ‘language is a form, not a substance’, Saussure admonished: ‘all the mistakes in our terminology, all our incorrect ways of naming things that pertain to language, stem from the involuntary supposition that the linguistic phenomenon must have substance’ (1966: 122). Yet he retained the material reality of language: ‘there are no linguistic facts apart from the phonic substance cut into significant elements’, and ‘abstract entities are always based, in the last analysis, on concrete entities’ (1966: 110, 138). Ambivalent too was his pronouncement that ‘word order is unquestionably an abstract entity, but it owes its existence solely to the concrete units that contain it and flow in a single dimension; to think that there is an incorporeal syntax outside the material units distributed in space would be a mistake’ (1966: 139) (cf. § 71). He didn’t see that this ‘materiality’ is a visual by-product of written language, whereas he himself had performatively disqualified ‘writing’ for ‘disguising’ and ‘obscuring language’ (1966 [1916]: 30).

20. Saussure’s assertion that the ‘material and mechanical manifestations’ of ‘language’ are all ‘psychological’ (1966: 6) supported a drastic short-circuit between mental and  material: ‘linguistic signs’ ‘are realities that have their seat in the brain’; and ‘the concrete object of linguistic science is the social product deposited in the brain of each individual’ (1966: 15, 23) The ‘brain’ became the material site for the ‘signified’ and the ‘signifier’, the ‘grammatical system’, and the ‘associative’ (i.e. paradigmatic) ‘co-ordinations formed outside discourse’,⁴  which ‘are not supported by linearity’ (1966: 65f, 13f, 123). Such performatives brandished a ‘social product’ whose reality was glibly moved inside the human anatomy to house any theoretical constructions Saussurian discourse might postulate. To challenge those constructions would seem to irrationally deny the hard material facts of biology (§ 45, 51f, 56).

21. Again repeating philosophy while ignoring it, Saussure suspected language of being ‘irrational’, although ‘the mind contrives to introduce a principle of order and regularity into the mass of signs’; indeed, ‘the whole system of language is based upon the irrational principle of the arbitrariness of the sign’, so we should ‘study language as its limits arbitrariness’ through ‘grammar’ (1966: 133). We see here the thematic strategy of linguistics to defend its own rationality by highlighting what seems rational about language and marginalising the rest (§ 28ff, 33, 35, 37ff, 50, 83, 93f, 97). Yet Saussure’s wistful notion that ‘language’ ‘could be studied independently’ if it were ‘entirely rational’ seems modest alongside Hjelmslev’s vision of a ‘real and rational genetic linguistics’ (1969: 6), wherein, as we saw, a ‘theory’ does not ‘refer to any existing texts and languages’ (§ 17)

22. When physicalism and materialism gained dominance with the support of ‘unified science’ (§ 10ff), scientists converged on the principle that ‘the phenomena of behaviour and mind are ultimately⁵ describable in the concepts of the mathematical and physical sciences’, as remarked by K.S. Lashley in a volume with the portentous title Cerebral Mechanisms in Behavior (Jeffress ed., 1951: 1). In the U.S. and the U.K., many linguists concurred, as when Bloomfield (1949) wrote on ‘linguistic aspects of science’ for the International Encyclopedia of Unified Science (cf. § 10). He also rewrote his mentalist Introduction to the Study of Language (1914) into his classic (and classical) book on Language, which performatively banished ‘mentalist theory’ and hailed ‘materialistic or, better, mechanistic theory’, with a behaviourist view of ‘speech’ being ‘human conduct’ within the ‘complex system’ of ‘the human body’, and ‘human actions’ being ‘part of cause-and-effect sequences exactly like those we may observe, say, in the study of physics’ (1933: 33). So ‘in all sciences like linguistics, which observe some specific type of human activity, the worker must proceed exactly as if he held the materialist view’ (1933: 38).

23. The dual reality principle created in the discourse of mentalist linguistics (§ 19) was now inverted: physicalists emphasised the primary material reality and declined to specify the secondary mental reality. For Bloomfield (1933: 23-26, 33), the material reality resided in ‘the speech event’ comprising ‘stimulus’ and ‘reaction’ chains as ‘cause-and-effect sequences of the material world’ and as ‘practical events’, and, in line with ‘the sciences of physiology and physics’ as a transaction wherein ‘the gap between the bodies of the speaker and the hearer — the discontinuity of the two nervous systems — is bridged by the sound waves’.

24. This radically materialist outlook fomented the impasse of requiring us to ‘know much more than we do about the external world’ (1933: 75). Once Bloomfield had ‘defined the meaning of a linguistic form’ not as a ‘mental event’, but as ‘the situation in which the speaker utters it and the response which it calls forth in the hearer’, he faced the prospect that ‘the situations which prompt people to utter speech include every object and happening in their universe’ (1933: 142, 139f). If so, ‘the study of speakers’ situations and hearers’ responses’ might be ‘equivalent to the sum total of all human knowledge’ (1933: 74). And to evade the demand ‘to give a scientifically accurate definition of meaning for every form in the language’ based upon ‘scientifically accurate knowledge of everything in the speaker’s world’, Bloomfield performatively vowed that ‘meaning cannot be analysed within the scope of our science’ (1933: 139, 161; cf. 1933: 93, 162, 167, 266, 268).

25. The performative feuding between mentalist versus physicalist-materialist discourse thus entrained early linguistics in implicitly irrational projects to study either language without speech, or else speech without meaning. Still, fieldwork linguistics achieved substantive progress by rationally investigating language in social reality, where mental and material richly interact. Instead of making ‘language stand apart from’ ‘the heterogeneous mass of speech facts’ (Saussure, § 17), fieldworkers — whether mentalists like Sapir or materialists like Bloomfield — derived a description of a language from ‘speech facts’ and showed them to be systematic. The fieldworkers’ own utterances in the language carried the subsidiary illocutionary force of requesting native speakers to confirm or disconfirm implied assumptions about the language. The final results merged the authority of rational discovery procedures with the authority of the community of speakers; yet this authority did not satisfy the aspiration of some linguists for a higher rationality (§ 21, 33).

26. The first and perhaps only complete success of linguistics in mastering the dual reality principle was achieved by describing ‘the phonic substance cut into significant elements’ (Saussure, § 19) through the ‘external’ science of phonetics and the ‘internal’ science of phonology. The sounds of language proved nicely amenable to ‘descriptive’ treatment in two parallel schemes: a material scheme of externally observable articulatory events and locations, which provides general and straightforward catalogues of minimal units, the ‘phonemes’ (like ‘voiced dental stop’); and a mental scheme of internally perceivable distinctions and features, which determines which units qualify as ‘phonemes’ in one specific language by differentiating words that also differ in meaning. Because the meanings did not have to be defined but only had to differ, linguistics could postpone the dilemmas regarding meaning which had baffled the materialism of Bloomfield (§ 24).

27. Due to the early double success of phonetics and phonology, linguistic theory instated a partly realistic and partly idealistic conception of an entire language being a uniform, stable, and deterministic system of mental-material units to be segmented and classified on their respective ‘levels’. Yet already on the ‘level’ of morphology, the conception was less rational. The meanings of the minimal units (the ‘morphemes’) needed to be defined and not just differentiated, and naturally proved far richer and more diverse than articulatory events and locations and were of course not so materially observable. Morphology restrained such problems by cataloguing the small, closed, and uniform classes of ‘morphemes’ having ‘grammatical’ functions and sparse meanings (e.g. the set of ‘noun inflections’), whilst placing the ‘morphemes’ with ‘lexical’ functions and rich meanings into large, open, and diverse classes (e.g. the set of ‘noun stems’) that were only illustrated rather than catalogued. The cataloguing could be assigned to the ‘level’ of lexicology, i.e., the investigation of whole words corresponding to ‘lexemes’.

28. However, lexicology attracted few linguists, no doubt because the ‘lexicon’ is plainly not a sparse, uniform, and stable system, and thus seemed dauntingly irrational. In Saussure’s view, the ‘lexical’ shared ‘a common principle’ with the ‘arbitrariness’ he pronounced ‘irrational’ (§ 21). Similarly, Bloomfield fretted that ‘lexical forms may belong arbitrarily or irregularly to a form-class that is indicated neither by their structure nor by a marker’, and glumly concluded that ‘the lexicon is really an appendix of grammar, a list of basic irregularities’ (1933: 269, 274).

29. Since linguists couldn’t reapply their prevailing conception of ‘ linguistic level’ to lexicology, they turned to ‘syntax’, which appeared reassuringly orderly and rational in its regular sequences and patterns of identifiable forms. But appearances were deceptive. Saussure himself had contemplated excluding syntax from ‘langue’ because ‘in the syntagm there is no clear-cut boundary between the language fact, which is a sign of collective usage, and the fact that belongs to speaking and depends on individual freedom; in a great number of instances it is hard to classify a combination of units because both forces have combined in producing it, and they have combined in indeterminate proportions’ (1966: 125).

30. Formal syntax, in contrast, has assumed that word-order can be most rationally accounted for by an independent system of deterministic ‘rules’ situated entirely on the side of ‘language’ (§ 69ff) and unaffected by the contexts of real speech, the motivations of real speakers, and perhaps even the meaning of utterances. If that assumption is untenable, then the whole enterprise of constituting formal ‘syntax’ on a separate ‘level’ (or, for later theories, in a separate ‘component’ or ‘module’) has been a capital mistake that must keep generating irresolvable disputes and overstated performatives. Linguists get boxed into attributing to ‘language’ an idealised rationality which it does not possess and which they must struggle to invent, whilst disqualifying the real rationality of language and discourse as irrational (cf. § 32).

31. Phonology and morphology had retained contact with the reality of language by matching their theoretical units (phonemes and morphemes) against the practical units (sounds, inflections, etc.) discovered in fieldwork data (§ 25). Syntax now disengaged from the reality of language to devise rule-systems called ‘grammars’ that supposedly ‘underlie’ and ‘generate’ all the ‘grammatical’ sequences of an entity which is still called ‘natural language’ but which is not ‘language’ in the sense understood by most people, including most scientists outside this mode of ‘theoretical linguistics’. In effect, this enterprise replaces real language with ideal language and seeks to ‘describe’ and ‘explain’ natural language as if it were formal language (cf. § 35). The enterprise could achieve rationality only through a ‘grammar’ with a single and complete rule-set that can be consistently and reliably applied to all ‘sentences’ of a language (and only those) by the community of ‘theoretical linguists’. What we have actually witnessed is the opposite: a turbulent proliferation of ‘grammars’, none of them complete, consistent, or reliable, many of them offering a handful of casual analyses (§ 87).

32. One striking consequence of the quest for the “perfect” order of the ideal “language by itself” (Saussure) is that some influential linguists have viewed real language and discourse as a massive disorder, witness Saussure’s own ‘heterogeneous mass’ (§ 17). A remarkable corollary would be that when the members of society use their language, it undergoes a special ‘catastrophe’ in the technical sense of ‘catastrophe theory’ (cf. Thom, 1989): a sudden transition from stable and integrative order to unstable and disintegrative disorder (§ 35). The corollary seems wildly implausible, indeed inverted: surely the transition would move from one mode of order to another mode of order. Linguists cannot grasp the transition if they attribute to language an idealised mode of order they have to construct for themselves.

33. The quest for this ideal order and the aspiration for a higher rationality (§ 25) has led away from fieldwork linguistics investigating real language toward homework linguistics (to coin the matching term) investigating ideal language. Instead of painstakingly gathering corpuses of data in the field, you stay comfortably at home (or in your office) and rationalise about ‘language’ as represented by handfuls of data which you invent in your role as a ‘native speaker’, and which you analyse and describe in your role as ‘theoretical linguist’. The dualism of roles ensures that the native speaker (you) and the linguist (also you) reach the same conclusions without the slogging and protracted processes of fieldwork constructing and testing hypotheses about a language you first have to learn (§ 25; but cf. § 65ff).

34. In parallel, analysis and description shift from the constative acts of fieldwork saying ‘here are data I recorded about real speakers using a real language’, to performative acts of homework saying ‘here are data I invented about the “underlying structure” of ideal “language” when nobody’s using it’. The fit between theory and data can be multiply enhanced at any stage of invention, analysis, and description, notably by admitting only brief, trivial sentences (like ‘the man hit the ball’, § 70) that seem to stand alone without the social contexts and speaker motivation that your theory disregards (§ 30). The operations of ‘derivation’, ‘transformation’, and so on purport to ‘describe’ or ‘explain’ data but actually get rid of them in favour of data whose ‘structures’ and ‘features’ the linguist is authorised to design.

35. These operations can make a simple analysis and description appear impressively complex by means of formalisations, such as formulas, mathematical symbols, and Greek letters, that are actually far simpler than real language utterances but appear complex because they are far less familiar. Through this further dualism, the replacement of real language with ideal language also replaces real complexity with artificial complexity as an insignia of superior rationality (§ 71). Insofar as the two modes of complexity are not compatible, and the shifts from the one to the other are casual and impressionistic, language and discourse again look like mere disorder (§ 32). In compensation, theoretical linguistics constructs its own mode of order and complexity which seems integrative only if you restrict your analysis and description to ‘a certain number of clear cases’ or to ‘some fragment of a language’ (Chomsky, § 70, 82), but which would become uncontrollably disintegrative for any substantive corpus of real data. Let us examine how these various issues are reflected in Chomsky’s own discourse on ‘theoretical linguistics’.

C. Chomsky’s performative dualisms⁶

36. If modern linguistics has often idealised language, Noam Chomsky has been a principal and perennial idealiser. His performative rhetoric has audaciously promulgate a ‘truly fresh and revolutionary approach to the study of language’ (back cover of Aspects), which was really just a cagey reanimation of traditional idealisations for fusing logic with grammar and imposing rationality upon language (§ 2, 5). Philosophers and logicians could now triumph as ‘revolutionaries’, marshalling a new generation of linguists against both science and linguistics, which had dissociated themselves from philosophy, as I remarked (§ 8, 18), just when linguistics was gaining expansive and lucrative prominence on the academic scene. Disregarding the admonitions of Saussure, Hjelmslev, and Firth, ‘linguistic theory’ reverted to ‘philosophising’ and to ‘basing grammar on logic’ equipped with ‘pretentious categories’ and ‘far removed from actual observation’; and ‘philosophy supplanted a great deal of general linguistics’ instead of the reverse (§ 18).

37. Chomsky’s discourse has excelled in managing the scenario of replacing real language with ideal language through performative dualisms whereby the irrational enterprise sketched in § 30-35 can purport to be the most rational one in all of linguistics. His key resource has been systematic ambiguities within the terminology (§ 62-65, 81), which almost constitutes a private language we might call ‘Chomskese’. ‘Language’ and the major terms defined with it in a closed and self-sustaining circle (§60, 80), like ‘speaker’, ‘language community’, and language acquisition’, all acquired a peculiar technical meaning whose relation to the ordinary meaning theoretical linguists can freely determine, manipulate, or deny. Explicit constative moves are paired with implicit performative moves having the main illocutionary force of conflating technical meanings with ordinary meanings, and the main perlocutionary force of instating this mode of ‘theory’ and ‘explanation’ as the most ‘rational’ and ‘interesting’ one. Validation is performed not by showing that specific predictions based on the ‘theory’ are confirmed by empirical evidence from real language, but rather by defining ‘language’ to be exactly what the ‘theory’ stipulates, unperturbed by evidence from real language (§ 59f, 76). Linguists who accept and use the terminology as a license for participating in the discourse of ‘theoretical linguistics’ are unwittingly helping to validate the ‘theory’, even if they harbour doubts about its validity (§ 62f, 76). And if your large claims about ‘human language’ are challenged, you can withdraw behind protective layers of technicality and dismiss the challenges as ‘unsophisticated’ or ‘unscientific misunderstandings’ (§ 87-91, 95) — even when, as we shall see Chomsky calmly admitting much later, the ‘impact of this linguistics’ was due mainly to ‘misunderstanding’ it (§ 87).

38. The irrational underside of a ‘theoretical linguistics’ that replaces real language with ideal language (§ 31) was ironically a rhetorical advantage for Chomsky in assuming the key role of a principal idealiser whose superior rationality sustains the whole enterprise and vanquishes prospective irrationality (cf. § 35, 42, 44, 50, 65, 68, 71, 78). He and his followers guard the key to an ‘underlying’ order and complexity we cannot expect to find in real language and observed discourse. He can declare that his own theory is ‘obviously’ or ‘logically’ the ‘truth’, and that other ‘theories’ or ‘explanations’ are ‘inadequate beyond any reasonable doubt’ or ‘grossly and obviously counter to fact’ (1965: 67, 204) (cf. § 42, 44, 47, 57, 64), because his superior rationality authorises him to divine the ‘truth’, define ‘adequacy’, and decide what counts as ‘facts’ which other people have seen differently or not at all (§ 57, ).

39. Interestingly, the performative dualisms in Chomskyan discourse constitute its basic consistency and opportunistically convert potential problems into illocutionary and perlocutionary advantages (cf. § 46, 58, 78, 97). Each irrational claim can furnish fresh opportunities for rhetorical assaults on the rationality of competing theories or linguists and can reinforce the dependency of his followers upon his own theory as they redouble their efforts to justify it. The less his own ‘theory’ depends upon real language, the more energetic and devout the leap of faith demanded to sustain it (§ 78).

40. With exquisite irony, he invoked ‘the position of the founders of modern general linguistics’ to be: ‘linguistic theory is concerned primarily with an ideal speaker-hearer in a completely homogeneous speech-community, who knows its language perfectly’ (1965: 3). Rendering the idealisation of language more explicit and programmatic than any ‘founders’ I know of nicely forwarded his plan to enthrone himself as the pre-eminent idealiser (§ 36, 38, 78, 88). And, instead of reasoning, like Saussure, that a ‘language’ which ‘exists perfectly only within a collectivity’ is therefore ‘not a function of the speaker’ (§ 17), he invented an ‘ideal speaker’ who is a hypothetical repository for ‘knowledge of the language’ and whose role was also destined for himself (§ 65ff). If it classical science ‘contains no reference to the observer’ (§ 7), this linguistics contains reference to a ‘speaker-hearer’ nobody can observe (§ 54, 81).

41. The same plan explains his eagerness to shift investigation from real language in a corpus of observed data over to ideal language in ‘internal representations’ — a move he later accentuated by distinguishing ‘externalised language’, which is ‘an epiphenomenon at best’, from ‘internalised language’ (or ‘e-language’ from ‘i-language’) (1986: 25) (cf. § 84). His key performatives sought to establish that the ‘internalised language’ with its ‘underlying’ mental reality is in principle not ‘reflected’ in real language with its social and behavioural reality (§ 77, 80). So he ordained that the ‘observed use of language’ ‘surely cannot constitute the subject-matter of linguistics, if this is to be a serious discipline’ (1965: 4), implying, as he often did, that the fieldwork of descriptive linguistics should not be ‘taken seriously’(§ 44, 49, 68, 72). When Newmeyer’s (1980: 249f) devout Linguistic Theory in America avowed that ‘the vast majority’ of ‘linguists’ ‘who take theory seriously acknowledge (explicitly or implicitly) their adoption of Chomsky’s view of language’, we must appreciate that ‘taking theory seriously’ was being equated with ‘adopting Chomsky’s view of language’ and with declining, like him, to take data seriously.

42. Similarly, Chomsky could vow that the ‘inadequacy’ of ‘modern taxonomic structuralist grammars’ (i.e., descriptive morphology and syntax) ‘for natural languages seems to me to have been established beyond any reasonable doubt’ because he had defined ‘adequacy’ expressly to ‘establish’ that (see § 57). And his contention that ‘descriptivist limitation-in-principle to classification and organisation of data, to extracting patterns from a corpus of observed speech’ ‘precludes the development of a theory’⁷ of actual performance’ (1965: 18f) was also circular after his performatives disconnecting ‘theory’ from ‘data’:  

there is no reason to expect that reliable operational criteria for the deeper and more important notions of linguistic theory [...] will ever be forthcoming; [...] knowledge of the language, like most facts of interest and importance, is neither presented for direct observation nor extractable from data by inductive procedures of any known sort (1965: 18f).  

So we must keep depending upon Chomsky to ascertain the ‘facts of interest and importance’ and the ‘deeper notions of linguistic theory’ which linguists had never required or even imagined before and which are cabalistically revealed to his superior rationality without requiring ‘observation of data’ or ‘inductive procedures’ (§ 35, 38).

43. His theory promised his followers immense savings of labour, if ‘sharpening the data by objective test is a matter of small importance’ for attaining ‘a new and deeper understanding of linguistic structure’ (1965: 20f). ‘Many important questions about the nature of linguistic structure’ cannot be answered if ‘linguistic theory’ is ‘required’ to ‘state methods of analysis that an investigator might actually use if he had the time’ and to stipulate a ‘discovery procedure’ ‘for actually constructing the grammar, given a corpus of utterances’ (1957: 51ff). Like his ‘deeper notions’, these ‘important questions’ were expressly formulated to keep ‘the nature of linguistic structure’ safely disconnected from ‘observed data’, and to accommodate linguists who possess neither the ‘time’ nor the ‘procedures’ for ‘constructing the grammar of a language from the raw data’ (1957: 52). They would placidly reject the requirement that ‘linguistic theory’ should ‘provide us with a discovery procedure’ for being an ‘unreasonably strong demand’ and — a consummate paradox — an incitement to ‘vicious circularity’ (1957: 50ff, 84). Chomsky vowed to ‘never consider the question of how one might have arrived at the grammar’, because ‘questions of this sort are not relevant to the programme of research we have outlined above; one may arrive at a grammar by intuition, guess-work, all sorts of partial methodological hints, reliance on past experience, etc.’ (1957: 56). He denied that ‘useful procedures of analysis’ could be ‘formulated rigorously, exhaustively, and simply enough to qualify as practical and mechanical’ (1957: 56). But surely the deliberately ‘unreasonable demand’ here is that the ‘procedures’ must be ‘rigorous, exhaustive’, ‘formalisable’, ‘simple’, ‘practical, and mechanical’ all at once, when these traits  are flagrantly incompatible.

44. Chomsky’s onslaught was keenly calculated to stupefy descriptive fieldwork linguists who had naturally ‘demanded’ that a ‘linguistic theory’ support ‘grammar discovery procedures’ (the title of Longacre’ 1964 [orig. 1958] Field Manual), and who had devoted their careers to developing and applying them in diligent fieldwork on previously undescribed languages (§ 25). Rather than expatiating upon ‘deeper notions’ and ‘unobservable facts’, they had subjected their ‘grammars’ to ‘operational tests’ in the field with real ‘speaker-hearers’ in a real ‘language-community’, who knew its language fluently, not ‘perfectly’. These linguists were astounded to be notified that their ‘grammars’ were all ‘inadequate beyond any reasonable doubt’ and that ‘explanatory adequacy’ could be achieved though ‘intuition, guess-work’, and ‘hints’. In a later interview, Chomsky restated his dismissal even more crudely and unfairly:

gross coverage of data is much too easily obtained, in too many different ways. There are a million ways in which you can give a kind of rough characterisation of a lot of data, and therefore you don’t learn anything about the principles (1982: 82f)

The terms ‘gross’, ‘easy’, or ‘rough’ describe not the procedures of fieldwork linguists, but those of homework linguistics (as defined in § 34), which can save the labour of ‘learning about the principles’ by using its superior rationality to invent the principles.

45. In Chomsky’s own version of the dual reality principle (cf. § 10, 19, 23, 26) ‘linguistic theory is mentalistic, since it is concerned with a mental reality underlying actual behaviour’, but ‘makes no assumptions about the possible physiological basis for the mental reality’ without ‘denying that there is such a basis’ (1965: 4, 193). Just as Saussure had made the ‘brain’ into the material site for ‘language’, ‘grammar’, and so on (§ 20), Chomsky endowed language with a biological and anatomical reality which gets hard-wired by genetic programming (§ 55). Again, challenges can be dismissed as irrational denials of the hard material facts of biology (§ 20).

46. The evasion of ‘assumptions about the possible physiological basis’ was briskly turned to performative advantage when Chomsky ‘guessed’ that ‘mentalist studies’ ‘will ultimately⁸ be of greatest value for the investigation of neurophysiological mechanisms, since they alone are concerned with determining abstractly the properties such mechanisms must exhibit’ (1965: 193, m.e. [= my emphasis]), much as Saussure sought the ‘principle of order and regularity’ which ‘the mind contrives to introduce’ into ‘the mechanism of language’ (cf. § 21). In a piquant twist of presenting linguistics as the model science (§ 16), Chomsky invited neurophysiologists to ‘value’ just those ‘studies’ that make their disinterest in the ‘physiological basis’ of ‘actual behaviour’ into a high-ranking methodological principle.

47. Another performative dualism with an ‘abstract mechanism’ resolved the problem of making it ‘obvious’ that ‘every speaker of a language mastered and internalised a generative grammar’ (1965: 8): how can ideal language get ‘internalised’ if children only encounter ‘externalised’ real language? Again outflanking descriptive fieldworkers, Chomsky recommended that ‘a theory of language’ be ‘regarded as a hypothesis about the innate “language-forming capacity” of humans’; and that ‘problems of linguistic theory’ be ‘formulated as questions about the construction of a hypothetical language acquisition device’ (1965: 30, 47). This ‘device’ was another mental reality whose ‘physiological basis’ was neither specified nor denied. And just as ‘mentalism’ had been praised for its ‘abstractness’, ‘rationalism’ was now praised for ‘eliminating’ ‘dogmatic presuppositions as to the nature of mental processes’ and advancing no ‘assumptions’ about ‘the internal structure of the device’ (1965: 207). Conversely, the ‘scientific’ status of ‘empiricism’ was impugned because ‘the empiricist approach to acquisition of knowledge has a certain dogmatic and aprioristic character’ in ‘stipulating that certain arbitrarily selected data-processing mechanisms’ ‘are the only ones available to the language acquisition device’ (1965: 207). Furthermore, ‘empiricist doctrine’ was charged with ‘attributing a complex human achievement to months or at most years of experience rather than to millions of years of evolution or to principles of neural organisation that may be even more deeply grounded in physical law’ (1965: 58f). Yet, apparently because Chomsky’s ‘mentalism’ should ‘make no assumptions about the possible physiological basis’ whereby ‘neural organisation’ performs this ‘achievement’, he ascended new heights of self-irony by attributing the achievement not to ‘months or years’ but to a single ‘moment’ in his ‘idealised “instantaneous” model’ of ‘successful language acquisition’ (1965: 36). This ‘attribution’ handily evaded the problem that, strictly speaking, an ideal language is not ‘acquired’ at all but rather constructed (§ 56). And if any theory allows us to construct one in a single moment, then it’s Chomsky’s.

48. These brash performatives purported to separate the natural sciences of biology and physics from ‘empiricism’ and tie them to ‘rationalism’ whilst repudiating any responsibility to specify the ‘physiological basis’. Just as he had faulted descriptive linguists for not addressing ‘deeper and more important notions’ they had never needed (§ 42), he now faulted ‘empiricists’ for ‘stipulating arbitrary mechanisms’ inside a ‘device’ they had never imagined, much like a tyrant proclaiming a new law and then prosecuting everybody who had not obeyed it before.

49. As one traditional mode of idealism, ‘rationalism’ is the “philosophic doctrine that reason alone is a source of knowledge and is independent of experience’.⁹  Yet the arguments invoking the doctrine in Aspects contained a fatal though rarely noticed flaw. In his ‘Notes against a Certain Programme’ of 1647, Descartes had declared: ‘nothing reaches our mind from external objects through the organs of sense beyond certain corporeal movements’, ‘but even these’ ‘are not conceived by us in the shape they assume in those organs of sense; hence it follows that the ideas of movements and figures are themselves innate in us’ (1965: 48). Similarly, the ‘Port-Royal Logic’ of 1662 asserted that every ‘idea in our minds has taken its rise’ from ‘those movements which are made in the brain through sense’; yet ‘these ideas have very rarely any resemblance to what takes place in the sense and in the brain’ (1965: 49f). The argument presupposes that ‘sense’ data from ‘movements and figures’ get processed by human ‘organs’ through reproducing the ‘shape they assume’, and yet do not retain that ‘shape’ in the ‘ideas in our minds’, the latter being therefore ‘innate’. But the presupposition is radically empiricist, and is certainly not held by modern empirical science. Instead, the sensory organs have been — empirically — found to organise and construct rich mental representations, including ‘ideas’ (Edelman, 1992). The rationalists had to invoke ‘innateness’ after idealising ‘ideas’ so radically that the latter seem impossible to derive from ‘external objects’ and ‘movements’. Chomsky in turn has idealised ‘language’ so radically that the latter seems impossible to derive from ‘actual speech’, and ‘innateness’ both does the job and confounds descriptive fieldworkers by suggesting that language simply cannot be ‘acquired’ from observed data unless the child or the linguist starts with a Chomskyan ‘theory of grammar’.

50. Again we need a rescuer with superior rationality, because ‘most data fail to help us attain insight into underlying principles and structure and are therefore uninteresting for the purpose of attaining rational understanding’; and because ‘the phenomena that surround us in the real world of ordinary experience’ ‘are too complex’ and ‘involve too many interacting factors’, and ‘our understanding is too limited’ (Chomsky, 1991b: 42). Such is exactly the image of ‘real world’ projected by Chomsky’s own theorising, which seeks ‘insights into underlying principles and structure’ without consulting ‘data’ or ‘phenomena’.

51. Like Chomsky, the rationalists sought authority for their idealisations by linking ‘ideas’ to biology and anatomy, despite denying the reliability of the organs of sensory perception (cf. § 20, 45, 56, 101). For Descartes, ‘ideas are innate in the sense that in some families’ ‘diseases like gout or gravel’ are ‘innate’ (1965: 49). For Chomsky, ‘there is a highly determinate, very definite structure of concepts and of meaning that is intrinsic to our nature, and as we acquire language or other cognitive systems these things just kind of grow in our minds, the same way we grow arms and legs’ (1991a: 66) (cf. § 81). Such claims unintentionally parodied the ambition in the rival camp of radical materialism, physicalism, and behaviourism to account for ‘speech’ in terms of ‘the human body’ and the ‘nervous system’ (Bloomfield, § 23ff).

52. Since he prided himself on holding ‘no assumptions about the physiological basis’ (§45f), Chomsky’s appropriations of biology and neurology have been episodic and oblique. To support the peremptory claim that ‘information regarding situational context’ need not ‘play any role in determining how language is acquired’ once the ‘acquisition device’ has been ‘set in operation’, he cited some ‘studies of animal learning’: although ‘it has been observed’ that ‘depth perception in lambs is considerably facilitated by mother-neonate contact’, ‘there is no reason to suppose that the nature of the lamb’s “theory of visual space” depends on that contact’ (1965: 33f). If the claim is that contact with the mother, which helps new-born lambs learn to see more easily or rapidly, does not endow them with the capacity for vision, it is indisputable, indeed trivial; but the biologists ‘observed’ the ‘facilitation’ and not any non-‘dependence’ upon a ‘lamb’s theory of visual space’. In fact, authoritative neurobiologists maintain that animal vision is not guided by any abstract ‘theory of space’ but by massive parallel interactions among networks of neurons. Moreover, Edelman’s (1992: 83f) well-known work on ‘neuronal group selection’ in ‘the formation of the neuroanatomy characteristic of a given species’ also verifies that any creditable theory of mental processing in either animals or humans vitally demands extensive and detailed assumptions about the physiological basis.

53. With equal obliqueness, Chomsky’s recommendation for ‘studying’ the ‘acquisition device’ avoided ‘preconceptions’ by hurrying over its ‘potential physiological base’ to ‘naturally’ arrive at ideal language:

Continuing with no preconceptions [sic], we would naturally turn to the study of uniformities in the output — formal and substantive universals — which we then must attribute to the structure of the device — or, if this can be shown, to uniformities in the input, this alternative rarely being a serious one in the cases that are of interest. (1965: 207)

Why we should ‘naturally study the uniformities in the output’ but not take ‘seriously’ or be ‘interested’ in those ‘in the input’?. A closely related recommendation shows why: ‘much information can be obtained about both the primary data that constitute the input and the grammar that is the “output” of such a device, and the theorist has the problem of determining the intrinsic properties of a device capable of mediating this input-output relation’ (1965: 47). So ‘input’ refers to the real language data — ‘primary data’ in Chomskese — perceived by prospective speakers who are still acquiring the language, whereas ‘output’ does not refer (as you might imagine) to the real language data produced by speakers after ‘acquisition’ but rather to the ‘grammar’, which belongs to ideal language. The job of ‘the theorist’ is thus to fabricate a mechanism that converts real language into ideal language, like an robot-linguist of the Chomskyan type, and to declare that ‘language acquisition’ has been thereby ‘explained’. Conveniently, this ‘output’ never gets put out where we can examine it for evidence about the ‘device’ or compare it to the real language constituting ‘input’. Conversely, real language would not display the ‘uniformities’ Chomsky required for his idealised ‘universals’ and is thus not ‘of interest’, just as for Saussure (1966: 19, 9) ‘speech cannot be studied for it is not homogeneous’ and ‘we cannot discover its unity’ (cf. § 17).

54. Both the ‘uniformities’ and the ‘universals’ — the two terms being equated here by apposition (cf. § 62) — follow directly from the notion of an ‘innate language acquisition device’. We can safely ignore Chomsky’s speculation that ‘universal grammar’ might explain the ‘essential property of language’ to ‘provide the means for expressing indefinitely many thoughts and for reacting appropriately in an indefinite range of new situations’ (1965: 6), because ,unlike Bloomfield’s ‘speaker’ and ‘hearer’ in § 24, Chomsky’s ‘ideal speaker-hearer’ never ‘reacts to situations’ at all (§ 77).

55. After ‘assuming that the genetically determined language faculty is a common human possession’ (1980b: 48), Chomsky mimicked physicalism (of all things) by linking materiality with ‘universality’ to claim that ‘universal grammar’

may be regarded as a characterisation of the genetically determined language faculty. One may think of this faculty as a ‘language acquisition device’, an innate component of the human mind that yields a particular language through interaction with presented experience: a device that converts experience into a system of knowledge attained. (1986: 3)

‘Ordinary experience’ may be ‘too complex’ for ‘our understanding’ (§ 50) but not for the wonder-working ‘device’. Notice how the performative of consigning ‘language’ to a ‘genetically determined faculty’ bypasses social reality and the social sciences and connects to the ‘natural sciences’ (cf. § 48, 78, 82, 94).

56. The same performative neatly evaded the problem, aired in § 47, that ideal language is not acquired. Besides, the vast complexes of formal rules and structures in a Chomskyan ‘grammar’ look patently unlearnable. The link to biology and neurology adroitly allotted to ‘language’ and ‘grammar’ a hard-wired and pre-programmed ‘mental reality’ (§ 20, 45, 51f) ensuring the ‘internalisation’ by the famous ‘ideal speaker-hearer’ and by the ‘completely homogeneous speech-community’ (§ 40). And counter-arguments against Chomskyan linguistics can again be handily rebuffed for irrationally denying biological realities (§ 20, 45).

57. If the ‘structure of the device’ is the same in all humans, we could rationally expect ‘uniformities’ among the ‘grammars’ of different languages. At one point, Chomsky ordained that ‘the main task of linguistic theory must be to develop an account of linguistic universals’ that ‘will not be falsified by the actual diversity of languages’, and ‘will be sufficiently rich and explicit to account for the rapidity and uniformity of language learning, and the remarkable complexity and range of the generative grammars that are the product of language learning’ (1965: 28, m.e.). If ‘a theory of linguistic structure that aims for explanatory adequacy incorporates an account of linguistic universals and it attributes tacit knowledge of these universals to the child’, and if ‘language learning would be impossible unless this were the case’ (1965: 27), then we are back to the conclusion that descriptive (‘structuralist’) linguists must have overlooked ‘the main task’ and have produced ‘grammars’ which are ‘inadequate beyond any reasonable doubt’ (cf. § 42, 44). Chomsky’s admission that ‘for the present we cannot come close to making a hypothesis about innate schemes that is rich, detailed, and specific enough to account for the fact of language acquisition’ might indicate why he later decided to consign ‘richness’ to the descriptivists and postulated

a tension between the demands of descriptive and explanatory adequacy. To achieve the latter, it is necessary to restrict available descriptive mechanisms so that few languages are accessible […] To achieve descriptive adequacy, however, the available device must be rich and diverse (1986: 55f)

The opposition between ‘explanatory adequacy’ versus ‘descriptive adequacy’ sweetly lined up here with the opposition between ideal language versus real language whilst saving labour by dumping ‘richness and diversity’ (cf. § 85).

58. Again, the most attractive advantage of Chomsky’s ‘main task’ was the labour it saves the linguist (cf. § 43f, 69, 71): ‘the grammar of a particular language, then, is to be supplemented by a universal grammar’ ‘expressing the deep-seated regularities which, being universal, are omitted from the grammar itself’ (1965: 6, m.e.). ‘Insofar as aspects of the base structure are not specific to a particular language, they need not be stated in the grammar of this language; instead, they are to be stated only in general linguistic theory as part of the definition of the notion “human language” itself’ and ‘presumably reflect what the mind brings to the task of language acquisition’ (1965: 117, m.e.; compare similar performatives in 1965: 35f, 112, 144, 168, 225). Stating the ‘structure’ of the ideal general language would automatically state most of the structure of any specific language like English and bring a huge savings of labour, especially if we remain on the general plane and postpone working on specific languages — a rising trend in Chomsky’s own work (cf. § 84-86).

59. In Chomsky’s ‘main task’, the ideal language would not be the idealised version of any one ‘particular’ real language like English but rather the idealised version of all possible languages or at least all known languages and therefore of no one real language. This ideal language would then itself be a ‘linguistic theory’, the universal blueprint proposed by Hjelmslev to ‘foresee all conceivable possibilities’ (§ 17, 21). Chomsky’s performative campaign to make the design of language fit the design of linguistic theory thus reached its relentless conclusion by making the two totally identical (§ 37). He has fulfilled his early vision of ‘the ultimate outcome’ being ‘a theory of linguistic structure in which the descriptive devices utilised in particular grammars are presented and studied abstractly’, without ‘reference to particular languages’ (1957: 11, 50).

60. The performative short-circuiting of the terms ‘language’, ‘linguistic theory’, ‘grammar’, and ‘linguistic structure’ is the great constant in Chomsky’s rambling enterprise and in its many imitations. Instead of ‘language’, ‘grammar’, and ‘linguistic structure’ being explained by the ‘linguistic theory’, they are all made identical with the ‘theory’, and each is defined solely by the others in a closed circle instead of by real language (§ 37, 80). So Chomsky also fulfilled Hjelmslev’s vision of a ‘linguistic theory’ that need ‘not be verified (confirmed or invalidated) by reference to any existing texts and languages’ (§ 17, 21).

61. If a language’ is also a ‘linguistic theory’, then learning a language would be like discovering a theory, and Chomsky’s performatives have exploited that very analogy. As befits the ‘device’ equipped with ‘universals’ and producing ‘uniformities’ (§ 53f, 57), ‘language acquisition’ becomes a transaction whereby the child encounters real language (‘primary data’) and converts it into ideal language (‘the grammar’) just like a Chomskyan linguist: ‘language acquisition is based on the child’s discovery of what from a formal point of view is a deep and abstract theory — a generative grammar of his language — many of the concepts and principles of which are only remotely related to experience by long and intricate chains of unconscious quasi-inferential steps’, and despite ‘the degenerate quality and narrowly limited extent of the available [i.e., primary] data’ (1965: 58, m.e.). Attributing the ‘universal’ ‘deep and abstract theory’ to just one language (to ‘his language’) must be an oversight. Also, the term ‘discovery’ jars with the ‘theory’ being prefigured in ‘the structure of the device’ (§ 53). A more consistent formulation might be: ‘the child’s specification and diversification of a deep and abstract theory’.

62. But the strategic point here was to equate the construction of a ‘deep and abstract theory’ with ‘discovery’, and of the ‘theory’ itself with ‘a generative grammar of his language’ — another move that asserted by apposition (§ 54) — and to equate the ideal language-acquiring child  with the generative linguist, thereby naturalising the latter’s proceedings and anticipating their success. The equation rode complacently upon ‘systematic ambiguities’:

Using the term ‘grammar’ with a systematic ambiguity to refer, first, to the native speaker’s internally represented ‘theory of his language’ and, second, to the linguist’s account of this, we can say that the child has developed and internally represented a generative grammar in the sense described. […] we are again using the term ‘theory’ — in this case ‘theory of language’ rather than ‘theory of a particular language’ — with a systematic ambiguity to refer both to the child’s innate predisposition to learn a language of a certain type and to the linguist’s account of this. (1965: 25)

The ‘linguist’s account’ gets slickly short-circuited with both the specific ‘theory of a particular language’ (e.g. English) and the general ‘theory of language’, thus yielding a third ambiguity. Nowhere is it more obvious how just using the terms in Chomsky’s ‘sense’ implicitly validates his ‘theory’ and his ‘account’ (§ 37).

63. The ‘acquisition device’ triggered yet another snub of real language. Just as the ‘observed use of cannot constitute the subject-matter of linguistics’ (§ 41), ‘many children acquire first or second languages quite successfully even though’ ‘much of the actual speech observed consists of fragments and deviant expressions’ (1965: 201, m.e.). ‘The child must have the ability to invent a generative grammar that defines well-formedness and assigns interpretations to sentences even though the primary linguistic data he uses as a basis for this act of theory construction may, from the point of view of the theory he constructs, be deficient in various respects’ (1965: 201, m.e.). As usual, we validate the ‘theory’ by invalidating the ‘data’.

64. The ‘ambiguities’ cited in § 62 entitled linguist-grammarians in producing ‘concepts and principles’ too that are ‘only remotely related to experience by long and intricate chains of unconscious quasi-inferential steps’ (§ 61), and in idealising the ‘degenerate’ and ‘narrowly limited’ qualities out of ‘the available data’ — in pointed contrast to those ‘undoubtedly inadequate structuralist grammars’ ‘limited in principle to classification and organisation of data’ (§ 42, 44, 57).

65. The ‘ambiguities’ also entitled Chomskyan linguist to represent not merely the ‘language-acquiring child’, but also the ‘ideal speaker-hearer’ whose ‘linguistic intuition is the ultimate standard that determines the accuracy of any proposed grammar’ (1965: 21), where the mismatch between ‘intuition’ and ‘accuracy’ is papered over with Chomsky’s pet qualifier ‘ultimate’ (more in § 83). The entitlement was nonchalantly awarded when the grammarian’ was assigned ‘the problem’ of ‘constructing a description’ and ‘explanation for the enormous mass of unquestionable data concerning the linguistic intuition of the native speaker, often himself’ (1965: 20, m.e.). Rather awkwardly, Chomsky himself had denied that the ‘speaker of a language’, who has ‘mastered and internalised a generative grammar, is aware of the rules of the grammar or even’ ‘can become aware of them’, and that ‘his statements about his intuitive knowledge are necessarily accurate’; since ‘a speaker’s reports and viewpoints about his behaviour and competence may be in error’, ‘a generative grammar attempts to specify what a speaker actually knows, not what he may report about his knowledge’ (1965: 8). If these denials were valid, then linguists who assume the role of ‘speakers of the language’ also could not ‘be or become aware of rules’, and their ‘reports and viewpoints about competence may be in error’. So, in the tradition of scientists  becoming ideal prototype of human beings (§ 11), the linguists became the ideal prototype of speakers, who command a superhuman rationality for ‘becoming aware of and reporting’ what real speakers cannot — indeed, for revealing the ‘competence’ of the ‘ideal speaker-hearer’ ‘who knows the language perfectly’ (§ 40)! So of course ‘discovery procedures’ cannot be ‘reasonably demanded’ (§43).

66. Chomsky’s early vow that ‘the major goal of grammatical theory is to replace this obscure reliance on intuition’ ‘about linguistic form’ ‘by some rigorous and objective approach’ (1957: 94) was opportunistically forgotten: the prospect that ‘the necessity for present-day linguistics to give such priority to introspective evidence and to the linguistic intuition of the native speaker excludes it from the domain of science’ got dismissed as a ‘terminological question’ with ‘no bearing at all on any serious issue’ (1965: 20), where we might recall ‘serious’ meaning: not having the ‘observed use of language’ as your ‘subject-matter’ (§ 41). His motives for keeping ‘terminological questions’ apart from ‘serious issues’ are obvious if his whole enterprise has been constructed upon opportunistic performatives for manoeuvring the terminology (§ 37, 76, 78ff, 84ff).

67. Besides, ‘introspection’ and ‘intuition’ raise methodological questions and not just ‘terminological’ ones. The discursive evidence indicates that ‘introspection’ and ‘intuition’ are prime channels for replacing real language with ideal language, which, rather obviously, cannot be done with ‘direct observation’,  ‘discovery procedures’, or ‘objective tests’ (§ 42f). Unless the methods are limited to ‘guess-work’ and ‘partial hints’, and Chomsky had good reason to ‘never consider’ how they operate (§ 43), analysis and description must be worked out in detail for reliable conversions between real and ideal, or  else there can be no ‘science’. Instead, Chomsky’s discourse continually withdraws from the real by disqualifying as it ‘degenerate’, ‘fragmented’, ‘deviant’, ‘deficient’, and so on (§ 61, 63).

68. We witness the same tendency when Chomsky’s discourse defended the ‘doctrines’ of ‘universal grammar’ against ‘modern linguistic and anthropological investigations’ by ‘claiming ‘universality’ for ‘deep structures’, which ‘may be quite distinct from the surface structures’, where he situated ‘diversity’  (1965: 118, m.e.) — a factor we saw him eventually dumping out of ‘explanatory’ into ‘descriptive adequacy’ (§ 57). In another jab at ‘structural (taxonomic) linguistics’, its ‘syntactic theories’ were criticised for ‘assuming that deep and surface structures are actually the same’ (1965: 16), as if doing so were not the obvious strategy for real language. Conversely, ‘the central idea of transformational grammar is that they are, in general, distinct, and that the surface structures are determined by repeated application of certain formal operations called “grammatical transformations”’ and ‘enter the phonological component to undergo phonetic interpretation’(1965: 16f), recalling the dualism between mental and material in early linguistics (§ 26). Predictably, his ‘concern’ in Aspects was ‘primarily with deep structure’, whereas he judged ‘surface structure’ ‘unrevealing’ or ‘irrelevant’, and warned that ‘surface similarities hide underlying distinctions’ which ‘no English grammar has pointed out’, nor are they ‘clear to the speaker’ (1965: 17, 200, 24, 22) — though they’re clear to Chomsky’s superhuman rationality (§ 65).

69. Linguists should now show how the ‘syntactic component’ ‘generates deep and surface structures for each sentence’ and ‘interrelates them’ (1965: 16f, 135). But Chomsky placidly announced that ‘the grammar does not, in itself, provide any sensible procedure for finding the deep structure of a given sentence’; ‘it merely defines these tasks in a precise way’ (1965: 141), where ‘sensible’ would be for real language, and ‘precise’ for ideal language. His breeziest evasion in Aspects was to make ‘no careful distinction’ ‘between the basic string and sentence itself’ and to proceed on the ‘simplifying and contrary-to-fact assumption that the underlying basic string is the sentence’, and that the ‘base phrase marker is the surface structure as well as the deep structure’ (1965: 18). So we were counselled to reject ‘syntactic theories’ which ‘assumed that deep and surface structures are the same’ and accept a ‘syntactic theory’ which asserted they are not the same whilst saving labour by treating them as the same. Such evasions were inescapable if  moving between ‘surface’ and ‘deep’  is also moving between ‘language-specific’ and ‘universal’, and so between real language and ideal language; and ‘surface structures’ are understandably ‘unrevealing’ about how to do so, just as ‘discovery procedures’ couldn’t be ‘reasonably demanded’ (§ 43, 65). The ‘grammar’ with its ‘rules’, ‘transformations’, ‘rewritings’, and the like, does in local detail what the ‘theory’ does in the global programme: fit the design of language to the design of the theory (§ 37ff, 59f). Conversely, real language data get ‘explained’ by ‘transforming’ them downwards into ideal language data, an operation whereby problematic aspects such as richness, diversity, and indeterminacy should get ‘abstracted away’. But how and why?

70. Chomsky’s own demonstrations of grammatical analysis in his two best-known books consist of improvised and casual manipulations of formal symbols, formulas, and diagrams in representations that only feebly capture the structure of English sentences. His early discourse proffered only a ‘weak test of adequacy’ on ‘a certain number of clear cases’, plus ‘direct applications to the description of English sentences’, witness his perfunctory ‘derivation of the sentence “the man hit the ball”’ (1957: 14, 34 26f):¹⁰

(i)     Sentence;

(ii)    NP + VP;

(iii)   T + N + VP;

(iv)   T + N + Verb + VP;

(v)    the + N + Verb + NP;

(vi)   the + man + Verb + NP;

(vii) the + man + hit + NP;

(viii)             the + man + hit + T + N;

(ix)   the + man + hit + the + N;

(x)    the + man + hit + the + ball

This ‘derivation’ was said to ‘apply rules’ like ‘Sentence è NP + VP’, ‘N è man, ball, etc.’, and ‘V è hit, took, etc.’ (1957: 26), which performed just two types of ‘rewriting’ operations: replacing a symbol with a string of symbols, and replacing a symbol with an English word, going from left to right. Of the intricate morphological and lexicological problems of cataloguing large word-classes such as ‘nouns’ and verbs’ (§ 27), nothing remained but ‘etc.’; and the verb neatly materialised in the past tense invented sentences so often have, no doubt to make them sound cut-and-dried. The formality of the plus signs was gratuitous, indeed misleading, by imposing uniformity upon the diverse connections of grammatical dependency (e.g. ‘the + man’) versus mere adjacency (e.g. ‘hit + the’).

71. This simple ‘derivation’ illustrates a major opportunistic strategy of formalist analysis: treating the graphic and visual representations of sentences and notations as the material reality of language in place of the ‘phonic substance’ of Saussure (§ 19) or the ‘speech events’ of Bloomfield (§ 23), whilst handling written sentences like a set of elongated real objects to be segmented, classified, rearranged, and so on — Saussure’s ‘material units distributed in space’ (§ 19) — whence the metaphoric spatial terminology (e.g. ‘insertion’, ‘deletion’, ‘left-branching’, ‘right-branching’, ‘subject-raising’). When Chomsky’s later programme grew more abstruse and portentous, his notations got ‘deeper’ and more ‘universal’ and were not so ‘directly applied to English sentences’, until eventually his definition of ‘syntax’ no longer referred to the word-order of sentences at all but rather to the ‘structure of mental representations’(§ 84). The rising artificial complexity supplied an insignia of superior rationality (§ 35), saved him the labour of working out the intricate details of sentence analysis, and retreated from the hopeless search for a ‘grammar’ of deterministic ‘rules’ to ‘assign structural descriptions to all the sentences’ of a language (§ 30).

72. The backlog of unaddressed methodological questions had been growing ever since Chomsky’s (1957: 13) primal performative resolving to ‘consider a  language to be a set (finite or infinite) of sentences, each finite in length and constructed out of a finite set of elements’, where the hopeful analogy to ‘a formal system of mathematics’ was explicitly drawn. However, the ratio between ‘infinite’ and ‘finite’ in his definition, though presented and hailed as an important constraint, is significant only in mathematics but empty and trivial for any real language like English: no linguist, whether descriptive or generative, would seriously affirm that all the possible sentences in a language like English have been or could be compiled and even less that such a language could have either an infinite number of phonemes or letters or any infinitely long sentences. The real illocutionary force, as usual, was to move toward the ideal language of Chomskyan linguistics and away from real language of descriptive fieldwork: ‘it is obvious that the set of grammatical sentences cannot be identified with any particular corpus of utterances obtained by the linguist in his field work’ (1957: 15). So where do you find the ‘set’? Answer: you breezily “assume that the set of sentences is somehow given in advance” (1957: 13, 85, 103, 18, 54). You do not turn to the very large set, finite but open, of discourses in real language, that disordered ‘heterogeneous mass’ of ‘fragments’ and ‘deviant expressions’ (§ 17, 25, 32, 63, 67).

73. If we used the terms strictly, an ‘infinite set’ would undercut rather than support the distinction between ‘grammatical’ versus ‘ungrammatical’. A truly infinite set would contain all combinations and sequences, including ones extravagantly unlikely to occur, just as the infinite typing of chimpanzees in the well-known philosophers’ example would, in infinite time, produce the works of Shakespeare. In its strict meaning, ‘infinity’ erases the borders not just between probable versus improbable, making our description random or aleatory, but also between possible versus impossible, making our description interminable (like Borges’ famous Library of Babel’).

74. Chomsky’s concept of the ‘infinite’ was plainly more pedestrian: a ‘grammar’ sets no limits on the longest sentence. He breezily proposed that ‘the creative aspect of language use’, which he enlisted  to motivate ‘universal grammar’¹¹ and which he rebuked the ‘grammars’ from descriptive linguistics for ‘not attempting to deal with’, can be ‘explicitly formulated’ in the ‘technical devices for expressing a system of recursive processes’, as developed in ‘mathematics’ (1965: 5 8). Here, ‘creativity’ got irrationally equated with mere recursion, e.g., producing ‘a sequence of adjectives longer than any ever before produced in the context “I saw a ____ house”’ (1957: 17). Creativity is the diametrical opposite of recursion , which can only keep churning out the same, and would not be demonstrated by a reasonable speaker trying to stick a world-record ‘sequence of adjectives’ into a trite little sentence. The real creativity of natural language resides in the delicate interactions among lexicogrammatical constraints during real discourse (Beaugrande, 1997).

75. The key role of the ‘sentence’ is Chomsky’s proverbial ‘set’ was to oscillate between ideal and real, and between a theoretical unit like a ‘string’ and a practical unit like an ‘utterance’, as when we ‘assume that the underlying basic string is the sentence’ (§ 69). So we read about a ‘grammar’ being ‘related to a corpus of sentences’ but also to ‘a corpus of utterances’; or about both ‘the deep structure’ of a ‘sentence’ and ‘the deep structure of an utterance’ (1957: 10, 50; 1965: 16, 131). ‘Utterances’ appeared in one egregiously facile analogy to material objects (cf. § 71): ‘chemical theory’ ‘might be said to generate all physically possible compounds, just as a grammar generates all grammatically “possible” utterances’ (1957: 48). By this time, we can savour the pungent irony and opportunism of a fervent ‘mentalist’ with an affinity for ‘natural sciences’ taking the laws of physics accepted by physicists all over the world as models for the ‘rules’ of  ‘grammar’ rarely accepted by more than one clique or claque of linguists — a divisiveness accentuated by Chomsky and his many epigones.

76. In yet another unwitting irony, Chomsky surmised that ‘once we have disclaimed any intention of finding a practical discovery procedure’, ‘certain problems that have been the subject of intense methodological controversy simply do not arise’ (1957: 56). With high poetic justice, just the opposite occurred: ‘methodological controversies’ ‘intensified’ when other linguists felt legitimated to devise theories of ideal language without applying ‘discovery procedures’ to real language. Still,  many of these ‘controversies’ had to be carried on in his ‘Chomskese’ terminology, whose use, as I have remarked, implicitly validates his ‘theory’, or at least his conception of ‘theory’ (§ 37, 62f). And, being unencumbered by  ‘direct observation’,  ‘discovery procedures’, or ‘objective tests’ (§ 42f), he can keep an edge by freely ‘revising’ or ‘extending’ his ‘theory’.

77. And ultimately, a ‘theory’ of the Chomskyan type is expressly designed to be irrefutable. How do you refute an ideal declared inaccessible to ‘observation’ or ‘discovery’ (§42f)? How can you disprove that the proposed ‘grammar mirrors the behaviour of the speaker who, on the basis of a finite and accidental experience with language, can produce and understand an infinite number of new sentences’ (1957: 15), when ‘finite and accidental’ apply to real language, and ‘infinite’ applies to ideal language? Do ‘producing and understanding’ imply real ‘behaviour’ or are they ‘infinite’ operations? How can you rebut a theory about ‘competence, the speaker-hearer's knowledge of his language’, when ‘performance, the actual use of language in concrete situations’ ‘obviously could not directly reflect competence’ and is replete with ‘deviations from the rules’ (1965: 4, h.e.)? We are only offered the ‘idealisation’ of the ‘speaker-hearer’ holding “perfect knowledge of the language’ (1965: 4).  This ‘speaker’ in theory knows everything about the language and in practice says nothing in the language, perhaps standing transfixed in ‘tacit introspection’ upon that wondrous ‘infinity of sentences’ he would be ‘competent’ to say. So we can’t find him, and even if we did, we wouldn’t hear a thing.

78. I have now examined a range of interlocking performative speech acts sustaining Chomsky’s opportunistic programme for replacing real language with ideal language (§ 31, 35, 37f, 67), and making himself the principal idealiser at the expense of descriptive fieldwork linguistics (§ 36, 38, 40). He has raised some trends toward idealisation in previous ‘theoretical linguistics’ to new intensities to suggest that only his superior rationality can lead us to ‘a new and deeper understanding’ and reveal ‘deeper and more important notions of linguistic theory’, which by definition cannot be ‘observed or extracted from data’  — otherwise, lesser mortals could find them. His constative speech acts about what ‘language’ is and how you ‘study’ it share the performative force of proving that ‘language’ is precisely not what rival schools of linguists might claim or assume, no matter how bizarre or irrational his own proofs might become. In this mode of confrontational discourse, irrationality is a rhetorical advantage for overreaching ordinary rationality in favour of superior rationality (cf. § 38), much as every occult faith promises to reward its loyal believers for to their devout leap of faith with a previously unreached level of ‘truth’ (§ 39). A ‘linguistic theory’ then becomes like a lens that inverts polarities: the irrational becomes rational and vice-versa; the obscure is suddenly ‘obvious’; ‘intuitions’ turn into ‘facts’; ‘introspection’ supplants ‘discovery’, and homework dislodges fieldwork; the linguist parades as the ‘ideal speaker-hearer’; the child is mechanised  into a ‘device’; and the meanings of terms flicker and oscillate. Polonius might  say: ‘if this be method, yet there is madness in’t’.

79. Perhaps the multiple connections between ideal languag