Rhetoric and Stylistics in Light of Large

1.1 So compendious a tome as the present work embracing both ‘rhetoric’ and ‘stylistics’ certainly suggests some affinities that make them appear related, and no doubt many contributors to this volume will have addressed the prospect, as I saw already in the editors’ original plan. Both are obviously dependent on specific selections and their combinations of expression in discourse, even where we may have some difficulty determining which ones are the more important or influential ‘indicators’, or whether the two factors belong to the discourse as a whole. According to Halliday (1973: 112, 103) — an exceptional linguist, to be sure — ‘there are no regions of language in which style does not reside’; the ‘central problem in the study of style’ is ‘relevance’: ‘determining whether any particular instance of linguistic prominence’ ‘is significant’ and ‘motivated’ (his emphasis). He might have said much the same about the study of rhetoric, which he would address as a factor in ‘registers’.

1.2 Complementarily, however, we might suggest some antinomies that make the two factors also appear distinct. The most original antinomy, at least in the ‘Western history of ideas’, is historical: to see rhetoric as a relatively ‘Classical’ enterprise (dating roughly from the 5^th century BC), and ‘stylistics’ as a relatively ‘modern’ enterprise (dating roughly from the 1930s or 1950s). In ancient Greece, ‘rhetoric’ was literally an enterprise — a profitable business — widely respected and deployed in training citizens (‘free’ males of ‘good families’, that is) in the high arts of fluency and eloquence as tools for attaining dignity and public office. It was eminently practical and was confidently expected to produce tangible results.

1.3 Yet ‘style’ was naturally one major topic within ‘rhetoric’, as attested for example by the compendious Aristotle. However, he seemed to view it with ambivalence, especially in contrasting ‘poetics’ (1) with ‘rhetoric’ (2), viz.:

(1) The perfection of style is to be clear without being mean. […] That diction […] is lofty and raised above the commonplace which employs […] strange (or rare) words, metaphorical, lengthened, [or] ornamental (Poetics, 9.4)[1]

(2) the virtue of style is to be clear, [whereas] poetic style [is] inappropriate to a speech; we should make little use of exotic, compound, and artificial (Rhetoric, 3.2)[2]

Demetrius’ treatise On Eloquence, despite its title being routinely mistranslated into English as On Style, is, in my view, rather a handbook for rhetoric in the Classical sense. Now, the ambivalence traceable in Aristotle’s writings became virulent, viz. (3-4) versus (5-6).[3]

(4) Diction [that is] grandiose, elaborate, and distinctly out of the ordinary […] possesses the needed gravity, whereas usual and current words, though clear, are unimpressive and liable to be held cheap. (§ 77)

(5) Overloading with figures […] betokens lack of taste and inequality of style (§ 67)

(6) Exuberant and inflated language must not be sought after in a style meant to carry conviction. (§ 221)

Even so, his favourites for citations were Homer and Demosthenes, whose fame rested (and still rests) on styles that are, to say the least, ‘distinctly out of the ordinary’. Demetrius was not too informative about just what ought to should count as ‘trivial’ or cheap’ (and cf. 1.14; 2.14).

1.4 Stylistics, in contrast, seems to have carved out no distinct historical outlines as a standing discipline. Perhaps it remained in limbo because rhetoric (and of course grammar) dominated the study of language even up into the 18^th century, largely as a static and ponderous laundry-list of rules, figures, or tropes, with portentous tongue-twisting names like ‘epizeuxis’ (uninterrupted repetition of words), ‘palilogia’ (emphatic or passionate version of the same), or ‘homoioptoton’ (the repetition of similar case endings in adjacent words). What I would consider the genuine concerns of style were dispersed among at least three diverging disciplines with diverging goals: rhetoric to master ‘persuasive’ language; poetics to master ‘beautiful’ language; and grammar to master ‘correct’ language (cf. 1.8).

1.5 I surmise that the major motors of the gradual historical shift away from Classical rhetoric in language study were exploration, trade, colonialism, and ‘missionarism’, all involving languages by no means constructed in the same mould as the ‘Classical languages’ nor the European languages which were being forced into that mould by language purists who, like Dryden, tested their English samples by translating them into Latin.[4] A leading role was occupied to institutions like the Propaganda Fide which helped to mask exploitation and slavery as ‘spreading the faith’ and ‘saving the souls’ of the ‘heathens’ (Firth 1937) — and logically enough, eventually translating the Bible into those languages, which proved a windfall for linguistics, but often a disaster for the ‘natives’ (cf. Hvalkov and Aaby eds., 1981). The priests or fieldworkers sent out on these ‘missions’ were mostly trained with the standard ‘linguistic’ topics like ‘pronunciation’ (or ‘phonetics’), ‘grammar’ (or ‘morphology’), and ‘vocabulary’ (or ‘lexicology’), rather than rhetoric and style as such.

1.6 A further traditional antinomy would fall between rhetoric highlighting purpose and style highlighting language. But I believe this antinomy is an artefact of unproductively disunified approaches to the study of language, such as trying to ‘do syntax’ without semantics, and trying to ‘do semantics’ without pragmatics (cf. Beaugrande, 1991, 1997). Language and purpose are inseparable, even when — as in more obscure poetic or academic discourse — the purpose is not easily defined, or is indeed expressly concealed, e.g., as a bid for academic power by deploying a needlessly complicated style that withholds information whilst pretending to transmit it (cf. 1.14).

1.7 A related and also traditional antinomy would fall between rhetoric highlighting text receivers (the audience to be impressed or persuaded), and style highlighting text producers (a speaker or writer in search of an individual and noteworthy mastery over language). But here too, we should be cautious, because the two sides of participation in discourse are not at all so neatly distinct as is typically implied by such stolid models of ‘communication’ as the ‘sender-message-receiver’ standby, which lured investigators to imagine we can catch the message in mid-air and treat it as the sole concern of linguistics or communication science — no rhetoric or style to be studied here, thank you. Far more plausibly, the producer must also act as co-receiver (‘how will this be understood or accepted?’), whilst the receiver must also act as co-producer (‘what would I mean by saying that?’) (Beaugrande 1984). Still, the balance of power is all too frequently skewed in favour of the producer (cf. 1.14-16).

1.8 A contemporary antinomy is institutional: when I consulted the Internet via the AltaVista search engine, I found 9,030 hits for ‘Department of Rhetoric’ at such institutions as the Universities of Minnesota, or of California at Berkeley, but just one ‘Department of Stylistics’ in Moscow, or maybe an agglomerate of several, since variously institutions are listed: Moscow State University, Moscow State Pedagogical Institute of Foreign Languages, and the Russian Academy of Sciences — all surely inspired by the redoubtable Olga Akhmanova (cf. Akhmanova 1976), whose fine work went virtually unnoticed outside the Soviet Union. The genuine concerns of style as I would view them nowadays manifest a modern dispersal, namely across Departments of English, Foreign Languages, Speech, Communication, Journalism and so on. Meanwhile, scant input has come from Linguistics, the official ‘modern science of language’, many of whose mainstream ‘descriptive’ and generative’ pundits were apparently blinkered by the arcane question of what language (‘langue’, ‘competence’, etc etc) is like when nobody’s using it (cf. Beaugrande 1991, 2004).

1.9 A closely related contemporary antinomy is intra-departmental: the split within most current ‘Departments of English’ between the ‘literature program’ aiming at acculturation versus the ‘language program’ aiming at standardization (Beaugrande 2004) The former has tended to presuppose the latter, aiming at a student audience already properly literate for ‘appreciating literature’. And until fairly recently, the staff in both ‘programs’ had ‘literature degrees’ and could make this presupposition without unbearable embarrassment. But as the student population grew far larger and more diverse in its language varieties, the staff were increasingly if not blatantly unsuited for tasks like ‘English composition’, much less ‘basic writing’. Many misprized their work and focussed relentlessly on (typically misinformed) fault-finding with ‘errors’ in ‘mechanics’ like grammar and punctuation, thus fomenting a high failure rate, with no regard or respect for the rhetoric and style of alternative varieties of English (e.g. African-American English) (cf. Shaughnessy, 1977). They pined to ‘move up to teaching literature’, where they could be revel in the discussion of ‘literary styles’, e.g. (7-8). Curiously, they have sometimes essentially suggested that style doesn’t really matter for some compulsory but mediocre author, e.g., one who nonetheless ‘reflected and criticized his time’ (9) or ‘gave a powerful illusion of life’ (10).

(7) Whatever its place in the literary league tables, After Silence is undoubtedly entertaining: [Jonathan] Carroll’s style has a high gloss, twinkling with near epigrams (lecture)

(8) The style of the novel [is] like the structure. […] His long sinuous phrases are designed to enclose the different levels and quirks of reality, just as liquid, spilt on a rough pavement, eventually seeps into every crack and cranny. (Ideas in Action)

(9) Criticized for his turgid style, […] Cooper is praised for having both reflected and criticized his time. (8 Classic American Novels)

(10) [His] canvas was broad enough to give full scope to Dreiser’s romantic compulsion to gather the facts and render the sheer density of experience. […] His cliché-ridden, pseudo-literary artificial style […] somehow gives a powerful illusion of life. (same)

‘Somehow’, I myself miss this ‘reflection’ or ‘illusion’, e.g., in styles like:

(12) The strong glare of the fire fell full upon his sturdy, weather-beaten countenance and forest attire, lending an air of romantic wildness to the aspect of an individual who, seen by the sober light of day, would have exhibited the peculiarities of a man remarkable for the strangeness of his dress, the iron-like inflexibility of his frame, and the singular compound of quick, vigilant sagacity, and of exquisite simplicity, that by turns usurped the possession of his muscular features (Last of the Mohicans)

But then perhaps unappreciative reactions like mine to styles that seem unappealing explain why Classic American Novels have to be ‘taught’ — tediously force-fed to young citizens who would get a much more engaging view of the ‘time’, say, from Michael Mann’s magnificently rhetorical and stylish film of Mohicans with Daniel Day Lewis and Russell Means.

1.10 Loosening this impasse in ‘language programs’ obsessed with ‘mechanical errors’ was a major motivation for creating ‘Departments of Rhetoric’ unburdened by such (un)professional elitism. There, the teaching of ‘writing’ and ‘composition’ was partially supported by re-animating Classical rhetoric in some suitably ‘modernised’ version (e.g., Perelman and Olbrechts-Tyteca, 1969; Corbett, 1971). But this laudable effort went ignored in most English departments, even those where I worked and warmly recommended it to the attention of my colleagues. Symptomatically, readers of my own books like Text Production (1984) — the book has 1,241 references, many from specialised sources like Cortex and the Canadian Journal of Behavioral Science — would ask me at congresses why someone so extensively engaged in synthesising empirical research would ‘bother’ actually asking to teach basic writing courses for first-year students. I always pointed out that the research, however theoretical or technical, may indeed be relevant to our practical tasks; and that unskilled writers were the ones where I could hope to make a real difference, provided I understood the wider nature of their challenge. How do people acquire rhetorical and stylistic skills quite different from those of the home language variety? Certainly not by being subjected to imperious fault-finding, which was often steered by a by misguided understanding of authentic English usage anyway (Greenbaum and Taylor, 1981).

1.11 One final antinomy might be attitudinal. Despite its long history of veneration, the term ‘rhetoric’ as used in contemporary corpus data such as the British National Corpus (hereafter BNC, 100 million words of authentic discourse, mostly 1980s and 1990s) is characteristically pejorative, e.g., indicating excessive elaboration (7-8), or a substitute for genuine action (9-10).

(13) In his work, there is an avoidance of subjective emotion, an abhorrence of rhetoric, an emphasis on simplicity (Leonard Cohen)

(14) The concrete image, unruffled by an adjective, was a thing Ezra would willingly have died for. Rhetoric was a thing he would gladly have murdered; (Ezra Pound)

(15) Relentless rhetoric, cynical misinformation and myopic stubbornness have coalesced to provide such a poisonous setting to the dispute that any compromise will be presented as defeat. (Independent)

(16) Grand gestures of defiance […] expressed resentment at contradictions between rhetoric and practice in an idiom which was sustained by modern technology. (Libyan Politics)

Adjectives collocating with ‘rhetoric’ in BNC data include ‘political’ and ‘official’ (by far the two most common), plus ‘dry’, ‘empty’, ‘hollow’, ‘inflated’, and ‘overblown’ — but never, say, ‘beautiful’, ‘lovely’, or just ‘pleasant’. Or worse: ‘rhetoric’ figures there as an idiom of intimidation or insult, with collocations including ‘heated’, ‘belligerent’, ‘incendiary’, ‘confrontational’, and ‘inflammatory’ — but never, say, ‘friendly’ or ‘conciliatory’.

1.12 In contrast, the usages of ‘style’ in the BNC are overwhelmingly ameliorative, e.g. (17), notably in praising goods and services, e.g. (18-20).

(17) in 1750, Worcester Lodge was built, in a sumptuous style, by William Kent at the end of an axial avenue. (The Perfect English Country House)

(18) Waitrose champagne […] has a glorious style of its own, light yet biscuity, classy with the muted perfume of hollyhocks and it makes a superb aperitif (Guardian)

(19) Our regulars look for ways to give your hair a new twist and express a different side of your personality. See in the New Year with a versatile style that's twice as nice! (Hair Flair)

(20) Dare to be different with an innovative style that will become a contemporary classic in shades of gold. (Wedding and Home)

Approval is accorded merely for ‘having style’ (21) or a ‘style of one’s own’ (22):

(21) Lord Carteron […] didn't do much for his king or his country. But he certainly had style. He knew how to enjoy life. He knew how to live well. (Ghost Stories)

(22) John Fowler took the romantic spirit of late eighteenth-century decoration, the simplicity of rural life with its celebration of nature, and fashioned it into a style of its own (Independent)

Adjectives collocating with ‘style’ in BNC data also included ‘modern’, ‘popular’, ‘pellucid’, ‘smart’, ‘handsome’, ‘lavish’, and ‘imperial’ — but never, say, ‘awful’, ‘terrible’, or ‘horrible’.

1.13 Moreover, the frequencies of occurrence differed immensely: ‘rhetoric’ at 952, but ‘style’ at 10,529, as if history were giving style its revenge for two millennia of academic domination by rhetoric. Of course, a large part of this phenomenon is due to the usages of ‘style’ in public media, especially adverts, fashion, and the arts. But all the same, the preference for ‘style’ over ‘rhetoric’ in such a large data bank may well signify a major disparity in attitudes. With that piquant irony peculiar to the history of ideas, the ambivalence we observed among Classical authorities about which styles should ‘win admiration’ (3) or be judged ‘overloaded’ (5) (1.3) seems to have been tacitly resolved in part by dumping certain disfavoured attributes into ‘rhetoric’.

1.14 Still, a somewhat different ambivalence in attitudes persists among views about style. Officially, ‘simple’ style is widely recommended and ‘complicated’ style discouraged:

(23) Priority is usually given […] to contributions using simple language. (Nature)

(24) To use words so true and simple that they oppose no obstacle to the flow of thought and feeling from mind to mind […] is the essence of style (John Galsworthy)

(25) It is important to avoid using jargon and complicated language. The principles of the course must be made to seem very simple (Anxiety and Stress Management)

Yet in numerous discourse domains, pointlessly complicated style is evidently favoured. In academic discourse, the rhetorical force may be a bid for power and authority at the expense of clarity, as in this ‘definition’ of a ‘dune’ (26) in the discourse of geography (unidentified Internet website).

(26) A dune is defined as a body of coarse sand shaped by ambient wind conditions and the grain-by-grain deflation of sand.

The style is marked by (a) obscure specialised terms like ‘ambient’ meaning ‘in the environment’, and ‘deflation’ meaning ‘erosion’; (b) gratuitous specifications like ‘ambient conditions’ (where else could ‘winds shape dunes’, in a tea cosy?) and ‘grain-by-grain’ (how else would sand move, in clumps?); and (c) inconsistency with ordinary knowledge, which would interpret ‘deflate’ as ‘remove the air from inside’ — hard to do with a grain of sand, even for William Blake. Besides, ‘body’ is an oddly organic term for an object that easily changes or disintegrates; and ‘coarse sand’ is plain wrong — the sand in dunes near my home in Arabia was fine enough to enter around the edges of closed windows. In the same over-complicated style, rainfall was a ‘precipitation event’; extreme dryness in a small area was ‘localized hyperaridity’; and sand grains that blow away again from their dune ‘become wind re-entrained’.

1.15 What citizens in general and students in particular need to develop is skills for critical rewriting (Beaugrande 2004), which transforms such styles into user-friendly counter-discourse, as in (26a). Other stretches from the same text could be rewritten as shown for (27-28) into (27a-28a).

(26a) A dune is defined as a mound of sand shaped by the erosion of wind and the motion of sand grains.

(27a) On medium-sized dunes, the waves get longer when the sand grains are bigger.

(28) Transverse dunes are characterized by low length:width ratios and marked asymmetry, where windward slopes are much gentler than the slip faces associated with lee slopes.

(28a) Dunes formed at a right angle to the wind are very long but very narrow. They rise gently on the side facing the wind and drop sharply on the other side.

1.16 Evidently, ‘academic’ style can empower the text producer whilst disempowering the text receiver. Even the most trivial statements can be dressed up in a dazzling academic style of deep wisdom:

(30a) The serial accumulation of elapsing temporality is significantly correlated with a linear increment in the human aging process.

(31a) The atmospheric encirculation of our planetary ambience imparts an ocular wave-length chromaticity between 450 and 500 nanometres.

The costs of impeded communication by complicated styles incurred by society as a whole must be staggering. But substantial profits can accrue to individuals, such as accountants handsomely paid to interpret horrendous bafflegab like (32).

(32) Financial products supplied by a financial supply facilitator are not financial supplies. However, a supply of an interest facilitated by a financial supply facilitator is a financial supply by the financial supply provider if the supply of the interest is one to which regulation 40-13 applies. (Australian Treasurer on the Goods and Services Tax)

Yet even English for business use, where costs surely should matter, is taught largely without regard for such problems. An ‘international survey’ of ‘on-site visits by staff of the Centre for Applied Linguistics (Washington, D.C.) at 55 English language training contractors in the Americas, Asia, Europe, and the Middle East, being considered for business training by [then still unindicted] Arthur Andersen’ found ‘only 38% using business-oriented instructional materials’ (Schleppegrell and Royster 1990: 8-9)

1.17 When I myself taught ‘Business English’ at the University of Botswana, the prescribed textbook, authored by the ‘Chief Examiner in Commercial Subjects, College of Preceptors’, allotted no role at all to rhetoric, whereas ‘style’ rated two pages each in ‘business letters’ and in ‘essays’, as did the hoary topic of ‘loose and periodic sentences’ (Gartside, 1990: 163f, 178f, 250f). In fact, I noticed a suspicious resemblance to ordinary ‘English textbooks’. ‘Speech and Writing Compared’, also equalling just two pages, warned against the ‘colloquial expressions’, ‘incomplete sentences’, and ‘grammatical blemishes’ of ‘speech’ (Gartside, 1990: 149-51). Moreover, students were asked to write ‘essays’ for no particular audience on topics which would demand some (doubtless unwanted) imagination to make them at all interesting and relevant to business, viz.: ‘when father papered the parlour’ (1990: 94) (e.g., he made a ghastly bodge-up by papering the doors and windows shut, and his family minutely negotiated the price of his rescue for two days); ‘a walk along the seashore’ (1990: 93) (e.g., you were kidnapped by pirates, who threw you overboard after they calculated your probable ransom value); and, yes, ‘a woman’s place is in the home’ (1990: 97) (e.g., where she manages a lucrative phone-sex business that pays no taxes). Never mind that Botswana homes have neither parlours nor wallpaper; that the nearest seashore is some 1,000 km away; and that the women do all the housework anyway — they even build the houses, I’ve watched them do it in Serowe, the ‘Village of the Rain Wind’ (Bessie Head).[5]

1.18 To quell my impulse for throwing the ‘textbook’ into the nearby Kalahari Desert and hopelessly polluting it, I went over to the ‘Department of Business Education’ to borrow a copy of the textbook they were using for the same students — and which boasted on its rear cover of wide use in British Polytechnics. But I found the style and rhetoric of real thing markedly pompous and obscure, witness discourse like this sample:

(33) Net realisable value means the estimated amount that would be received from the sale of the asset less the estimated costs on its disposal. The term ‘exit value’ is often used as it is the amount receivable when an asset leaves the business. A very important factor affecting such a valuation is the conditions under which the assets are to be sold. To realise in a hurry would often mean accepting a very low price. Look at the sale prices received from stock in bankruptcies — usually very low figures. The standard way of approaching this problem is to value as though the realisation were ‘in the normal course of business.’ This is not capable of an absolutely precise meaning, as economic conditions change and the firm might never sell such an asset ‘in the normal course of business.’

The difficulties of establishing an asset’s net realisable value are similar to those of the replacement value method when similar assets are not being bought and sold in the marketplace. However, the problems are more severe as the units of service approach cannot be used, since that takes the seller’s rather than the buyer’s viewpoint. (Wood and Sangster, 1990: 19)

This passage sets out to tell you what ‘net realisable value means’, but ends up instead asserting it has no ‘precise meaning’ and seems to confirm this itself. You are to pretend you are selling the asset ‘in the normal course of business’ and in no ‘hurry,’ even though the asset might in fact ‘never be sold’ or else only under non-‘normal’ and hurried conditions such as ‘bankruptcies.’ Moreover, you should not refer to ‘buying’ a new ‘replacement’, because of course you couldn’t get the same price as for a used asset. Nobody but me has noticed how obscure the passage is — the author, the editors, or (since the book has been reissued or ‘updated’ four times) the unhappy readers.

1.19 Such is our modern ambivalence: advocate simple styles and reward complicated ones with unmerited respect because we don’t command the skills to see through them, let alone defend ourselves. Compare now my own ‘critical rewriting’ of the same passage:

(33a) Estimating the net realisable value or exit value is done by calculating the sum that would be left after selling an asset and subtracting the costs of the sale process. Because the conditions for selling can affect the price in unpredictable ways, as when stock gets quickly sold off after a bankruptcy, you can base your estimate on the value you would presumably receive selling the asset at a time when the business is operating normally.

Estimating the replacement value is done by calculating the sum that would be spent to actually buy the asset again. Problems naturally arise if the asset is currently not on the market. Also, you cannot calculate by units of service, which would be adopting the seller’s viewpoint rather than the buyer’s. (127 words)

My version is uses one-third fewer words than the original and eliminates the redundant series of partial synonyms for ‘selling’ noticeable in (33). I also dispensed with the repetitions of ‘net realisable value’, ‘estimated’, ‘low price/low figures’, ‘in the normal course of business’; and I replaced ‘not being bought and sold in the marketplace’ with ‘currently not on the market’.

1.20 More significantly, I have treated ‘replacement value’ as a second topic in its own right whose explanation is part of the purpose of the passage rather than a confusing sidelight. I have highlighted this equal status by italicising the respective terms and by using parallel choices of vocabulary and grammar in the opening sentence of the respective paragraphs. Not at all difficult, but my students confirmed that they were being trained for the first time in their lives.

2. The new ‘light’ of large corpora

2.1 I ventured to present my brief commentary on attitudes towards the very terms ‘rhetoric’ and ‘style’ in 1.11-13 only after consulting large sets of authentic data, thus leading into my main contention: the collocations (preferred lexical combinations) and colligations (preferred grammatical combinations) as represented in a large corpus of authentic discourse constitute neglected factors eminently relevant to rhetoric and style, and to their systematic study.

2.2 Most work I have encountered has tended to make a wide emblematic leap. On one side, rhetoric and style are approached as the set of resources available to the entire ‘linguistic community’. The underlying assumption of homogeneity has been egregiously influential in linguistics, philosophy, language planning, and so on — as when linguists dressed themselves up as the ‘ideal speaker-hearer in a completely homogeneous speech-community, who knows its language perfectly’ (Chomsky, 1965: 3) — and now the ‘English Only’ movement demanding that it be written into law (cf. Pennycook, 1992; Macedo et al., 2003), but has rested on the mere intuition or self-interest of privileged groups. Now, very large corpora have laid it to rest for anyone who takes a genuine interest in them (which the aforementioned groups have strong reasons not to).

2.3 And on the other side of the leap, homogeneity radically evaporates. Rhetoric and style are approached as attributes of the individual single text or author — chiefly ones admitted to the ‘canon of great works’, though this latter restriction has been much relaxed in recent years (cf. Gates, 1992). But how individual are they in respect to the ‘linguistic community’, and which selections or combinations are more or less preferred and where? Here too, very large corpora can point us toward some answers.

2.4 Classical rhetoric at least tried to bridge the leap because it was required to provide practical training. But aside from Aristotle and Demetrius, too few works have survived — say, of Gorgias, Isocrates, Thrasymachus, Lycophron, Alcidamas, and Licymnius of Chios — to assess the actual methods, outcomes, or success rates. Moreover, most trainees had had since childhood personal tutorship in language and ‘classic works’, well before they entered these ‘schools of rhetoric’.

2.5 My contention now is that a large corpus can help us bridge the leap between communal and individual by examining major cross-sections of the language in action. The role of the linguist shifts dramatically from adducing to unaided ‘intuition’ as the royal road to ‘theories of language’ argued from invented data over to applying and reconsidering intuition about authentic data enriched by contexts and attitudes.

2.6 Even fairly unassuming locutions can indicate or support styles. For example, in the BNC, the colligation ‘a receipt for’ occurs in a range of ordinary styles with the straightforward rhetorical intent for a document affirming that payment has been made for some purpose, where the rhetorical attitude may be ameliorative (34) or (less often) pejorative (35).

(34) You do not have to worry about keeping track of your spending. You will get a receipt for each transaction and full details will appear on your next statement (financial leaflet)

(35) It was all pretty sordid. She found a receipt for a hotel in Hampshire in my suit. I'd told her I'd been away filming in Scotland that night. (Taped)

But the colligation ‘in receipt of’ indicates official styles, with the far more narrow rhetorical intent of stipulating the established criteria whereby persons in general are or are not entitled to payment from some institutional source, again with a characteristic distinction between ameliorative (36) and (much less common, mostly referring to the French!) pejorative (37).

(36) My Bill proposes that people in receipt of this benefit and in receipt of poll tax benefit, invalidity benefit, long-term sickness benefit, the severe disablement allowance, and mobility and attendance allowances should be exempt from prescription charges. (Hansard Parliamentary Debates)

(37) It would appear that much of France's corporate aggrandisement has been undertaken by state-owned companies, many of which have been in receipt of illicit subsidies. (Against a Federal Europe)

The simple participle ‘receiving’ would carry much the same meaning, but without these distinctive rhetorical and stylistic tendencies.

2.7 Corpus data can also help to uncover the rhetorical and stylistic bridge between the community and the individual evolving over time. In my own British and American Writers Corpus (herafter BAWC, currently 65 million words of authentic texts), the lexical expression ‘overcome’ figures prominently in the sentimental styles of 19^th-century novels to denote ‘suffer adverse effects of overly intense emotion’, viz.:

(38) the tall soldier seemed to be overcome by a terror. His face turned to a semblance of gray paste. […] He now hung babelike to the youth's arm. His eyes rolled in the wildness of his terror. (Red Badge of Courage)

(39) Mr Edgar was taken with a nervous trembling, and his countenance grew deadly pale. For his life he could not avert that access of emotion; mingled anguish and humiliation overcame him completely. He leant on the back of a chair, and covered his face. (Wuthering Heights)

Here again, rhetorical attitudes emerged: just a few ameliorative collocates like ‘joy’, ‘admiration’, ‘delight’, and ‘happiness’, against a host of pejoratives: along with ‘terror’ and ‘anguish’, ‘humiliation’ (38-39), I found ‘despair’, ‘gloom’, ‘grief’, ‘misery’, ‘nausea’, ‘pain’, ‘repugnance’, ‘shame’, and ‘sorrow’. The rhetorical force was presumably to encourage empathy in the reader to share the emotion or even be vicariously ‘overcome’, as Miss Morland in Northanger Abbey was while reading Mrs Radcliffe:

(40) Catherine was then left to the luxury of a raised, restless, and frightened imagination over the pages of Udolpho, lost from all worldly concerns

She afterwards claimed to have ‘finished it in two days — my hair standing on end the whole time’.

2.8 But in contemporary data from the BNC, mostly from the 1980s and 1990s, the same expression chiefly indicates the rhetorical force of denoting mechanical causalities unrelated to emotion, e.g. like noxious fumes (41), physical forces (42), and sports events (43).

(41) A disabled woman was rushed to hospital after she was overcome by gas fumes at her council house. (Northern Echo)

(42) The centrifugal force at the edge of the nebula overcame the gravitational force exerted by the mass, and a ring of material broke away. (Fate of the Dinosaurs)

(43) Milton United overcame their Reading Senior League opponents at Fairmile in the Berks and Bucks Intermediate Cup (Radio Oxford)^BNC

Here, the rhetorical force has flattened into an explanatory link in a series of events. Meanwhile, the impact of Emotions has been taken up by quite different means in the fervid style of Mills and Boon, most obtrusively by endowing the vital organs of the heroine with an astounding mobility (43-45). In the data, not one is ever said to be ‘overcome’.

(43) His voice held a caressing note which flipped her heart over in her chest. (Calypso’s Island)

(44) The kiss […] was a sensual onslaught that made her brain whirl. (Healing Fire)

(45) her stomach leapt inside her, remembering that dance (Conspiracy of Love)

Even if, like me, you are nauseated by Mills and Boon, you must admit that the consistent style of its whole series is unmistakeably recognisable.

2.9 But corpus data show that the historical contrast for ‘overcome’ is not so simple after all, because Passive and Active Transitivity diverge in style. The causal use is indeed found in older data, but mainly in the Active and in more technical styles, viz.:

(47) We had to call out the fire department; and they came down and put a stream through the window. That let all the fumes and chemicals out and overcame the firemen (Edison: His Life And Inventions)

(48) To bring the ore to their converters necessitated a long voyage by water and rail, with several reloadings. They overcame these obstacles by developing machinery for handling ore and by acquiring the raw materials and the connecting links of transportation (The Age of Big Business)

Mind you, it is hardly reassuring for the researcher to find, as I frequently do in corpus data, that the very same lexical item can vary among styles in authentic discourse right within its repertory of forms. We begin to envision just how much hard work lies ahead.

2.10 Linguistics would probably declare the rhetorical or stylistic tendencies of a given sampling of corpus data to be flukes or isolated ‘epiphenomena’ with ‘no bearing on serious linguistic theory’ (cf. Chomsky 1965: 127). ‘Computational linguistics’ has occasionally used computers to look into style (cf. contributions in Doležel and Bailey eds., 1969), but mainly for statistics on frequencies of words or word-classes. In my view, frequencies are only symptoms of something deeper, such as the attitudes I diagnosed from the data on ‘rhetoric’ and ‘style’ (1.13). And they are elusive because a rhetorical or stylistic tendency may assume a variety of surface forms, such as the whole roster of pejorative events that ‘warrant’ a significant response, as I shall show in a moment.

2.11 In truth, authentic data are so thoroughly pervaded with rhetorical or stylistic tendencies as to hardly make sense if we leave those out of account. For another example, the data on the infinitive of the verb ‘to warrant’ (198 attestations in the BNC) revealed a style that is generally official, whilst the rhetoric is authoritative, viz. (48-49). The dominant rhetorical attitude is heavily pejorative, suggesting something bad or at least irregular has occurred which is now being declared by some (often unnamed or else self-proclaimed) authority, to ‘warrant’ (or not) a corrective response or a penalty, e.g.:

(49) Abdominal symptoms or haematological changes of hyperspleenism can be severe enough to warrant surgery. (GUT)

(50) Indeed, the legislation as a whole conveys the message that marital rape in and by itself is of insufficient gravity to warrant the intervention of the criminal law. (Consumer and Commercial Law)

Further collocates in the BNC include ‘proof of danger’, ‘damage to the structure’, ‘crimes of violence’, ‘plant poisoning’, and ‘horror stories of rape, murder, and robbery’. In contrast, the ameliorative data were scarce indeed, e.g.:

(51) It was a wonderful drive on an afternoon balmy enough to warrant lowering the electric windows (advert)

(53) they offer such exciting prospects as possibly to warrant major investment. (Pipeline)

The colligation ‘I warrant’ or ‘I’ll warrant’, common enough in older data — 392 occurrences in my BAWC — has virtually dropped out of usage — only 23 in the BNC, all of them representing old-fashioned speech (e.g. in The Child Bride), after we discount those in Richardson’s Pamela, whose presence in a contemporary corpus frankly baffles me.

2.12 As one possibly helpful exercise, we might manipulate a style to sort out some types of indicators. Thus, we might start Milton’s strenuous and formal original (53), reducing first grammatical complexity (53a), then lexical complexity (53b), and finally imposing a markedly casual informality (53c).

(53) thee chantress oft the woods among I woo to hear thy evensong. (Il Penseroso)

(53b) I often lure you in the forest, nightingale, so that I can hear your evening song.

(53c) Lotsa times I decoys you in da boonies, birdie, to dig you evenin’ chirps.

Milton’s ‘style’ too is unmistakeably recognisable despite a host of epigones, but his stylistic tendencies are elegantly articulated and can be meaningfully assessed in relation to Spenser and Shakespeare. Corpus data lack such grand appeal, but they can lead us toward understanding the rhetoric and style that help to weave the texture of social life.

2.13 Another exercise could be applied to questions about disputable authorship. In The Two Noble Kinsmen, my intuitive response to its mixed styles is that (54) (pleading for the Duke to give the ‘slain kings’ a proper burial, a ‘grace’ even granted to suicides) is definitely by Shakespeare, whilst (55) (pleading for the audience not to ‘hiss’ the play, whose ‘story’ came from ‘admired Chaucer’s’ Canterbury Tales) is definitely not.

My response to both samples is based on prosodic flow, and on lexicogrammatical combinations which are creative in the one sample (e.g. ‘knolls in the ear’, ‘drams’ precipitance’) and trite in the other (e.g. ‘hear a hiss’, ‘shake the bones’).

2.14 The rhetorical force of diverging styles is naturally also essential, e.g., in the famously unadorned style of ‘Honest’ Abe Lincoln that was certainly not held ‘trivial’ or cheap’ (1.3), as when he effectually defended Ulysses S. Grant against critics:

(56) What I want and what the people want is generals who will fight battles and win victories. General Grant has done this, and I propose to stand by him.

(57) The proclivity of our illustrious populace, as well as of my humble self, esteems generals who superintend forcible bellicosities and achieve supremacy therein. These encompassments having been effectuated by General Grant, my resolve is to tender him my adamant allegiance.

the rhetorical force would have been undermined by giving the impression of craftily throwing a stylistic smoke screen around the bellicose General, or maybe of Abe losing his sanity — if indeed the populace could figure out what in thunderation he was talking about.

3.1 The daunting prospect now impends of fully reassessing our conceptions of the relations between language and discourse. Linguistics has a long history of dividing and dichotomising and then sweeping problematic issues and factors under the rug, thus keeping rhetoric and style out of the ‘mainstream’, which they will almost certainly come to occupy during this reassessment.

3.2 My own work on these issues is necessarily still at an exploratory stage. In my latest book (Beaugrande 2004), I have expounded a revised view of (real) language and (real) discourse being based on three factors: lexicogrammar, prosody, and visuality. As a corollary, all three are strongly implicated in questions in rhetoric and style, without which any empirically supportable view of language and discourse (and not a ‘theory of language’ as some abstract system) is at best blurred and at worst irrelevant. For the Shakespeare data (54), lexicogrammar and prosody easily single him out; we are not fooled by the visually similar appearance of the shoddy scribbling (55).

3.3 Recalling the meagre results of linguistics attempting to define language a priori from the top down by abstract theory should caution us against attempts to approach rhetoric or style by the same strategy. Rather, we need practical theories, such as: in appropriate contexts, identifiable selections and combinations are relevant for rhetoric and style. We shall require large teams of investigators reacting to numerous and varied sets of data and comparing their reactions. We can expect to deal in gradations, not clear-cut categories. And we will find a ‘widening effect’ whereby small discoveries point us outward toward highly significant issues.

3.4 Luckily, a large minutely ‘tagged’ corpus like the BNC can allow us to distinguish rhetorical and stylistic factors according to registers, media, genders, age groups, and social classes. For example, the item ‘illiberal’ never occurs there in spoken discourse, imaginative writing, or the discourse of women; it occurs only in the written discourse of men of (self-appointed) high status on the political far right to accuse ‘liberals’ of hypocrisy, e.g.:

(58) It is the Civil Liberties people who are being illiberal and intolerant by presenting evolution as fact when it is only a theory. (In Good Faith)

Ironically, ‘liberals’, the apparent antonym is the adored target of accusation among the ‘New Right’ — or better, the ‘New Rant’ — enough to make up a mini-corpus, which I have begun. There, ‘liberals’ get branded as ‘nothing less than Communists, Socialists, and Dictators’, who ‘despise freedom because they despise morality’ (huh?); and they ‘seek to disarm individuals and render them powerless before the thugs, thieves and murderers who rule the inner cities’ (Rush Limbaugh). Also, they ‘take sheer joy in telling lies’; they ‘hate America’ and ‘society’; and they ‘want to take more of our money, kill babies, and discriminate on the basis of race’ (Ann Coulter).

3.5 The lexical item (mainly occurring for Agents in the Plural) thus abets the rhetorical force of viciously contested power by debasing the style of public political discourse to the level of snarling junkyard dogs. Lacking rational, let alone humane, arguments for their tireless machete attack on welfare, civil rights, consumer health, worker safety, and the environment, the ‘New Right’ simply drowns out dialogue by means of ‘pre-emptive character assassination’ of their opponents. Massive funding by right-wing think tanks — or better, stink tanks — and well-paid leverage in mass media, this rhetorical and stylistic absurdity has hoisted its producers into positions of influence upon the US White House, Congress, and the judiciary system, and supported an agenda of economic and military world domination, by ‘nucular’ weapons if the US ‘admininsitration’ of chickenhawks[6] so decides.

3.6 So there we are: the ‘widening effect’ I predicted, and in spades. Far-reaching studies of rhetoric and style, effectively transmitted to sensitise the citizenry, were never so urgent. We are at the final stage of the history of the rhetorical contest between money power and people power, and we must grasp who is speaking for and to whom and how and why — or most of the people in the world will have no tomorrow worth living for.

1997 New Foundations for a Science of Text and Discourse (Greenwich, CT: Ablex)

2004 A New Introduction to the Study of Text and Discourse (Published on the Internet at www.beaugrande.com)

1981 ‘The recognition of usage errors by instructors of freshman composition’, College Composition and Communication, 32, 169-174.

1981 Is God an American? An Anthropological Perspective on the Missionary Work of the Summer Institute of Linguistics (Copenhagen: International Work Group for Indigenous Affairs).

1994 The Cultural Politics of English as an International Language. (London: Longman)

1969 The New Rhetoric: A Treatise on Argumentation (Notre Dame: Univ. of Notre Dame Press)

1990 ‘Business English: An international survey’, English for Specific Purposes, 9, 3-16.

[1] Quoted from the translation by S.H. Butcher.

[2] Quoted from the translation by Hugh Lawson-Tancred.

[3] Quoted from the ranslation of W. Rhys Roberts, who does not provide the original Greek data Inevitably, translation obscures the stylistic indicators specific to Classical Greek, such as the ‘length of vowels’ and the use of ‘particles’.

[4] Which led him to ban the ‘split infinitive’ which Latin cannot produce.

[5] ⁷ To be fair, I should mention that these insipid topics were lifted from ‘examining bodies’ like the ‘London Chamber of Commerce and Industry’.

[6] According to Tam Dalyell, longest-serving member of the British Parliament. ‘chickenhawks ‘ are ‘men who have never seen the horror of war but are in love with the idea of war’ (quoted in Kurt Nimmo, ‘The Committee for the Liberation of Iraq: PR Spinning the Bush Doctrine’, Counterpunch, November 19, 2002.) None of the Bush regime but Colin Powell have seen combat. Compare Jim Lobe, The Chicken Hawk Factor, at www.alternet.org, September 9, 2002; and the ‘Chicken Hawk Database’ of the New Hampshire Gazette at www.nhgazette.com/chickenhawks.html.