Back to Previous Page

Download this Article

by Edward Jayne
October 4, 1998

Many statistical studies have been devised to prove that grammar bears little or no value in helping students to read and write. These studies demonstrate that students who are taught composition without any knowledge of grammar perform at least as well in their written prose assignments as students forced to study grammar. At best the time spent learning grammar deducts from the writing experience that actually brings improvement, and at worst grammar's prescriptive demands inhibit students from expressing themselves as effectively as possible. In any case, the argument goes, students compartmentalize their grammar from their writing skills, so its clutter of arbitrary rules and prescriptions does not not contribute to their growth as writers. Therefore, grammar can and ought to be excluded from their education, and, in fact, it should even be deleted from the college education of future teachers to discourage the temptation to teach it. This is not exactly acknowledged, but today in fact it is a prevalent strategy of English teachers' education programs based on the assumption that the best way to prevent teachers from teaching grammar is to keep them from knowing it themselves. As the invention of Protagoras, Plato, Aristotle, and Dionysius Thrax, grammar is ancient and hoary--it's past time to junk it. Let students learn how to write by exclusively linking their study of verbal expression with their spontaneous communication of the feelings and experience they think important.

However, a different viewpoint continues to haunt many educated Americans beyond the age of fifty who have addressed themselves to the ongoing task of putting their ideas into written sentences. As well as they can recollect, their youthful encounter with grammar, whether from having learned to diagram sentences in the seventh or eighth grade, or from having studied a foreign language, or from having themselves taught grammar at one point or another in their careers, has been extremely useful to their later development as writers. Moreover, these individuals are often disturbed to find that, indeed, the presumably competent basic prose of students without any background in grammar might be intact--correct, even--but very little more, and they suspect its inadequacy may at least partly be traced to a lack of grammatical sophistication. Somehow, somewhere, these mossback traditionalists conclude, a solid exposure to grammar early in one's education very likely turns out sooner or later to foster the emergence of one's mature writing skills if one possesses adequate potential to cultivate these skills at an advanced level. For the average or "challenged" student, grammar might be irrelevant, or an impediment, or even a source of discouragement, but for better students it truly does help in their effort to cultivate a mature style adequate to their thinking process as it expands on itself.

As one of these mossback traditionalists with ample experience in teaching a variety of approaches to composition between the early sixties and late eighties, I share these feelings despite the presumably hard statistical data that seemingly proves otherwise. If statistics doesn't bear out my views, I argue, this is because of the inadequacy of statistics: (a) because its sampling does not fully take into account differences in verbal ability levels whereby advanced students are more likely to benefit from grammar than "challenged" students; (b) because longitudinal studies are not sufficiently longitudinal (more time is needed--decades even--to be able to measure the ultimate results); and/or (c) because the traditional grammar investigated has usually featured the pursuit of correctness rather than a cultivation of what I describe as syntactic risk.

By syntactic risk I mean at least the appearance of a free expression of ideas with more emphasis on capturing their flow than on the final correct organization of one's sentences. Syntactic risk entails the use of new and unfamiliar constructions to emphasize one's thinking with seeming effortlessness. To misappropriate the words of Robert Frost, good syntax rides easy in harness. Authors capable of syntactic risk can say what they please, totally confident of their ability to bend, stretch, and reconfigure the current syntactic context in order to carry forward the ideas they are trying to articulate without going overboard and losing both their readers and themselves in their elaborations. They are never fully aware how their sentences might turn out or even, word for word, or where they come from, but they are always confident of their adequacy in making their transition from one end to the other, even when anacoluthon spells disaster preliminary to renewed syntactic effort to clarify their failure to contextualize their meaning in correct sentences. They likewise know when to be concise. This they can do both to offset their more elaborate sentences and to emphasize the simple truth of what they are saying when it seems appropriate. All in all, they possess enough confidence in their verbal skills to navigate the difficult straits between excess and deficiency. Simplicity they can put on display, but they can also augment--indeed, engulf and all but dissolve--its obligatory two-step subject-predicate nexus (Frankenstein's monster trapsing through mud: subj-pred, subj-pred, etc.) with an adjunctive use of words, phrases, and clauses that keep their words and locutions vitally relevant to what they are thinking 1. Of course the predicative noun-verb interaction continues to dominate, but without exerting totalitarian authority, since its implicit this-equals-that (or this-happens-to-that) equation has been absorbed and provided flesh within a more inclusive, and more challenging, verbal context.

It is syntactic risk that I want to emphasize here, since the marvelous triumphs in the use of language throughout the history of British and American letters seem almost completely irrelevant to the barren plain style that is now treated as a virtue in too many of the composition courses taught in our schools and universities. The value of syntactic risk was entirely taken for granted in the heyday of British civilization, when grammar was stretched to the limit by thousands of published authors who depended on Ciceronian excellence. Such major British figures as Shakespeare, Bacon, Milton, Burton, Browne, Hobbes, Dryden, Locke, Fielding, Sterne, Gibbon, Hume, Johnson, Burke, Hazlitt, Lamb, De Quincey, Macaulay, Carlyle, Ruskin, Newman, Mill, Dickens, Arnold, Pater, and Thomas Huxley unabashedly resorted to extravagant syntactic constructions in their prose style in order to capture and elaborate the fullest nuance of what they were trying to say. They wrote as if talking in a monologue, as if engaged in uninterrupted reverie. There was no compromise whatsoever to the challenged reader: these authors said exactly what they wanted, confident that their audience was fully capable of grasping and benefitting from its meaning. Grammar they shared with their readers not as a Procrustian torture machine that chopped all sentences to pasta segements between ten and thirty-five words in length, as many as possible with three or fewer syllables, but as bricolage that granted them their marvelous liberties in declaring their ideas. Among Americans comparable freedom was exercised by Irving, Holmes, Emerson, Hawthorne, Whitman, Lowell, Henry Adams, the James Brothers, Santayana, Mencken, Faulkner, the Wolfes (both Thomas and Tom), Hoagland, Lewis Thomas, and, at their best, Mailer and Vidal. This style also survives in our intellectual and literary journals such as The New York Review of Books, Diacritics, Critical Inquiry, and New Literary History. Granted, Ciceronian freedom has always been limited to relatively small audiences well enough educated to sift its implications. And granted, the average student in today's composition classes is hardly likely to join such an audience, much less write for it. Nevertheless average needs should not serve as justification to stunt excellence. Moreover, at least a few "average" readers might actually be taught to stretch their capabilities once liberated from the demotic prescriptions of gnomic brevity ("clear and simple, dummy!") that are almost incessantly promoted today as the paramount standard of excellence.

The use of syntactic risk to articulate complex ideas seems plain, for example, in John Stuart Mill's classic defense of intellectual freedom, On Liberty, in which sentences may be almost randomly quoted whose multiply imbedded phrases and clauses completely supercede the subject/predicate nexus with word counts in excess of sixty, eighty, or even a hundred words apiece. For example, chosen at random:

The ancient commonwealths thought themselves entitled to practice, and the ancient philosophers countenanced, the regulation of every part of private conduct by public authority, on the ground that the State had a deep interest in the whole bodily and mental discipline of every one of its citizens--a mode of thinking which may have been admissible in small republics surrounded by powerful enemies, in constant peril of being subverted by foreign attack or internal commotion, and to which even a short interval of relaxed energy and self command might so easily be fatal that they could not afford to wait for the salutary permanent effects of freedom.

Here Mill gives an elaborate compound predicate a single direct object and uses two elongated subordinate elements, a prepositional construction beginning with on the ground that and an elaborate appositive beginning with a mode of thinking which, to assert his important concession that it might have been appropriate to curtail liberty in earlier societies. With shorter sentences, perhaps half of what Mill says here would need to be eliminated, consigned to a footnote, or spelled out in additional sentences that unnecessarily interrupt his essential line of argument.

Again picked randomly, a sentence from John Ruskin's Modern Painters, explains with far greater economy than any grouping of shorter sentences could the painterly texture in Turner's paintings:

For who but this consummate artist [Turner] would have had the courage, even if he had perceived the laws which required it, to undertake in a single small space of water, the painting of an entirely new picture, with all its tones and arrangements altered,--what was made above bright by opposition to blue, being underneath made cool and dark by opposition to gold;--or would have dared to contradict so boldly the ordinary expectations of the uncultivated eye, to find in the reflection a mockery for the reality?

The rhetorical question is a long one, since Ruskin resorted to amplification to convey in full the expressive freedom he found in Turner's canvasses. For me, at least, there is substantial aesthetic pleasure here simply in watching the sentence unravel and dissolve into abstractions as if spoken while one's eyes glide across a Turner painting.

Then again, syntactic complexity can be marvelously dramatized in periodic constructions, for example in Henry Fielding's Tom Jones, whose style obliges one to hold in mind all the phrases and clauses that accumulate until everything is finally brought into the fold, often with the subject/predicate nexus of the dominant clause (not necessarily the main clause) saved until the very end to heighten the reader's anticipation:

Reader, I think proper, before we proceed any farther together, to acquaint thee that I entend to digress, through this whole history, as often as I see occasion, of which I am myself a better judge than any pitiful critic whatever; and here I must desire all those critics to mind their own business, and not to intermeddle with affairs or works which no ways concern them; for till they produce the authority by which they are constituted judges, I shall not plead to their jurisdiction.

The same effect may also be observed in the prose style of Sterne, Dickens, and Thackeray, as opposed to the essentially cumulative, but no less challenging, prose style of Henry James.

In poetry the subject-predicate nexus of the main clause is often almost entirely overbalanced by an adjunctive use of subordinate structures, as, for example, in the tortuously overwrought sentences of Milton's Paradise Lost, one of which can be quoted here formatted as prose in order to emphasize its behavior specifically as a sentence:

I fled [Death]; but he pursued (though more, it seems, inflamed with lust than rage) and swifter far, me overtook, his mother, all dismayed, and in embraces forcible and foul engendering with me, of that rape begot these yelling monsters that with ceaseless cry surround me, as thou sawest, hourly conceived and hourly born, with sorrow infinite to me; for when they list, into the womb that bred them they return, and howl and gnaw my bowels, their repast; then bursting forth afresh, with conscious terrors vex me round, that rest or intermission none I find. [italics added]

Here Sin's laboured explanation of having been raped by her son Death dramatizes the extent to which Milton's imagery couched in Latinate syntax can overwhelm the simple noun-verb-object nexus, in this instance, he me overtook and begot monsters. Buried in syntactic excess, the presumably dominant main clause's accusation of incestuous rape enlarges to baroque monstrosity.

At times the predicative nexus in the sentences of poetry may be all but forgotten, as, for example, in the first stanza of John Keats's "Ode to Autumn," which I also take the liberty of formatting as prose:

Season of mists and mellow fruitfulness, close bosom-friend of the maturing sun; conspiring with him how to load and bless with fruit the vines that round the thatch-eves run; to bend with apples the moss'd cottage-trees, and fill all fruit with ripeness to the core; to swell the gourd, and plump the hazel shells with a sweet kernel; to set budding more, and still more, later flowers for the bees, until they think warm days will never cease, for summer has o'er-brimm'd their clammy cells.

Where is the predicate--the sentence's verb and its complement? There isn't any, so the subject of the fragment (season of mists) must be treated as an inverted appositive of the direct object thee in the first sentence of the next stanza: "Who hath not seen thee in thy store?" What is the syntactic effect? That everything said in the first stanza serves as an elaborate introductory apostrophe telling what the personification of autumn meant to Keats. Syntactic freedom gave Keats the opportunity to declare this relationship with wonderful specification, but with so much risk that most students uncertain of grammar are likely to be befuddled by what he meant. They can immerse themselves in the flow of images, but they have no clear understanding how and to what extent this flow describes Autumn's personification as a passive spirit with whom Keats can identify. Since they overlook the function of the first stanza in its entirety as a sentence fragment, they cannot fully grasp its subordinate role in anticipating what is told in the next stanza.

One last example, the most extravagant of all, is the convoluted second (and final) sentence buried in Section 25 of the 1789 Judiciary Act that first established the authority of the Supreme Court of the United States to veto state laws and state supreme court decisions. This single sentence, probably drafted by Oliver Ellsworth, was perhaps the most important in American constitutional history, since it first imposed effective federal sovereignty over state governments by outlining the specific procedures by which this sovereignty could be exercised. The Constitution had skirted the issue by promising in Article 3 that the court system would be organized later by Congress, and now, with this one almost impenetrable sentence buried in the Judiciary Act, the very first legislation launched by the Senate in 1789, the Supreme Court was granted its authority to review and deny state laws and court decisions. Why did this crucial function of the Supreme Court pass muster with the nation despite its absence from the Constitution? As much as anything because this enabling sentence was so Byzantine in its legal specifications, and included so many adjunctive phrases and clauses dispersed over 307 words, that few contemporaries knew exactly what it said:

[Be it enacted,] That a final judgment or decree in any suit, in the highest court of law or equity of a State in which a decision in the suit could be had, where is drawn in question the validity of a treaty or statute of, or an authority exercised under, the United States, and the decision is against their validity; or where is drawn in question the validity of a statute of, or an authority exercised under, any state, on the ground of their being repugnant to the constitution, treaties, or laws of the United States, and the decision is in favour of such their validity, or where is drawn in question the construction of any clause of the constitution, or of a treaty, or statute of, or commission held under, the United States, and the decision is against the title, right, privilege, or exemption, specifically set up or claimed by either party, under such clause of the said Constitution, treaty, statute or commission, may be reexamined, and reversed or affirmed by the Supreme Court of the United States upon a writ of error, the citation being signed by the chief justice, or judge or chancellor of the court rendering or passing the judgment or decree complained of, or by a justice of the Supreme Court of the United States, in the same manner and under the same regulations, and the writ shall have the same effect as if the judgment or decree complained of had been rendered or passed in a circuit court, and the proceedings upon the reversal shall also be the same, except that the Supreme Court, instead of remanding the cause for a final decision as before provided, may, at their discretion, if the cause shall have been once remanded before, proceed to a final decision of the same, and award execution. [italics added]

One hundred and fifty-six words separate the compound subject (final judgment or decree) from its compound verb (may be reexamined, and reversed or affirmed). If all intervening qualifications are eliminated, a much simpler sentence emerges to the effect that the favorable review of a state law by a state supreme court may be reversed by the Supreme Court of the United States. Thus, an elaborate adjunctive grammar achieved exactly what was needed at the time--a use of syntax that obscured the extraordinary powers bestowed upon the federal Supreme court in the very act of granting, specifying, and qualifying these powers. Section 25 was probably drafted in this fashion to draw as little attention to it as possible in order to establish judicial review without provoking the opponents of a strong federal government. And with good cause. Its potential opposition might have been formidable, yet without judicial review the national government established under the Constitution would have been just as devoid of genuine sovereign authority as it had been under the Articles of Confederation. Grammar to the rescue: to recognize this major accomplishment of the Judiciary Act required a skill and patience with grammar beyond the resources of most educated citizens, Founding Fathers and constitutional historians included 2. Only hard-core grammarians can be expected to fathom what happens here.

Obviously, as in the case of Ellsworth's sentence, adjunctive complexity can be taken too far, but a knowledge of its full potential actually teaches one a better command of syntax at all levels of complexity. As Fielding maintained in Tom Jones, it is easier to trim excess than to augment deficiency. This might not seem true of the verbal expression of small children, but it certainly applies to adults in their thirties, forties and fifties. Without ever having been taught to cope with advanced syntactic demands, our plain-style (and plain-minded) fellow citizens are all the more likely to flounder when confronted with prose demands at every level beyond the very simplest. Why? Because the hypotactic skill from imbedding modifiers in longer sentences actually helps to refine one's paratactic use of short sentences. Length subsumes brevity; brevity segments length. Moreover, the sequence among shorter sentences can be more effectively employed to imply conjunction ("It is summer. [and] The sky is blue."), cause (It is raining. [therefore] I am wet."), and effect ("I am wet. [because] It is raining.") Imbedded modifiers continue to play a role in all but the shortest sentences, and the use of cadence, or cursus, to end longer sentences is no less important in the attainment of attic simplicity as opposed to the haphazard plain style treated as divine law in our composition courses. Sentences may be short and to the point, as our preceptors of simplicity insist, but with far better control.

We must also recognize that one's grammar inevitably expresses logical propositions whose basic components are simple and flexible enough to accommodate a very broad range of thinking. According to Hegel as quoted by Lionel Trilling, "grammar is the work of thought, which makes its categories distinctly visible therein," and its "visible" categories, I would argue, comprise the words, phrases, and clauses linked with the more basic equation declared by a sentence's subject-predicate nexus 3. Is some kind of an equation somehow involved? For sure. When a subject declares its verb, some kind of an equivalence between the two ("this is that," this is what does that," etc.) is implied that subsumes everything in the sentence. Just as Whitehead and Russell demonstrated in Principia Mathematica that all mathematics translates into logic (or sentential calculus), all sentences likewise draw upon logic rooted in the subject-predicate equation, even those the most distended by syntactic risk. The logic might be bad or seemingly scrambled, but despite its liberties it utilizes a consistent (i.e., habitual) set of operations more or less appropriate for communicating one's intended meaning.

The subject-predicate relationship of sentences acts as the primary vehicle of this logic both by establishing the predicate as the primary modifier of the subject and by asserting an equivalence (grammar's law of identity) between the subject and predicate of sentences. Parallel constructions likewise equate the ideas connected by conjunctions (or by asyndeton, the obvious use of parallelism without conjunctions)--such, for example, that the simple parallel construction, happy and sincere, declares the equivalence between these two adjectives in the context in which they are uttered. Let eight words or phrases be used in a "runaway" parallel construction, and the implication remains clear--all eight of them share this parity relative to the context as a whole. Conjunction (and, and, and) might be told or implied, but what is logically established is equivalence among the words as well as a "plateau" effect that gives readers the opportunity to take a brief vacation from fathoming syntactic linkages with the rest of the sentence. This is important to recognize, for the adjunctive style typical of our Anglo-American essayistic tradition normally features parallel constructions, especially the compound predicate, the compound direct object, and imbedded parallel elements within other parallel elements.

As opposed to parallelism, subordinate constructions incrementally specify the meaning of what they modify (A > AB > ABC, etc.), as illustrated by the accumulation of attributive adjectives: dog > brown dog > old brown dog > little old brown dog, etc. The same principle applies to subordinate constructions in the appositive position, of course excluding the appositive itself, which very directly behaves as a parallel element, the addition of one or more new words to identify once again the substantive already described. The function of incremental specification thus extends to include both phrases and subordinate clauses that begin with subordinating elements that help to specify the context as a whole. The preposition in, for example, imposes the concept of "inside," or, more generally, "relative to, in an inside manner"; and the subordinate conjunction because signifies either cause or enthymemic logical consequence (this, as demonstrated by that). These linking words imbed phrases and clauses in particular sentences, and they can be treated as part of its basic equation that contributes to the validity of its subject-predicate nexus: Subject [in X, which Y] = Predicate [because A, in the event that B after C, unless D], etc. By logical multiplication, the more complex the sentence, the more intricate its implied equation, and the more refined the idea it expresses. The sentences already quoted by Mill, Milton, Keats, and Ellsworth may all be interpreted on this basis.

In other words, the subordinate connections within a sentence, no matter how stretched or convoluted, spell out the writer's sense of relatedness among the ideas expressed. The bigger and more ambitious these subordinate connections become, the more attenuated their truth value, since each subordination compounds the claims of accuracy for the basic equation declared by the subject-predicate nexus. Hence the importance of verbal intelligence as one of the most accurate indicators of one's overall intelligence (or I.Q.), once described by Charles Spearman, one of the founders of psychometrics, as the "eductive" ability to draw generalizations from particulars and to use particulars to exemplify generalizations. Like Chinese boxes, the "nested" subordinate constructions of sentences particularize their implied central equations asserted by the subject-predicate nexus, and an author's verbal intelligence may be understood to manifest his/her success in making this happen. Traditionally, of course, vocabulary tests have been used to establish our level of verbal intelligence, since our familiarity with words reflects the level of precision we need to express our ideas. However, our use of syntactic structures to bring our vocabulary into play seems at least as significant.

As seems confirmed by K. Hunt's theory of the T-unit (the main clause plus all its modifiers, the predicate included), the bigger the T-unit, the better its organization of thought must be to retain its credibility 4. Those who do not take their ideas very far usually (but not always) depend on a simplistic sentence logic, as reflected by their unavoidably categorical syntactic habits--short and not exactly to the point. Hence their problems when they try their hand at writing longer sentences, and hence their impatience when exposed to the more ambitious sentences of others. Their ideas are plain and simple, and might even be true, but too often they remain inadequate as compared to what might or ought to be said.

But is the lack of syntactic risk in student compositions altogether impervious to remediation--that's the question. Can greater syntactic risk be encouraged in students, or would it be a total waste of effort except for the few who may be expected to take up careers as writers later in their lives? Isn't it a total waste of time to teach skills to students who lack the capacity to use them? Perhaps so, but then again many students--possibly more than we realize--might actually benefit from the knowledge of grammar, whatever their level of ability. Too often students curtail what they might want to say because of their uncertainty whether the phrases and clauses they might need for this purpose are "correct" in the opinion of language authorities, for example their hard-core make-it-simple parents and teachers. As a result, verbal deprivation occurs at every level, including what's said, the structure of sentences that organize what's said, and, one step further (a very terrible step), the nascent conscious activity whose stirrings challenge one's ability to bring this organization into play. Just as less thinking diminishes the capacity for verbalization, less thinking results when fewer thoughts can be generated by this capacity. A cycle of inadequacy prevails that might be somewhat ameliorated, I suspect, by using grammar to teach syntactic risk.

But what improvements, exactly? A fair question, in answer to which one may list among the grammar operations almost totally beyond the verbal resources of today's average college student, complex parallel and antithetical constructions, extended appositive constructions, ambitious noun clause constructions, a sophisticated use of complex attributive modifiers, full and truncated absolute constructions (their participles absent), elongated parenthetical constructions, effective sentence fragments, and the like. Without any clear idea of the correctness of these syntactic devices, and with no experience in using them, students may be expected to avoid trying them out as potential liabilities to their correct--or, even worse, "clear and simple" (i.e., slow and simple-minded)--expression of their ideas. As a result, their articulation is limited to a modest "syntactic lexicon" of relative and adverb clauses as well as the tried and true assortment of prepositional, infinitive, and participle phrases. They even steer clear of punctuation based on the restrictive-nonrestrictive distinction rooted in grammar. As a result, their written prose style languishes at the most basic narrative and expository levels, often even more syntactically barren than the sentences they use in their daily conversation. Unaware of grammar except for the assortment of prescriptive avoidances they have acquired from teachers, friends, and relatives, they cannot avail themselves of its extraordinary versatility in the hands of anybody fully able to put it to use. Like basic piano students, they plink out chopsticks on an instrument that for others can accommodate the likes of Bach, Beethoven and Chopin.

True, our ignorance of grammar as a theoretical discipline does not particularly matter in our habitual use of sentences, since we may freely draw upon them whether or not we can explain their function. However, our conceptual ignorance of formal grammar too often reinforces the limitations of our habitual grammar, as for example when trying to read authors who stretch its use (Keats, Ellsworth, etc.) or when forced to stretch our own grammar skills to express relatively complicated ideas. In basic student prose, for example, this inadequacy almost inevitably precludes the use of inversions, appositives, absolute constructions, elliptical constructions, and multiply imbedded phrases and clauses. Students need not be able to parse all the syntactic constructions they use, but, all in all, I am confident that those with genuine authorial potential do benefit from being able to risk syntactic constructions that let them bring all the words they think they need into their sentences. As opposed to their classmates who typically underwrite (and who find comfort in being taught to view this inadequacy as a virtue), these more gifted students can benefit from venturing to overwrite, resorting to syntactic amplification because they know how to prune back their excess verbiage to manageable proportions. From excess rather than deficiency they best achieve their full adequacy as writers. And perhaps almost everybody benefits one way or the other. A healthy majority of students can be expected to become slightly more articulate, their verbal intelligence enlarged if ever so slightly. Surely they feel better educated, and this alone is a plus not to be disregarded.

If anything, the knowledge of grammar is even more important in revising our prose when we submit our initial drafts to further revision on either the typewriter or, with even greater success, on the computer. When we work up our sentences in memos, letters, assignments, manuscripts for publication, etc., we inevitably revise what we're saying with quick displacements from one grammatical construction to another, then perhaps yet another, and still yet another--for example by converting a relative clause into a participle phrase, then perhaps by scrapping everything except the nouns, whose synonyms we can later use in an infinitive phrase to be expanded into an entirely new sentence leading to two or three additional sentences as well. We enlarge our sentences to incorporate everything we think potentially relevant, then contract them again to trim elements we have decided are dispensable--then go through this process of amplification and contraction once, twice, or many times again, and of course by repeatedly making new syntactic conversions to achieve relative idiomatic ease (just as I am doing right now in writing this paragraph). At times we resort to parturition by dividing overloaded sentences, and this obliges us to shape and expand one or both of the separated components, often to the extent that entirely new syntactic possibilities must be explored. This facility can only benefit from the formal knowledge of grammar rooted in the ability to identify and generate syntactic constructions. Of course anybody can "do" a limited grammar on a strictly habitual basis, but I cannot reject the notion that the ability to explain it eventually enlarges one's ability to do it, at least among those who can bridge the gap between usage and knowledge to be able to say exactly what they want.

But how does one teach such a grammar? Apparently sentence combining bears excellent results, and without forcing students to identify any of the grammatical categories by name. However, the question arises whether a clear understanding of what happens in combining sentences might not require a knowledge of grammar that would encourage even greater freedom in sentence construction. Again it is a mistake in my opinion to assume that the evidence is in--that grammar can be discarded in the educational curriculum. No matter how it is treated, grammar takes place when we speak and write--why not know what we do?

To test the benefits of teaching syntactic risk among graduate students, I myself once took what might be described as a pedagogical risk by organizing a graduate advanced composition class as an experiment both to measure and expand my students' syntactic versatility. I taught my class enough basic grammar to be able to parse and tabulate their prose, then applied statistics in almost three dozen categories to measure the frequency, mean, and standard deviation of almost three dozen syntactic devices in their prose samples as compared to anthologized American prose essayists admired by the class--such figures as White, Didion, Thomas, Dillard, etc. We next converted raw scores to z scores and graphed our results as grammar "profiles" of individual syntactic reliance probabilities. My students were then able to work against these profiles in whatever direction seemed appropriate, most likely by expanding their syntactic variety to a level they still found comfortable. Unfortunately, I was teaching a summer session class, so I could not obtain the solid statistical comparisons needed to complete the experiment, but, based on admittedly subjective retrospective assessments, it seems that almost all my students--predominantly teachers, editors and published authors---were able to refine their style with genuine success. From the results we could "eyeball," we felt that very substantial changes had probably been achieved.

Hardly to our surprise, we discovered that in several categories a stark bimodal distribution occurred between established authors on one hand and the graduate students themselves on the other. Despite their relative sophistication, my students were expressing themselves in their prose compositions on a significantly different plane from the authors they admired, and the differences could be statistically calculated based on our grammatical tabulations. In at least a couple of instances we found almost an entire SD between the two means (the average difference from the average, which amounts to roughly thirty percent for scores approaching the mean). And how very surprising these differences turned out to be--for example, with established writers having used more lengthened appositives, many more parallel constructions, and, most significant of all, many more prepositional constructions (often in doublets and triplets) instead of the participle and relative constructions that typically dominate ambitious student (and faculty) writing. The professional essayists also tended to use more inversions (anastrophe) and more elaborate noun clauses, and, though we didn't calculate percentages in these particular categories, they seemed freer in their use of interjections, rhetorical questions, sentence fragments, and paratactic leaps from one sentence to the next. We also found that they keep their subject/predicate interface averages closer to the beginning of sentences, but with much greater variability from one sentence to the next. Their average sentence length was not more than a word or two longer, but, more important, there is much greater variety in length among adjacent sentences. In other words, the established authors can mix their prose more effectively to afford a livelier and more accurate expression of their ideas. It should be mentioned that percentages for relative, participle, and absolute constructions were roughly the same between the two sample populations, but I am confident that for these categories significant differences would have been found between graduate students and professional writers on one hand and undergraduate students on the other.

As confirmed by daily quizzes, most of my graduate students started out almost entirely ignorant of formal grammar, and they were only able to parse sentences well enough to tabulate their prose samples because of the crash course I gave in grammar at the beginning of the semester--an introductory unit that lasted a week or two longer than I had hoped. A couple of students never did grasp the principles of grammar, and, as to be expected, they were less than enthusiastic about the experiment. However, the rest of the class caught on quickly enough, and for these our findings bore excellent results, since they themselves could make the necessary adjustments by revising their syntactic profiles in whatever direction they wanted. And they did, based on the zero-sum assumption that the habitual use of any particular grammatical construction necessarily crowds out other constructions, so that a highly variable balance (or profile) emerges, but one with a "signature" that like a fingerprint is unique for any particular author--the more so, the more sophisticated the use of grammar. It was also my assumption that a profile can be revised simply by taking the necessary steps to reorganize the habits already in use. "Too thick with participles and subject-predicate interfaces that are too predictably short and without enough variety," a student could decide (as actually happened in one instance), and then work to augment the subject zone a little more effectively and to diminish participle constructions by shifting more to prepositional constructions as well as a better sampling of relative, absolute and adjective constructions. As became plain on her profile, this particular student, whose emphasis was in creative writing, had been excessively dependent on participles for poetic effect, and she needed to explore other syntactic options more aggressively.

I made it plain to the class that I was not trying to enforce normative standards based on the avoidance of errors, but rather to liberate their use of grammar by resorting to statistics to compare their limitations with the liberties taken by exactly the essayists they admired the most. I encouraged exploring Christensen's theory of sentence modifiers, but the approach could be expanded, I felt, by the use of elongated terminal appositive constructions, "runaway" parallel constructions, elliptical constructions, stretched periodic constructions, and antithetical symmetries---even the Johnsonian parison (balance between like structures) verging on parisosis (perfect symmetry). It did not bother me when students resorted to free associations or fell victim to the despised anacoluthon (any sentence structure that wanders from its original intent). Even the anacoluthon? And why not, for in stretching syntax one must be willing to accept the consequences of occasionally stretching it too far, confident that the problem can be easily fixed in later revisions. Let worthy and ambitious solecisms thrive that they might be reexamined in subsequent revisions.

What I wanted was for students to be able to draw upon the full lexicon of syntactic devices needed to capture their thinking in the very process of its being thought. Once exposed to the total possibilities of grammar, students could of course fall back on whatever assortment of constructions they found the most comfortable, but with somewhat better syntactic flexibility than before. Whatever they are trying to say in their sentences, I wanted them to possess sufficient confidence in their syntax to be able to bring into play the words and phrases that best communicated what they thought exactly while in the process of thinking it. My purpose then, as now, was to facilitate an easier articulation of ideas otherwise lost in the flow of thought--ideas at the brink of being spoken but too often abandoned because they don't readily make themselves available to what seems the correct selection of words.

Just as important, I also believe that composition teachers should have a solid background in grammar, whether or not they intend to teach it to their student in basic English composition courses. Those who are competent in parsing and analyzing syntactic constructions will almost inevitably be better able to judge the value of student writing and to help students revise their compositions, and without necessarily mentioning the relevant grammatical concepts to their students while discussing their work with them. Teachers, for example, need not fully explain a dangling participle phrase to their students to be able to show how to eliminate it, but they can explore possibilities with students much more effectively if they themselves know the appropriate function of participles and how the constructions in which they "dangle" can probably be modified in several ways. And the same goes for the distinction between restrictive and nonrestrictive modifiers in the use of commas. Professors aware of this relatively simple difference can at least be confident of their suggestions to students who misuse commas because of their ignorance of the rules involved. Like lawyers expert in the legal code and garage mechanics expert in the parts and construction of automobile engines, teachers will do a better job if they can be expert enough to be able to spell out in simpler terms exactly what happens in a sentence in their own particular metier--written prose. They need not flaunt their specialized knowledge, but they should be able to simplify what they know to make it plain to their clients. Just as cancer specialists should possess a sufficient knowledge of pathology for them to be able to explain to their patients what's wrong with them, teachers should possess sufficient knowledge of grammar to be able to tell their students what changes seem needed in their prose compositions, again without making a show of the theory and specialized nomenclature with which they themselves are familiar.

Unfortunately, the rejection of grammar for students seems to have led to its neglect by teachers as well, and, one step further, by even the professors who teach the teaching of composition at the college level. They do not need to teach it, they think, so they feel free to forego its study--even to forbid its study by others. What amounts to pedagogical censorship is imposed--grammar is declared taboo and rejected from the curriculum at every level, so teachers ignorant of syntax can be expected to teach their own level of ignorance. They are hazy about cookbook grammar prescriptions and ignorant of the monumental contribution of such established traditional grammarians as Jespersen, Curme, and Quirk. From the top down, the study of grammar has thus been neglected, and from the bottom up prose style has become utterly simplified and homogenized. Almost everybody wallows in feel-good expressiveness submitted to inspection by feel-good teachers and professors with feel-good pedagogical strategies. As a result, the field of composition has been paraprofessionalized, most obviously by the substitution of classroom verbal group therapy tactics for the hard task of diagnosing exactly what has been written and how it might be improved. The use and study of grammar has atrophied at all levels, and, not surprisingly, its decline has accompanied the collapse of our students' ability to write--dare one suggest, even to think. Ideas are simpler, sentences shorter, and the verbal attention span utterly blown to the winds.

I remain dubious of any theory or methodology that makes a virtue of ignorance, and the neglect of grammar (which does fully manifest itself when we use sentences, whether we like it or not) is tantamount to just that--self-imposed ignorance verging on what might be described as pedagogical Stalinism--not because it it entails leftist politics (which it doesn't), but in the sense that an elite cadre of anti-elitists limits to itself the choice what must be excluded from the purview of the unwashed masses presumably for their own benefit, and with the inevitable result that even the cadre itself gradually sinks into ignorance. What concerns me at this point is that an ideological commitment to a holistic (and thus exclusionary) theory of composition might hinder or altogether impede this gradual seasoning that should otherwise take place in the experience of any normal composition teacher. Regardless of the many fads in composition theory that temporarily enjoy the status of holy revalation, good mature composition teaching depends as much as anything on heavy writing assignments, judicious corrections, frequent revisions, personal conferences, classroom editing sessions, and the teacher's enthusiasm for whatever approach(es) he/she is using, but also, last but not least, the teacher's possession of a very solid background in as many theories and methodologies as possible bearing upon composition, including the study of grammar.

We do not want to be boxed in by any particular theory that forcibly imposes ignorance. How ironic, then, that the so-called holistic approaches to composition that are popular today tend to do just this, and that all or most of the categories listed above might be granted by proponents of this approach except professional expertise relevant to the writing process itself. Indeed, how can such approaches be described as "holistic," since all three categories of the medieval trivium are effectively excluded--grammar, rhetoric, and logic--each of which bears a direct and obvious relationship to the actual generation of words in sentences. If writing is to be revered as a process, why, exactly, is the study of the syntactic operations that actually take place when words gather into sentences banished from the education of both students and their teachers?

I suspect grammar has been abolished in composition classes as much as anything, not because it is an impediment to written expression, or because it deducts too much time from the teaching of composition, but because too many of our students are unable to cope with it at even the simplest level. They can barely read and write--how can the study of grammar play any kind of a role? Conveniently, educational statisticians have been enlisted to demonstrate its irrelevance, and, voila, grammar is proscribed. Education reaches down even further to compensate for our students' inability to reach up. But at a loss, I think, as demonstrated by the basic limitations of too much of the prose encountered, taught and encouraged on college campuses today. It therefore seems about time to reverse the trend and encourage the teaching of grammar not as a system of prescriptive avoidances, but as an encouragement of syntactic risk by acquainting those students who might benefit, and ourselves as well, with exactly what happens when we immerse ourselves in the flow of words. And all the rest--fewer than one might realize--can at least learn to respect what they cannot understand.

GLOSSARY

ABSOLUTE PHRASE (or construction)--a noun or substantive modified by a participle phrase: She blinked, her eyes filled with wonder; he turned, his arms swinging by his side. The TRUNCATED absolute phrase implies the participle being: She stared, her mouth [being] taut with anger. The absolute phrase acts as a sentence modifier and usually occurs either at the end or (less frequently) at the beginning of a sentence.

ADJUNCTION--Jespersen's term for any use of subordinate modifiers--phrases and clauses--as opposed to the NEXUS between the subject and predicate.

ADVERB CLAUSE--any clause that modifies the verb and that begins with a subordinating conjunction: I ran because I was happy; It was time to leave before we knew it, etc.

AMPLIFICATION--expanding a construction. This usually applies to the rhetorical expansion of a topic, but it can also refer to the enlargement of a sentence or syntactic unit.

ANACOLUTHON--a sentence that ends with a different organization than originally planned: I wanted his car was not new. It is usually to be found in a long sentence whose original construction is forgotten.

ANASTROPHE--any syntactic inversion, but usually between the subject and predicate: This house I think I know. But it also occurs between modifiers: Yeats's "terrified vague fingers" instead of "vague terrified fingers."

ANTITHETICAL CONSTRUCTION--parallelism that plays upon syntactic resemblances to stress differences: They wanted a pleasant but elaborate picnic, but I preferred a cheap and exciting afternoon at the amusement park.

APPOSITIVE--any word or group of words that helps to describe the object of an adjacent noun: my father, a strict disciplinarian; the house, a ramshackle construction that must have been built fifty years ago, etc. An INVERTED APPOSITIVE occurs before the noun it modifies: A strict disciplinarian, my father often spanked me. What I describe as a TERMINAL APPOSITIVE at the end of a sentence may stretch out much longer than the main clause itself: We lived in an old house, a ramshackle construction that must have been built fifty years ago, when wood was cheap, labor was plentiful, and carpentry was still practiced as a skill.

APPOSITIVE MODIFIER--any modifier that follows the word(s) it modifies: The dog which I love; the dog in the manger; the dog big and friendly, etc. See ATTRIBUTIVE MODIFIER.

ASYNDETON--the omission of conjunctions for parallelism: I threw the ball, [and] caught it in the harsh [and] glaring sunlight.

ATTIC STYLE--a relatively simple style but without altogether eliminating the syntactic features of the Ciceronian style. The first major English author who perfected the Attic style is generally considered to have been Joseph Addison.

ATTRIBUTIVE MODIFIER--any modifier that precedes the word(s) it modifies: The large green jaw-snapping turtle. See APPOSITIVE MODIFIER.

BRICOLAGE--French for a random assortment of tools. Literary critics have used bricolage to describe a useful assortment of critical theories for explicating texts, but I am extending its use here to describe the full assortment of syntactic devices that one can draw upon--for example, appositives and absolute and participle constructions might be the bricolage of one author, but not another.

CICERONIAN STYLE--an ornamental style modeled after Cicero's that was cultivated during the Renaissance and revived in the mid-eighteenth century. It features balance, antithesis, an emphasis on cadence, and a cumulative expansion in the use of phrases and clauses. Fielding's passage in the text illustrates the style. Also see ANTITHESIS, PARISON, PARISOSIS, and CURSUS.

CURSUS--the attainment of a smooth ending (or "run") in a sentence when the rhythm (or cadence) of its prose style can finally be dictated by meter. Medieval rhetoricians distinguished three types of cursus: the CURSUS PLANUS, an "even run" which combined the dactyl and trochee (/-- /-), the CURSUS TARDUS, a "slow run" which combined two dactyls (/-- /--), and the CURSUS VELOX, a "quick run" which combined a dactyl and two trochees (/-- /- /-).

DOUBLETS AND TRIPLETS--phrases and clauses used, respectively, in twos and threes. A triplet of prepositional phrases: I ran in the house, up the stairs, and into my bedroom. Clauses and phrases used in longer Ciceronian sentences may be conveniently described in this fashion.

ELLIPSIS--the omission of words and groups of words that can easily be supplied: I knew [that] she was coming; They wanted to eat cake, and I [to] eat ice cream.

ENTHYMEME--An informal syllogism which omits either the major or minor premise: Socrates was human, so he died; Socrates died because all human beings do. In the first instance the major premise, All humans die, is missing; in the second instance, the minor premise, Socrates was human, is missing. Whenever the words because, since, and therefore are either used or implied, there is probably an enthymeme involved.

GRAMMAR, formal and habitual--one's FORMAL grammar consists of the rules one knows about the structure of sentences, whereas one's HABITUAL grammar consists of the rules one follows without necessarily knowing why.

HYPOTAXIS--the HYPOTACTIC arrangement of words, phrases, clauses, and sentences is based on the use of explicit dependent or subordinate connections. For example, sentences are combined by the use of pronouns, the repetition of words, and the repetition of the subject from one sentence to the next: John is a carpenter. He works hard. His specialty is roofing. etc. See PARATAXIS.

IMBED--an IMBEDDED phrase or clause may be understood as a fragment of one sentence incorporated into another. In the sentence, "I prefer movies that surprise me," the relative clause that surprise me may be treated as a potential sentence they surprise me imbedded in first sentence.

INFINITIVE PHRASE--Any phrase which begins with an infinitive (to jump, to ponder, to have discerned). An infinitive phrase may be used as an adjective ("They had a plan to buy a new car"), as an adverb ("They were ready to buy a new car"), or as a noun ("They wanted to buy a new car").

INTERFACE--My own term for the site in a sentence where the first word of the predicate is adjacent to a word of the subject: Oscar / was happy; for once his ideas / seemed valid. Arbitrarily, I include introductory adverbial modifiers as being part of the subject. A command begins with the interface: / Go away. An inversion postpones the interface: Why they came / I'll never know. Prose typified by delayed interfaces tends to be turgid, while prose typified by quick interfaces tends to be superficial.

INVERSION--see ANASTROPHE.

NEXUS--the subject-predicate core of a sentence. See ADJUNCTION.

NOMINALIZATION--any word or combination of words that plays the role of a noun in the context of a sentence. Included are gerunds (participles used as nouns: Seeing is believing), infinitives (To see is to believe), sometimes adjectives (I admire the bold and the free), gerund phrases (Seeing God is believing in God), infinitive phrases (To see God is to believe in God), and noun clauses (That she believes in God depends on whether she sees God).

NOUN CLAUSE--a clause used as a noun, usually introduced by a subordinating conjunction: That they think right doesn't necessarily dictate what they do. The noun clause may be used as the subject or direct object of a sentence (What they do is harmful; I enjoy what they do), or as the object of a preposition (I think little of what they do).

PARALLELISM--the use of two or more words, phrases, or clauses in sequence: The tall and flowering trees sway back and forth their leaves turned backwards and their limbs rapidly swinging in the wind. By RUNAWAY parallelism, I refer to any use of parallelism with more than three units in sequence: His face was large, bulbous, red verging on purple, and splattered with pimples, blackheads, moles, and warts.

PARATAXIS--any sequence of sentences without explicit syntactic connection: The sun was shining. The sky was beautiful. We were in a glorious mood. See HYPOTAXIS.

PARISON--grammatical symmetry achieved through balance between like structures: The carpenter reached in his tool box and pulled out a hammer.

PARISOSIS--making two members of a sentence equal in the use of imbedded subordinate constructions, as for example in Ciceronian prose: Ye who listen with credulity to the whispers of fancy, and pursue with eagerness the phantoms of hope, etc.

PARSE--To parse a sentence is to identify the grammatical category of every word in it as well as its function in the phrase and/or clause to which it belongs.

PARTICIPLE--a verb used as an adjective: the playing children; the baseball game played at night. In the PARTICIPLE PHRASE the participle continues to serve as an adjective but carries a predicate exactly as if it were a verb: the children playing games at the park. A DANGLING PARTICIPLE PHRASE is one that does not follow or directly precede the noun it modifies: Playing games at the park, the intention of the children was to have fun. [here intention, not children, the obvious subject of the participle, is the subject of the sentence.]

PARTURITION--my term for dividing into two a sentence that has become too long and building up both portions, each as a separate sentence. Parturition often occurs when revising one's prose, and just as often it permits the addition of new words and phrases.

PARENTHESIS--any word, phrase, or clause that interrupts the immediate flow of thinking, indicated by commas and dashes as well as parenthesis marks: I love Shakespeare (his genius is obvious to all), but sometimes I'm bored by his plays.

PERIODIC SENTENCE--a sentence with an elaborate arrangement of phrases and clauses that remains incomplete until the very end in order to build suspense. Typical of the CICERONIAN style. See the Fielding passage in the text.

PLAIN STYLE--a simple, straight-forward style with no obvious effort to feature cadence and stylistic embellishment. The first major text in English that used the plain style effectively was Bunyan's Pilgrim's Progress.

RELATIVE CLAUSE--a clause that modifies a noun and that is introduced by a relative pronoun: a woman who smokes, a cat that purrs, a creed whose adherents can be trusted, a creed in which I can put my faith.

RESTRICTIVE and NONRESTRICTIVE MODIFIERS--restrictive modifiers help to identify the words they modify whereas nonrestrictive modifiers add incidental information. As a rule restrictive modifiers are not separated by commas ("The dog in the yard barked too much"), whereas nonrestrictive modifiers are ("The dog, its tail wagging, barked very little.")

SENTENCE FRAGMENT--a sentence lacking a subject, predicate, or both. Sometimes sentence fragments are effective, especially if used in response to previous sentences.

SIGNATURE--the unique pattern of frequencies in syntactic usage that I anticipate can and will be found in the syntactic profile of any particular writer--much like a finger print. See SYNTACTIC PROFILE.

SYNTACTIC PROFILE--a chart representing the usage frequencies for all the syntactic devices that are tabulated for a prose sample by a writer. The profile may calculated based on raw scores or zero-sum percentages, but, best of all, by determining their syntactic reliance probabilities based on the deviation from the mean for the use of each device established for a more inclusive sample of writers (students, established authors, whoever one chooses). See SIGNATURE, SYNTACTIC RELIANCE PROBABILITY, and ZERO-SUM.

SYNTACTIC RELIANCE PROBABILITY--the probability of a writer's dependence on a particular syntactic device instead of others, as for example when an individual concentrates on prepositional phrases instead of relative clauses. See SYNTACTIC PROFILE and ZERO-SUM.

T-UNIT--K. Hunt's term for an entire clause inclusive of all its modifiers. A complex sentence that consists of a single clause and all its modifiers comprises a single T-unit; a compound sentence that consists of two or more clauses and their modifiers comprises as many T-units as main clauses, whatever their use of subordination. According to Hunt, the complexity of our prose expression is more or less quantified by the length of the T-units we use.

ZERO-SUM--the base value by which syntactic frequencies may be calculated, on the assumption that the habitual preponderance of one or more syntactic devices can only be at the sacrifice of others, thus establishing the standard for calculating one's syntactic profile. See SYNTACTIC RELIANCE PROBABILITY and SYNTACTIC PROFILE.

1. I borrow the word adjunction from Otto Jespersen's Essentials of English Language (University, Alabama: University of Alabama Press, 1964), pp. 91-96.

2. I explore the central importance of Section 25 in my article, "An Accidental Conspiracy: The Early History of Judicial Review from the Constitutional Convention to the 1789 Judiciary Act and Marbury v. Madison," in The American Constitution at the End of the Twentieth Century, ed. by Ralph Chandler and William Ritchie (New Issues Press, 1996), pp. 159-251, esp. pp. 222-29.

3. Lionel Trilling, "The Meaning of a Literary Idea," in The Liberal Imagination (New York: The Viking Press, 1951), p. 290.

4. K Hunt, Grammatical Structures Written at Three Grade Levels (NCTE: Champaign, IL, 1965).