For I must be allowed to say, that when we are employed, after a proper manner, in the study of composition, we are cultivating reason itself. True rhetoric and sound logic are very nearly allied. The study of arranging and expressing our thought with propriety, teaches to think, as well as to speak, accurately. By putting our sentiments into words, we always conceive them more distinctly. Everyone who has the slightest acquaintance with composition knows, that when he expresses himself ill on any subject, when his arrangement is loose, and his sentences become feeble, the defects of his style can, almost on every occasion, be traced back to his indistinct conception of the subject: so close is the connexion between thoughts, and the words in which they are clothed. (952)

These words, written by Hugh Blair and published in 1783, express an attitude toward the study and exercise of composition that pervaded the next century, and which we still embrace in ours. Good writing is good thinking, and writing is necessary to education because it accomplishes what other means of study do not: it reveals, more glaringly than any multiple-choice exam, the writer's "indistinct conception of the subject," and necessitates the forging of a more distinct one. Yet somehow, this primary truth has often been obscured in composition pedagogies. The connections between writing and thinking, and logic and arrangement-the very crux of composition-are often glossed over, to the point that John T. Gage, in his 1986 essay "Why Write?", felt it necessary to exhort composition teachers to dare to judge between good ideas and bad ones in evaluating student essays, just as they do in professional writing; and further, to value good ideas over "proper" grammar and structure.

All too often, the composition instructor hopes for a certain type of essay-one which reveals independent thought-but teaches and even evaluates with another sort of essay in mind, one where technical considerations, particularly correct structure and grammatical correctness, are at the fore. This split-personality evidenced in many composition classrooms is encouraged by most of the composition rhetorics on the market, as they teach form-such as paragraph development-while offering writing models that supersede those forms; and teach the application of rather rigid structures, while suggesting writing assignments that encourage exploratory and independent thought, which do not easily fit into such forms.

In turn, this split-personality of composition handbooks and freshman English rhetorics reveals a parallel split in contemporary composition theory, which is symptomatic of an incomplete break from current-traditional rhetoric, at least in the instructional advice that many contemporary freshman rhetorics and handbooks offer. Current-traditional rhetoric, developed in the late 19th and early 20th centuries, can be defined as an approach to teaching composition that emphasizes prescripts of structure and style. As such, current-rhetorical theory is widely criticized as a pedagogy that encourages the separation of form from content. James A. Berlin calls current-traditional rhetoric "the manifestation of the assembly line in education" and "the triumph of the scientific and technical world view" (62).

In this essay, I will attempt to consider the cultural, historical, philosophical, and practical forces that have kept the current-traditional approach entrenched in the composition classroom, and to propose the need to reexamine paragraph theory in two contexts. First, we need to question how the prevailing paragraph pedagogy affects student writing, and second, we need to examine its effects on how society at large considers the act of and purposes for writing. For if we believe that the development of ideas progresses within the language act, then composition must take on a much more central educational role throughout the various disciplines; and if composition is to do so, we must reconsider how to teach it, on a practical as well as an ideological level. Finally, I intend to suggest that in an age immersed in Thomas Kuhn and Nietzche, it is pedagogically irresponsible, rather absurd, and counter to our educational goals to continue to follow a pedagogical approach that distinguishes form from thought, and which reduces writing instruction to set formulas and matters of superficial correctness.

Bain's Paragraphing Rules in the Classroom

The development of current-traditional rhetoric, as well as the reactions against it, can be framed by the history of paragraph theory. When a writing instructor turns to that part of most freshman composition rhetorics or handbooks that deal with essay structure, she finds that the architecture of the paragraph holds a prominent place; in fact, the composition is commonly conceived of as a structure of paragraphs. The paragraphing rules set forth in such textbooks are in most cases very much like those formulated by Alexander Bain in his 1866 text, English Composition and Rhetoric, which conceived of the paragraph as a rhetorical structure with fixed requirements (Lewis 20). His rules have become so entrenched in our understanding of paragraph and essay structure that it is surprising to realize that the paragraph was not a subject of any detailed rhetorical consideration until Bain. Bain's rules have become our rules, and most contemporary freshman rhetorics restate them in more or less the same fashion.

Bain conceived of the paragraph's structure as analogous to the sentence. He likened the topic sentence to main clause, asserting that "[t]he opening sentence, unless so constructed as to be obviously preparatory, is expected to indicate with prominence the subject of the paragraph." The other sentences are conceived of in analogy to subordinate elements, and serve to flesh out the meaning of the topic sentence. Furthermore, Bain asserted that the paragraph is also characterized by three features, coherence, unity and development. Bain spent the most ink on the subject of coherence, taking care to list the various conjunctions and connectives that can be used to show the relationships between the clauses and sentences within the paragraph; and emphasizing the importance of parallel structure in sentences that illustrate the same idea. Unity, "which implies a definite purpose, and forbids digressions and irrelevant matter," is another prominent consideration. The third is development (which subsequent current-traditionalists renamed mass or proportion), meaning that the paragraph topic is expanded upon sufficiently to satisfy the reader, and the mass of the paragraph is in proportion to its importance to the composition's overall purpose.

Yet if we perceive the paragraph rules we encounter in our rhetorics as as self-evident as gravity, the evidence of many composition theorists since the 1960s shows that our sense of paragraph theory is a bit confused-Bain did not discover some inevitable laws of the paragraph. Instead, he prescribed a particular style of paragraphing, which proceeds deductively. Bain's rules went virtually unquestioned for a century, until a number of composition theorists investigated the actual English paragraph, and discovered that professional essayists have transgressed Bain's standards to excellent effect, both before and since 1866.1 At the most extreme, Richard Braddock concluded that paragraphs beginning with topic sentences were surprisingly rare in professional contemporary expository prose-comprising only 13 percent of paragraphs he examined. William Irmscher's graduate students repeated Braddock's investigations and found topic-led paragraphs to be more common, between 40 and 50 percent, although their frequency varied among individual writers (Irmscher 98). A.L. Becker identified a variety of paragraph patterns beyond Bain's topic-led one.2

With some exceptions, the majority of composition theorists since the 1960s have viewed Bain's highly teachable prescriptions as actually detrimental to the development of student writing. The title of Virginia M. Burke's 1967 essay, "The Paragraph: Dancer in Chains," says it all. For Burke, the current-traditional rhetoricians are very like the critics that Fielding criticized in Tom Jones: "men of shallow capacities" who "mistake mere form for substance" (Burke 37). Philosophically, the new rhetoricians object to the current-traditionalist's division of form and thought. As James A. Berlin encapsulates their theoretical stance, "[s]tructure and language are a part of the very formation of meaning; they are at the center of the discovery of truth and are not simply the dress of thought" (91). Or, as John T. Gage asserts,

The forms cannot be given to students, whose task is then to fill them up with ideas, since in such cases the ideas will not have a purpose in themselves other than to satisfy the demand of the form. It is ideas which come first, in writing, and forms which satisfy the demands that face a writer who has them. (729)

If your freshman composition classes are anything like the ones I have taught, your students are not particularly inspired, or helped, by the current-traditional explanation of the paragraph. Usually such advice elicits groans-or worse-glazed-over eyes. It's no wonder: they've had this advice drilled into them since grade school. Most of them are competent at this structure, but it doesn't seem to be helping them to fulfill that part of your assignment that asks them to deal with a thought-provoking issue, to consider something in a new light, to consider an unfamiliar concept. In other words, to think, for God's sake.

First of all, there's the rather embarrassing fact that the model essays that we encourage our students to study and emulate do not always follow the paragraph prescriptions we teach. Further, our paragraph prescriptions, which seem so concise on the page, become problematic when put into practice. The paragraph is supposed to confine itself to one main idea, but so is the essay, the sentence, and even the word. Upon close examination, it's very difficult to ascertain exactly what an "idea" is, as many critics of current-traditional paragraphing rules have pointed out. Paul C. Rodgers Jr., in his 1966 essay, "A Discourse-Centered Rhetoric of the Paragraph," listed the amendments that have been made to Bain's rules, based on the study of the paragraph as it is actually written:

--A proper paragraph always has a single central topic idea, except when it has two, three, or more.
--Development of the topic is always limited to the paragraph in which the topic is broached, except when the topic requires that exposition continue in the next.
--The topic sentence always expresses the topic idea, but the work of expression may be disposed of in a minor segment of the sentence; or, on the other hand, a complicated topic may take several sentences, and these sentences may be widely separated in the paragraph.
--There is always a topic sentence, yet it may not actually be stated. In this case, it is "implied," and serves as a sort of offstage influence directing the action in the paragraph.
--A paragraph by definition is a series of sentences, but now and then it turns out to be one sentence only. If the sentences-series seems too long for presentation as a unit, it can be subdivided into several paragraphs without loss of unity. Conversely, a series of short paragraphs can be combined into a single unit, sometimes with the original components identified by number or letter.
--Moreover there are certain very useful and common paragraph types that show little interest in amplifying topics: transitional, introductory, directive, summary, and concluding paragraphs. (40)

Such exceptions render paragraph rules absurd, and point to the fact that the paragraph is a much more flexible structure than freshman composition texts tend to admit. In fact, it is rather odd that although the paragraph is a much more complex unit than the sentence, current-traditional rhetoric defines it by more rigid prescriptions. It is perhaps even odder that it took a century for rhetoricians to question the validity of Bain's prescriptions, and I suspect that the reason for this oddity can be found in the current-traditionalist's implicit attitude toward both audience and invention, both of which are formally defined to a large extent by the insistence on a topic-led paragraph.

When Bain likened the paragraph to the sentence, he was thinking of a specific type: the loose sentence. Just as the loose sentence begins with the main clause, which is supported by subordinate elements that follow, Bain's description of the paragraph begins with the main idea, which controls the paragraph. Such a structure, on both the sentence and the paragraph level, can be labeled deductive, or analytical. Rhetorically, the deductive form, whether on the sentence or paragraph level, implies an expectation that the audience will readily accept the idea proposed at its most general level (the topic sentence). If the audience is expected to resist one's conclusions or generalities, the composer will proceed inductively, or synthetically; that is, she will take care to begin at a lower level of generality, offering reasons for her conclusions before she states them. Thus the deductive paragraph is useful in rhetorical situations wherein the composer need not consider audience resistance.

In addition, the deductive quality of the current-traditional paragraph model poses some problems when we ask our students to use their writing as an occasion to think through ideas. We think inductively, proceeding from the particulars to a generalized conclusion. We explain deductively, or analytically, and the ability to present an idea deductively presupposes a good deal of inductive thinking, which is undertaken in the invention phase. In an attempt to resurrect invention in the composition classroom, process theorists emphasize such invention heuristics as brainstorming and freewriting, as well as dialogue between students and instructors. But invention goes beyond pre-writing strategies, continuing throughout multiple drafts, and the writing process itself is seen as an inventive act.

For a process pedagogue, writing in the classroom is primarily exploratory rather than explanatory. Bringing a concept to distinctness is the work of writing, at least according to Gage, William Irmscher and others who have opposed the competency model, and writing can be used to this end more effectively than other means because, by the very nature of the writing act, thoughts are slowed, examined, and reassessed in a way that they rarely are otherwise, either in conversation or in our own internal thoughts. The act of writing forces us to slow down and examine our ideas, test them for validity and logical connections to other ideas. As Irmscher noted in 1979, some two hundred years after Blair but in perfect concord, "Because writing is so much more deliberative than talking, it helps us determine what we know and what we don't know. In our minds, we can fool ourselves. Not on paper. [. . .] Mental fuzziness translates into words only as fuzziness or meaninglessness." (20)

Thus for the process pedagogue, the real work of invention occurs in the actual composition, not before it, and fuzzy thought on the page is often the beginning of independent thought, simply because of a natural desire to understand what we are forced to acknowledge that we don't yet understand. Well developed, original and logically structured writing rarely springs full-formed onto the page, unless the ideas that inform it have come pre-packaged from without. Most commonly, it requires a great deal of work to develop a germ of a thought to its final fruition, and that work will proceed inductively, from particulars that lead up to a conclusion.

In the contemporary composition handbook, process pedagogy has found its place, listed under such titles as "pre-writing," or "generating" or "discovering" ideas, but the inquiry-led nature of process pedagogy is not carried into the drafting of the paper. Instead, drafting is still characteristically taught using the current-traditional model. The Bainian paragraph is the standard and the basis of the total form, which is often taught as a structure of paragraphs which, like the paragraph itself, proceeds deductively. As a consequence, despite our hope that our students will think synthetically, we expect them to draft analytically.

This is fine as far as it goes, for the essay should often proceed from inductive inquiry in early drafts to deductive presentation in the final product. However, the bottom line is that the final product, the essay turned in for a grade, and even the rough drafts, are often evaluated primarily by their success in deductive presentation, as well as by their adherence to grammatical and usage norms on the sentence level. Because these are the evaluation priorities, they are what count for the student. As long as structure and form are dominant considerations in the evaluation of papers, we cannot expect our students to value the quality of their content, and thus to see composition as an exercise of independent thought and inventiveness. We ask one thing, but grade another. We dutifully reward the dull, predictable paper with a decent grade, noting that it is coherent in purpose, with a clear thesis statement and paragraphs that begin with topic sentences and demonstrate unity, coherence and proportion, more or less.

Unfortunately, in a classroom informed by the competency model, the work of developing a germinal thought to its fruition is often never undertaken. This is the danger that Gage expressed well in his criticism of the competency model, and which any number of writing instructors can relate to in trying to encourage students to explore and clarify their ideas.

[I]f students have been taught to view success on such assignments as fulfillment of the technical requirements, however these may be defined, then it will no doubt occur to them that the best way to ensure success is to keep the ideas as simple and meaningless as possible. If successful writing is defined as technical skills only, then students may be learning an unspoken lesson that is unintended by the pedagogy, namely, that ideas do not matter. (721)

If an instructor notes that a paragraph is unclear, illogical, or undeveloped, the student can do one of two things: develop and clarify it, or cut it. With the pressure of a deadline and an imminent grade, the student more often than not chooses the latter path, and more than a bad paragraph is at stake. Students prefer simple and relatively meaningless ideas because they are relatively easy to explicate in a coherent, well-structured form; a complex idea, after all, can make a linear pattern of thought-and of writing-difficult, or even impossible, to attain. A complex idea, if pursued, could necessitate a divergence from the thesis statement as originally conceived. An unclear, undeveloped or illogical paragraph is sometimes more than a sign of improper structure, or a call to return to the topics or modes to develop it; such a paragraph can be a sign that the thinking that has prompted the act of writing needs to be reconsidered. In other words, as Blair noted, loose arrangement is a sign of indistinct conception. The question is, are we as writing instructors making clear that the real work of writing is to bring concepts to clarity?

The Historical Context

No doubt we judge professional writing primarily by the quality of thought behind it, and agree with Blair that rhetoric and logic are inextricable. How have we gotten to a place where writers such as Gage feel it necessary to suggest that we make the same judgments concerning our students' compositions? For Gage, as for Blair, the tantamount skill learned in writing is the skill of thinking. Writing instruction is valued because it develops independent thought, or as Gage dares to call it, philosophy, or rhetoric in the Aristotelian sense-that is, the discovery of good reasons in situations where conclusions can never be more than provisional. Where did we get off track, to the point that we often fail to make the connection between good writing and good thinking; and to the point that university administrators, students, and even composition instructors believe that in the course of one or two composition classes, students will have "gotten" writing, that writing is a skill akin to following a cooking recipe?

Bain's rules persist because they imply that learning to write is much like following a recipe, and also because the conditions that ushered in their acceptance are still in place. An investigation into the pressures that came to bear on the American university in the late 19th century can help to explain why Bain's paragraphing rules came to be seen as undisputed truth, despite all evidence to the contrary in actual writing samples; and can help us to better see our own situation.

Bain's students in Aberdeen, Scotland, were not especially well read, many came from rural areas, and many spoke in nonstandard dialects; and perhaps the most immediately obvious reason that Bain's theories took hold on American soil was that he had developed easily teachable, prescriptive lessons in essay structure. Bain designed his rhetoric in order to quickly bring poorly educated students up to a writing competency standard (Ferreira-Buckley and Horner 200). In America, universities were increasingly challenged with just such students. Following the American Civil War, enrollment in the nation's universities mushroomed, and universities were increasingly serving not just an elite student body whose classical education could be assumed, but a rapidly growing population of students who aspired to join the middle class. Many of these students possessed little in the way of classical education or practice in composition. Composition had formerly been guided by the rhetoricians of the late 18th and early 19th century, including Hugh Blair, George Campbell and Richard Whately, whose tenants were derived from classical rhetoric and assumed an education steeped in the classics. The American universities of the late 19th century needed to find another way. For these students, the study of Blair's belletristic rhetoric was untenable, and universities were responding to the needs of their students by developing English departments that could not only emphasize English literature, but answer to the needs of a student body largely unschooled in the classics.

Consequently, nascent English departments were faced with a sharply rising student body-all in need of composition instruction, but essentially unequipped with the intellectual and literary background which had formerly been assumed. The student-teacher ratio was crushing. For example, in 1894 the University of Michigan's English department served nearly 1,200 composition students with four faculty members and two graduate students. The student-teacher ratio in Harvard's composition courses was 100 students per teacher (Berlin 60)-somewhat more manageable, but as those of us who have attempted to attend to the compositions of 100 students can attest, still extremely taxing. Clearly composition could not be taught on a student by student basis, and some form of pragmatic and time-efficient basis for evaluating student writing needed to be employed.

Doubtless, interest in the paragraph in composition theory also coincided with a large-scale shift from oral readings in the classroom to silent reading, and concomitantly, from composition as intended for oratory to writing intended mainly for the page, all made possible by increasingly available and affordable writing tools and books. Bain was innovative in developing a rhetorical model that answered to needs of a textual, as opposed to an oral, rhetoric (Corbett and Conners 525). Although paragraphing as evidenced by indentation has been found in manuscripts dating as far back as the sixth century,3 a need for a theory that could govern paragraph development was not perceived until the emphasis in composition shifted toward silent reading. That is, as compositions were more frequently received by the eye instead of the ear, the appearance of text on the page increased in importance. An extremely long paragraph could make a reader hesitate to plunge in, even if the same discourse might not spark the same foreboding if presented orally.

While composition shifted away from an emphasis on oratory, the mid-19th century American university was also experiencing a major transformation in educational purpose, as James Berlin has emphasized. Prior to the Civil War, American universities had been dominated by clergymen, who taught rhetoric as a central component in the education of future clergy and an elite class expected to hold political power. The major figures in rhetoric through the late 18th and early to mid-19th centuries, including Hugh Blair, George Campbell and Richard Whately, were clergymen, all of whom adapted classical rhetoric to the requirements of moral leadership and the development of refined taste. A central task of education was to develop the art of oratory, as it would be used to inspire and instruct from the pulpit and to persuade in the political realm.

The centrality of classical rhetoric was gradually displaced as universities increasingly responded to the practical needs of a burgeoning industry. Education was opening up to all members of society, and the majority of new students were not interested in becoming clergymen. Neither were they preparing for political leadership, as were the elite students of the early 19th century. Instead, they were seeking entrance into the middle class through careers in industry and agriculture, and their interests were decidedly secular. In response to the needs of such students, American universities adopted the German model of education, wherein scientific and technological information took precedence over religious and moral concerns.

The aims of composition shifted as a result. When composition was geared to the clergy, all the elements of classical rhetoric came into play, as emotional and ethical appeals were as important as logical ones. The purpose of oratory was persuasion. The rhetoric of the late 19th century, as it pertained to composition classrooms, had no need for all of that. Sharon Crowley notes that consideration of arrangement in classical rhetoric in large part ensues from the need to persuade an audience, and the skill in this arrangement lies in predicting the audience's attitude toward the message. The emotional disposition of the audience was one of utmost importance, and the composer shaped his discourse to appeal to the emotions and ethics of his audience, as well as to logic.

This is an essential point, because Aristotle's rhetoric, and classical rhetoric after him, dealt only in probable truths, for these are the ones that cannot be demonstrated beyond refutation. The realm of rhetoric, as it was classically defined, concerned how humans should act in a situation that presents alternative possibilities; and the purpose of rhetoric, in Aristotle's model, was for the composer to discover the best possible reasons for provisional actions, and to persuade his audience to accept his point of view. However, the rhetoric developed by the current-traditionalists was not interested in probabilities, for the subjects of compositions were perceived as essentially factual; and the goal of the writer was not to persuade her reader to any provisional conclusions, where other conclusions could be countered. Crowley notes that this had a marked effect on the tone, and arrangement, of the compositions written in the current-traditional classroom, both of which arose from its pedagogical attitude toward the audience:

Current-traditional discourse theory [. . .] painted listeners and readers as curiously docile. They were never hostile or inattentive-they were just interested. Writers needed only to arrange their discourse, then, in a fashion that would ease the reading process-that would, in fact, reflect the way any reasonable person might have written it, according to the natural dictates of the rational mind. (267)

Arrangement in such a situation is deductive, and easily managed by rigid formulas. The writer is not seeking to convince her audience of her conclusions, or seeking to prompt her audience to a particular course of action in the moral or political spheres. She is not encouraging her audience to follow the train of thought that would lead to these conclusions, and is expecting no resistance, no alternative possibilities. She is merely arranging and reporting preconceived, unassailable facts, by the most efficient means possible, to an interested audience in need of instruction. In such a writing situation, consideration of audience shrinks considerably, with the main consideration becoming the reader's knowledge base, which indicates the level of explanation required in the composition. The expectations and requirements of an educational system geared toward science, technology and business encouraged a composition based on reporting. As Berlin notes, composition topics assigned during the late 19th century encouraged a composition based on "either close observation, in the scientific sense, or the use of research material, the thinking of others. In both cases, the student was asked to report on [. . .] either empirical data or the work of better observers than he himself" (68).

Neither Bain, nor the modern-traditional school as it developed in American colleges, had much to say about invention. For Blair, invention presupposed and involved a protracted intellectual inquiry guided by belles lettres-for which the student preparing for a practical career serving industry had little need. Corbett and Conners suggest that the neglect of invention in composition was a "conscious move away from the complex or mechanical invention systems that were a necessary part of trying to use the old abstract rhetorical assignments in a world where wide reading-especially in the classics-could no longer be assumed" (525). While invention, as well as attendance to audience response, was of great significance to students of theology, it was simply unnecessary to students preparing for careers in business or technology. The conclusions that formed the outline of their compositions had been developed for them, in the scientific inquiries that had coalesced into unquestionable dogma, or in the equally unquestionable cultural assumptions that prevailed in the middle class, to which they aspired. Essentially, the work of invention was outside of their area of responsibility, and the worth of their composition rested in large part on their comprehension of and assent to the conclusions of their culture, whether received from the scientific textbook, the political speech, or the pulpit. If, as Crowley notes, the reader was assumed to be docile and merely interested, the writer was expected to be so also.

In such a writing situation, the paragraph as described by Bain, with its unwaveringly deductive logic, could easily be construed as the superior, and even the only, form for the paragraph. To ease comprehension, and to facilitate quick reading, the paragraph must begin with the declaration of the main fact, which is then clarified by one or more of the modes. The flow of attention in such a paragraph, from the conclusion in its generality to the specifics leading to that conclusion, meets the criteria for the expository essay designed to inform. Issues of audience and invention proved largely unnecessary to the student of current-traditional pedagogy. Both the reader's and the writer's attitudes toward the material for the composition were largely a moot point, as the content was unassailably true, as long as it aligned with the accepted scientific fact and middle-class moral norm. As Berlin puts it, composition was reduced to technical writing (63).

Rules of paragraphing flourished as a primary rhetorical focus under these conditions, and Bain's prescripts were adapted in America, most prominently by John Genung of Amherst College and A.S. Hill and Barrett Wendell of Harvard, forefathers of the current-traditional approach to writing instruction. All the factors discussed above paved the way for current-traditional paragraph theory, but their practical necessity was perhaps most obviously caused by the Harvard Reports of the 1890s, the first of a continual wave of "why Johnny can't write" studies that persist to this day.

Beginning in 1891, the writing abilities of Harvard's incoming freshmen were investigated by a committee of three men unschooled and inexperienced in writing instruction, Charles Francis Adams, E.L. Godkin, and Josiah Quincy. Under the direction of this committee, Harvard's composition instructors assigned their students themes concerning their preparatory school training in composition. The committee read these students' essays, as well as freshmen entrance exams, and concluded that students were coming into Harvard with poor writing skills. The fault, and responsibility, was laid on preparatory schools; as a result, entrance requirements related to writing skills stiffened at Harvard and at colleges across the country, and secondary schools began to take English composition more seriously. As Berlin notes, the Harvard Reports were beneficial in that preparatory schools began to take writing skills more seriously; on the other hand, the criteria used in deciding the value of compositions tended to accentuate details of style, increasingly construed as matters of superficial correctness, rather than content. (61). Bain's rules had laid out a clear rubric by which structure could be identified nearly as easily as a comma splice.

In investigating the ascendancy of the paragraph as the significant unit of composition structure, one cannot omit mention of Paragraph-Writing (1891), written by Fred Newton Scott of the University of Michigan and Joseph V. Denney of Ohio State University. Answering to the pressures of an overburdened faculty and generally unskilled student body, this text proposed the practice of composition through the writing of paragraphs, and thus encouraged the concept of the paragraph as something very close to a mini-composition. Between 1900 and 1930, according to Corbett and Conners (535), over 90 percent of textbooks used some version of the paragraphing theories developed by Scott and Denney. Given the practical conditions of the time, it is not surprising that Paragraph-Writing proved to be seminal: the paragraph was easier for the beginning writer to handle, and easier for the instructor to grade.

Some Alternative Voices

To assert the above is an unfortunate and misleading oversimplification of Scott's contribution to composition pedagogy-contributions that were largely ignored as current-traditional pedagogy took hold. While Scott doubtless saw the exercise of writing paragraphs as a lightening of the burden of teaching composition, and thus presented the paragraph as a discrete unit of discourse, he emphasized the development of the paragraph in the context of the entire essay; and of the essay in context to the rhetorical situation. Paragraph-Writing encouraged the student writer to experiment with options in developing paragraphs-options guided by the content which the writer intended to express and the context of that expression.

As such, Scott's paragraph theory is strikingly similar to those proposed by process theorists during the 1960s, when such writers as A.L. Becker and Paul C. Rodgers, Jr., first challenged the current-traditional concept of the paragraph with significant force. Like the process theories, and unlike Bain and the current-traditionalists, Scott saw form as inseparable from content, and subordinate to the rhetorical situation of the moment. William Irmscher expresses the difference between process and current-traditional concepts of arrangement thus: "we discover as we write; it means that the writer writes more like a sculptor who finds form while sculpting than like a bricklayer who piles bricks to construct a wall" (99); in looking at Scott's concept of the paragraph, it is clear that he would agree with the first analogy rather than the second.

Scott's paragraph theory thus holds strong affinities to the tagmemic concept of the paragraph, as developed by Becker, Kenneth L. Pike, and Richard Young, who also understood paragraphing in terms of the larger rhetorical situation. In tagmemic theory, any behavior needs to be examined from three perspectives: as a discrete unit of behavior, or particle; as part of an unsegmentable flow, or wave; and in context to the surrounding situation, or field. Tagmemic analysis reveals that the paragraph cannot be considered merely in its existence as a particle, or unit, as current-traditional theory tends to do. The paragraph also needs to be considered in the context of the larger situation, both the essay and the larger communication-situation that prompted it. In doing so, the relationship between form and meaning is respected, and Becker expresses the significance of this approach: "this means that a whole is not the sum of its parts (if by 'parts' we mean only the isolated segments), but only of its parts plus their relationships" (33).

Scott's analysis was also very like Rodgers' theory of "stadia of discourse," or semantic structures that can operate independently of paragraph indentation. In fact, regarding Rodgers' stadia theory, affinity is too weak a word, as Scott had identified such stadia by nearly the exact definition, and for the exact reasons, that Rodgers did some 70 years later.4 Both Rodgers and Scott viewed the paragraph as it was originally conceived, as a punctuation device more emphatic than the period, but which, like the period, can be inserted in a variety of places within the flow of the discourse. The decision to punctuate with a period is not dictated by grammatical mechanics, as a writer may choose to group any number of ideas within a sentence by means of coordinate and subordinate clauses, and in doing so interprets that group of ideas as unified. So it is with paragraphing. Rodgers stated that the paragraph works, "as does all punctuation, as a gloss upon the overall literary process [. . .]. To compose is to create; to indent is to interpret" (43), and we might add that all punctuation fulfills a rhetorical role of interpretation, aside from its grammatical role. Both Rodgers and Scott perceived the essay as a flow of thought, which is interrupted at certain points through paragraphing in order to signal what Rodgers called "a noteworthy break in the flow of discourse" (41), and which Scott explained thus:

The essay is the result of a sustained movement of the writer's thought toward a definite goal, but within this large development several intermediate steps are discoverable. The thought, on its way to the main conclusion, passes through many stages of transition, attains many minor conclusions, pauses for many retrospective glances. (28) (italics mine)

I italicize "discoverable" in this quote because, for Scott, segmentation of the flow of discourse through paragraphing is not governed by the discourse itself, any more than the decision to end a sentence is necessarily governed by the end of one idea and the beginning of another. The writer discovers where he wants to indent, given his interpretation of the level of import of each stadium of thought, just as an orator pauses in order to emphasize a point he considers especially cogent to his larger intentions. That is, he wishes to give certain ideas emphasis over others, and this decision is not always based on any logical necessities in dividing ideas into discrete units, as Bain and his followers would have it, but rather on the importance of the content within it, as it pertains to the purposes of the larger composition. Thus Rodgers reaches this conclusion:
About all we can usefully say of all paragraphs at present is that their authors have marked them off for special consideration as stadia of discourse, in preference to other stadia, other patterns, in the same material. "At this point," the writer tells us with his indentation, "a major stadium of discourse has just been completed. Rest of a moment, recollect and consider, before the next begins." But his decision to indent may be taken for any one (or more) of at least half a dozen different reasons. (42)

Thus the function of the paragraph is to signal what both Scott and Rodgers termed "stadia" of discourse, but as both noted, these stadia do not necessarily correspond to indentation. As Rodgers noted, the reasons to indent are manifold, and can be located outside of the decision to mark prose with paragraphing according to the logical laws of a "unified idea." They can be guided by the physical requirements of a print-era rhetoric, such as the need to break up a long unit of discourse into shorter paragraphs to satisfy the reader's eye, who might be put off by the sight of a long paragraph; or alternatively, a number of short stadia might be grouped into one paragraph, in order to prevent the development from appearing anemic. Rythmic considerations may also prompt an author to indent, as may shifts in tone. In addition, Rodgers definition makes clear the original concept of the paragraph mark as it appears in Greek manuscript-a mark intended to show closure, a place to pause, rather than a beginning.5

Scott made clear that paragraphing is used to mark the closure of "intermediate steps" within the flow of thought (which he also referred to as "natural articulations"). However, as did Rodgers, he also noted that "[t]he mechanical paragraphing does not always represent every joint in the structure of the essay" (28). It is up to the writer to decide just which "joints" he wishes to accentuate by indentation, just as it is up to the orator to decide when to accentuate a point by a changing his intonation or pausing for emphasis. Scott illustrated the paragraphing options open to the writer with the following diagram and explanation:
______________________________________

(will place diagram here)
______________________________________

A, B, and C here represent the more important stadia of the developing thought, the small letters, the partial conclusions. The vincula above show the three methods of paragraphing. Many variations of the third method might of course be adopted, according to the kind of discourse and the varying degrees of subordination of the minor articulations. (29)

This examination of the discourse stadium reveals the nature of the paragraph as more than a discrete unit that can be analyzed via "some Procrustean formula for governing the behavior of sentences between breaks, and to insist upon applying it over and over again throughout written discourse," to quote Rodgers' assessment of Bain's paragraphing rules (41). Instead, the paragraph is developed according to the needs of the larger composition, or as Scott expressed it, "[i]t is the business of the paragraph, as a section of the essay, to develop a specific subject by bringing particular facts into their due relation to the theme of the whole essay" (29).

Ironically, although Paragraph-Writing served to cement the current-traditionalist curriculum, Scott can be seen as a forebear of the process view of composition. As one of the founders of the National Council of Teachers of English (NCTE) in 1911, Scott fought against the practice of issuing entrance exams, which had been enacted as a result of the Harvard Reports; and criticized the rigid pedagogies that such exams encouraged, insisting that an emphasis on error-hunting and technical competency elevates matters that are subsidiary to the central role of composition. Unlike the current-traditionalists, who saw technical correctness as the litmus test for middle class status, Scott saw the development of thought and communication through composition as central in preparing students for participation in a democratic society. As Corbett and Conners remark, "Scott drew a strong distinction between a system that tests or grades a composition for administrative purposes and that which evaluates it as a stage in a pupil's progress"; and Scott asserted that the "lay[ing] hold of the student as an individual is, for composition work, simply indispensible" (Corbett and Conners 534). His preferred solution to the exceedingly high student-teacher ratios of his time was to lower the ratio, but given the realistic situation of composition instruction, offered Paragraph-Writing as a pragmatic instructional path.

The end of Scott's participation in rhetoric essentially curtailed any significant intellectual interest in composition theory. Scott's rhetoric department was dissolved only two years after his retirement, and English departments did not seriously attempt to develop a rhetoric doctorate for some time Instead, the task of composition instruction fell mainly to graduate students and service employees, a situation which persists in many universities to this day (Corbett and Conners 535).

Fittingly, the current-traditionalist view of the paragraph did not meet any serious scrutiny until the 1960s, when academia's acquiescence to the needs of corporate America was becoming uncomfortable, and political dissent was increasing. At this time, the humanities were steeped in existentialism, rendering the practice of extracting ethical maxims from literature and philosophy exceedingly problematic. And, this was the Cold War era, and a time when increasing globalization was taxing communication among disparate cultures. Meanwhile, the secular pieties of scientific dogma were being undermined by Thomas Kuhn, who proposed that science is hardly pure of cultural and political concerns.

Composition theorists were hardly immune to the tenor of these times, and if the scientistic university of the late 19th century prompted a change in teaching composition, so did questions raised against it in the 1960s. The precepts of current-traditional rhetoric-that the composer is in possession of unassailable scientific and ethical truths, which she can set forth to a merely interested audience-were increasingly perceived as untenable. The two theorists who most powerfully shaped the philosophy behind process theory, Kenneth Burke and Kenneth Pike, reassessed the classical considerations of rhetoric that had been ignored by the current-traditionalists. Burke and Pike's concerns regarding rhetoric were political and social, and as such they took seriously the need to develop communication techniques that could bind together peoples of disparate concerns and cultures. Both perceived the need for a new rhetoric capable of addressing the communication challenges of an increasingly global, post World War II climate, and both perceived the need for a rhetoric capable of surpassing even the Aristotelian goal of persuasion. Burke was highly suspicious of the prevailing notion that science is objective and pure of social and political agendas. Burke named the goal for a 20th century rhetoric "identification" rather than persuasion, and the difference between the new rhetoric and the classical one can be located in the need to not just move an audience, but to move the speaker as well. That is, the goal of communication, and thus composition, became one of inquiry, and as such, invention took on greater import.

The catch-word of the 1960s rhetorician was "discovery," as opposed to the reporting role that Berlin notices in current-traditional writing assignments of the late 19th century. The rhetorical role of the writer in the discovery process is altogether different from that of the reporter, and his audience is not the passive audience looking to be informed. Nor is the audience conceived of as in classical rhetoric, that is, as an other to be persuaded. In process rhetoric, the audience can often be identified as none other than the writer herself, who looks for the best possible reasons for reaching a provisional truth, as in Aristotelian rhetoric, or even as a Platonic quest for truth, as Gage asserts. As such, the writing, as an investigation, becomes increasingly inductive by necessity; and the role of writing shifts from a finished product designed to inform, persuade, or argue a point with an audience as other; and toward a synthetic, or dialectical, search for truth in the author's own mind.

The attack of Bain's paragraphing rules in the 1960s resulted, on one level, from the long overdue recognition that his rules do not adequately describe the varieties of paragraphs that are used in good writing. On another level, however, they can be perceived as a recognition that an insistence on the deductive paragraph is unsuitable to a rhetoric that emphasizes identification between composer and audience, and which views the act of writing as inquiry-led, rather than setting forth. In an "identification" rhetoric, the insistence on deductive paragraphing, and thus deductive form, is ill-suited for two reasons. For one, inductive form is more persuasive to the audience, particularly for an audience who does not share the social and political assumptions of the author. For another, induction is the process by which the writer discovers her own meaning, and such meaning is increasingly seen as something that is developed through the writing process, not prior to it. In such a view, writing as process is writing as invention; it is epistemic in itself, rather than the vehicle for setting forth knowledge.

The Contemporary Situation

From the perspective of our times, current-traditional theory can thus be seen as a sort of dark age in rhetoric, and since the 1960s, a number of approaches have been developed to address writing as a more vital, epistemic activity. In his Writing Instruction in the 19th Century American College, Berlin identified three contemporary approaches to composition theory which attempt to overstep the current-traditional approach, and to move writing instruction away from the limitations of teaching form. They share an awareness, according to Berlin, that "[w]hen we teach students to write, we are teaching more than an instrumental skill. We are teaching a mode of conduct, a way of responding to experience" (86). Each in their own way work to derail the paragraph prescriptions that continue to dominate freshman rhetorics. These three approaches are 1) the classical, 2) the expressionistic, and 3) the new rhetoric.6

The first of these approaches-the classical-attempts to return writing to its aims in classical rhetoric; specifically, as an art of public discourse that readies the student to participate in the decision-making processes of a democracy. While the current-traditional approach considers the audience only in terms of its rational capacities, classical rhetoric considers the audience's emotional stance and social reality (Berlin 88). Corbett's Classical Rhetoric for the Modern Student7 is a composition rhetoric based on the classical approach, which analyzes choices in writing according to their effect on the audience rather than accentuating correctness in grammar and structure, and as such differs radically from most composition textbooks on the market. This difference extends to the consideration of the paragraph. For one thing, its section dealing with arrangement does not break down formal requirements by paragraphs, as most contemporary rhetorics and handbooks do. The modern-traditional prescripts of unity, coherence and development are conspicuously absent. Instead, paragraphing is relegated to a small subsection of style, and is explained in its original, pre-Bainian sense; that is, as a typographical device-a punctuation mark more pronounced than the period-and like all typographical marks, conceived as the written equivalent of intonation and used to facilitate reading, or to "mark the shifts in the development of thought and indicate the relationship of the parts" (369). Thus paragraphing functions as a guide to interpretation, recalling Rodger's dictum, "To compose is to create; to indent is to interpret" (Rodgers 43).

For Corbett, paragraph length, or "density," is primarily a stylistic consideration. "Many considerations, of course, dictate whether paragraphs will be long or short-the subject matter, the occasion, the audience" (369). The need for paragraph development, however, is not ignored. While Corbett attends to the various justifications for the one- or two-sentence paragraph, such as the need for transition or emphasis, he does note the probable reason for many short paragraphs in student essays: "Many of the one- and two-sentence paragraphs that students write have no rhetorical justification whatever. Such short paragraphs simply reveal that the students have not developed their thoughts adequately" (369). His answer is to return to the topics, which is, of course, one point on which the current-traditionalist and the classicist agree. And as such, paragraph pedagogy becomes an occasion for invention in both approaches. The difference lies in Corbett's concept of the paragraph as a rhetorical marker, with the level of development dependent on the occasion it serves, rather than on any requirements of the paragraph per se.

While Berlin distinguishes between the "expressionist" and "new" rhetoric approaches to writing instruction, he does admit that the distinction can become tenuous at the ideological level, and imperceptible in the classroom (91). Both are concerned primarily with writing as a vehicle by which the writer discovers her own truths, and as such both are less concerned with the audience's response to the composition than is classical rhetoric. Adherents to the new rhetoric as identified by Berlin have all been identified as process pedagogues, and he names Becker, Pike, and Young as proponents of this approach. Here, the writing process is seen as a dialectical "means of arriving at truth," as Berlin puts it, and the role of language in this arrival is crucial. (91) In the expressionistic approach "[t]he focus is on the individual's private struggle to arrive at truth" which is "conceived as a result of a private vision that must be constantly consulted in writing" (88-89).8 The difference between the two in Berlin's schema is the role of language: in the new rhetoric, truth is essentially constructed within the language act; in the expressionistic rhetoric, language leads us to a truth that exists outside of language. But the pedagogical ramifications of both are similar, and for my purposes they can be united in that, first, composition is seen as essentially inductive, or a search for conclusions; and second, the primary audience is the writer. As such, the two approaches tend to disregard the prescriptive teaching of structure, and have done little to address such issues as paragraphing rules.

These three composition approaches, the classical, new, and expressionistic, offer ideological stances regarding the teaching of composition that return to the emphasis on good writing as good thinking. But as anyone who teaches composition knows, current-traditional pedagogy is still very much alive. The circumstances in which the topic-led, or deductive, paragraph became the norm are still in place. The "why Johnny can't write" inquiries that solidified the current-traditional approach to writing in the late 19th century still abound and are the subject of political debates and pressure on secondary schools in the form of proficiency exams, and thus students entering the university have usually been trained via a current-traditional approach. Prospective employers, and even many professors both inside and outside of English departments, still expect the same outcome from composition classes: that the students be prepared to expound upon bodies of information that they receive more or less passively, and to organize that information into a given, largely formulaic structure. The scientistic model of education still prevails, with education seen largely as an accumulation of knowledge. As long as this model persists, composition will remain marginalized.

Even in the university English department, the advantages of the current-traditional approach are much the same now as they were in 1900. Composition classrooms, often the only places where freshman university students are treated as individuals, are still led mostly by graduate students or service employees, who are given the task of attending to a large body of student writing for very little money or prestige, and who are expected to raise their students' writing to a competency standard in the course of a semester or two. The current-tradition's prescriptive rules are highly teachable, and freshman English textbooks still include them. Maxine Hairston's Contemporary Composition, for instance, contains a 37-page chapter on paragraphs that follows current-traditional rules quite faithfully.

Yet however much Hairston's composition textbook emphasizes a current-traditional approach to paragraph development, she also understood the problems inherent in the approach. In an 1986 article, she addressed the split personality evident in many composition classrooms by identifying two different types of writing, which she calls Class II and Class III writing (Class I writing being "short memos and brief notes"). She defines Class III writing as "[e]xtended REFLECTIVE writing in which the writer discovers much of his or her thought during the writing process." Class II writing includes "research reports, technical papers, laboratory reports, case studies, or summaries and analyses of assigned readings"; and such writing is "SELF-LIMITING; that is, before the writer begins to write, she already knows most of what she is going to write or she can easily retrieve the content from memory or known sources" ("Different" 95). Clearly, Hairston is making a distinction between the current-traditional and the new (or expressionistic or process) approaches.

Hairston notes that while composition instructors often prefer to assign Class III writing, it is inevitably less teachable. In Class II writing, one can "show students how to outline a paper for which they know the content, show them how to start a paragraph with a topic sentence and downshift into examples [. . .]"; however, she believes that this sort of advice is not applicable to Class III writing: "Class III writing is much harder to teach because we have no working prescriptions for it and often cannot describe [. . .] just how one goes about discovering, organizing, and then restructuring material" (ibid. 100-101).

As Hairston correctly notes, many students prefer Class II writing. It is more practical, as it will be used throughout their academic and professional careers, and it follows, after all, the structuring advice of their composition rhetorics. Further, as Hairston also notes, students don't like Class III writing because it's risky. Moving from inductive investigation to deductive presentation is much more time consuming, and there's no guarantee that the writer will make it to interesting and coherent conclusions, no guarantee of a good finished product at the end. It's a lot more work with a lot less surety of a decent grade.

While it is true that sometimes, when a student ventures to write well by thinking well and reaching her own conclusions, the writing structure actually improves, perhaps because her thinking has deepened, and the connections made among ideas are often more carefully and strategically considered. However, this is not always, or even often, the case. Papers appear that seem to be honest attempts to fulfill the assignment, but the work is, on the whole, inchoate and structurally incoherent. The writer has attempted to think through the topic and develop a stance, but has not been able to, in the week or two that she has been given to complete the assignment, create a clear and linear explication of her position. One might say that the writing has succeeded as a process, but failed as a product; or that the student has made some progress in regard to invention, but has failed to reach any certain conclusions, and has therefore been unable to proceed to effective arrangement.

In speaking of Class III writing, Hairston uses adjectives such as "romantic," "original, "expressive" and "creative"; as with the term "expressionistic," such language makes it sound like something very artsy is going on in the classroom, and we cannot blame our students, few of whom are interested in becoming artists, for shunning such work. Given Irmscher's analogies of the writer as either "a sculptor who finds form while sculpting" or "a bricklayer who piles bricks to construct a wall," most students prefer the bricklaying approach. Compared to Class II writing, the whole process of bringing Class III writing to completion is described in mystifying terms.

Perhaps for this reason, Berlin proposes renaming "new" writing "epistemic," and "expressionistic" writing "Platonic" ("Rhetoric" 88-89)-that is, the point of such writing is not to produce some beautiful work of art, after all, but to return writing to a central role in an educational process that emphasizes the ability to decide the best reasons for a particular stance on issues that have alternative stances. The lingering association of "Class III" writing (to sidestep the various and constantly proliferating terms for writing that is essentially reflective and investigative) with artistic, or literary, writing is perhaps the most debilitating point against it. Most likely the association persists because Blair conflated composition with the study of belles lettres, and because composition instructors are trained in literature more than anything. This association prevents writing teachers, as well as professors throughout the university, from recognizing such writing as central to the educational process, and as central to critical thinking skills as it is to "creativity"-at least to creativity as defined in a narrow, artistic sense.

Instead of viewing composition instruction as a "how to" course to produce a competent final product, which is, essentially, the purpose behind current-traditional instruction, composition can be viewed as a workplace in which students develop stances on issues that have no clear-cut answers. In such a classroom, reading and classroom discussion is not geared to imparting facts, but to identifying controversies and exercising critical thinking, which the students work out in the writing process. Evaluation of writing is geared less to the final product, and more to its role in the student's progress in developing and refining independent thought-an insight that we can attribute to Scott, or for that matter, Blair or Aristotle, or even and more particularly to that ancient enemy of rhetorical study, Plato. Thus instructors such as Gage would evaluate a thick stack of drafts instead of a thin stack of finished products, and writing could be assessed as documentation of the student's progress as a thinker first, and a writer second.

When such an approach is proposed, the final product is devalued by necessity. First of all, accentuating the final product booby-traps the process: if the bottom line is to find a stance and explicate it, and yet evaluation emphasizes structural considerations, few students are going to go after the hard work of developing one to any extent: easier to take a ready-made stance, whether that be the instructor's or the student's own knee-jerk response. And second of all, what is the practical purpose of developing such critical thinking into a finished product? After all, in the real world of industry, where most of our students are going, they won't be asked to produce finished products that entail much "Class III" writing. They'll be writing memos and progress reports and technical reports, and all that is covered under technical writing, and rightly so. Class II writing is also largely sufficient for research reports written for classes outside of the freshman English class. If the purpose of freshman English is to develop such skills, then we should be having our students do technical writing, as composition theorists such as Linda Flower propose. In all such writing, Bain's paragraph is ideal, and the current-traditional approach is sufficient to the task.

So why ask for Class III writing? It's the same question that Gage asked when he titled his essay "Why Write?" One answer is that the scientistic model of education is showing signs of wear. As Stanley Fish notes in "Rhetoric,"

As I write, the fortunes of rhetorical man are on the upswing, as in discipline after discipline there is evidence of what has been called the interpretive turn, the realization (at least for those it seizes) that the givens of any field of activity-including the facts it commands, the procedures it trusts in, and the values it expresses and extends-are socially and politically constructed, are fashioned by man rather than delivered by God or Nature. (128)

Composition was once the cornerstone of education because it developed the ability to enter into the decision-making responsibilities of citizens in the social and political spheres, who needed to form stances in situations where knowledge could only be provisional. Rhetoric was central because it dealt with knowledge that is socially and politically constructed, and education involved the consideration of such provisional knowledge. As even scientific knowledge is increasingly seen as provisional and culture-bound, and created within the purview of language, university departments outside of English are increasingly emphasizing the value of writing in their own disciplines, evidenced by movements toward across-the-curriculum writing programs. As education in all departments increasingly accentuates critical thinking over the accumulation of information, the connection between writing and learning must, or at least should, strengthen. Holding to the view that composition has a primarily expository purpose, taught by current-traditional prescriptions that emphasize style over content, is insufficient to the task that writing instruction needs to hold throughout the university. What we do in freshman composition classes could well define the position of English departments in the future.

Works Cited

Bain, Alexander. English Composition and Rhetoric. 1871. Facsimile. Delmar: Scholars' Facsimiles & Reprints, 1996.

Becker, A.L. "A Tagmemic Approach to Paragraph Analysis." College Composition and Communication XVI.5 (Dec. 1965). Rpt. in The Sentence and the Paragraph. Urbana: National Council of Teachers of English, n.d. 33-38.

Berlin, James A. Writing Instruction in the 19th Century American College. Carbondale: Southern Illinois UP, 1984.

---. "Rhetoric and Ideology in the Writing Class." College English 50 (1988). Rpt. in Rhetoric: Concepts, Definitions, Boundaries. Ed. William A. Covino and David A. Jolliffe. Boston: Allyn and Bacon, 1995. 734-751.

Blair, Hugh. Lectures on Rhetoric and Belles Lettres. 1783. Excerpts rpt. in The Rhetorical Tradition: Readings from Classical Times to the Present. 2nd ed. Ed. Patricia Bizzell and Bruce Herzberg. Boston: Bedford/St. Martins, 2001. 950-980.

Burke, Virginia M. "The Paragraph: Dancer in Chains." Rhetoric: Theories for Application. Ed. Robert M. Garrell. Champaign: National Council of Teachers of English, 1967.

---. Introduction. The Paragraph in Context. Ed. Virginia M. Burke. Indianapolis: Bobbs-Merrill, 1969. 5-9.

Christensen, Francis. "Symposium on the Paragraph" College Composition and Communication, XVII.2 (May 1966). Rpt. in The Sentence and the Paragraph. Urbana: National Council of Teachers of English, n.d. 49-55.

Corbett, Edward P.J. and Robert J. Conners. Classical Rhetoric for the Modern Student. 4th ed. New York: Oxford UP, 1999.

Crowley, Sharon. "The Methodical Memory on Display: The Five-Paragraph Theme." The Methodical Memory on Display: Invention in Current-Tradition Rhetoric. Board of Trustees, Southern Illinois University, 1990. Rpt. in Rhetoric: Concepts, Definitions, Boundaries. Ed. William A. Covino and David A. Jolliffe. Boston: Allyn and Bacon, 1995. 265-281.

Ferreira-Buckley, Linda and Winifred Bryan Horner. "Writing Instruction in Great Britain: The Eighteenth and Nineteenth Centuries." A Short History of Writing Instruction: From Ancient Greece to Modern America. Ed. James J. Murphy. 2nd ed. Mahwah, NJ: Hermagoras P, 2001. 173-212.

Fish, Stanley. "Rhetoric." Critical Terms for Literary Studies. Ed. Frank Lentricchia and Thomas McLaughlin. Chicago: U of Chicago P, 1990. Rpt. in Rhetoric: Concepts, Definitions, Boundaries. Ed. William A. Covino and David A. Jolliffe. Boston: Allyn and Bacon, 1995. 122-140.

Gage, John T. "Why Write?" The Teaching of Writing. Ed. Anthony Petroskey and David Bartholomew. The National Society of the Study of Education, 1986. Rpt. in Rhetoric: Concepts, Definitions, Boundaries. Ed. William A. Covino and David A. Jolliffe. Boston: Allyn and Bacon, 1995. 715-733.

Hairston, Maxine. Contemporary Composition. 4th edition. Boston: Houghton Mifflin, 1986.

---. "Different Products, Different Processes: A Theory About Writing." College Composition and Communication 37 (Dec 1986). Rpt. in Against the Grain : A Volume in Honor of Maxine Hairston. Ed. David A. Jolliffe et al. Cresskill, N.J: Hampton Press, 2002. 91-103.

Irmscher, William F. Teaching Expository Writing. New York: Holt, Rinehart and Winston, 1979.

Lewis, Edwin Herbert. The History of the English Paragraph. Chicago: U of Chicago P, 1894.

Rodgers, Paul C., Jr. "A Discourse-Centered Rhetoric of the Paragraph." College Composition and Communication, XVII.1 (Feb. 1966). Rpt. in The Sentence and the Paragraph. Urbana: National Council of Teachers of English, n.d. 39-48.

Scott, Fred N. and Joseph V. Denney. "Laws and Theory of the Paragraph." Paragraph-Writing. 2nd ed. Boston: Allyn and Bacon, 1894. Rpt. in The Paragraph in Context. Ed. Virginia M. Burke. Indianapolis: Bobbs-Merrill, 1969. 23-29.

Notes

1 As Virginia M. Burke pointed out in her 1967 essay, "The Paragraph: Dancer in Chains," expository prose from the 17th through the 19th century shows an unclear distinction between the paragraph and the sentence on the one hand, and the paragraph and the section on the other. In the 18th century, writers such as Daniel Defoe and Samuel Johnson often composed in one to two sentence paragraphs; along with earlier prose stylists such as Lyly, Spenser and Walton, they often devised paragraphs of one extremely long sentence, and it is perhaps this practice that led Bain to consider the paragraph as analogous to the sentence. However, other writers, including Richard Hooker of the 16th century through James Russell Lowell of the 19th, tended to conceive of paragraphs as analogous to entire sections, and wrote extremely long paragraphs-sometimes over 2,000 words in Lowell's case (38).

2 See "A Tagmemic Approach to Paragraph Analysis," in which Becker extends the tagmemic approach to sentence analysis to the paragraph, and identifies a variety of patterns possible in paragraph development.

3 In her introduction to The Paragraph in Context, (5) Virginia M. Burke cites E.H. Lewis, the late 19th century historian of the paragraph, who found a sixth century manuscript that included paragraphing in quotations. However, according to Lewis's findings, the modern paragraph, with the printers em marking the beginning of the paragraph, was not in evidence until 1482.

4 Interestingly enough, Rodgers did cite Scott and Denney, along with a great many other paragraph theorists in his essay, "A Discourse-Centered Rhetoric of the Paragraph," but he did not cite Scott as the originator of the concept of stadia of thought as distinguishable from the paragraph. Yet Rodgers' theory of discourse stadia is identical to Scott's, and this fact is obvious when one compares Scott's diagram, p. 23 above, to one illustrating Rodgers' stadia of discourse theory, as explained by his colleague, Francis Christensen (53). (In this diagram, "S" groups stand for stadia and "P" groups stand for paragraphs.)
______________________________________
Case I SSSSSSSSSSSSSSS
PPPPPPPPPPPPPPP

Case IIa SSSSSSSSSSSSSSSS or IIb SSSS SSSSS SSSSSS
PPPPP PPPPPPPPPPP PPPPPPPPPPPPPPPP

Case III SSSSSSSSSSSSSSSSSSS SSSSSSSSSSSSSSSSSSSSSSS
PPPPP PPPPPPPP PPPPPPPP PPPPPPPP PPPPPPPPPPP
______________________________________

5 Virginia M. Burke points this out in her introduction to The Paragraph in Context (5): "The paragraph graphic mark or signal, the oldest in Greek manuscripts, first appeared as a horizontal stroke, sometimes with a dot over it, just below the first two or three letters of a line to indicate that a sentence or some larger unit of discourse was closing in the underscored line."

6 In his 1988 essay, "Rhetoric and Ideology in the Writing Class," Berlin renames "new" rhetoric "social-epistemic," and emphasizes in this approach the writer's realization of her standing in her economic and social situation.

7 The fourth edition, which I am using, is co-authored by Robert J. Conners.

8 Berlin cites William Coles, Jr., Ken Macrorie, James E. Miller and Stephan Judy, and Ronald Stewart, as proponents of such an approach.