The Grammar of Coherence

Author(s): W. Ross Winterowd

Source: College English, Vol. 31, No. 8 (May, 1970), pp. 828-835
Published by: National Council of Teachers of English
Stable URL: http://www.jstor.org/stable/374228
Accessed: 11-07-2018 11:27 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

National Council of Teachers of English is collaborating with JSTOR to digitize, preserve

and extend access to College English

This content downloaded from 160.45.152.6 on Wed, 11 Jul 2018 11:27:04 UTC
All use subject to http://about.jstor.org/terms
W. Ross WINTEROWD

The Grammar of Coherence

JUST AT THE POINT where it could best serve rhetoric transformational generativ
grammar fails: it does not jump the double-cross mark (#) that signifies "sentenc
boundary" or, more accurately, "transformational unit boundary." The signifi-
cance of this limitation is underscored by the inability of grammarians to write
rule for the simplest of all transformations: clause coordination.

Since the number of sentences that can be conjoined in this way is, theoretically
at least, unlimited, it is not immediately obvious how to write a constituent-
structure rule to permit the generation of compound sentences. . . . It is clearly
unsatisfactory to have to postulate an infinity of rules. ... .

As a result, transformational generative grammar has been tremendously usefu

in the study of style, but it has had little application (except metaphorically) t
invention and organization. That is, it has cast only dim light on concepts of form
and coherence.
The following discussion will argue that there is a grammar of coherence (or
form, for in the following, the two terms are virtually synonymous.) If one per-
ceives form in discourse, he also perceives coherence, for form is the internal set of
consistent relationships in any stretch of discourse, whether poem, play, essay,
oration, or whatever. This set of relationships-like the relationships that rules of
grammar describe-must be finite in number; otherwise: formlessness, for the
very concepts of form and coherence imply a finite number of relationships
that can be perceived. (A generative grammar implies a finite number of rules,
some of which may be applied recursively.) Following the model of grammar,
one might look for some sort of "constituent structure rules" that underlie
coherent utterances beyond the sentence, and then for the equivalent of "lexical
rules," and finally for something approximating "transformational rules." In a
very rough, loosely analogous way, the following discussion concentrates only
on the "phrase structure rules" of coherence and, as a result, excludes "lexical"
data which is undoubtedly significant. For instance, one reason that a paragraph
"hangs together" or is a convention is that chains of equivalent words run through
it. A switch in equivalence chains signals: new paragraph.2 The present discussion

TV. Ross Winterowd has published many books and articles on style and rhetoric. He teaches
in the Department of English at the University of Southern California.

ID. Terence Langendoen, The Study of Syntax (New York, 1969), p. 31.
2A. L. Becker, "A Tagmemic Approach to Paragraph Analysis," The Sentence and the Para-
graph (Champaign, Ill., 1966), pp. 33-38.

828

This content downloaded from 160.45.152.6 on Wed, 11 Jul 2018 11:27:04 UTC
All use subject to http://about.jstor.org/terms
 The Grammar of Coherence 829

will ignore everything but the abstract configurations or sets of relationships

that constitute coherence. (This, of course, is not to say that any one component
of the whole body of discourse is unimportant.)
Modern grammar nicely describes the first two stadia in the hierarchy of dis-
course relationship sets that make up coherence. The first set of relationships is
those that can develop from the application of rules to S and then to all con-
stituents that develop from S. The result (after lexical rules have been applied)
will be a sentence, divided into two parts: Modality and Proposition. As Charles
J. Fillmore explains:

In the basic structure of sentences . . . we find what might be called the

'proposition,' a tenseless set of relationships involving verbs and nouns (and
embedded sentences, if there are any), separated from what might be called the
'modality' constituent. This latter will include such modalities on the sentence-
as-a-whole as negation, tense, mood, and aspect.3

Each noun in the proposition stands in a case relationship with the verb, thus:

Sentence

Modality Proposition

Tense Verb Agentive Dative Obective Instrumental

I I A A
Past pay by Jones to Smith 0 the money with a check

This deep structure can have the following surface manifestations, all of them
synonymous:
(1) Jones paid Smith the money with a check.
(2) Jones paid the money to Smith with a check.
(3) The money was paid Smith by Jones with a check.
(4) The money was paid to Smith by Jones with a check.
(5) The money was paid by Jones to Smith with a check.
(6) Smith was paid the money by Jones with a check.
And with the cleft sentence transformation: A check is what Jones paid Smith
the money with. It is worth pointing out that syntactic relationships in these
sentences change, but case relationships ("who did what and with which and to
whom") are invariable. Thus, in 1 and 2, "Jones" is the grammatical subject of the
verb; in 3, 4, and 5, "the money" is the grammatical subject; in 6, "Smith" is the
subject. But "Jones" is always in the agentive case, "the money" is always in the
objective case, and "Smith" is always in the dative. That is, we never lose sight of
the relationships among the noun phrases or of their relationships with the
verb. It is also worth noting-in fact, crucial to this discussion-that certain
"particles" which are represented in the deep structure diagram may or may not
appear in the surface structure. Thus, the agentive "by" does not appear until
3"The Case for Case," Universals in Linguistic Theory, ed. Emmon Bach and Robert T.
Harms (New York, 1968), p. 23.

This content downloaded from 160.45.152.6 on Wed, 11 Jul 2018 11:27:04 UTC
All use subject to http://about.jstor.org/terms
830 COLLEGE ENGLISH

after the passive transformation has been applied, and dative "to" disappears with
application of the indirect object inversion transformation. These signals of cas
relationships may or may not be in the surface structure.
The first "layer" of relationships that make up coherence, then, is cases.
The second "layer" might well be called syntax (in a somewhat specialized
and restricted use of the word). The relationships of syntax are described by thos
transformations that have to do with inserting sentences within other sentences
by any means but coordination. Thus, the relationships characteristic of syntax
(as I use the word) are, for instance,

complements:
It is strange. He is here.
It is strange that he is here.
It is strange for him to be here.
His being here is strange.

relatives:
The banker owned the town. The banker was rich.
The banker who was rich owned the town.
The rich banker owned the town.

subordinates:
He chews tobacco. He likes it.
He chews tobacco because he likes it.

absolutes:
The airport was fogged in. The plane circled for an hour.
The airport being fogged in, the plane circled for an hour.
and so on.

This is the cat. The cat chased the rat. The rat ate the malt. The malt lay in
the house. Jack built the house.
This is the cat that chased the rat that ate the malt that lay in the house that
Jack built.

And this, of course, is just the point at which grammar ends-that very point at
which inventio and dispositio really begin.
I argue that there is a set of relationships beyond case and syntax and that this
set constitutes the relationships that make for coherence-among the transforma-
tional units in a paragraph, among the paragraphs in a chapter, among the
chapters in a book. I call these relationships transitions, and I claim that beyond
the sentence marker, the double-cross, we perceive coherence only as the con-
sistent relationships among transitions. All of this, of course, is more easily il-
lustrated than explained, and illustration is forthcoming. For the moment,
however, I should like to underscore my claim that the relationships I am about

This content downloaded from 160.45.152.6 on Wed, 11 Jul 2018 11:27:04 UTC
All use subject to http://about.jstor.org/terms
 The Grammar of Coherence 831

to describe constitute the grammar of coherence for all units of discourse beyond
the level of what I have called "syntax."4
In another place, I will detail the method whereby I arrived at the following
conclusions. But for the time being, I will concentrate on results and their ap-
plications.
Analysis of thousands of transformational units in sequences reveals that there
are seven relationships that prevail among T-units and, I would argue, in any
stretch of discourse that is perceived as coherent. I have called these relation-
ships (1) coordinate, (2) obversative, (3) causative, (4) conclusive, (5) alternative,
(6) inclusive, and (7) sequential. These relationships can be either expressed or
implied. They are expressed in a variety of ways: through coordinating con-
junctions, transitional adverbs, and a variety of other moveable modifiers. Just
how they are implied remains a mystery.5 However, the relationships are easily
demonstrated.

Coordination can always be expressed by and. (Synonyms: furthermore, in

addition, too, also, again, etc.)
Boswell was a Rousseau-ite, one of the first of the Romantics, an inveterate senti-
mentalist, and nothing could be more complete than the contrast between his
career and Gibbon's.-Lytton Strachey
They almost hid from us the front, but through the dust and the spaces between
running legs we could see the soldiers in the trench leap their barricade like a
breaking wave. AND then the impenetrable dust shut down AND the fierce stab-
bing needle of the machine guns sewed the mighty jumble of sounds together.
-John Reed

... Marat is, in most of his speeches, tinsel, stage scenery, or an element in a
great painting. AGAIN, the Brechtian songs are touching, but ironically and
allusively touching; Charlotte Corday, the mad, beautiful country girl mouthing
her lines, is AGAIN an element in a picture, an aesthetic contrivance.-Stuart
Hampshire

Obversativity can always be expressed by but. (Synonyms: yet, however, on

the other hand, etc.)

4The reader who is familiar with modern logic will immediately perceive the similarity be-
tween what I am about to outline and the relationships among propositions listed in logic. They
are initial, additive (and), adversative (but), alternative (or), explanatory (that is), illustrative
(for example), illative (therefore), causal (for). I would urge the reader, however, to be more
conscious of the differences between the two systems than of the similarities. What I call
transitions are not merely an adaptation, but, it seems to me, are manifestations of some of the
most basic properties of language.
5When I first began working on these ideas, I communicated my findings to Charles Fillmore
of Ohio State. His comment on my tentative conclusions is revealing. I was talking strictly
about the relationships in the paragraph, and he said, "Your ideas about paragraph structure are
appealing, but it's hard to see, as you admit, how they can lead to any clarification of the prob-
lems of coherence on the paragraph level. The 'coherence' of clauses in a sentence is just as
unsolved an issue as ever, but to the extent that your proposals are right you can at least claim
to have demonstrated that what might have appeared to be two separate mysteries are reducible
to one and the same mystery." The fact that coherence among clauses in a T-unit and co-
herence among T-units are reducible to the same mystery is, of course, the point here, not that
coherence is mysterious. In general, I am indebted to Professor Fillmore for a great variety of
insights.

This content downloaded from 160.45.152.6 on Wed, 11 Jul 2018 11:27:04 UTC
All use subject to http://about.jstor.org/terms
832 COLLEGE ENGLISH

It has been ambitious and plucky of me to attempt to describe what is inde-

scribable, and I have failed, as I knew I would. BUT I have discharged my duty to
society... .-E. B. White
And Johnson, as Kennedy has often acknowledged, was a man of force and decision
to whom, in case anything happened, the government could responsibly be
assigned.
ON THE OTHER HAND, the designation of Johnson would outrage the
liberal wing of the party.-Arthur Schlesinger, Jr.

Causativity can always be expressed by for. It is interesting to note that

among the transitional adverbs commonly used (nevertheless, however, moreover,
hence, consequently, nonetheless, accordingly, then, besides likewise, indeed,
therefore), none expresses the causative relationship.

Now, on that morning, I stopped still in the middle of the block, FOR I'd caught
out of the corner of my eye a tunnel-passage, an overgrown courtyard.-Truman
Capote

Conclusivity can always be expressed by so. (Synonyms: therefore, thus, for

this reason, etc.)

She has a rattling Corsican accent, likes Edith Piaf records, and gives me extra shrimp
bits in my shrimp bit salad. SO some things change. Last time I heard no Edith Piaf
and earned no extra forkfuls of shrimp.-Herbert Gold

Alternativity can always be expressed by or.

Now such an entity, even if it could be proved beyond dispute, would not be
God: it would merely be a further piece of existence, that might conceivably not
have been there-OR a demonstration would not have been required.-John A. T.
Robinson

Inclusivity is often expressed with a colon.

In the first century B.C., Lucretius wrote this description of the pageant of Cybele:
Adorned with emblem and crown . . . she is carried in awe-inspiring state....
-Harvey Cox

The inclusive relationship is that of the example to the generality or the nar-
ration of the case to the statement of the case. Often, inclusivity is expressed by
the transformational possibility of complementization:

He realized that their discovery [Aristotle's discovery of the statues of Daedalus]

would shatter his own "natural" law: Managers would no longer need subordi-
nates, masters could dispense with slaves.-Michael Harrington

With the last two clauses complementized, the sentence reads like this:

He realized that their discovery would shatter his own "natural" law, that man-
agers would no longer need subordinates, and that masters could dispense with
slaves.

The sequential relationship is expressed by such transitions as "first ... second

This content downloaded from 160.45.152.6 on Wed, 11 Jul 2018 11:27:04 UTC
All use subject to http://about.jstor.org/terms
 The Grammar of Coherence 833

... third," "earlier ... later," "on the bottom ... in the middle ... on top," and
so on.

Three types of relationships, then, constitute coherence

transitions (with the transitions either expressed or implied
obvious that transitions can be implied, for it is common to f
formational units with no expressed transitions.

It is possible to love the theater and to revel in theatricality, t

and unreality of the stage wholly absorbing in its own right. I
that most actors and directors, if left to their own tastes and im
after theatrical effects above all else. The satisfaction of
interest might be quite secondary.-Stuart Hamnpshire

I read this as an "and ... and" series, but another interpretat

... for." That is, there can be ambiguity in transitions as we
syntax.
The interesting possibility, however, is that the seven relationships that prevail
among T-units also prevail among larger elements of discourse. For instance, ap-
plied at the level of the T-unit, the seven constitute a series of topics that will
automatically generate discourse, for the second T-unit must stand in one of the
seven relationships to the first, and the third must stand in one of the seven rela-
tionships to the second, and so on. Therefore, transitions are topics for a genera-
tive rhetoric. But a rhetoric that will generate only paragraphs has limited useful-
ness. If, however, the seven topics isolate the relationships among any segments of
discourse (except those related to one another by the grammar of the T-unit),
then they might well be the basis for a true generative rhetoric.
Shakespeare's sonnets have proved to be useful models for my purposes, and
they will serve here to demonstrate that the seven relationships prevail in "whole
works," though, of course, one might argue that a sonnet is, after all, just another
kind of paragraph. Expressed transitions in the following will be in capitals; im-
plied transitions (according to my reading) will be in bracketed capitals.

Sonnet XVII
Who will believe my verse in time to come,
If it were fill'd with your most high deserts?
THOUGH YET, heaven knows, it is but as a tomb
Which hides your life and shows not half your parts.
[BUT] If I could write the beauty of your eyes
And in fresh numbers number all your graces,
The age to come would say 'This poet lies;
[FOR] Such heavenly touches ne'er touch'd earthly faces.'
SO should my papers, yellowed with their age,
Be scorn'd, like old men of less truth than tongue,
And your true rights be term'd a poet's rage
And stretched metre of an antique song:
BUT were some child of yours alive that time,
You should live twice, in it and in my rhyme.

In his cycle, Shakespeare upon occasion needs two sonnets rather than one t

This content downloaded from 160.45.152.6 on Wed, 11 Jul 2018 11:27:04 UTC
All use subject to http://about.jstor.org/terms
834 COLLEGE ENGLISH

press his complete idea. In such cases, he supplies the proper transition. The re-
lationship between V and VI is conclusive, expressed as then. (So is the minimal
transition to express conclusivity.)

Sonnet V
Those hours that with gentle work did frame
The lovely gaze where every eye doth dwell,
Will play the tyrants to the very same
And that unfair which fairly doth excel:
FOR never-resting time leads summer on
To hideous winter and confounds him there;
Sap check'd with frost and lusty leaves quite gone,
Beauty o'ersnow'd and bareness every where:
THEN, were not summer's distillation left,
A liquid prisoner pent in walls of glass,
Beauty's effect with beauty were bereft,
Nor it, nor no remembrance what it was:
BUT flowers distill'd, though they with winter meet,
Leese but their show; [FOR] their substance still lives sweet.

Sonnet VI
THEN let not winter's ragged hand deface
In thee thy summer, ere thou be distill'd:
[BUT] Make sweet some vial; [AND] treasure thou some place
With beauty's treasure, ere it be self-kill'd.
[FOR] That use is not forbidden usury,
Which happies those that pay the willing loan;
[FOR] That's for thyself to breed another thee,
OR ten times happier, be it ten for one;
[FOR] Ten times thyself were happier than thou art,
If ten of thine ten times refigured thee:
THEN what could death do, if thou shouldst depart,
Leaving thee living in posterity?
[SO] Be not self-will'd, FOR thou art much too fair
To be death's conquest and make worms thine heir.

To apply this test to a series of paragraphs that make up an essay, for in-
stance, is too cumbersome a job for the present discussion and is, in any case,
unnecessary. The reader can make his own test. "What relationships prevail
among the sections-paragraphs or other-of an extended piece of discourse?"
is the question. If the seven outlined here are the answer, then the system has
stood the test. (By the way, the question transformation might be viewed as a
transition in itself. That is, it predicts some kind of answer.)
Finally, it is necessary to clarify the exact sense in which I take these seven
relationships (they might be called "topics") to constitute a generative rhetoric.
The term "generative" is of itself productive, for it exactly designates the process
whereby discourse-at the sentence level and beyond-comes into being. An
oversimplified explanation of the language process is to say that at any level of
generality, one unit has the potential for generating other units and of com-
bining these units in some meaningful way. Any set of topics is merely a way of
triggering the process. Thus the student, say, who has difficulty with the invention

This content downloaded from 160.45.152.6 on Wed, 11 Jul 2018 11:27:04 UTC
All use subject to http://about.jstor.org/terms
 The Grammar of Coherence 835

of arguments, can use the seven-item list to tell him what might come next-not
what content, to be sure, but what relation his next unit must take to the previous
one. There are only seven possibilities.
Inability to write sentences stems not from the writer's lack of subject
matter (everyone is the repository of an infinitude of subject matter), but from his
not knowing how to get the subject matter into structures. The problem at levels
beyond the sentence is, I think, exactly the same. The seven relation-oriented
"topics" that I have outlined name the structures that can hold the writer's ideas.
A generative rhetoric, a heuristic model, even a grammar of form-whatever
it might be called, the schema of these seven relationships ought to be easily ap-
plicable in the classroom. But equally important, they should have wide ranging
theroretical possibilities, for instance, in explaining the disjunction of schizoid
language, in identifying "the eighth ambiguity" (that which takes place between
units larger than the sentence and results from the inability to perceive transitions)
and in dealing with form in literature.

This content downloaded from 160.45.152.6 on Wed, 11 Jul 2018 11:27:04 UTC
All use subject to http://about.jstor.org/terms