Using Transformational Grammar as an Editing Tool

(The text version of a paper given at IPCC 95, Savannah, Georgia, on September 29, 1995)
Abstract: Reviewers and editors of technical prose often encounter sentences that seem needlessly opaque. To revise such sentences for readers' benefit without disrupting meaning, you need to be able to analyze their structure. Once you discover what the actual underlying structure is, you can see how to change it for greater clarity and continuity. For analytic purposes, a tool superior to traditional grammar is transformational grammar. Using it, one can describe sentences according to a limited number of types of transformations and embeddings of simple sentences. In this presentation, I'll show how its analytic techniques can help editors and reviewers help readers.

What's "Technical" about Technical Reviewing and Editing?

With online documentation, hypertext links, collaboration, and usability tests getting so much attention, one could reasonably think that the "technical" in technical communication (TC) refers to the content or the means of production. If the "technical" part of an editor's or reviewer's background includes linguistic science, it's unusual. A browse through Academic Programs in Technical Communication (Geonetta 1993)suggests that only about a third of the schools offering bachelors degrees or minors in TC require coursework in editing. Many fewer appear to require coursework in grammar or linguistics.

People with formal TC backgrounds make up only a small portion of all who review and edit documents. Thus, one might guess that most editors and reviewers are operating on intuition and whatever sense of grammatical structure they learned in grades K- 12. Where in their educations are technical editors supposed to gain technical skills for analyzing language? Editors now learn to work with language mostly by reading literary works. Since we can produce and understand sentences of marvelous complexity, we should therefore also be able to analyze complicated sentences, right?

Not only not right, not even logical. True, linguistic theory states that all of us acquire as toddlers the generative rules allowing us to produce idiomatic sentences in our native language. Generating sentences, 19th century faculty psychology might say, requires the use of the synthetic imagination. But analytic reason comes into play for editing sentences. Jim Hanna (1994), in comparing legalese and layperson versions of a legal notice, has used fruitfully traditional grammar's sentence diagraming as an analytic tool.

A better "tool-set" for editors and peer reviewers comes from transformational grammar. Over the past 40 years, one variety of descriptive grammar, transformational, has become standard. Some recent graduates of technical communication programs may have studied linguistics enough to understand transformational theory, but those of us of a certain age learned only "traditional" grammar, and (if my own students are any indication) many younger people get no exposure to grammar at all. Even in technical editing texts, some ( Coggin and Porter 1993, Samson 1993) seem to assume that readers have already picked up the tools for analyzing language somewhere else; others ( Stoughton 1989,Rude 1991, Eisenberg 1992, Mancuso 1992, Tarutz 1992) offer examples of editing without explaining the theory behind it (C.P. Campbell 1993a-d, 1994). One-- Dragga and Gong (1989)--treats stylistic editing as a rhetorical element, using linguistic descriptors (C.P. Campbell 1991), but not syntactic ones. In a recent listing of 20 characteristics of a good editor (Grove and Johnston 1994), 17 are attitudinal. Only three (attention to detail, passion for clarity, sense of logic) depend at least in part on technical skills with language.

And editing does require technical skills. The crucial ones produce continuity in documents, continuity being achieved by the producers' using techniques of cohesion that make the discourse seem coherent to a reader. For a succinct synthesis and catalog of continuity devices, visual as well as verbal, see K.S. Campbell (1995). Her synthesis derives from linguistics, mainly from the branches concerned with the workings of discourse in larger units (pragmatics) but also partly from syntactics. For developing continuity and clarity in text, Joseph M. Williams' Style (1994) is useful to editors. I've written a bit on that subject myself (C.P. Campbell 1992,Bush and Campbell 1995).

Higher-order editing skills need an underlying basis in syntactic skills. Around 95 CE, Quintilian counseled that students must master the skills of the grammarian before they can study with the rhetorician. In modern linguistic terms, the ability to achieve pragmatic effects depends in large part on the ability to maneuver at the syntactic level.

"Grammar," for most, means "the rules I'm supposed to follow to avoid looking like a dolt in print." Many of these rules, as Finegan (1980) and Baron (1982) have shown, were made up by verbal critics and prescriptive grammarians who, as experts, were usually self-appointed. Their made-up rules are the least observed, according to Williams (1994). Williams distinguishes between such faddish rules, as well as those that distinguish Standard from nonstandard speech (such as avoiding "he don't"), and the real rules, those observed by every native speaker.

Understanding the Real Rules with Transformational Grammar

The revelation of the real rules is the project of the several varieties of descriptive grammar. See, for example, the articles "Grammar" and "Twentieth-century linguistics" in Crystal (1987). The transformational variety has been around long enough by now to provide a stable analytic paradigm. Chomsky (1957) and, for example, Akmajian and Heny (1975) and Kaplan (1995) pretty much agree on what constitutes a canonical sentence in English and on the kinds of transformations that produce more complicated sentences. Morenberg (1991) gives perhaps the most direct and readable account of transformational methods. Morenberg's orderly "doing grammar" approach makes it fairly easy to learn the basic transformations.

Transformational grammar, like Freudian psychology, assumes an unconscious. In it reside the transformational rules that produce the sentences we speak. Underlying every utterance, or surface structure, is a deep structure. The deep structure, which can only be revealed by analysis, contains canonical sentences. Canonical sentence structure varies by language; in English it is noun phrase plus verb phrase (S=NP+VP). The basic canonical sentences are few, being permutations of four basic kinds of VP.

Transformations of canonical sentences produce other surface structures, such as questions, commands, and passive sentences. Other deep-structure transformations embed sentences within other sentences as modifiers--restrictive relative clauses, further reducible to participial phrases, prepositional phrases, and adjectives; nonrestrictive modifiers, including absolute phrases, subordinate clauses, participial phrases, and relatives.

Often, people try to fix sentences by dealing with the surface structures. They try to make do using hyphens and commas as Band-Aids. But if you run into a phrase like "first draft software documentation usability test procedure," hyphens can't help much. You can either consider the whole thing a noun phrase, maybe reducing it to an initialism ("The FDSDUTP showed a need to revise . . ."), or you can trace the deep structure, understand the relationships there, and then revise the surface structure.

In this space, it's not possible to do more than sketch the theory, but it may be possible to show how the theory could help a reviewer deal with a difficult sentence.

Applying Transformational Grammar

Here's the kind of sentence that TG can help unscramble. This one came from a draft report.
Visual observation of the waterflood displacement mechanisms after asphaltene precipitation are shown to be similar to the mechanisms of displacement of the unprecipitated crude oil.
Such a sentence looks like two men in a horse costume--except that they're wearing half-costumes from different horses. One TG principle says that every sentence has to have a matrix clause, one to which other embedded clauses attach themselves as piglets to a sow. To find the matrix clause, we can use a crude process of elimination to reveal the basic structure: bracket off modifying phrases and clauses.
Visual observation (of the waterflood displacement mechanisms) (after asphaltene precipitation) are shown <to be similar {to the mechanisms [of displacement (of the unprecipitated crude oil)]}>.
The matrix clause turns out to be this:
Visual observation are shown to be [something].
A full tree diagram of the sentence appears in Figure 1. Besides being grammatically incorrect, the sentence's structure suggests that it's about "visual observation," though its real topic is the similarity of mechanisms. To fix it, you rebuild a matrix clause, one in which the verb phrase (predicate) is consistent with the noun phrase (subject). Because we recognize a passive transformation, are shown, we know there's an agent lurking nearby. From the -ion ending, we recognize observation as a smothered verb:
[Somebody] observes [something].
The original sentence is in passive voice, suggesting that an agent got suppressed during the transformation. Transformed back into a canonical sentence, it would look like this:
[Somebody] shows [some] mechanisms to be similar to [other] mechanisms.
If "somebody" were "petroleum geologists," we might make that phrase the subject of a new sentence. More likely, though, our topic is "mechanisms," so the passive transformation could stand without the redundant "visual observations":
[Some] mechanisms [under some conditions] are shown to be similar to [other] mechanisms.
If agency really is unimportant, we could eliminate the passive transformation, then re-embed the original subject as a sentence modifier, whose placement or inclusion is optional:
[Some kind of] mechanisms, observed visually, are similar to [other kinds of] mechanisms.
All that's left is to re-embed the modifying phrases:
The waterflood displacement mechanisms after asphaltene precipitation are similar to the displacement mechanisms of the unprecipitated crude oil.
"Mechanisms of displacement" is reduced to "displacement mechanisms," parallel to the subject noun phrase. Our revised sentence may still be less than perfectly clear. That's because of complicated embeddings in the technical jargon. Were we to account for all the transformed canonical sentences embedded in the revised version, the list might look like this (reduction-transformations indicated by arrows):
Mechanisms(1) are similar to mechanisms(2). (Matrix clause)
Mechanisms(1) displace crude oil by [a means]. (Verb->noun->adjective)
Water floods [something]. (Noun+verb->noun)
A means is waterflood. (Noun->adjective)
[Somebody] precipitates [something] with asphaltene. (Clause->adverbial sentence modifier)
Mechanisms2 displace crude oil. (Verb->noun->;adjective)
Crude oil is unprecipitated. (verb->past participle->adjective)
This list suggests why people who aren't petroleum engineers or geologists may still be bemused by the restructured sentence. That innocent-looking prepositional phrase "after asphaltene precipitation" is the residue of a whole clause that might have explained who was doing the precipitating and what was being precipitated.

An editor or reviewer, by analyzing a sentence thus, could hazard an educated guess about the knowledge the sentence assumes. Somebody (oil company crews?) injects asphaltene into a geological formation that contains crude oil. The asphaltene causes some of the oil to settle to the bottom of the formation. This oil can be forced out by waterflooding, a mechanism for displacing it. But the sentence suggests that precipitation by asphaltene may not be necessary, since the displacement mechanisms for the precipitated and the unprecipitated crude are similar.

Transformational Techniques in the World of Work

In the case just presented, transformational analysis would give a reviewer or editor a tool for negotiation with a writer. Even if the interpretation by the reviewer was wrong, it should call attention to a need for rephrasing.

Obviously, no one could use this sort of full-blown analysis very often. It's not much fun to do, except in the way that solving calculus problems is fun. And it wouldn't exactly be a pleasure to receive such an analysis. But the method is worth learning, like calculus, because it's an active knowledge rather than a passive one.

Passive knowledge is at best semioperational. It requires memorizing rules ("Avoid the passive.") and learning to recognize when to apply them ("Is 'Ed was retired' a passive?"). An active knowledge is always operating in the background, activating itself when it's needed.

As active knowledge, transformational technique comes to the fore in response to sentences such as I've discussed here. It alerts intuition from background, then brings tools to bear. Testimonial: the "displacement mechanism" sentence came to me from an editor who couldn't quite pinpoint what was wrong with it. That was several years ago, before I'd started teaching an elective grammar course. Then, using my trusty bracketing heuristic to isolate main sentence elements, I could tell him about the subject-verb mismatch. Now, transformational theory helps me see why technical prose is so dense and so complicated-- and why it so easily becomes impenetrable. And, like Ariadne's thread, it offers a way out of the labyrinth.

References

Akmajian, Adrian, and Frank Heny. 1975. An Introduction to the Principles of Transformational Syntax. Cambridge, MA: MIT Press. Back to article

Baron, Dennis E. 1982. Grammar and Good Taste: Reforming the American Language. New Haven: Yale UP. Back to article

Bush, Donald W., and Charles P. Campbell. 1995. How to Edit Technical Documents. Phoenix: Oryx. Back to article

Campbell, Charles P. 1994. Review of Samson's Editing Technical Writing (Oxford, 1993). Technical Communication 41 (May): 306- 308. Back to article

_________. 1993a. Review of Coggin and Porter's Editing for the Technical Professions (Macmillan, 1993). Technical Communication 40 (November): 733-734. Back to article

__________. 1993b. Review of Tarutz's Technical Editing: The Practical Guide for Editors and Writers (Addison Wesley, 1992). Technical Communication 40 (May): 278-279. Back to article

__________. 1993c. Review of Mancuso's Technical Editing (Prentice Hall, 1992). Technical Communication Quarterly 2 (Spring): 232-234. Back to article

__________. 1993d. Review of Eisenberg's Guide to Technical Editing (Oxford, 1992). Technical Communication 40 (February): 121-122. Back to article

__________. 1992. "Engineering Style: Striving for Efficiency." IEEE Transactions on Professional Communication 35 (September): 130-137. Back to article

__________. 1991. Review of Dragga and Gong's Editing: The Design of Rhetoric (Baywood, 1989). Technical Writing Teacher 18 (Fall 1991): 251-253. Back to article

Campbell, Kim Sydow. 1995. Coherence, Continuity, and Cohesion: Theoretical Foundations for Document Design. Hillsdale, NJ: Laurence Erlbaum. Back to article

Chomsky, Noam. 1957. Syntactic Structures. The Hague: Mouton. Back to article

Coggin, William O, and Lynnette R.Porter. 1993. Editing for the Technical Professions New York:Macmillan. Back to article

Crystal, David. 1987. The Cambridge Encyclopedia of Language. New York: Cambridge University Press. Back to article

Dragga, Sam, and Gwendolyn Gong. 1989.Editing: The Design of Rhetoric. Amityville, NY: Baywood. Back to article

Eisenberg, Anne. 1992. Guide to Technical Editing: Discussion, Dictionary, & Exercises. New York: Oxford. Back to article

Finegan, Edward. 1980. Attitudes Toward English Usage. New York: Teachers College Press. Back to article

Geonetta, Sam C. 1993. Academic Programs in Technical Communication, 4th ed. (Draft) Alexandria, VA: STC. Back to article

Grove, Laurel K., and Beverly V. Johnston. 1994. "What Does It Take to Be a Good Editor?" IEEE-PCS Newsletter 38 (July/August): 9. Back to article

Hanna, Jim. 1994. "Diagram for Analysis." Technical Communication 41 (May): 327-336.Back to article

Kaplan, Jeffrey P. 1995. English Grammar: Principles and Facts. 2nd ed. Englewood Cliffs, NJ: Prentice Hall. Back to article

Mancuso, Joseph. 1992. Technical Editing. Englewood Cliffs, NJ: Prentice Hall. Back to article

Morenberg, Max. 1991. Doing Grammar. New York: Oxford.Back to article

Samson, Donald C., Jr. 1993. Editing Technical Writing New York: Oxford. Back to article

Stoughton, Mary. 1989. Substance and Style: Instruction and Practice in Copyediting. Alexandria, VA: EEI. Back to article

Tarutz, Judith. 1992. Technical Editing: The Practical Guide for Editors and Writers Reading, MA: Addison Wesley. Back to article

Williams, Joseph M. 1994. Style: Ten Lessons in Clarity and Grace. 4th ed. New York: HarperCollins. Back to first reference Back to second reference

© 1995.
Charles P. Campbell, PhD
505-881-2861, E-mail: cpc@nmt.edu
New Mexico Institute of Mining and Technology
Humanities Department
Socorro, NM 87801