Style and Effectiveness in
Technical Communication

This paper was originally presented at Style Council 91 (Melbourne, November 1991). It was subsequently published in Australian Style into the Nineties (ed. P.H. Peters; Macquarie University, 1992)

Although this paper is included in the "Norms and Standards" part of the program, it has to be said that Technical Communication is notably lacking in any general standards. The field is too diverse for any such generality. So rather than talking about something that doesn’t exist, I want to explore the normative influences that I think do exist.

I want to discuss first exactly what Technical Communication is. Once we have a common understanding of that, we can usefully talk about what style means in the genre, and from there I’ll go on to consider how to get it right (or wrong). But the last will necessarily be far from rigorous, and it’s worth pausing for a moment to ask why.

Technical Communication is perhaps the most pervasive genre in modern society. Yet it has only recently begun to excite the level of academic interest that it really warrants. Perhaps it has been seen as too sordidly commercial. Whatever the reason, there is no significant corpus of Australian examples: perhaps, given the erosion of our manufacturing industry, there never will be.

Let me illustrate this point by referring to the paper which Pam Peters presented at Style Council 88. Out of 500 samples taken up by the Australian Corpus project, as shown in the table which was included in that paper, there were apparently no samples of user manuals, product descriptions, or indeed any work normally considered the province of technical writers.

A profile of the Australian Corpus
Taken from The Australian Corpus Project: Word Punctuation in Newspapers,
P.H. Peters, Style Council 88

We might think that this doesn’t matter much: that the results of research done in the US and the UK can be used here. So, to some extent, they can. But the makeup of the Australian Corpus is based on the two most important international corpora, so perhaps the situation is not very different elsewhere.

In fact, though, there is a lot of work being done, especially in the US. Carnegie Mellon University, the University of Delaware, Rensselaer Polytechnic, the Colorado School of Mines, and probably many other institutions, are putting a lot of energy into this line of research. Even so, we cannot accept their findings uncritically. We speak a different variety of English, and we have a need to accommodate a non-Anglophone public which doesn’t really match anything found overseas.

To illustrate, there can be no argument that safety-related instructions MUST be clear. But, given that a domestic user of a microwave oven might be Turkish, Italian, or Vietnamese, with little English, how can we be sure that the important messages are getting through?

Clearly, non-Australian material can be of some value, but equally clearly it cannot be taken as wholly adequate for our local research needs. The only real solution is to develop our own corpus. Without one, we can either use overseas data and hope for the best, or be anecdotal. For today’s purposes, I will choose the latter.

So much, for now, for our research problems. Let me return to the question of just what it is we are trying to research into.

Technical communication is often taken to be synonymous with technical writing, and I took advantage of that common view a moment ago. But this common view misses the fact that "writing" is a solitary act, while "communication" is an interactive process Technical communication is not the creation of documentary or textual resources, such as specifications: it is the creation of documentary or textual instruments, such as user manuals.

Because any communicative process must be interactive, the receiver of technical communication has an active role. It is the task of the initiator of technical communication to make the receiver’s role as easy as possible. It is not the reader’s job to go mining for nuggets of useful information: it is the writer’s job to present the information in a way that is suited to the reader’s needs. (To refer to jobs here is entirely appropriate: technical communication is necessarily and unashamedly commercial.)

Our first normative influence is the way that technical communication as a genre can be categorised according to purpose. It includes three types, which we can characterise like this:

Descriptive texts are concerned with "What it is";

Explanatory texts are concerned with "How it works";

Tutorial texts are concerned with "How to use it".

Descriptive and explanatory texts have something of the nature of resources, being concerned with "it"; tutorial texts are almost wholly instrumental, being concerned with the reader’s use of "it". (What "it" is, of course, is very material to any instance of the Technical Communication process, but is not material to our understanding of the process.) What I will be saying from here on, therefore, is oriented towards the tutorial subgenre, which presents the most complete combination of problems and issues. Examples of tutorial, or "how-to", texts are abundant; they range from kitchen recipes to computer software manuals.

As our second normative Influence, there is widespread agreement that a technical text must have four attributes. The information it contains must be easy to find, easy to understand, accurate, and sufficient for the intended purpose. The first is determined by document structure & organisation; the third and fourth are determined by (or rather are attributes of) content. The second is determined by style.

Halliday and Hasan pulled together some discourse classifications arrived at by several previous researchers, and found some significant parallels. Their findings suggest that there are three possible orientations of discourse: toward content, toward the writer/speaker, and toward the reader/hearer. Apart from the category of imaginative uses (which one would hope to have little bearing on technical discourse!), these orientations have apparently been accepted — at least by systemicists — as universally relevant.

(Adapted from M.A.K. Halliday and Ruqaiya Hasan: Language, Context and Text, Geelong 1985)

Note that these classifications do not make any distinction between matter and manner — they assume no difference between what is discussed and how it is discussed. This has not been seen as a problem because, in general, the orientation towards matter and that towards manner do correspond. We expect to find content-centred matter delivered in a content-centred manner, and so on.

But the reader of a "how-to" text is usually in search of a solution to an immediate and essentially personal problem: "how do I get this thing to do X?" As I said previously, the nature of a "how-to" text is "tutorial", leading the reader by the hand from a problem (an information need) to the solution (the required information). The text must be written from the point of view of the reader’s need, and therefore from the point of view of the reader; yet it must focus on the functions and facilities provided by the product That is, the writer must deliver content-centred matter in a reader-centred manner. My contention is that this is achieved by the selection of an appropriate style.

This easy assertion raises the question: ‘"What is Style?" Rather than reciting any of the conventional definitions, I want to explore the nature of style, in terms of its contribution to the communicative process. This means, of course, that we are talking about "stylistics", rather than about conventions of orthography or punctuation.

Joseph M. Williams, in his admirable book on Style (Chicago, 1990), provides a valuable stating point for our exploration. Williams treats style as consisting of two aspects, or dimensions: clarity and grace. Sadly, he nowhere defines these, so I offer these working definitions:

Grace is the appropriate use of consistent lexical, syntactical, and structural forms;

Clarity is the presentation of facts and ideas in appropriate relationships.

Grace is what makes a text readable; clarity is what makes it comprehensible. But there is a third dimension, which I shall call tone:

Tone is the adoption of an appropriate relationship between the writer and the reader.

Tone is what makes a technical text immediately relevant to the reader’s needs. It combines mode of address, which is the way the communication is directed to the reader as a person, and point of view, which is the way the text is directed to the reader’s "problem" as a topic. Getting the tone right — or rather, that aspect of tone which I have called "mode of address" — is probably the hardest part of the technical communication process. To get it right, the writer must have a clear idea of the real live human being at the other end of the communication channel. Without careful and thorough audience analysis, we might produce text that has a condescending flavour or places too many demands on the reader’s knowledge. With superficial analysis, we finish up talking to stereotypes. But audience analysis can be very difficult, especially for products destined for overseas markets. I will return to this point later.

We might characterise grace as the text’s inherent character, tone as its orientation towards the reader, and clarity as its presentation of content. Clearly, all three must be "right" to achieve maximum communicative value; though undoubtedly clarity is the most important. If I could legislate to ensure that every technical communicator had a "credo" pinned up above his or her desk, I would institutionalise this misquotation of St Paul:

And now abideth grace, tone, clarity, these three;
but the greatest of these is clarity.

The cumulative effect of these three dimensions of style is the extent to which the text is easy to understand and use: the extent to which it is readable, applicable, and comprehensible.

The need for these three factors, then, is our third normative influence. How are these factors realised in a text?

Grace, I have suggested, is the appropriate use of consistent lexical, syntactical, and structural forms. Lexical forms include the use of appropriate terminology — jargon for specialists; everyday words for the rest of us. Syntactical forms include active voice and imperative mood. Structural forms include syntactical structures such as the use (or rather the avoidance) of dependent clauses and nested phrases, as well as larger structural features like lists and tables. The aim of technical communication is not to impress, but to express: syntactical and structural complexities which need time or effort to unravel are an obstacle to the flow of information. This is not to suggest that we should descend to the "See Spot run" level, but that we should use the simplest and clearest forms appropriate to the meaning being expressed.

Tone is the expression of the relationship between the writer and the reader. A comfortable, unobtrusive working relationship will result in a tone that is conversational without being chatty, and that maintains a focus on the reader’s task or activity. It will be based on extensive use of the second person, and will link to the dimension of grace through the use of active voice and imperative mood. That use of the second person will only work of course, if we address that second person in a way that makes the reader feel comfortable. As I said earlier, this is very dependent on good research.

Clarity is the presentation of facts or ideas in appropriate relationships. While logical relationships should never be obscured, the emphasis is on perceptual relationships. Typical relationships in the ordering of facts are cause-to-effect, known-to-unknown, simple-to-complex. Chronological relationships are important in explaining processes or procedures; spatial relationships are important in describing physical appearances. Clarity of presentation will link to the dimension of grace by its emphasis on the reader’s perception, so that the reader finds textual structures which reflect the structure of the information. It will link to the dimension of tone by its focus on the reader’s activity, so that the reader finds information presented in the order of use. Clarity, too, is dependent on good research: this time, on analysis of the reader’s probable tasks or activities.

Links between the dimensions of style

A well-known quotation serves to epitomise the three virtues of effective technical communication. Indeed, rather than have it pinned up above all technical communicators’ desks, I would have this one engraved on their hearts. Isabella Beeton is alleged to have written:

First, catch your rabbit.

Grace is achieved through the elegance of austerity: there is not a wasted word, not a redundant phrase. Tone is achieved through the direct manner: nothing could be more briskly business-like or more conversational. Clarity is achieved through the emphasis on the task: the reader is left in no doubt about the order of activities, nor about who should perform the activities. It is instructive to compare what she wrote with what she might have written: "First, catch a rabbit." Grace and clarity would not be compromised, but the tone would be cold and impersonal.

Cookery books in general are a useful model for the technical communicator. Most recipes are presented in a form that clearly separates the shopping task from the cooking task, by putting a list of ingredients at the start. As many cooks have found to their cost, though, some writers of recipes do manage to bury "day-before" preparations in the details of final cooking.

However, there are two very highly regarded cookery books, written by Gretta Anna Teplitzky (a Sydney cookery teacher), which use a quite different approach. It is interesting to compare two similar recipes, one from a "Gretta Anna" book and one from a more conventional source. In my view, the Gretta Anna book is a typographical disaster; worse, it lacks page numbers and even an index. Yet, as I said, her books are very highly regarded. Clearly there are no normative influences here!

A recipe from "The Gretta Anna Recipes" (Sydney, 1978)

A recipe from "Cork, Fork and Ladle" (Melbourne. 1975)

Still on the domestic front, the instruction manuals for many appliances are good examples of "how not to do it". Video recorders (VCRs) are notoriously difficult to use, and demonstrate that a poorly-designed product cannot be rescued by its documentation. More to the point, perhaps, they demonstrate the difficulty of producing effective documentation if the reader’s tasks are made difficult by the product’s design. The common failing of most such documentation is a focus on "what button X does" rather than on "what buttons you must use to achieve Y".

Even the best instruction manuals sometimes get it wrong. Here is a quotation from the (generally excellent) instructions for a Philips steam-iron:

You can fill this iron with normal tap water.

If the water is very hard (hardness higher than 17ºDH), it is advised to use distilled or demineralised water only.

This guarantees the maximum lifetime of your steam-iron.

While the lack of grace in the second element can perhaps be attributed to an infelicitous translation, the lack of clarity can only be put down to a failure in audience research: I am not aware of any community in Australia or Britain where the inhabitants know how water hardness is expressed — never mind the numerical rating of their own supply.

The problems I have outlined here are not confined to instruction manuals. Here is the text of a label I found on a folding bed:

WARNING

CAUTION SHOULD BE TAKEN WHEN
OPENING OR CLOSING BED AS TENSION
EXERTED ON SPRINGS MAY CAUSE THE
BED TO CLOSE SUDDENLY.

FOR MODELS WITH WHEELS, THESE
HAVE TO BE INSERTED FIRST FOR UNIT
TO STAY OPEN.

In just a heading and two sentences, this writer has achieved a high degree of obscurity, a remote and pompous tone, and a total absence of grace.

In these domestic examples, we have seen that the three dimensions of style — grace, tone, and clarity — are important determinants of effectiveness in technical communication. Even the best style cannot compensate for poor organisation, nor for incomplete or inaccurate content. Poor style, though, can act as an impenetrable barrier to the transfer of information.

Obviously, the anecdotal material I have presented here has only scratched the surface. It has highlighted three very broad normative influences: the three categories of technical text, the four attributes of effective technical texts, and the three dimensions of effective technical style. What must we do to learn more?

To begin with, of course, we need a corpus of Australian-sourced technical texts. We need to determine how the normative influences are realised — not in general terms as I have done here, but in precise and algorithmic terms. Then we need analytical tools which can identify critical features in a text. We need, too, a deeper understanding of how audience characteristics determine readership responses; and we need to develop some better ways to assess those audience characteristics -better tools for audience analysis. Finally, of course, we need resources — time and money, people and computers — to put all this together.

I referred earlier to the erosion of our manufacturing industry. Despite exhortations to be patriotic, Australians evidently prefer to buy imported goods. Perhaps, by making local products better documented and therefore easier to use, Australian manufacturers might be able to recapture some of the local market. Perhaps informed research into technical communication might be a very practical way of promoting "the clever country". My own company has helped a major Japanese corporation to improve Australian customer satisfaction by improving the quality and useability of its English-language documentation. (I am pleased to say that this is in an industry where there is no local competitor.) It should not be impossible for academia and industry, with a little governmental support, to achieve as much for Australian manufacturers.

If you have questions or comments about this web site, please send them to us.
Copyright © Brandle Pty Limited, 1997-2006
Last modified: March 24, 2002