A print version of this article appears in IEEE Transactions on Professional Communication, September 1995. Info about ordering reprints: Specify Article Log 13331D.
Though Pirsig uses "classic" and "romantic" idiosyncratically, his comment about technical prose resonates even for us who read it and write it constantly. Technical prose, we might say, just lacks character.
Pirsig ascribes the classic-romantic split to a falling away from Quality. This prelapsarian state supposedly existed before Athenian philosophers wielded their analytic knives, creating such divisions as truth/probability, mind/body, philosophy/rhetoric--divisions so deeply ingrained in Western culture that they appear "natural." One consequence of the resulting dualism, as Daniel R. Jones points out in an insightful interpretation of Pirsig's book , is a "spectator attitude" toward technology. Spectators are alienated from technology, whether they consume its products or even work with it.
Pirsig notes, early in the book, that the technical manuals he worked on as writer and editor were spectator manuals:
This stance, according to my colleague Lynn Deming, comes from the difficulties writers in technical fields seem to have
The problem that Pirsig ascribes to the absence of Quality and Deming to problems in narrative stance looks somewhat different from another perspective, that of Aristotelian rhetoric. The Aristotelian perspective has been used forcefully by Arthur E. Walzer and Alan Gross  in their analysis of accounts of the Challenger disaster. They found these accounts conforming to one of two perspectives: positivist or postmodernist. Positivist accounts emphasize a failure in communication--the disaster occurred because engineers and managers failed to transmit or receive all the facts. Postmodernist accounts stress differences in interpretive frameworks that cause engineers and managers to draw different conclusions from the same data. "The positivistic perspective . . . attempts, in effect, to minimize deliberation and circumvent ethics. . . .Since the postmodernists have been forceful critics of the positivists, it seems even more surprising that their analyses of the Challenger case all but dismiss ethics as irrelevant and virtually deny the possibility of rhetorical deliberation reaching consensus"[4, p. 431].
From a rhetorical perspective, there are two sorts of deliberation that could have taken place: those contributing to technical knowledge for making engineering decisions, and those contributing to normative understanding for making ethical decisions. When the technical arguments cannot produce a clear consensus but a decision must be made anyway, deliberations need to become normative--to ask, as Walzer and Gross do for Challenger, "What rules ought to guide our decision in the absence of knowledge of how the O-rings will perform?" [4, p. 427]
In such circumstances, neither the spectator status conferred by positivist focus on "objective facts" nor the insulation afforded by immersion in a particular profession's interpretive framework serves the public interest. With Walzer and Gross, I regard Aristotelian rhetoric as offering a method for conducting public discourse, one that could improve the quality of public decision-making. Yet a part of Aristotelian rhetoric, the artistic proofs, appears to support the radical individualism that is part of the positivistic culture of professionalism. I think in fact that such support is only the result of a fallacy of translation. Aristotle's ethos is usually translated as "character," a quality we conventionally ascribe to individuals. But ethos is a richer concept, involving the individual in a deliberative community and thus having, as the Greek root implies, an ethical dimension as well as a transactional one.
Information as message packets is another manifestation of spectatorhood, which stands in the way of effective communication so long as documents focus solely on the technical. Fortunately, the emphasis of technical writing has been shifting late in this century from "technical" to "writing." This history can be explained through the paradigm of the "communication triangle" implicit in Aristotle's artistic proofs (Figure 1), which regard not only reasoning (logos) as a persuasive element, but also the condition of the audience (pathos) and the character of the speaker (ethos).
Before there was a professional field called technical communication, technical writing was done by technical people. Like Pirsig's classicists, they cared mostly about such matters as design and function. In terms of Aristotle's triangle, these folks thought only of logos, content and reasoning.
Did their writing therefore escape the claims of pathos and ethos? No: often their prose created such obstacles to understanding, through its jargon and density, that it created the ethos of the expert: one whose esoteric knowledge makes him-- and it still was mostly him--a member of a priesthood, a wizard whose mysteries aren't supposed to be understood by the uninitiated. The pathos dimension could be summed up in the sentence, "If you're smart enough to understand this, fine; if not, too bad." Call this dominance of logos the objectivist model (Figure 2).
The perils of the objectivist model are summed up well in a 1946 book written for technical specialists, ironically with a point of view that objectivizes its readers:
As the field gained momentum, two complementary things happened, both on the pathos corner of the triangle. First, books on technical writing, such as Houp and Pearsall's  and Mathes and Stevenson's , began stressing audience needs and ways to analyze them; such books, used in college technical writing courses, began to chip away at the "objectivist" model. Second, with the growing use of computer software, the field of usability testing developed. Bringing the audience into the design phase brought much new and productive thinking about how documents should be organized, text chunked for easy access, and graphics used for information and access.
The tyranny of audience may be seen most clearly in the current fashion of considering it as "users." Users are implicitly conceived of as being somewhat infantile: they want specific bits of pragmatic knowledge (e.g., "How can I get my system to stop crashing when I have three programs open at the same time?") and they want it now. How this tyranny might be undermined has been shown humorously by Marilyn Cooper , who put together a short manual conflating cartoons from Michael Paul McLester's Beset by Demons--a bloated spider singing, "Every blade of grass you see, every flower and every tree . . . Everything belongs to ME!--and parts of a chain saw manual.
Cooper's "manual" makes it hard for the user to be a spectator: "Yes, indeed, your new chain saw is not only a precious commodity but is also instrumental in the effort to transform the natural world for the market. In purchasing and using this chain saw, you have chosen to participate in this effort." Cooper also forces the user to be aware of being part of a community with experiences in common: "Perhaps because of its awkward position, the spider has cut off part of its left foot. Never allow any part of your body to touch the rotating chain. You know chain saws are dangerous. You've seen chain saw massacre movies."
The dimension Cooper brings in is the ethos still missing in Figure 3. It's an ethos situated within the cultural context of the late 20th century in America, not floating disembodied outside time and place.
In late Roman times, and again after the rhetorical reforms of Peter Ramus in the sixteenth century, when rhetoric became reduced mostly to questions of style, style became the means of creating persona--of giving the prose personality. Merrill Whitburn wrote an article almost twenty years ago  that discussed personality in technical writing. Like T.A. Rickard in his 1908 Guide to Technical Writing , Whitburn was mainly interested in improving technical documents by better prose style.
Style and authorial voice are strongly linked, but strong persona is not necessarily a plus in technical writing. For example, humor might not be appreciated by the harried computer user as she consults "help" to figure out why a program function isn't working, and it seems almost unimaginable in an operating manual for a nuclear power plant.
Ethos is not, however, at all the same thing as persona, described thus in the last two editions of the venerable Reporting Technical Information by Houp and Pearsall:
Aristotle too treats ethos and pathos as fungible, and character as something that can be crafted for particular audiences and situations:
What I am describing as fungibility strongly resembles Kenneth Burke's principles of identification and consubstantiality . As one works as an engineer or a banker, one identifies his or her interests to a large extent with those of the group, and through habituation acquires the group ethos. Consubstantiality comes into play when the engineer, while remaining an engineer, becomes "substantially one" with a banker in creating a common sphere of interest through a business proposal. Or does in successful proposals.
To the extent, then, that one is not born a banker or an engineer, the ethos one has is partly acquired and partly invented. Under the circumstances shown in Figures 2 and 3, ethos in technical prose was largely unconscious. Yet, as James S. Baumlin notes, "More than an expression of individual psychology or an intersection of social forces, ethos is, as Aristotle himself suggests, quintessentially a linguistic phenomenon" . Such a view of ethos seems to justify Houp, Pearsall, and Tebeaux's treating, in the passage quoted earlier, the relation of persona and audience as something that can be done consciously--that is, one can invent a persona appropriate to a particular document's intended audience.
Now, there's a virtual industry that depends on persona. It provides software instruction to the uninitiated by writing in a breezy, you-oriented style with humorous authorial asides. An example chosen at random : "That 8 MB of extra memory that Quattro Pro would like could cost you as much as $400. (That's almost $3700 in dog dollars.)"
So, though persona is part of ethos, it is not the whole. Killingsworth and Gilbertson summarize well the relatively few articles on the primary character in technical writing, which, they say, follow three general theoretical trends:
But in technical writing, the narrative voice is also a construction, not just a transparent window on truth. (I suggested something of the sort in an earlier article on engineering style .) It is even more obviously a construction, in that it is likely to be either a corporate or a generic voice. Killingsworth and Gilbertson  assert that the poststructuralist notion of an author submerged in a network of intertextuality applies even better to technical writing than to literary works. That is, text is the medium by which ideas are mediated and compromises reached. The narrative voice of corporate documents, how they are developed and maintained and how they sustain a corporate ethos, is a subject worthy of study in itself.
Construction of ethos is the flip side of writers' constructing audiences. Ethos stands in relation to persona as Mary B. Coney's constructed audiences stand in relation to the empirical audiences of the technical writing textbooks. Coney has argued that writers usually require readers to read in a variety of roles, so that a reader-in-the-flesh has to adapt to the role theorized for her by the writer. Coney's article  anatomized the ways technical readers construct the meaning of technical texts. Coney provides an alternative to regarding readers as static empirical subjects, analyzable in terms of their roles, backgrounds and biases. Instead, she looks at the roles readers are called upon to play as they are reading--how the text call the reader to fill the roles it requires.
This way of looking at readers reading technical texts represents an advance: it treats reading as an active part of a rhetorical transaction that has the potential to change both the writer and the reader: "roles are always transforming themselves in the course of the reading process: what started out as a naive user on page 2 of a manual is transformed by the very act of reading into a more sophisticated chooser of options on page 72" [5, p. 61, emphasis Coney's].
Many people of a certain age could vouch for such transformation by a manual. John Muir's How to Keep Your Volkswagen Alive , now available in a 25th anniversary edition, illustrates very well both Aristotle's adaptation of speaker to audience and of Coney's providing a role for the reader. Muir hypothesized an audience who, like him, wanted to be able to fix their Volkswagens in out-of-the-way places and who, like him, were willing to get their hands dirty. And he found that audience. The authorial persona was prominent but not dominant; mainly it established a dynamic with the reader. As a result, people otherwise technically innocent became technically adept, unafraid of adjusting their valves or even replacing the whole top end under a shade tree.
In this Rogerian form of argument (derived from Young, Becker, and Pike ), logos comes first as a neutral statement of the facts of the case, then comes pathos as an acknowledgement of the reader's perspective. Finally comes ethos, which conveys not only the desire of the writer but the relation in which reader and writer may stand to one another after the reading. This method of argument is particularly useful when it's as important to establish a relationship as to convey information, as is the case in proposals, job-application letters, and customer-relations letters.
For example, suppose that a customer has ordered a specific hydraulic valve from a wholesaler. The wholesaler has sold all of them, and the manufacturer has substituted a design that is functionally the same but physically different. A faxed letter giving the customer the bad news might go like this:
Part of the ethics problem are the very codes that are supposed to describe ethical behavior in professional settings. Like the manuals Pirsig complained about, codes of ethics objectify behavior, alienating it from the people who are behaving through their writing. This problem is discussed well by Brown , who notes that the STC Code for Communicators' "vague and abstract language . . . makes adherence a problem" and that the Code's "metaphor of 'bridge' turns the writer into a conveyance, one implicitly neutral." Brown worries that writers in institutional settings "implicitly assume the privatized sphere of an organizational ethos rather than that derived from personal ethics or the public good."
If writers do implicitly assume an organizational ethos (which would be like trying to adhere to an objectified ethical code), it may be because they and their organizations still operate under a model in which organization is hierarchical and power flows down from the top. Yet even in such organizations, documents that convey the organizational ethos emerge from a process of negotiation--the organization can speak to its customers or the public with one voice only after it has resolved differences among many internal voices. That may account for the character of much organizational prose: it represents a lowest common denominator, and those who are supposed to turn out the prose--writers and editors--are often the people with the least influence in shaping the message.
Those with more clout, the scientists and engineers and managers, often insist on writing in the same way no matter what the audience. Scientists may think it unethical to present their work in anything less than the full complexity it has for them, or to tailor presentations to nonspecialist audiences. Engineers often want to drag the audience through all the details of a technical analysis. Managers may think it undignified to use ordinary language and instead prefer an emptily sonorous, abstract prose.
One might well ask, though, how ethical it is not to be aware of, nor to adapt to, the audience's prior understandings-- to risk either snowing them or talking over their heads. Here, we might take a hint from Coney and look at organizational prose in terms of the kind of audience it seems to imply. Then we could compare the description of that audience with the characteristics of the audience we're actually trying to reach.
Another dimension of this exercise is to look at organizational prose in terms of its implicit attitude toward its audience. Attitude is the aspect of ethos defined partly by purpose or motive and partly by choices of content and style.
Purpose is the guise of ethos in composition books in this century, borrowing Alexander Bain's 19th-century ideas of purpose as narration, description, exposition, and argument. Purpose is usually stated quite explicitly in technical prose. Motive, however, is another matter. As Kenneth Burke suggests, motive may be conveyed by tonality, described as
Attitude may show through content and style in several ways. Prose may pander to the audience by telling it only what it seems to want to hear. It may ignore audience by paying no attention to its level of understanding or its needs. It may convey an attitude about the audience through the choice of words and information, like online help that, by repeatedly belaboring the obvious, evidently considers its audience dimwitted.
In summary, it is time we explicitly theorized the ethos component of technical writing. We need to recognize that ethos is a construction, not a function of personality, and that writers can ethically construct a variety of ethoi just as readers, according to Mary Coney, customarily read in a variety of roles. Until we pay the kind of explicit attention to ethos that we've given pathos and logos, our technical rhetoric will continue be a spectator's rhetoric.