What's Right About Scientific Writing

Illustration: A. Canamucio In recent commentaries in Nature,1 the New York Times Magazine,2 and the American Scholar,3 the scientific article has come under attack because it has allegedly degenerated into a thick "molasses of jargon and academic code."2 Furthermore, according to scientist Peter Medawar in a 1964 Saturday Review essay,4 because the structure of the scientific article seriously misrepresents the way science happens, authors of scientific articles are also mendacious, perpetrators of a fraud. In other words, we have it on good authority that scientific prose is execrable, scientific communication dishonest. What's wrong with this picture? It cannot account for the fact that science is a successful enterprise, an enterprise whose success depends crucially on the efficacy and honesty of its communicative practices. Contemporary scientists convey their discoveries in the writing conventions of their time, just as Newton, Franklin, Darwin, Mendeleev, Curie, and Einstein did in their time. While Nature, the American Scholar, and the New York Times Magazine may be right in the specific examples they criticize, they are wrong in making generalizations like "Pleas for scientists to write readably have failed for at least 300 years."1 And because the scientific article is about persuading readers of the importance and trustworthiness of new knowledge claims, not about accurately re-creating the practices that spawned them, Medawar is also wrong. We contend that the scientific article has been since its inception in the 17th century, and remains today, an effective medium for honestly communicating the cognitively complex knowledge scientists create--despite the varying linguistic abilities and ethical predilections of individual scientists. That, indeed, is its achievement, an achievement to be celebrated at the University of Chicago as part of our forthcoming exhibit at the Regenstein Library (www.lib.uchicago.edu/e/spcl/curex.html) and also in a book we are preparing for Oxford University Press. Emphasized in the exhibit is the rich variety of scientific communications over the past four centuries. Emphasized in the book is the claim that the scientific article has evolved as a master narrative that makes visuals central to its arguments, is organized as a master finding system, and is written in a style appropriate to its subject matter and its audience. The scientific journal article debuted in Paris and London, 1665. The first articles are typically brief reports describing an individual's encounter with nature through observation or experiment. They tend to be mainly descriptive and offer little in the way of theoretical explanation, mathematical or mechanical, for the gathered facts. Over the past three centuries, however, a master narrative has evolved. This narrative represents a tribute to the efficacy of the experimental method as a means of exploring nature. It opens with a title and abstract intended to minimize the time and effort needed to uncover the article's gist. That is followed by an introduction placing the reader in the scientific context in which the authors are working. Next comes the main body: a section on methods and materials that outlines the procedures and materials used to acquire new facts, a results section that displays the facts so generated and the intellectual context of their acquisition, and a discussion section that offers an explanation for the new facts. A conclusion section reiterates the central facts and explanations and, perhaps, also addresses future work that would confirm or extend the original investigation. This narrative is not a straitjacket, but a flexible prototype out of which authors fashion numerous variations. An often-overlooked complement to this verbal narrative is the visual evidence presented in establishing facts and their explanations. Scientific visuals perform multiple tasks. They depict relationships visible in nature, such as those between insects and plants; relationships in nature not visible to the naked eye, such as crystalline structures and underground geological sections; relationships posited by a theory, such as the Feynman diagram; or tables and graphs organizing masses of data in support of lawlike relationships. Because of their obvious communicative utility, visuals have greatly increased in number and proportion taken up within the text. As a result of this increased visual component, the scientific article has now become almost as much about interpreting information in figures and tables as reading straight text. In conjunction with the master narrative, the modern scientific article has also evolved a master finding system. This system compartmentalizes the essential features in articles through the use of headings and subheadings, tables and figures integrated into the text with numbered captions describing their contents, and citations that supply additional context for statements at any point in the text. It also permits scientists to read articles selectively rather than sequentially, opportunistically scavenging the various components in search of useful bits of method, theory, and fact. With headings in place, for example, scientists not interested in methodological details do not need to read the Methods section. Alternatively, with the swarm of data segregated into tables and figures, scientists are able more easily to focus on them or on their commentary in the text or to alternate freely between the two. In early scientific articles, one can find citations, headings, and numbered graphics and tables, but only on occasion and without uniform format. Overall, this elaborate system of verbal and visual presentation evolved to make the comprehension of the underlying arguments as trouble free as possible, given the inherent complexity of the main message. Let's now turn from presentational features to style. In general, the early scientific articles seek to establish credibility more by means of reliable testimony than technical details, more by qualitative experience than by quantitative experiment and observation in support of theory. Hence, a more personal voice is appropriate, and notions of relevance are given a wide berth. In a Philosophical Transactions article, Isaac Newton mentions the Great Plague interrupted his thoughts on an optics problem for two years. In the midst of maritime observations reported in the same journal, a Mr. Stubbs informs us his ship sailed off course "to avoid the Spanish fleet." Today, such detail would be regarded as extraneous. These early articles also largely rely on everyday language and are written in a style conforming to the compositional norms of expository prose generally, not norms specific to science. For the most part, the targeted audience is a loosely knit community united by a curiosity about how nature works. The modern scientific article, in contrast, has evolved a specialized language and prose style adapted for efficient communication to other professionals engaged in similar research. The impersonal style of these articles is designed to focus the reader's mind on the things of the laboratory and the natural world, rather than to draw attention to the text itself or its author. As some critics have argued, the impersonal, compressed style of modern scientific writing has alienated the general reader and even the amateur enthusiast that was, at one time, part of the assumed audience. At the same time, we would argue that it has ably served the specialized reader that is the primary audience for a literature with article titles such as "The Free Energy Relations among Paraffin and Olefin Hydrocarbons." In writing for journal publication, scientists seek to achieve not literary uniqueness, but literate typicality. What typifies this style is impersonality and meticulous attention to technical details, qualities clearly inappropriate in communications meant for general readers. True, individual scientists like Roald Hoffmann, Richard Feynman, Stephen Jay Gould, and Lewis Thomas write with flair in their popular science works. When they write up their research for journal publication, however, they seldom deviate dramatically from the conventional style. In a popular work, for example, Thomas radiates linguistic geniality:"We noted that the rabbits, for all their display of good health, looked different and funny." But in a scientific article reporting the same research, he writes unwaveringly in the lingua franca of science: "The earliest visible change in the gross behavior began to occur 3 hours after papain, and consisted of drooping and increased flexibility of the ear tips." Each style serves a distinct purpose and audience. One provides new information to the intellectually curious. The other seeks to argue persuasively in favor of a new claim to those with similar training and qualifications. To be sure, the scientific literature has its fair share of stilted and badly written documents. Because of the abstract and technical language so prevalent in modern science, scientists can easily fall into the habit of making the simple sound unnecessarily complex. Nicholas Wade, science editor for the New York Times, quotes an example from a recent Science issue: "Vision plays an important role in notifying animals of imminent danger, such as an impending collision with a predator or an environmental surface." Wade rephrases that sentence more concisely and vividly as "Without eyes, you'd soon crash into a tiger or a cliff."2 But the same trick can be played on discourse in the humanities. Here is a recent example from the Quarterly Journal of Speech, the flagship journal of the National Communication Association: "Thus, there could be partisan preference for the utilization of persuasive attack, with Democrats favoring acclaiming and Republicans showing an affinity to attacks." The author means that Democrats prefer to attack by praising their past good deeds, Republicans by criticizing their opponents. We are not asserting that the scientific article is perfect and cannot be improved, or that there are not individual specimens with syntax so mangled that it makes one wonder how they made it through the review process. We are asserting that, whatever its failings, the present scientific article is an effective means for securing ownership of knowledge claims and an efficient vehicle for communicating the knowledge science creates. Those who write scientific articles for a living should take pride in participating in this long and distinguished tradition. Alan Gross (grossalan@aol.com) teaches and writes about the rhetoric of science at the University of Minnesota. Joseph Harmon works as a senior technical editor for the Chemical Technology Division, Argonne National Laboratory. References 1. M.W. Gregory, "The infectiousness of pompous prose," Nature, 360: 11-2, 1992. 2. N. Wade, "Racing to the top," New York Times Magazine, Dec. 10, 1995, page 28. 3. D. Locke, "Voices of science," American Scholar, 67[3]:103-14, 1998. 4. P. Medawar, "Is the scientific report fraudulent? Yes, it misrepresents scientific thought," Saturday Review, 47:42-3, 1964.

What's Right About Scientific Writing

Interested in reading more?

Become a Member of