The Craft of Scientific Communication

Joseph E. Harmon, Alan G. Gross

• Publisher: University of Chicago Press
• Year: 2010
• ISBN:
  9780226316635
• Topic: Biological Sciences
• Subtopic: Science General
• Index: Biological Sciences

The ability to communicate in print and person is essential to the life of a successful scientist. But since writing is often secondary in scientific education and teaching, there remains a significant need for guides that teach scientists how best to convey their research to general and professional audiences. The Craft of Scientific Communication will teach science students and scientists alike how to improve the clarity, cogency, and communicative power of their words and images.

In this remarkable guide, Joseph E. Harmon and Alan G. Gross have combined their many years of experience in the art of science writing to analyze published examples of how the best scientists communicate. Organized topically with information on the structural elements and the style of scientific communications, each chapter draws on models of past successes and failures to show students and practitioners how best to negotiate the world of print, online publication, and oral presentation.

PART I

The Scientific Article

1
Introducing Your Problem

In a seventeenth-century article, the gentleman scientist Robert Boyle opened with the following brief introduction to his experiments on the respiration of animals:

Nature having, as Zoologists teach us, furnished Ducks and other water-Fowl with a peculiar structure of some vessels about the heart, to enable them, when they have occasion to Dive, to forbear for a pretty while respiring under water without prejudice: I thought it worth the tryal, whether such Birds would much better than other Animals endure the absence of the Air in our exhausted Receiver. The accounts of which tryals were, when they were made, registered as follows.

Since Boyle’s time, introductions have evolved. Modern readers expect more information than Boyle gives—at least something about recent research on the subject bolstered by apt citations, and maybe a hint at the conclusion from the “tryals.” While the science in the best modern scientific articles is never conventional, over time science has adopted a conventional form for the introduction that relies less on the stylistic artistry of individual authors and more on their ability to manipulate three simple components. The subject of this chapter is what those components are and how to bend them to your own expository purposes.

The Structure of the Typical Scientific Introduction

In his classic Rhetoric, Aristotle makes the obvious point that all introductions, whether a prologue to a poem, a prelude to a musical composition, or a preface to a speech, pave the way for what is to follow. According to linguist John Swales, modern scientific introductions conventionally accomplish this purpose in three stages:

1. Define a research territory. This stage normally summarizes the state of knowledge in the scientific research front being studied.

2. Establish a limited problem in that territory, one at the leading edge of a research front. In this stage, authors point out a contradiction or inconsistency or gap in that state or propose to build upon a neglected, undeveloped, or misunderstood aspect of it.

3. Suggest or summarize your solution to this problem. This stage typically focuses on the solution to the problem or an approach for solving it. It might also deal with why we should care. In long articles, scientist-authors sometimes end the introduction with a roadmap to the rest of the article. The first two stages of the introduction provide a context for the third. In so doing, they tell readers in what way the conclusions represent a significant contribution to new knowledge: readers are being told why they should read on.

Imagine you are a newly minted child psychologist. You want to conduct a research study of children in a social setting. “How children behave in the presence of others” is too vague a question to begin a research project. Would you be addressing any kind of behavior whatsoever? Who else would be present? What would the participants be doing in the setting? After further deliberation, you decide you want to study children imitating aggressive adults, a topic of no small concern to society at large. Your research territory has narrowed considerably. But that is not enough. You must first know what others have published on this topic. You do not want to reinvent the wheel; if you did, the resulting manuscript would likely be rejected for publication. Even if not, should a priority conflict arise after publication, you would lose. So you google “aggressive behavior,” “children,” and “adult models” and also search relevant publication databases.

From your search, you discover that others have reported that when children watch an adult behave aggressively with a doll, they imitate the same behavior in the same setting with the model adult present. You decide you want to extend that published research, to take it one step further: do such children still behave aggressively in a different setting with no model adult present? And as side issues you will ask: Are boys more prone to such behavior than girls? And are male adult models more likely to induce aggressive behavior than females? You now have in hand some original research problems in behavioral psychology. At a scientific conference during the winter in Hawaii, you check with some senior colleagues just to make sure your new problems really are new and worth pursuing. Encouragement received, you begin thinking about how to solve those problems.

Those are the research problems addressed by Albert Bandura, Dorothea Ross, and Sheila A. Ross in a classic psychology study published in 1961. We quote their first three paragraphs, inserting italicized headings to remind you of the prototypical three components:

Research territory

A previous study, designed to account for the phenomenon of identification in terms of incidental learning, demonstrated that children readily imitated behavior exhibited by an adult model in the presence of the model (Bandura & Huston, 1961). A series of experiments by Blake (1958) and others (Grosser, Polansky, & Lippitt, 1951; Rosenblith, 1959; Schachter & Hall, 1952) have likewise shown that mere observation responses of a model have a facilitating effect on subjects’ reactions in the immediate social setting.

Problem

While these studies provide convincing evidence for the influence and control exerted on others by the behavior of a model, a more crucial test of imitative learning involves the generalization of imitative response patterns under new settings in which the model is absent.

Approach to solving problem

In the experiment reported in this paper children were exposed to aggressive and nonaggressive adult models and were then tested for the amount of imitative learning in a new situation in the absence of the model. According to prediction, subjects exposed to aggressive models would reproduce aggressive acts resembling those of their models and would differ in this respect both from subjects who served as nonaggressive models and from those who had no prior exposure to any models. This hypothesis assumed that subjects had learned imitative habits as a result of prior reinforcement, and these tendencies would generalize to some extent to adult experimenters (Miller & Dollard, 1941).

This selection does not complete the authors’ introduction. It continues for several paragraphs, presenting subsidiary research problems along with hypothesized outcomes. By the end, readers are well prepared for the next section—the details of their method for solving their stated problems.

The next time you read a research article by a leading figure in your field of interest, pay particular attention to the introduction. In some form you will likely find the three stages given by Swales: research territory, problem with that territory, some hint as to the solution to the stated problem. They make sense in all learned communications centered on solving a research problem and defining it in terms of current knowledge. You do not need to be fully cognizant of these three stages to write a good introduction, any more than you need to know the mechanics behind walking to walk. But writing is not as natural as walking. And just as intimate knowledge of the mechanics behind many challenging activities from skiing to poker to computer programming can lead to substantial improvements in performance, so too with writing introductions. This knowledge is also helpful in more efficiently reading and critiquing introductions by others. Finally, it is helpful in modified form for writing proposals to win research funds—a topic we cover in chapter 10.

First Stage: Research Territory

Let’s look more closely at the first introductory stage, summarizing the state of knowledge in a research territory. Robert Boyle, writing about respiration in animals in the late seventeenth century, formulates a problem that had probably occurred to most zoologists of his day, indeed to any student of nature. But we are left in the dark as to what others had done, if anything, on his particular problem. In contrast, Bandura and colleagues clearly place the reader in the context of their research front. They do so by summarizing appropriate articles whose bibliographic details will appear in the references section at the end.

In the two following short introductory paragraphs, we also see this principle in action. The first is from an article by physicists Lise Meitner and Otto Frisch (1939). The authors begin by summarizing recent experimental research on what happens when you bombard the heavy element uranium with neutrons:

On bombarding uranium with neutrons, Fermi and his collaborators¹ found that at least four radioactive substances were produced, to two of which atomic numbers larger than 92 were ascribed. Further investigations² demonstrated the existence of at least nine radioactive periods, six of which were assigned to elements beyond uranium, and nuclear isomerism had to be assumed in order to account for their chemical behavior together with their genetic relations. . . . Following up an observation of Curie and Savitch,³ Hahn and Strassman⁴ found that a group of at least three radioactive bodies, formed from uranium under neutron bombardment, were chemically similar to barium and, therefore, presumably isotopic with radium. Further investigation,⁵ however, showed that it was impossible to separate these bodies from barium . . .

The superscript numbers of references in the first sentences of this introduction are a tribute to the cumulative achievement that is the essence of an advancing science, in this case, nuclear physics in the 1930s. These sentences lay the intellectual foundation for the authors’ theoretical explanation to follow, their theory of nuclear fission, the splitting of uranium atoms into much smaller elements.

How far back in history should these introductory references go? References seldom exceed ten years of age. For example, here is a heavily referenced introductory paragraph by Nobel laureate David Baltimore (1970) on the subject of viruses and infection:

DNA seems to have a critical role in the multiplication and transforming ability of RNA tumor viruses.¹ Infection and transformation by these viruses can be prevented by inhibitors of DNA synthesis added during the first 8–12 h after exposure of cells to the virus.^1–4 The necessary DNA synthesis seems to involve the production of DNA which is genetically specific for the infecting virus,^5,6 although hybridization studies intended to demonstrate virus-specific DNA have been inconclusive.¹ Also, the formation of virions by the RNA tumor viruses is sensitive to actinomycin D and therefore seems to involve DNA-dependent RNA synthesis.^1–4,7 One model which explains these data postulates the transformation of the infecting RNA to a DNA copy which then serves as a template for the synthesis of viral RNA.^1,2,7 This model requires a unique enzyme, an RNA-dependent DNA polymerase.

Baltimore’s six sentences contain sixteen citations to seven sources, none older than seven years. In principle, Baltimore’s introduction could easily have extended much further in time: to the discovery of the principle of inheritance, to the first detection of genes, DNA, and RNA, or to the unraveling of the structure of DNA and RNA and their roles in forming proteins. But experts in the biomedical field would already know that history well. Baltimore touches upon only pertinent earlier published theories and experiments necessary for his readers to appreciate the significance of the problem or question addressed in the subsequent paragraph. Indeed, a common failing of scientific introductions is that they give far more information related to the first component than the reader needs to appreciate the forthcoming problem.

Second Stage: Research Problem

The second introductory stage presents the problem the author’s research will solve. Scott Montgomery (1996) puts his finger on the nature of these well-formed problems within subdisciplines when he speaks of the job of researchers as “an unending attempt to create the conditions for new work, to find gaps or instabilities in existing [intellectual] structures.” The first stage only paves the way for the second. Without the problem, there would be no research to report.

Research problems arise from a variety of sources. In the introduction mentioned earlier, for example, Baltimore (1970) argues that something is missing in current molecular biology—its processes require an as-yet-undetected enzyme:

No enzyme which synthesizes DNA from an RNA template has been found in any type of cell. Unless such an enzyme exists in uninfected cells, the RNA tumor viruses must either induce its synthesis soon after infection or carry the enzyme into the cell as part of the virion [our emphasis].

In another example, Meitner and Frisch (1939) formulate an as-yet-unanswered question: Why does bombarding uranium with neutrons produce a much smaller element, in contradiction to current theory?

. . . Hahn and Strassmann were forced to conclude that isotopes of barium (Z [atomic number] = 56) are formed as a consequence of bombardment of uranium (Z = 92) with neutrons.

At first sight, this result seems very hard to understand. The f...