Writing an Effective Abstract: An Audience-Based Approach

Most journals require authors to submit abstracts along with their articles, as do both of the journals we edit, ARIEL and Narrative. This requirement has two main rationales: an abstract offers readers a helpful, succinct summary of the longer argument developed in the essay, and it identifies keywords that will make it easier for search engines to find the essay.

Notice that these rationales presuppose the publication of both abstract and essay and, in so doing, assume that the main audience for the abstract is prospective readers of the published essay. However, from the perspective of an author submitting work to a journal, there is another important audience to consider: the journal editor(s) and the external reviewers to whom the editor(s) send it.

This audience looks at your abstract with their most pressing question in mind: is this article publishable in this journal? A good abstract tilts them toward an affirmative answer by leaving them well-disposed toward the longer argument in the article. A bad abstract won’t by itself cause this audience to reject an article, but it does incline the audience toward an initial negative answer. In that way, an ineffective abstract becomes an obstacle that your article needs to overcome.

How do you produce a good abstract for this audience?  In a process of reverse engineering, we’ve identified a set of recurring questions that underlie the strong abstracts that we have published over the years. You do not need to answer these questions in the order in which we list them here, and you do not need to give them equal time and space, but a good abstract will address all of them.

• What is the central issue or question or problem driving your inquiry? You might not state the question or problem in an explicit sentence or two in the essay, but you should articulate it in your abstract.  
• What is your answer to this question or problem? Again, you might not state this answer in a single sentence in the essay, but you should state it explicitly in your abstract. Furthermore, you should closely tie the answer to the question. Your abstract is not a teaser but a spoiler.
• What steps does your article take to get to this answer? What is your method of analysis, and how does your argument proceed? In the course of explaining these matters, you should mention the key concepts, theories or texts you rely on to make your case.
• How does your article contribute to an existing scholarly conversation? In other words, what’s your answer to the “so what?” question? Effective abstracts often begin by addressing this question, characterizing the state of the scholarly conversation about the problem or question and highlighting how the article intervenes in that conversation. Your intervention may be to revise, extend or even overturn received wisdom. It may be to bring new evidence and insights to an ongoing debate.  It may be to call attention to some objects of study that previous scholarship has neglected and whose significance for the field you will elucidate.  And that’s just a partial list.  But whatever your intervention, your abstract should express it clearly and directly. We can’t overstate how important this element is: it is the one from which everything else -- in both abstract and essay -- flows.

Our reverse engineering of effective abstracts has also led us to identify some common types of ineffective ones:

• The abstract that announces the topic(s) the essay examines or considers or meditates on without revealing the conclusions that have been drawn from this activity or how those conclusions bear on a larger scholarly conversation. This kind of abstract mistakenly privileges the what (those topics) over the so what (those conclusions and why they matter).  
• The abstract that goes through the article chronologically, describing what it does first, second, third and so on. This kind of abstract focuses on the trees and ignores the forest. Good abstracts give their audience a clear vision of the forest.
• The abstract that simply repeats the article’s first paragraph. Such an abstract assumes that the purposes of first paragraphs and abstracts are essentially the same, but a little reflection reveals the inadequacy of that assumption. The purpose of the first paragraph is to launch the argument, while the purpose of the abstract is to provide a comprehensive overview of it and its stakes. Both the abstract and the first paragraph may include the thesis of the argument, but the first paragraph can’t offer the bird’s-eye view of the whole essay and why it matters that an effective abstract does.

A Tale of Two Abstracts

In order to illustrate these general points, we offer two abstracts of an essay that, one of us (Jim) has recently contributed to a collection of essays on Narration as Argument, a volume designed to address debates about the efficacy and validity of stories in argumentative discourse. (The collection is edited by Paula Olmos and forthcoming from Springer.)

The title of the essay is “Narrative as Argument in Atul Gawande’s ‘On Washing Hands’ and ‘Letting Go’”  As the title suggests, much of the space of the essay is devoted to the analysis of Gawande’s two essays, which become case studies in the larger debate to which the collection is devoted. The two abstracts handle those case studies in very different ways.

Abstract 1: This essay demonstrates how Atul Gawande uses stories in the service of his arguments in two of his essays, “On Washing Hands” from Better (2007) and “Letting Go” from Being Mortal (2014). In both essays, Gawande works with a problem-solution argumentative structure and uses narrative to complicate that structure.  In “On Washing Hands,” he does not construct a straightforward argument with a straightforward thesis. Instead, he uses several mini-narratives in combination with exposition and with thematizing commentary to alter his audience’s understanding of both the problem and the solution. Indeed, he uses the ending to the central narrative as a way to temper his audience’s enthusiasm for the solution. “Letting Go” is longer and more complexly organized than “On Washing Hands,” but Gawande’s use of a central story threaded throughout the essay and his representation of himself are crucial to his adaptation of the problem-solution structure. Furthermore, Gawande uses narrative to raise an important objection to his solution and responds to the objection not with a counternarrative but with a counterargument.

Abstract 2: This essay responds to scholarly skepticism about narrative as argument, due to its reliance on hindsight effects (because such and such happened, then so and so must be the causes), and its tendency to develop inadequate analogies or to overgeneralize from single cases. The essay contends that, while some uses of narrative as argument display these problems, they are not inherent in narrative itself. It offers warrants for that contention by (a) proposing a conception of narrative as rhetoric and (b) using that conception to analyze two essays by Atul Gawande, “On Washing Hands” (2007) and “Letting Go” (2014), which rely heavily on narrative as part of their larger problem-solution argumentative structure. The analysis leads to the conclusion that a skillful author can, depending on his or her overall purposes, use narrative either as a mode of argument in itself or as a means of supporting arguments made through non-narrative means -- and can even use both approaches within a single piece.

Which abstract is stronger? Abstract 1 adopts the strategy of offering a general statement about the larger argument and focusing on what the essay says about the case studies. Abstract 2, in contrast, backgrounds the details about the case studies and foregrounds the larger issues of the argument. Not surprisingly, in light of what we have said so far, we find Abstract 2 to be far more effective than Abstract 1.

Abstract 1 is ineffective because it is almost all trees and no forest. Indeed, its opening acknowledgment of the forest does not do much more than restate the essay’s title. In addition, Abstract 1 does not address the “so what?” question, does not say anything significant about its methodological commitments, and does not even do a good job of explaining how the analyses of the two case studies relate to each other. Its fundamental flaw is its unexamined assumption that each of its two cases studies is in itself a significant contribution to conversations about the relation between narrative and argument.

Abstract 2 is much more effective because it backgrounds the trees and foregrounds the forest. It starts by addressing the “so what?” question, it explicitly announces its methodological commitment (to a conception of “narrative as rhetoric”), and it clearly states its conclusions in a way that situates them in the larger debate. Rather than assuming that the analysis of the case studies is a worthwhile project, it focuses on articulating its worth. The abstract itself does not provide sufficient grounds for an editor or reviewer to decide that the essay should be published, (after all, the execution of the argument sketched in the essay could be unsuccessful), but it is likely to make those readers interested in examining that execution.  

In closing, we note that we’ve been assuming that the article out of which the abstract is formed is a publishable one. But that might not be the case. If you’re having trouble writing an abstract that articulates what it contributes to an existing scholarly conversation, the reason could be that the article doesn’t do that either.  If you find yourself in that situation, you can use the exercise of writing an abstract to detect the raw or undercooked elements of your article.

In fact, we recommend that you not wait until you’re finished with your article to write your abstract. Try writing it at different stages in the composing and revising process and use what you discover about what’s easy and what’s hard to articulate about the project to clarify and strengthen your argument in the essay itself. In other words, yet another important audience for your abstract is yourself.