Formal Features of Abstracts: Length, Word Choice, and Grammar

This section of the chapter will examine the formal features of abstracts, including length, word choice, and some grammatical issues. Please note that this is not a comprehensive guide but rather a few highlights of issues that we feel are particularly relevant to novice researchers.

Length

In their book-length work on abstracts, Swales & Feak (2009)^[1] suggest that the first step to writing an abstract is to find out the word or character limit. Indeed, word/character limits are an especially noteworthy aspect of abstracts because they are so often used as a gateway into acceptance — especially for conferences. Swales and Feak present several guidelines for writing abstracts, and the ones relevant to our discussion are as follows:

1. Abstracts in journals are usually between 150-200 words without subheadings (unstructured) or a bit longer (~250 words) if structured (with headings).
2. Disciplinary differences may allow for longer (~500 words or more) abstracts; specifically, those for conference submission are generally longer.
3. The IEEE^[2] uses shorter abstracts (~50 words) for “short communications,” which is a type of article in many of the journals that it publishes.

Although length is often a challenging component of writing an abstract because it’s difficult to summarize an entire study in only a few hundred words, it is a fairly straightforward aspect of this section of a research article. Each journal has a requirement, so you can look up their maximum word lengths on their website.

Word Choice

The primary determinant of the language used in an abstract is its function/purpose. The purpose drives the decisions you will make about what kinds of words and information to use or avoid. Prescriptive guidelines for abstracts are usually dictated by either the discipline or the journal/editor, and, as Swales & Feak (2012) note, there are often “not rules to follow, but rather choices you can make”^[3] about what language to include or exclude. The next sections provide some guidance to help you make those choices, and although we are including these in the chapter on Abstracts, the information can be applied to any aspect of research writing. However, as the abstract is a prominent way for others to be introduced to your research^[4], it is particularly important that the language be clear and appropriate. Several research writing guide books claim that because of the heavy reading requirements that researchers face, they must be extremely selective about what they choose to read. Many choose to read abstracts as a summary of an article either as a preview or a replacement. ^[5][6]One way to begin drafting your abstract is to consider that both its placement and function within an RA is to bridge the title and the text of the manuscript. Therefore, one best practice is to include every important word from the title in the abstract^[7]. This will help you consider the content of the abstract, but there are other considerations as well. For instance, generally abstracts present information in the same order as the research article: IMRD/C. As noted previously, the final section may or may not be included based on what stage the research is in, but regardless, the abstract will flow in this general order/pattern.

Next, we’ll explore some particular word/grammatical categories in an effort to provide you with some ideas. These are not rules, but rather, are suggestions based on linguistic research and general academic writing standards.

Pronouns

The use of personal pronouns (e.g., I, me, my, you, s/he, you, they, etc.) in academic writing are a common source of confusion for many graduate students and other novice academic writers. Are pronouns acceptable in formal academic writing? Should they be avoided? If you are writing about someone else’s work, the third person “they” or “s/he” is widely accepted. In contrast, the use of the second person “you” is typically avoided ^[8].

In general, there is much variation in the use of personal pronouns in academic writing even among expert writers^[9]. Most writing style guides advise against the use of first and second pronouns and the use of the singular masculine he ^[10]. Often the suggestions about alternative structures include the use of passive voice, the combined s/he (which can be formed as she/he, he/she, or with the conjunction and between the two), or they regardless of gender and number. In fact, studies show that the singular pronoun they is the most frequent personal pronoun, and has been for at least two decades^[11][12][13].

Given this information, we suggest that the use of they is very likely the best choice regardless of the number (singular or plural) or gender. That is, the pronoun they can have antecedents that are male, female, singular, or plural.

Verbs

Tense

Generally, the use of verb tenses varies by section within an abstract. Swales and Feak’s comprehensive guide to academic writing^[14] asserts that abstracts tend to begin and end in the present tense but vary significantly in their mid-sections. While there has not been a large, cross-disciplinary study of verb tenses in research article abstracts, there have been some studies about specific disciplines such as applied linguistics^[15] and medical research^[16]. The latter study found that past tense is often used to write about the purpose/background of the article, to explain the Methods, and to highlight the most significant Results. However, that study also notes that when a writer wants to focus on the generalizability or boost the significance of the findings, authors will often opt for present tense. Note the difference in the strength of these two sentences:

1. Our results showed that there were significant differences between the two types of X.
2. Our results show that there are significant differences in the two types of X.

The second sentence presents a stronger stance simply because it is in the present tense, which is typically used for facts, general truths, or fixed circumstances. Simple past tense is used for actions started and finished in the past but not necessarily continuing into the present. Therefore, choose your tense carefully because it can be indirect evidence of your stance.

Modality

Modal verbs indicate stance; in other words, they allow the writer to strengthen or weaken a claim. In English, there are nine modal verbs as shown in the table:

It’s important to note that each verb carries a level of certainty or doubt that is a really useful way to hedge or boost your claims about findings. Since the abstract is such an important part of representing your work, the careful use of modals is essential to striking the right tone in terms of your level of confidence. Of course, there are other words you can use that also indicate the tentative nature of your work (e.g., adverbs and adjectives such as possibly/possibly, obviously/obvious, etc.). You can explore those on your own as they are outside the scope of this book. The important point to remember is that you always have a choice about presenting your level of certainty, and modal verbs are important tools in your writing toolbox.

Clauses with that

In 2005, linguistic researchers did a study of over 200 abstracts from six disciplines. The most important finding from the study was that writers tended to use that clauses when writing about their results. For example, “The study’s results indicate that …” or “The findings confirm that …”. In the full research article, you’ll probably be using these clauses in the Introduction to review literature (e.g., “Other studies have shown that …”) as well as to report the findings in the Results section. However, in an abstract, you will not be reviewing literature at such a fine level of detail, so it’s more likely that you’ll use such a clause to highlight your findings.

Reporting verbs are often used with the word that to form such clauses. So, it’s important for you to familiarize yourself with some of the more common reporting verbs. A few highly-frequent verbs of this type are presented in the table below:

Each verb has a level of certainty of its own (e.g., assert is more confident than indicate), so be sure that you are familiar with the exact meaning of each verb. On top of that inherent meaning, such verbs can be combined with adverbs (e.g., possibly, certainly, etc.) and/or modal verbs that can strengthen or weaken the claim as well.

Citations in Abstracts

To cite or not to cite, that is the question. If you do a Google search “Can I cite sources in my abstract?” you’ll find an overwhelming number of negative responses (i.e., No, you should not.). However, as we’ve noted in other sections, academic writing is often not black and white, wrong or right, no or yes. So, while the internet majority may have very sure opinions, it’s less clear than it seems.People provide several reasons why citations in abstracts should be avoided, namely the following:

• To focus on your work, not someone else’s.
• To present a self-contained work without the need for outside reference.
• To reduce unnecessary words contributing to the small word count limit.

On the other hand, there may be some instances in which it is incumbent upon you to cite the work of others. When you are doing research that is a direct response to another researchers’ findings, theories, or claims, then it’s important to cite them. Likewise, if you are using a little-known term that was coined by someone in particular, then it’s important to reference that person. And undoubtedly, when you are doing a replication study or some meta-analyses, it’s necessary to have citations.

On the whole, though, the best practice is not to cite sources in an abstract unless it is absolutely necessary^[18]. As with other topics in this book, there are different approaches within disciplines, journals, and writers, who may all have variable stylistic approaches to this issue, so read the Instructions for Authors for your target journal. If there are no guidelines, then examine the abstracts of previously published papers in that journal to see what the common rule seems to be. If all else fails, you can contact the editor.