Version classiqueVersion mobile

Writing and Publishing Scientific Papers

Gábor L. Lövei

Part II. Writing the paper

8. Abstract and Keywords

Texte intégral

1The Abstract (sometimes called the Summary) is always printed near the start of the paper, usually immediately after the title, authors and addresses. This is, one can argue, the most significant part of a paper, because:

  • this is the part that is read by most people, even by those who will, ultimately, not read the whole paper.
  • most readers, including the first reader of your manuscript submission — the editor — will also start reading here. First impressions are important. Moreover, the editors know from experience that a bad abstract is rarely followed by a brilliant paper. Consequently, after reading the abstract, the editor will be close to forming a first opinion.
  • an abstract is often reproduced by itself in various databases. This dictates that the requirement that the summary should be self-explanatory — it must be understood without reference to other parts of the paper. For many readers of the summary, the full paper will not even be available.

2Occasionally, an abstract gains additional importance. Conference invitation, participation, and even financial support to attend, can depend on the abstract of a proposed contribution. In these cases, the conference organisers must make their decision based on the abstract only. So, a good abstract can influence organisers to offer conference acceptance and/or funds to support participation.

3A good abstract is a mini-review of the paper. It states, briefly, the question/problem, the method(s) used, followed by brief results and the main conclusions. Some journals follow a system of numbered statements, or headings within the summary. A few provide detailed instructions (Box 7). However, even if the journal in question does not indicate the main aspects with headings, you have to follow the same structure.

Box 7. Nature’s abstract-writing template

Nature’s abstract-writing template,​nature/​for-authors/​
formatting-guide. © 2021 Springer Nature Limited. All rights reserved. Permission for further reuse must be obtained from the relevant holder of the exclusive rights.

4Because an abstract is often reproduced separately from the full paper, it has to be self-explanatory. Consequently, you should avoid using abbreviations, because they are understandable only by reference to the full article. Similarly, references to figures or tables are not allowed, because readers who only have access to the abstract cannot check or see the figure mentioned. For the same reason, references to published articles should be avoided. If unavoidable (for example, your paper is a direct reply to a published paper), a short version of the full reference should appear in the abstract. This short version should include the surname of the first author, the year of publication, the abbreviated name of the journal, the volume number, and the number of the first page only.


5Abstracts always have a word limit, usually 200-500 words. You cannot go beyond this limit, but it is not mandatory to use all of it. If you can clearly summarise your study in 150 words, you do not have to use 200. The important thing is that you should not list what was done, but concentrate on the results. The abstract centres on your own results, so it should be mostly written in the past tense.

6The abstract is a summary of the paper, and there should be no statement or conclusion that is not in the paper. One should be careful not to include information that is not in the text (a surprisingly common error!). A good abstract is not a set of carefully cut-and-pasted sentences from the full paper; you must rephrase the same facts or statements that are present — usually in more detail — in the paper itself. At the end, the conclusions can be mentioned. These, however, should be meaningful. The statement that “The consequences are discussed” is neither very original, nor does it say much. This is the purpose of the discussion, after all. “More research is needed” is another meaningless conclusion. Avoid “throwaway sentences” such as these.

When to Write?

7I suggest that the abstract is best written once the manuscript itself is finished. This is only a personal recommendation as I do this myself. If it helps you to structure the paper, you can start with drafting the main points but I find it unlikely that you can write an effective summary of your work before it is completed.

Research Highlights, Graphical Abstract

8Several journals now require a tightly structured set of “research highlights”. The relationship between these and the full summary is a little like that between the running title and the full title. Concentrate on your main results, and consider, carefully, how to shorten them because this section is limited by the number of spaces, i.e. letters and punctuation.

9If required, a graphical abstract accompanies the research highlights. Both will appear on the website of the journal but not, usually, in the final paper. When facing such a task, think about creating a new figure rather than repeating one of your figures from the full paper. However, this is not a rule: in cases when your main results can be effectively presented on a graph, this graphical abstract can be identical to a figure that is also in the paper.


10Keywords serve to assist those who use various databases and search engines to find your paper. They are usually single words that mention some important concept or aspect of your study. The number of keywords is always limited, usually to 6-10 words (double words and, exceptionally, triple words are also allowed). These key words will be entered into databases and keyword lists.

11It is a good idea not to use words that are already contained in the title, because effective title words are always used for the same purpose. Given the limited number of possible “pointers” to your paper, it would be a wasted opportunity to use the same word twice: once in the title, and again among the keywords. Several journals do not allow title words to also be keywords. An effective keyword is a word with a specific meaning or significance; words such as study, change, or experiment are not effective words in this context.

12When considering keywords, imagine yourself as someone searching for your paper. What aspects are relevant? These can be locations, organism names, concepts, or method terms. Avoid fashionable, or too general, keywords — a reader searching for information on a certain topic will probably disregard the results of a search with hundreds of “hits”. Thus, if you include a keyword that is frequently used, your article may be included, but in so large a result list that it does not help the searcher. Still, you can mention broader ideas, or concepts. If your title contains a species name, mention the family or higher taxonomic association. If there is only a scientific name in the title, include the common name. If it contains a location name, add the name of a wider region. Likewise, if you indicated a kind of habitat, add the more precise location as a keyword.

13However, use this option in moderation — you do not have to use the maximum number of keywords allowed. You can write fewer keywords if you want, but you cannot include more.

14Keywords are usually placed after the abstract — but check the journal instructions for precise placement of the keywords.

Table des illustrations

Fichier image/jpeg, 95k


Le texte seul est utilisable sous licence CC BY 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search