Is it wrong to repeat yourself in the abstract, introduction and conclusions?

Question

I find that when I writing a paper, I use the same phrases and sometimes the same sentences in the abstract, introduction, and conclusions.

When I compare this to non-fiction writing (e.g. Malcolm Gladwell or Michael Pollan), my impression is that they do not repeat themselves in the prologue and epilogue.

However, I realize that they are writing non-fiction books/articles, whereas we are writing scientific articles, for a different audience and a different purpose. My questions are:

1. Does using the same phrases and sentences in the abstract, introduction and conclusions make my paper boring to read?
2. Or is it easier to read and understand a paper if it repeats itself?

Answer 1

The three things are doing pretty different jobs, and so should be pretty different in content.

The abstract is an advert for your paper. Someone might read a list of titles and abstracts (for example on arXiv) and decide on that basis whether it's worth their while downloading and reading the rest of the paper. The abstract has to be attractive, but not fraudulent ('truth in advertising'!). When you read an abstract, you're trying to find out: (i) is this paper in my area?, (ii) is it asking a question I care about? (iii) does the conclusion sound interesting? You don't have to summarise the whole paper, but you do have to answer those questions.

The introduction gives your reader a map of the paper. By the end of the introduction, your reader should have a pretty good idea of what they're in for, and where they're going to end up. They should know what sections they're probably going to skip, and which ones they look forward to disagreeing with. A paper is not a detective novel: you don't need cliffhangers or mysteries.

After that, the conclusion doesn't have a lot of work to do (at least in my experience). It might be the last thing the reader reads of your paper (that is, unless they care enough about its contents that they'll re-read it), so this is an opportunity for you to frame their memory of it. "We have shown that ..." is the sort of thing you'd find here.

Key points:

• Remember the reader is a human being, just like you – talk to them, as you'd like to be talked to (far far too many people bizarrely forget this).
• Make it easy for them to read the paper, and easy to agree with you (that's why the introduction has a map of the paper, to ease the reader's way through your golden prose).
• Your reader is asking themself "why am I reading this paper?" Make sure they have a good answer to that question by the end of the introduction. If you get a colleague to read your draft, get them to answer that question at that point. If they don't give the answer you want them to, it's your fault; so edit.

There will probably be some overlap in text in the three parts. That's OK, but it probably shouldn't be cut-and-paste.

Answer 2

There are many examples of papers where there is repetition in these part. While this is not wrong, it is not something one should adopt as a way of writing. It is important to write each part so that the text is adapted to be included seamlessly in that part.

The discussion should lead up to the conclusion(s) of the paper and from this perspective it is the origin of the content that will be repeated (in context) in the other parts. The conclusion should summarize the main findings of the discussion. Since the point is to provide the important points in a brief way, one should at least try to write it with the summarizing in mind. The abstract should summarize the entire paper and lead up to the main conclusion(s) of the paper. This is most often focussing on a subset of what is summarized in the conclusions.

Copy-pasting sentences from discussion to conclusions and to abstract means you build in sentence structures that may not be optimal for the different parts. I therefore suggest at least trying to rewrite the information for each part so that they become optimized in each case. If a sentence you have already written fits very well, then its use is fine but if you simply try to force sentences in the resulting text might not be as clear as it could be. So, in general, write everything as a new text based on the story you are trying to convey rather than inheriting structure from what you have written already stemming from a different context.

Answer 3

I'm not bothered by the repetition of sentences if they are isolated sentences, that serve as a link. However it is inelegant, you can always put a cross reference and reword what you say in some other place, for several reasons:

• Style in writing, you should show some knowledge about the language and the capability of paraphrasing.
• Fit better the contents to the context where they are.
• Writing the same thing in several different ways helps people to understand what is written there. English, as most (probably all) other languages has an inherent level of ambiguity, some sentences can have several meanings. If you rewrite some sentence with some different words by doing so you can make clear that the meaning of both sentences is in the intersection of the potential meanings of both sentences.

I find the third reason is the most important and other answers were missing this aspect, so I had to drop another...

Having said this, the repetition of a paragraph is unacceptable for me. If there is some paragraph in some part of the paper and you want to refer to this same piece of information you should summarize the contents of the paragraph in a sentence (sometimes in a term, e.g. when the paragraph is a definition).

This answer isn't very constructive and I feel obliged to provide constructive answers so...

Take the abstract as a reference. It will summarize the paper, as the introduction does and also the conclusions.

1. In the abstract you provide the shortest summary and describe what have you done.
2. In the introduction you extend that summary and describe the context of your work (some people include the motivation). It is different from the abstract because you don't say what you have done, but why it was missing, why you have done it and where in the paper you explain into more detail these aspects.
3. In the conclusions you evaluate what you have done, how it serves to fill the gap that you identified in the introduction and the gaps that remain (future work).

All these are similar but not equal.

BTW: related question: https://tex.stackexchange.com/questions/181080/automate-the-description-of-the-paper I should get back to that...