"Retrofitting" sources

Question

There seems to be a sort of craze about having as many references as possible. Among the students at my university, it has become common practice to write a passage of text and later figure out if someone else has done something similar and attribute it to them, changing the original text if necessary.

Is this in any way, shape or form appropriate? It seems rather contradictory to me, since you are attributing your own work to other people, even if they came up with it first. The author did not have prior knowledge of this.

I suppose you could simply claim it as your own and note that someone else has come up with it, as well. Is this necessary? It is your own work, after all, even if somebody else came up with something remarkably similar.

Many problems have straight-forward solutions that you can easily come up with. I am asking this now, because I am writing my bachelor's thesis and have come up with some of those straight-forward solutions myself. I just now found, by chance, papers on exactly that subject and, reading them, realise just how similar our approaches are.

What would be the best way to handle situations like these?

Meta: I'm not happy with the title. If you can think of a better one, feel free to edit and remove this notice.

Answer 1

Among the students at my university, it has become common practice to write a passage of text and later figure out if someone else has done something similar and attribute it to them

Indeed, that seems to be a constant part of paper writing: A portion of what you have made up has been done by others the same way before. In my opinion, this is often because it is simply impossible to always know all details of previous work (that is by no means always "related" to your own work in its overall topic) by heart, and because it is often more straightforward to just do something oneself rather than dive into the literature that may or may not contain something vaguely similar.

Sometimes, it gets apparent even only afterwards that an aspect of one's work is non-obvious enough to be mentioned elsewhere.

In the end, it is simply a question of benefits vs. drawbacks:

Drawbacks

• You do attribute your own work to someone else. Thereby, even though you may have spent some energy to develop or define something, you
  • falsely pretend to have relied on someone else's work
  • misrepresent the effort you may have invested for your work

Benefits

• You secure yourself against claims of plagiarism. While you cannot possibly copy anything you do not know, the suspicion that you did actually know some previous work but failed to mention them can arise when certain of your ideas resemble those mentioned in other papers. As such, it is less cumbersome if you just pretend you are replicating someone else's work.
• Usually, the statements subject to this phenomenon do not nearly seem complex enough to warrant the mention as a contribution. Hence, you do not lose much by not mentioning you were the actual source for a specific aspect of your work.
• Citing in such situations strengthens your work. While the contribution of the cited aspects would be minor, showing that you build your actual contributions on top of proven practices makes your actual core contribution more convincing.

In particular, the last point is a major benefit compared to the drawback. This is reflected in your observation that some are

changing the original text if necessary

Given that the original ideas are not a significant contribution that needs to be promoted, anyway, replacing them with a possibly better method that is truly taken from another source is fully reasonable and can help to further improve the overall document.

EDIT: To put this answer in relation to jakebeal's answer: I mainly interpreted the question to refer to the "third branch of possibilities":

• They are writing down original text that they think is novel. Only later, they realize or find out others have done the same (possibly, in a completely different field or context, or with a completely different terminology, so finding the other work is likely a result of pure chance rather than an expected event).

Answer 2

I see two issues here:

1. Is it misleading to cite a reference you didn't know about until after your work was done? In particularly, does it misrepresent history by falsely suggesting that this reference influenced your work?

2. Is it unfair if you have to give credit to someone else for something you independently did yourself, without even knowing about their work?

The answer to both questions is no.

Citing sources is primarily about assigning credit and indicating where one can find details or more information. It's not about describing your personal influences. In particular, everyone is aware that some references may only have been found after the fact, so nobody will interpret a citation as necessarily indicating direct influence while your work was being carried out.

As for the second issue, academic credit is awarded for making a novel contribution that advances the state of the art, rather than for rediscovering what was already known. Being credited is a form of recognition for telling the world something new, rather than a reward for being clever or having worked hard. (The person who rediscovers something may be just as clever and hard working, but hasn't done as much to increase the world's store of knowledge. Of course, there may be borderline cases, where the original discovery was really obscure or its full generality had not been recognized. I'm talking in broad generalities here.)

If you independently discover something at roughly the same time as someone else, then you can appropriately claim some of the credit for it. However, you can't show up substantially later and expect to share the credit.

I suppose you could simply claim it as your own and note that someone else has come up with it, as well. Is this necessary? It is your own work, after all, even if somebody else came up with something remarkably similar.

There are times when noting that you did it independently could make sense, for example to explain why your approach differs in some ways from what is usually done. However, you should be very careful about how you do this, since it looks bad if you seem to be claiming more credit than you deserve. In most cases it's better not to point out your independent rediscovery.

When you ask "Is this necessary?", are you asking whether you need to include the citation at all? Yes, you almost certainly do. Omitting a citation is allowed only for things that have become common knowledge in your area (e.g., just about nobody cites Newton for the laws of motion) or are so trivial that they aren't worth any credit at all. This citation is evidently not common knowledge, since you didn't know about it. It might in principle be a triviality, but the previous discoverer must have thought it was worth publishing. Declaring it trivial and unworthy of citation is a very risky move, which you should make only if you are confident in your judgment and expertise.

Answer 3

I find it difficult to understand what exactly you say that people are doing, so let me split it into two branches of possibilities that I think you may be saying:

• Perhaps they are writing down background they think they know, but aren't bothering to look up citations while at an early stage of drafting. e.g., I might write in an early draft: "The lower bound on comparison sort is O(n log n) [find citations]", then later go back and fill in the citations, adjusting as necessary. This is entirely reasonable: there is no requirement to write a paper in any particular order, only that the result be scientifically honest and ethical.

• On the other hand, if they are writing original text or results and trying to put the words into somebody else's mouth, that's clearly unethical and also completely inscrutable to me in its motivations. Unless they're trying to follow certain medieval practices where you have to pretend your original thoughts are all just interpretations of the wisdom of the ancients?

Answer 4

When producing a publishable piece of work (this includes your bachelor's thesis) it is your responsibility to check whether the methods you've implemented were done so before. If so compare the results and steps of your experiments with the one(s) previously published. Otherwise there is no point reproducing the same matter in theory.

I can understand your claim of attributing your findings to another person. But this cannot be differentiated from plagiarism. It is equally possible to include another's content without citation to claim it as your own. That is what the definition of plagiarism is about.

It is true for many to elaborate on the reference section and reviewers look for this too. This is to prove to the reviewers that enough work has been done to assess your field of application before proposing your own work which may be randomly novel or not.