Using reference as a subject in writing [duplicate]

Question

Is it acceptable to use references as a subject in a sentence (in the context of academic writing, preferably in computer science)?

For example:

An extensive research has been done on eye capture by [Bérard et al. 2014].

Versus:

An extensive research has been done on eye capture by Bérard et al. [Bérard et al. 2014].

I feel like it removes tautology and makes the text easier on eyes. Moreover, I have seen published papers where authors write like this (although not many). However, it is considered to be bad style by a couple of people who I have asked. Their reasoning is that a reference should be an additional information and if you remove all references from the text, the text should still be complete and make sense. I understand this reason, but I feel like there are no real cases when this can be needed.

Answer 1

Is there any document that regulates your citation style? For well-established citation styles, like Chicago or APA, there are extensive manuals that cover all corner cases of usage. Maybe try consulting this kind of document first?

As far as personal taste and experiences go, when editing papers in computer science according to IEEE requirements, I tend to use forms like:

Blah was done by Smith [1]

or

Blah was described in [2]

The first one is similar in idea to your second example, but use of numeric reference removes aesthetically unpleasant repetition. In your case I like your first example more, but whether it is acceptable should really depend on your citation style guidelines (for example, some allow for the form this was done by Smith (2014)). Nevertheless I don't subscribe to argument that text should be readable without references. After all, they are integral part of your paper and you wouldn't want anyone to remove them, thus reducing the credidibility of your work.

Answer 2

I don't know about computer science, but we would do

An extensive research has been done on eye capture by Berard et al. (2014).

Answer 3

An extensive research has been done on eye capture. [Bérard et al. 2014]

would seem to be a good compromise, which, though perhaps technically ambiguous, would likely be read in the spirit intended by any reasonable reader.