the semantics of references in academic papers

[gusl]

If you see a quote like:

A: "The problem of P has been previously treated in context C (Bovik, 2005)."

or

B: "The problem of P has been previously treated in context C [1]."

[1] Bovik, Harry Q, 2005 - TheBook.


is the author saying that Bovik has treated problem P in TheBook, or that Bovik has stated this in TheBook (i.e. the author is giving his source for the claim above)? Unfortunately, English has no evidentiality.

Also, when should you use style A vs style B?

Comments

[roseandsigil.livejournal.com]

Presumably the first.

The difference between the two options is, I believe, purely stylistic, and will depend on where this sentence is being published. Different disciplines, journals and geographic regions have different style guidelines. I generally see B in U.S. computer science papers. A tends to be for humanities, I think.

[gustavolacerda.livejournal.com]

I guess my Master's thesis was in the humanities then.

[roseandsigil.livejournal.com]

I'm assuming that's sarcasm, in which case, I think it's really a per-journal or per-publisher thing, and what I said was only broad tendencies, and there, probably just in the U.S. (where did you do your Master's work?).

[gustavolacerda.livejournal.com]

It was friendly semi-sarcasm. I agree with your observations.

[gustavolacerda.livejournal.com]

I did my Master's at the ILLC in Amsterdam.

[simrob.livejournal.com]

Well, you know those crazy europeans ;-)

[bhudson.livejournal.com]

It's not entirely just stylistic: in the age before automated citation programs, numbers were stupid because they all needed to be redone if you added any citations. Unfortunately, and perhaps surprisingly, we're still in that age for a number of people.

Most of computer science (though less so in systems), math, and physics use TeX or LaTeX because of how easy it makes it to write equations. Having bought into LaTeX, we then flip the page to the chapter on bibtex, and thus use [1] because that's the style used in the examples, and hardly anyone really understands how the typesetting programs really work so we just build off examples. I'm not sure that counts as a stylistic 'choice'...

[gregh1983]

If it's saying "was treated by," then I expect that the reference includes some actual work on the problem. "Was formalized by," on the other hand, might indicate that the problem was just stated.

There are many LT papers that use the (Bovik 2005) style, but maybe that grew out of the linguistics side of LT rather than the technical? I like it better than [1] because you can immediately know something about the paper without having to look to the appendix. That way you start to recognize citations or papers you've read before.

[bhudson.livejournal.com]

Knuth (I think -- it might actually be Lamport) has an extended rant against (Bovik, 1964) references, based on the fact that they take up so much space and break up the flow a lot more than [1]. The former is the Chicago Manual of Style recommendation, so lots of people use it. One of the main counterarguments the Chicago Manual uses against [1]-style is that you have to *gasp* renumber all the references by HAND! Which is, of course, bogus nowadays.

My favourite bibtex style is the 'alpha' style, which uses [Bov05]. Both short, yet informative enough that I only have to look at the bibliography once, at most.

[mdinitz.livejournal.com]

I agree about the alpha style. It's far and away my favorite, followed by [1], followed by parenthetical references.

[simrob.livejournal.com]

That is so hard to read because I keep trying to read all the references as references (meta-syntax) and not objects (object syntax)

[gustavolacerda.livejournal.com]

yeah, me too!

[bhudson.livejournal.com]

Maybe I'm dense, but I have no idea what that sentence means.

[gustavolacerda.livejournal.com]

He is saying that when you write "(Bovik, 1964)", it's very hard for him to interpret this as the style. Instead, he's imagining what Bovik could possibly have said on this subject.

Likewise when he reads "My favourite bibtex style is the 'alpha' style, which uses [Bov05]", his reaction is that the sentence is incomplete: "use" is a transitive verb!

[bhudson.livejournal.com]

is the author saying that Bovik has treated problem P in TheBook, or that Bovik has stated this in TheBook (i.e. the author is giving his source for believing the claim he states)?

It could be that Bovik talks about P, or Bovik talks about C, or Bovik talks about P in the context of C, or that Bovik once said something in passing that was somehow vaguely related to how there's a fly in my coffee. Really, the standards for what goes into the brackets are pretty lax.

[easwaran.livejournal.com]

I would assume that Bovik actually talked about P, rather than talking about people talking about P, though the latter is also acceptable. However, I would prefer to phrase the latter as:

"As (Bovik, 2005) mentions, the problem of P has been previously treated in context C."

And style A is used in philosophy - I've only seen B in occasional math things I've read. My explanation for a preference for A is that it makes it easier to know who said it when (and also, generally, in what publication), but I can see why people would complain about the amount of space it takes up.

But yeah, you just use whichever one the journal uses, which tends to be fairly standard across a discipline. (Though The Journal of Philosophy does the crazy thing where it gives the full citation in the first footnote, rather than a bibliography.)