While adding references to a .bib file the other day I found myself putting the title field first as it provided more information about the citations than the author names. After some thought, I wondered why most bibliography formats don't do that to begin with; if the primary goal of references is to provide support for the topic at hand while also giving proper attribution to existing works and to provide avenues for further research by readers, would it not make more sense to put the titles of papers/articles/etc. first and then the author names? It seems to me that the important thing (in terms of the work being written) is not who wrote the cited works but the works themselves, and if citing multiple papers by the same person putting the name first provides somewhat of a distraction from the actual content of the citations. Particularly in the case of collaborative efforts, where a work may have been written by a multitude of people, the title of the work can get pushed aside by the multitude of names (at least when using a format that doesn't just collapse the names to "et al"). What are the underlying reasons for this, or is it just a social thing that happened because researchers like having a significant effect in their fields/their careers depend on it?

  • 6
    If I am looking for a certain paper in the bibliography, I know the author name, not the title of the paper (in most cases). If I want to discuss the choice of a textbook for a certain course, I refer to them by author name, since the titles are pretty much interchangeable.
    – GEdgar
    Apr 19, 2016 at 14:20
  • 1
    @GEdgar Your first statement assumes the paper is cited by author name (and possibly year) rather than by number (the latter seems to be more common in computer engineering) or that, if citations are numeric, the name is explicitly mentioned in the text (which again seems to differ by field).
    – JAB
    Apr 19, 2016 at 14:28
  • 1
    So many titles starting by "The impact of...". Longer to read than author's name
    – Emilie
    Apr 19, 2016 at 15:07
  • 3
    What @GEdgar says. Most people find it a lot easier to remember ideas by the names of those from which they've learnt them rather than by the (often uninspired and random) titles that have been chosen to convey them. Apr 19, 2016 at 15:12
  • 4
    Usually people will write multiple papers exploring one overarching theme, each paper building on the previous ones and looking into a new aspect. The papers will be connected by one or more common authors, not by their titles. I know a couple of papers on certain topics that I recommend over and over - I know their authors, but not the titles, and I always go back to my .bib file to find the exact references. Apr 19, 2016 at 16:17

4 Answers 4


This Question makes sense in practical terms, but includes a misconception in the area of academic culture, when you say, 'if the primary goal of references is to provide support for the topic at hand [...]'. That is not quite right.

Also, disciplined academic writing tends to take the form of a dialogue (or in simple cases a dialectic), where differing perspectives are chucked into an arena and required to fight it out. None of them necessarily has to 'win': finding them all to be wrong is a positive result.

It is very often useful to label a view primarily as emerging historically from the body of some identifiable person's work, rather than hovering it vaguely in conceptual space.

Referencing has two primary goals.

The first is to help interested readers to identify and follow-up the material that you have recruited for your argument, so that they can, if they wish, follow your evidence in detail and, perhaps, find the stuff that you left out. The idea is (partly) that you have openly saved a reader time in either embracing or challenging your position. Under present archival conventions, the tracking-down of sources for this purpose (or any other) is almost always most efficiently done by starting with author name. Scrambling for titles on the web, although I now do it every day, does not (yet) count as an archival convention.

Second, referencing is part of the process of locating yourself in the critical discourse of whatever discipline you are working in. Your writing is in some sense an act of positioning your view in relation to others', even (or perhaps especially) if some other person's view has shifted over time.

For example... In film studies it would seem rather ignorant to talk about 'the male gaze' as a convention of camerawork and editing without (a) attributing the expression to Laura Mulvey and (b) noting that she changed her view over time (see Wikipedia, for shorthand).

Merely referring to ideas from published articles, blog posts or even comments by anyone is fine in its place (e.g. my doctoral thesis, here and there). For the most part, however, it usually works best to say something like, 'Aristotle proposes [an impressive idea]. Hitler disagrees [here, because of this].
Colonel Sanders inadvertently resolves that conceptual conflict by suggesting [that surprising thing].

That is also why we use the present tense: 'Roland Barthes says [something]. Michel Foucault (while never identifying Barthes) responds by saying [something that works far better in the real world].' You are taking part in a present debate, with points of view that are most efficiently located by identifying their authors.

  • I apologize for the confusing wording; I didn't really mean "support" in the sense of supporting an argument but more of supporting a concept and providing background information, which I guess would most closely match your self-locating aspect. That said, your response is nicely enlightening for someone still relatively new to academia.
    – JAB
    Apr 19, 2016 at 16:45
  • 2
    Your mention of film studies makes me wonder: Do academic texts in film studies refer to films primarily by title or by their director, actors, etc.? Apr 19, 2016 at 18:06
  • @O.R.Mapper Referencing films is something of a fraught issue. (You tell me which version of Metropolis we are talking about... and who 'created' it... and when...) TV episodes are usually a bit easier because they are (usually...) more reliably identifiable. Fan films, YouTube vlogs and other relatively informal material introduce all sorts of complications. (I am suddenly a bit swamped, but nag me at the weekend and I will give a fuller answer.) Apr 19, 2016 at 19:37

If I compile a bibliography and ask myself what to cite my thoughts are "There was this work by X in 1995..." and sometimes "There was this work by X in journal Y..." but really never "There was this paper Z...". I guess many people's minds work like this. In fact in conference chat people usually refer to the "Smith paper from the 90s" or something like that.

So when I read a paper and look up a reference in the bibliography and see that the respective work is by Smith and Jones I usually know the paper from the context. So I don't wonder why the authors often come first but why the year often comes last at the same time because these two things are what I find most useful. At second thought this is not the worst solution since beginning and end are somehow more easily to spot than first and second place.

I'd like to add that I read some papers in Physics where the bibliography did not contain the title at all, but only author, journal, number, pages and year.

  • If X wrote more than one interesting paper in 1995, do you just add on the subject?
    – JAB
    Apr 19, 2016 at 19:06
  • 1
    @JAB If you cite Smith's two 1995 papers, exact referencing details vary but in general the alphabetical first is Smith (1995 a), and the second is Smith (1995 b). Your university/publisher/employer will always be able to tell you the adopted citation format. Many are available, but all are quite well understood and fairly easy to follow. Making up your own will just make it hard for people, and you might have to go back and amend your work in future. This Question is resolving into a matter of academic style, where seeking local advice on preferred format is always the way to go. Apr 19, 2016 at 19:24
  • @CaptainCranium I was referring to Dirk's mental model, not to the citation format (which I prefer to generate rather than manually format these days now that I've been working with LaTeX for a while and don't have to try to mess with the limited bibliography management in most WYSIWYG word processors.)
    – JAB
    Apr 19, 2016 at 19:29
  • 1
    @Dirk Not if you have referred to two, three or twelve papers/books/songs by Smith from 1995. Forcing the reader to guess and hunt is basically lazy and rude, when you obviously imply that you have the information right in front of you... or you might be trying to hide the fact that you are guessing yourself and your 'research' did not really occur at all. Referencing is always about helping an imagined reader to replicate your project shelf for this stuff (whatever it is), as far and as clearly as possible. Apr 19, 2016 at 19:47
  • 3
    @captaincranium I can follow... You seem to be writing about citations in written text while I wrote about how I remember articles.
    – Dirk
    Apr 19, 2016 at 19:55

I have often wondered about the same question and cannot help but speculate that it is mainly due to historical developments. Maybe, in times when research was only a dialogue between a handful of people in each field, rather than hundreds to thousands of researchers worldwide, as it is nowadays, strongly tying concepts and findings to people seemed like a good idea.

Indeed, referring to papers by authors is quite impractical in various ways:

  • Often, the same group of authors publishes several papers on different stages or aspects of the same topic. Thus, the author names are always the same and distinguishing the papers based on their author names is impossible. The titles usually serve better to highlight which parts of the work are described in the respective paper.
  • Another usual occurrence is for the same group of authors (e.g. from the same lab) to produce joint papers on several different topics. Once again, using author names to distinguish the works is pointless. In fact, it could actively contribute to misunderstandings, as readers may know the respective group of researchers, but they might think of a different project by the same researchers than what the paper author is referring to.
  • Projects and concepts tend to be gradually extended by different groups of authors. Such papers on the different extensions are usually connected by their titles, but not by the names of their authors. Just compare how many papers building upon treemaps mention treemap in their title (along with some specification of the novel extension of the concept), and how many mention Shneiderman in their title.
  • Titles are chosen, whereas names are not. Sane authors in a given field will not intentionally try to pick a paper title that is the same as another paper title very closely related to their work. On the other hand, it is absolutely possible that several authors in the same subfield share a common surname. (I have more than once read papers where I had no idea (based on just the name) which (w.l.o.g.) Mr. Chen and which Ms. Li of the various possible ones the paper was talking about.)
  • In general, of course, an author name can only carry information if the reader has heard about the person's research before. A (well-chosen) title carries information in itself, and thus lowers the hurdle to understanding the text.

So, the question remains why many bibliography formats nowadays still put authors into the foreground. I can see several possible reasons:

  • Despite the aforementioned drawbacks, the benefit of changing (e.g. the formatting of established publication channels) is not worth the perceived risk of abandoning traditions. Established venues might also simply fear unnecessarily creating some chaos until authors have adapted to and become used to the change, whereas new venues may be more likely to replicate established practices as closely as possible in order to appear "serious", comparable to the established venues.
  • While they carry less information, names are usually shorter than titles. Many titles can be shortened to a few words, but not in a systematical way that would work equally well for all titles.
  • Bibliography entries are often not referenced by full names, but sometimes by abbreviations of author names. While titles are more meaningful than author names, the first letter of each word in the title (i.e. including articles and such) is rather less meaningful than the first letter of each author name. Furthermore, at least in fields with contribution-based author ordering, using the first letters of the first few author names is guaranteed to also include the "main" authors, whereas the first few words of the title are not guaranteed to be more important than the last ones.
  • I like this answer and its alternate viewpoint in addition to Captain Cranium's; it's unfortunate that choice of "official" answer can't be split between multiple ones.
    – JAB
    Apr 19, 2016 at 22:23
  • To put it another way, you could cite 'What Is An Author?', but that would not give me much of a clue as to the thrust of the article (which might well be the opposite of what I first assume). If you refer in-text to 'Foucault (1969)', however, then you have located the piece in a way that helps me to remember its relationship to 'Barthes (1967)' -- a piece that happens to be entitled 'The Death of the Author'. The point seems to be to create a universe or continuum of identifiable viewpoints (not just ideas) and thus position oneself in the dialogue, to help the reader make sense of it all. Apr 20, 2016 at 10:49
  • @CaptainCranium: But then, the mention of Barthes (1967) and Foucault (1969) might just as well refer to two unrelated works, which just happened to have been published in close succession. If you do not happen to know already what the respective authors wrote about in the stated years, your guess is as good as mine. In contrast, if The Death of the Author (1967) and What Is An Author? (1969) are cited, all readers - including those who did not know one or both of the referenced works beforehand - can immediately see the connection and thus make sense of the statement. Apr 20, 2016 at 20:26
  • @O.R.Mapper Apologies for my huge delay in responding with an important subpoint. I still feel that participation in an academic discourse is primarily a matter of locating one's present view in the existing debate. By far my preferred supporting convention has two parts. First, brief references with titles and specific page numbers in numbered footnotes, which also gives an opportunity to include a note if that might help some less informed readers to stay there on the page. Second is an alphabetical gathering of all references, with full publication information, at the end of the piece. Nov 15, 2016 at 17:23

In mathematics, in each of the several general areas of interest to me, the possible titles are so poor in information-content that they are nearly worthless, while the number of authors making the most meaningful contributions is small-enough so that author+year (with possible suffix "a,b,...") is vastly more effective than giving titles.

Coming at it from the opposite end, "new" authors who've written relatively few things are still far better distinguished by name+year (in terms of information content with the goal of distinguishing them, etc) than titles (once one knows the general field).

Thus, operationally, I look at author first. If recognizable, then I look at year. If author is not recognizable, then I look at the title as guide for general subject area, then year (for an idea of where-in-the-progression it is).

(Also, the bibliographic style of [number] is really not helpful in reading, since the number itself has essentially zero information content...)

  • I find that the number style works fine when using rich text documents that include hyperlinks between citations in text and the full references at the end, such as those generated by hyperref. Of course, we're not yet at the point where hard copies offer the same functionality, so that only really works for electronic versions.
    – JAB
    Apr 19, 2016 at 22:29
  • 1
    @JAB, oh, indeed, rich-text possibilities are another thing entirely! Apr 19, 2016 at 22:32
  • I assume that the [number] format would also be potentially confusing in mathematics papers, as it looks a little too much like maths.
    – TRiG
    Sep 28, 2018 at 15:55

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .