Nota Bene

by Kieran Healy on May 7, 2007

Via Andrew Gelman, a post by Aaron Haspel about the evils of poorly-done endnotes, and endnotes in general. This is something John has written about before, too. Endnotes really are a problem in scholarly books. In general, footnotes are better. Both are better than author-in-text citations (Healy 2006).

Haspel also arues that

2. Each endnote page should be headed by the page numbers of the notes it contains, to facilitate easy flipping. … 3. Notes should not be numbered. Numbers tax the reader needlessly, especially when they reach three figures. They should be marked by a symbol in the text … It would be especially helpful to use two symbols, to distinguish substantive comments from simple citations … 4. The notes must be indexed. … 5. The text should contain as little scholarly detritus as possible.

I agree with 2, disagree with 3 and also endorse 4 and 5. I used endnotes my own book, but did some work to keep the system friendly, especially for backreferences to works cited earlier in the text. The insane scholastic and legal conventions of ibid, idem and loc. and op. cit are especially to be avoided. I used a system where any work cited within a chapter was given a full reference in the notes, and then an abbreviated reference for any subsequent citations in that chapter (e.g., Quiggin, I Hate Endnotes).

I also did my bit to revive an older tradition, the analytic table of contents. Each chapter has a little paragraph summary on the Contents page, so that a casual reader can get a sense of the entire argument right away. Some libraries—like the Library of Congress—record tables of contents in their catalogs, so someone doing a search can get that information as well. I’d like to see more of this in other books.

{ 1 trackback }

Outside The Beltway | OTB
05.07.07 at 2:15 pm



chiasmus 05.07.07 at 2:10 am

All of these are excellent principles, and I agree wholeheartedly. Down with endnotes! Death to author-in-text citation! But what I really need are some tips on getting nice footnote citations with BibTeX–I’ve played around with jurabib, but can’t get it to produce anything nice looking.


A. G. Rud 05.07.07 at 2:26 am

I liked the style used in _The Courtier and the Heretic_, where a page number is given and a brief quote of the text under discussion, then any discussion. At least that is what I remember, and though the book is downstairs, I will not get up to crack it open to verify my memory. I very much like number 1, and find the practice of putting endnotes in by chapter title or number very aggravating. I guess I prefer footnotes too, and yes, the in text citation used mostly in APA but Chicago too (Spinoza, 1677; H. Thompson, 1975; T. Leary, 1968; G. Steinem, 1975) is so jerky and ugly. Alas, I have gotten so used to it in social science texts that my eyes just glaze over.


Neil 05.07.07 at 2:33 am

What’s wrong with author, date? When the point is simply to refer to someone’s work (rather than to make an aside) it works perfectly well while cutting down on clutter. It’s especially suitable for specialist work, where there’s an expectation that the reader will have at least a passing familiarity with most of the names mentioned. Then the bottom of the page can be used for substantive footnotes alone.


Richard 05.07.07 at 2:57 am

I love the analytic table of contents, and I really cannot think why it ever went away (except that it’s more work for the author & publisher). It’s a huge pleasure to read those massive Victorian travelogues partly because of this feature. Bravo for reviving it.

On notes, IMHO, footnotes, or even sidebars, are the way of righteousness: I hate using a book the way you have to use it if you actually pay attention to the endnotes; I can think of few things more disruptive to the act of reading. And really, is it so useful to have all the notes together in one place – if you have to know the page number of the citation anyway?


Kieran Healy 05.07.07 at 3:41 am

chiasmus, I used jurabib to format my book’s notes, but it entailed some hacking. The latest and greatest thing in this field — which I haven’t used in any serious way — is BibLatex, by Philipp Lehman, which looks like the ground-up redesign that latex bibliography styles have needed for a while.


Kieran Healy 05.07.07 at 3:43 am

What’s wrong with author, date?

I tend to think for scholarly articles, not very much. For many kinds of books it’s a harder sell, though still defensible. But it’s very easy for it to spiral into unreadability.


Kieran Healy 05.07.07 at 3:52 am

Here’s a page with some citation style files for biblatex that implement author/title footnote citations.


Peter Hollo 05.07.07 at 4:37 am

I like [Author YYYY] myself – it’s compact and unobtrusive. Works for me in non-academic books like Dennett, and I was very annoyed when he went for a different system in Breaking the Spell.

And any actual notes on the text should be numbered footnotes on the pages themselves. Numbered endnotes, where at least you can keep up because you can see the little superscript numbers in the text, are allowable but annoying; unnumbered notes not even mentioned in the text are insanely annoying. It’s what Douglas Hofstadter’s doing in his new book, and I could kill him – especially when some of his notes are discursive anecdotes or elaborations of stuff in the main body, and others are just citations. Madness!


Peter Hollo 05.07.07 at 4:43 am

Alternatively, I’ve often seen “scholarly” books which use endnotes solely for citations, thus keeping author/date out of the flow, but making them eminently ignorable if you don’t want to look up the source.
Footnotes are then marked with asterisk, paragraph mark etc – hopefully encouraging the author not to go overboard with them; one or two per page is plenty after all…


Brandon del Pozo 05.07.07 at 5:43 am

Author and date makes little sense, IMNSHO.

Okay, so you know an author, and, of all things, a year (aside from referencing author’s mulitple works, it seems to promote a strange bias that more recent years are “better” cites), and you get to know it right there in the middle of the text you are supposedly reading.

You still, however, don’t know the title or the gist of the work cited, its format, where to find it, or where it was published. If a statement piques your interest, a small, unobtrusive number should send you directly to the bottom of the page, where you can read the statement’s citation. If not, then you read on, uninterrupted.

Author, year citations are the flotsam of the APA.


tom hurka 05.07.07 at 5:59 am

Worst is when the endnotes use the author-date system, so that having gone to the back pages to get the note you have then to go to the Bibliography to see which of X’s works ‘X 1997′ is.

One justification publishers had for preferring endnotes, I think, is that footnotes were harder and more expensive to produce. (Technology has changed that; I insisted on footnotes for my last book and it wasn’t a problem.) But another is that text with massive numbers of footnotes at the bottom of the page is ugly. So if you’re going to use footnotes (and agreed they’re better), keep their numbers down. Lots of them aren’t necessary.

Over ten years as a journal editor I saw the number of footnotes in submitted articles climb dramatically; one revision I often asked authors to make was to cut the number of notes by, say, half.

And watch out for Analytical Tables of Contents: I’m reading scholarly work on a Victorian philosopher which uses bits of his Analytical Table as evidence for this or that interpretation.


conchis 05.07.07 at 10:44 am

My ideal solution, which would only work on a computer, would be: (Author, Date), but with a fully expanded cite that appears if you run a mouse over it. Captures the primary benefits of author-date (a minimal amount of info that doesn’t require you to look anywhere else, and which may be sufficient if you already know the reference) and footnotes (minimal amount of effort to get the rest of the info if you want it).

Endnotes are just insane: even if large numbers of footnotes are ugly, that’s nothing compared to the inconvenience of flicking back and forth to look up endnotes.


dw 05.07.07 at 11:32 am

The most annoying endnotes are those headed “Notes to Chapter Six”, etc — where the number of the chapter doesn’t appear at the top of every page! In order to find the content of such an endnote, I have to:

Go to the table of contents to figure out the number of the chapter I’m currently reading
Go to the end of the book to find the notes for that chapter
Find the exact endnote I’m looking for

I’d like to strangle the person who invented this system!


Barry 05.07.07 at 12:18 pm

Kieran, anything which makes scholarly work more readable is a work of grace. I keep hitting sentences which have more references than ideas, and just can’t continue. It’s like a road which consists of 90% craters.


bza 05.07.07 at 12:23 pm

The author-date style can look absurd when dealing with historical material. I cringe whenever I see a citation like “Kant 1781″. Or even worse, when authors use the year of modern republication, leading to citations like “Hume 1963″.


Tim O'Keefe 05.07.07 at 1:54 pm

Re: comment 15. Yeah, people need to use the standard scholarly reference system (if one exists) when dealing with historical figures like Kant, not author (date)–especially as their works come in lots of different translations and editions. So Euthyphro 10a, not Plato (1997) 9, which (unfortunately) I’ve seen. But doing so still allows you to use author (date) when referring to books by Daniel Dennett and the like.


joseph hill 05.07.07 at 2:10 pm

all notes are for the insecure…do not use them. Ever.


nick s 05.07.07 at 4:53 pm

what I really need are some tips on getting nice footnote citations with BibTeX

I managed to wangle something around jurabib that worked pretty well: it also allowed for old-style footnote numbers and new-style footnote markers. It does involve a bit of Plain TeX for redefinitions, though: drop me a line [nick at nonspace dot org] and I’ll send you the .tex file I worked from.

[Kieran: what was the general style package you used for no-nonsense MS-submission LaTeXing? My search of the archives is failing me.]

For my thesis, I used a long-form citation for the first reference, and a short-form for future references. I’m more confined by house style right now, but I do like the idea of separating citations from ‘asides’.

That said, the trend in EngLit monographs appears to be towards demoting as much lit-review to the back-matter as possible, especially w/r/t contemporary readings: you use endnotes to cite modern scholars X, Y and Z on author A or topic B, while devoting a few sentences to why they’re [worthy but flawed | clueless hacks | taking the piss].

As for author-date, it’s generally worse than useless in the humanities when dealing with long-form works, or for primary sources that are non-discrete (i.e. existing in a number of editions or forms) or more than a hundred years old. Best practice, it seems, is to state up front which editions you’ll be working from, and then use standard scholarly references: lines for poems, act/scene for plays, chapter/section for philosophical texts.

Having proofed my wife’s dissertation, I feel increasingly certain that APA style is a form of psychological torture.


stm 05.07.07 at 7:19 pm

Yeah, endnotes are awful. They must add an extra 15% or so to my reading time for works that I want to pay attention to the citations for.

One virtue of the author-in-text citation is that it makes a bibliography obligatory. Yes, they are clunkier to read through and harder to ignore than footnote numbers. And, yes, when poorly deployed, they’ll give you cites like (Descartes 2005). But they provide the gist of the citations without necessitating eye-jumps to the bottom of the page (or god forbid to the back of the book!), and for the reader who wants to track down references, they provide a handy bibliography.

I’m a social scientist who also reads a lot of history, as well as a decent amount of philosophy and literary theory, so I regularly work from author-in-text citations and footnotes. I find that the former, while imperfect, provide a better balance of unobtrusieness and ease of use.


eszter 05.07.07 at 8:40 pm

[Tried to comment earlier as #11, but my PDA lost the note.]

I like Author Year in text, I find it very helpful. It is true that if there are too many in one sentence and inserted in several locations within one sentence that can make text hard to follow. But that is rare. Otherwise, I find that format helpful, because often if you’re reading something in a field with which you are familiar then simply glancing at the Author Year tells you that you already know the citation and don’t have to go searching for it. It’s annoying to search for a citation only to find that you already know that particular paper/book.


Another Damned Medievalist 05.08.07 at 1:32 am

MLA and APA are incredibly lame. Footnotes rather than endnotes, preferably in Chicago/Turabian or similar. Why not author, date? One, I find parenthetical references distracting — if they only exist for academic honesty, there’s no reason for them to distract me. Historians and others use footnotes (and evil endnotes) because we actually engage with our sources — and many of our sources are not secondary. Passages quoted in original languages, or translations, go in the notes. Tangential but important references to scholarly debate go in the notes. Entire historiographies of minute points can go in the notes. This makes it possible to write something fluid and sensible without sacrificing scholarly rigour. Footnotes rule.


sara 05.08.07 at 2:31 am

Have you already written a post about evil indexes?
Those useless that only collect capitalized proper names are worse than useless in a book about (e.g. — I won’t name it, as I am on good terms with the author) social history.


sara 05.08.07 at 2:32 am

“Those useless indexes,” sorry.


Alan Bostick 05.08.07 at 3:08 am

I assume, Kieran, that your disagreement with Haspel’s point 3 is strictly about numbering, and not about distinguishing between citation from commentary.

I hate, hate, hate reading works that mix commentary and citation in endnotes. I hate having to maintain a separate bookmark for the notes and flipping back and forth just to see whether or not the author is actually saying something or just citing a reference. My sense of how things ought to be is that citations should go in the back and commentary at the bottom of the page.


josh 05.08.07 at 4:09 am

I’m generally, but not wholly, opposed to endnotes. They should be used, I think, only for very long notes. Some people are of course against long notes entirely; I disagree (and my practice, I fear, reflects this). Sometimes one has a great deal to add to a statement, but putting it in the text will break up the flow of the argument; yet one DOES want to say it, and somehow link it to or anchor it in the part of the text to which it’s connected. In that case, long end-notes seem to me an ideal device.
For this reason, I think the ideal would be to have BOTH footnotes — for citations — AND endnotes, for commentary/excursions that can’t be elegantly worked into the text itself. The second best solution is to go with footnotes, and try to keep commentary w/in the notes fairly short — while I think mixing citation and commentary is fine, I don’t think that it’s fine for footnotes to take up a quarter of the page or more.
One thing that should not be done is to mix footnotes and in-text citations willy-nilly (one might mix them in the same way that I’ve proposed mixing footnotes and end-notes — but I think that’s a bad idea). Unfortunately, this is precisely what was imposed on one of my own published pieces (On the other hand, the editors of the volume in question adopted the policy of referring to all works in the in-text citations by their initial publication dates — so no cases of ‘Hume 1963′, thank god).


magistra 05.08.07 at 6:40 am

Sara – evil indexing is common because subject indexing is *hard* and in a cost-benefit analysis possibly not worth it. I’ve just indexed a book (not my own) for the first time and it was seriously difficult work. It took me 50-60 hours of work to index a collection of historical essays of around 300 pages. (Whether or not the result is good is not for me to say, but I did index the endnotes). The problem is that I’ve made all that effort, working to a fairly tight deadline, to produce something that most readers and book reviewers will then take for granted. I could probably have written most of a new article in that time: which is the better move?


Eszter 05.08.07 at 4:59 pm

Oh yes, I forgot to add that I also hate it when citations and substantive comments get mixed up in either footnotes or endnotes. Maybe that’s another reason I like the Author Year style. That way you know that a footnote/endnote is not simply a citation.


mollymooly 05.09.07 at 5:14 pm

Publishers favour endnotes for books where most readers are never going to bother reading the notes. In that case, minimizing the clutter for the lazy majority outweighs the inconvenience for the motivated minority. Of course, as Al Franken points out wrt Ann Coulter, some authors count on readers’ not bothering to check the endnotes.


Henry (not the famous one) 05.10.07 at 9:00 pm

As for footnotes: I commend to all the first footnote in the article “The Common Law Origins of the Infield Fly Rule,” published in the University of Pennsylvania Law Review in the late 1970s. The rest of the article–and the footnotes–are pretty good too.

Comments on this entry are closed.