Endnotes, again

by John Quiggin on July 28, 2009

I really, really hate endnotes. But now that I am writing a book I have to decide whether I have to swallow my pride and use them, and if not, what alternative to adopt.

To start with, I want to distinguish between explanatory notes, spelling out a point that is marginal to the main text and references giving authority for some claim made in the text, or examples or a person making a claim that I may endorse or criticise. In academic work, I’m used to the Harvard format where explanatory notes are placed as footnotes, and references cited in the body of the text as “Quiggin (2009)”, then listed in full at the end. This is much better than the all-footnotes system used, for example, in legal writing.

For a popular book on a technical subject like “Zombie economics”, there are a few options, which can be mashed up in various ways.

  • The standard endnotes setup with explanatory notes and references listed at the end of the book
  • Footnotes for explanation only: this leaves open the question of how to deal with references
  • A further reading section at the end of each chapter, in place of references
  • A book without references, but with an online hypertext version in which readers who want to chase references can find them.

Any thoughts?



Peter Hollo 07.28.09 at 9:07 am

Missing a quotation mark at the end of your first a href, John. Which is rather stuffing up the rest of the post :)


Peter Hollo 07.28.09 at 9:08 am

Read this on the RSS feed though, and yeah, I hate endnotes and hate reading books with endnotes. I like footnotes for explanatory stuff and Harvard-style references. Clear and easy to use.


Justin 07.28.09 at 9:46 am

You can also see the text that Quiggin meant to present if you look at the source code for this page. That’s option-command-U in safari, and similar commands in other browsers.


Justin 07.28.09 at 9:49 am

Substantively, I’ve seen some books use the equivalent of sidebars to display expository notes or background information. You have to think pretty carefully about what information goes in those boxes to make the presentation work, but it’s an option worth considering.


Katherine 07.28.09 at 10:26 am

I too hate endnotes – the endless flicking backwards and forwards is really annoying. I don’t mind copious footnotes (although some legal articles, especially American ones, get silly). I think I agree with what Peter Holbo said though – it seems the best compromise between explanation and readability.


Phil 07.28.09 at 10:54 am

For my book – which includes sizeable chunks of translated text – I used Harvard referencing, chapter endnotes for the original text quoted (for anyone who wanted to check my translation), & footnotes for explanatory notes. Then the publisher said they didn’t like footnotes, so I absorbed all those notes into the main text. This meant I couldn’t get as De Selbyesque as I would have liked, but that’s probably all to the good.


Peter Hollo 07.28.09 at 11:11 am

Hey, much though I am honoured to be associated with John H, my surname is one that would be more familiar to Eszter… No “b” :)


conchis 07.28.09 at 11:13 am

Endnotes bad. Harvard good. Online hypertext excellent (though I wouldn’t remove refs from the physical book).

Is there any particular reason you can’t just use Harvard?


ajay 07.28.09 at 11:15 am

Antony Beevor uses neither footnotes nor superscript numbers in the body of the text – quotes and facts are given without attribution. If you go to the back of the book, there are pages of endnotes in the format

154: “Would the fall never…” Lewis Carroll, “Alice in Wonderland”, p. 72.

and the full bibliographic details are given in the bibliography.
This seems a fairly good way to go for a popular book: you don’t break up the flow of the text, but if you want to find out where a reference comes from, you can.


J. Otto Pohl 07.28.09 at 11:16 am

It has been my experience that the publisher not the writer dictates what form the citations will take. Generally, they prefer endnotes to footnotes. I just finished proofing an academic journal article which required endnotes and it was quite a pain flipping back and forth to verify them. My guess is that like Phil states that there is a good chance the publisher will force you to use endnotes rather than footnotes. I have written one non-academic book and for that I went with a modified version of option three. I had a single “furhter reading section” at the end of the book. Not having to write endnotes made it a much much easier book to write than my first two.


Salient 07.28.09 at 11:45 am

Insofar as you do have the choice (I don’t know how correct J. Otto Pohl is about this), I agree that the Harvard format yields a much more readable text. I appreciate the inline citations (though for some books, when I know it’ll be desirable to go read the cited work and then reread the book’s claims about that work, I immediately flip to the references and write in the page number where the reference was cited — the result is a reference index, the categorical opposite of endnotes, where the reference in the back tells me where to go find the citation in the text).

Also, whatever happened to the idea that a writer could put together a website where readers may find additional references and explanatory notes? Best anti-endnotes idea to date, I think.


tom s. 07.28.09 at 11:57 am

I don’t mind endnotes, but please please please avoid the stupid stupid stupid practice of having the chapter title at the top of each page of the main text, and the endnotes marked by chapter number. I am sick of looking back through the pages, one finger keeping my current position and another keeping my place in the endnotes, looking for the chapter number of the chapter I’m reading. I can’t understand why any publisher would do this, but it’s not uncommon. And it’s really stupid.


dr 07.28.09 at 12:16 pm

For popular books, I like the system ajay proposed.

But I’m curious about the term ‘harvard format’. Is that really what this style is called? I have a vague recollection of it being referred to as ‘social science style’ or ‘APA style.’


Phil 07.28.09 at 12:28 pm

Tom – agreed. A running head of “References for pages nnn-nnn” is ideal, but sadly rare.


Nick 07.28.09 at 12:38 pm

I recently ran across a book (title escapes me) that used the horrible system that Tom S. describes, and then it added insult to injury by not including the complete citation in the endnote. The endnote, once I located it, would say something like Doe, J (2009). Then I’d have to flip to yet another section to get the complete reference.


Another Damned Medievalist 07.28.09 at 12:51 pm

Footnotes. Chicago style or sim. Love them. There is no substitute. I find endnotes annoying, especially as there is no good reason for them now that typesetting is done on computers. I hate in-body parenthetical references, because I find them distracting. Text is text, references and notes should be separate, but easily accessible, i.e., at the bottom of the damned page!


Donald A. Coffin 07.28.09 at 12:53 pm

Certainly one reason for the (publishers’) preference for endnotes was that it made typesetting easier. There’s no longer any excuse for it, given that (almost) everything now goes to press from computer-prepared text, in which the use of footnotes becomes essentially no work at all for the publisher.

I also strongly prefer the inline citations, with complete references in a bibliographt, with footnotes used for tangential explications, asides, authorial wanderings, etc.


jacob 07.28.09 at 1:07 pm

Not surprisingly (since I’m a historian) I agree with Another Damned Medievalist. I find in-text citations clunky and distracting. In history, they’re impossible, since it’s hard to cite archival material in text. Endnotes are a pain, as everyone here has already noted–and I wish to echo the people who object to footnotes with any header other than “notes for pages xxx-yyy.”

A side benefit to footnotes is that it will discourage you from having too many discursive notes, which are really not very good practice. I’ve always felt that the material should either be in the main text–because it’s important–or not be there at all. Which is not to say that my work doesn’t often have such notes, only that I recognize they’re indulgent and a bad idea. I’m currently reading a book with end notes, and nearly every one has a paragraph worth of extraneous material. I’m constantly flipping back and forth and spending a fair amount of time in the endnotes. It’s like reading a David Foster Wallace novel! It takes discipline as an author, but things that aren’t relevant enough to be in the main text really ought just to be cut.


jacob 07.28.09 at 1:08 pm

Sorry: the last sentence of my first paragraph should read “…the people who object to endnotes with any other header…”


Brock 07.28.09 at 1:12 pm

Footnotes for explanatory stuff, Harvard-style references. I hate endnotes.


The Raven 07.28.09 at 1:13 pm

I go along with the Medievalist; I like the Chicago footnotes with short references, and a full bibliography at the end of the book. These have the additional advantage of being much easier to follow when reading the document as a PDF.


apthorp 07.28.09 at 1:18 pm

notes: in footnotes. If really long: 1. are you sure it isn’t a section in the text? 2. inset box, even if it takes up the whole page. Notes are really part of the text for careful reading. In an online version something like an inline drop down would be appropriate.

citations: if not there you are clearly writing a novel or self-help drivel. I like the superscript numbers, particularly if there are a lot. Citations are a list each of which will be used separately to do the look up and therefore not in the text flow.


other_jacob 07.28.09 at 1:22 pm

I vote for sidenotes! You have nice wide margins, right?


Barry 07.28.09 at 1:33 pm

I second Another Damned Medievalist. To me, the classic sin of academic writing is putting long cites in the test. It breaks the flow of reading every other minute.

It’s like listening to somebody try to tell a story, when they keep interrupting themselves every other sentence.


deliasmith 07.28.09 at 1:39 pm

dr @ 13:

The author–date system is very often (mis)named the Harvard system.

The Harvard system is one typographical variant of author–date; the Ivy League name prevails, no doubt, because of academic authors’ fondness for the recondite, reflexive status-seeking and craven submission to blood-soaked US imperialism. These same forces are responsible for the widespread misuse of the word ‘font’ when ‘typeface’ is what is meant. (For a more general, measured view, see George Orwell on ‘antirrhinum vs ‘Snapdragon’.)

When you reach the production stage, I suggest that you start your dealings with the publisher’s copy editor by explaining your system, whatever it finally turns out to be; inviting him or her to comment; and encouraging him or her to point out flaws in the system and errors, if any, in its application.

And authors everywhere: read Butcher’s Copy-editing: The Cambridge Handbook for Editors, Copy-editors and Proofreaders 4th Edition


richard 07.28.09 at 1:51 pm

I’m in the pro-footnotes, anti in-line citations group for print, because of disciplinary boundaries/habits. So it depends on the disciplines you most want to address, I guess.

BUT: for all etext I support endnotes or, much better, links/tooltip hovers, partly as a matter of style and legibility, but mostly because (machine-readable) etext can be accessible to sight-impaired readers, who can listen to it. Unless you have a really smart reader, and format for it, footnotes are the worst violence you can do to a text for listeners: breaking flow, adding number strings the readers can’t parse etc. In-line citations do this too, especially if you’re in the habit of bunching them up together.

Re avoiding discursive footnotes as bad writing discipline, I think it depends what sort of book you’re writing: Goitein’s A Mediterranean Society simply would lose much of its charm and value treated this way, because it surveys a field of resources, Marx’s Kapital or Geertz’s Negara, being all argument, require nothing but references. For casual/popular reading, I really like books that aren’t afraid of an interesting digression or titbit and for this I love sidebars, but that implies something else about the kind of book you’re writing.


Tim Wilkinson 07.28.09 at 2:11 pm

Not ‘Further Reading’ – or not just that, anyway. It’s lazy, fairly useless for anyone with a specific point in mind and gives the impression (with some justification) of a lack of rigour.

The only good reason for endnotes is for comments that are too long for a sensible footnotes – in which case some sort of appendix or sidebar might be appropriate. (Or choose between incorporating into the text or dropping it, as noted above.)

Glad to see others are similarly counfounded by the practice of making endnotes insanely difficult to find – and I’d also repeat here another comment from the last ‘bookblogging’ post: please don’t use ‘ibid‘ or ‘op. cit.’ if there is any chance that it won’t be entirely and immediately clear what they refer to. It hasn’t taken up quite so much of my time and patience as the chapter number/name ludicrosity but still, having to leaf back through page after page of footnotes to find the original ‘op cit’ is a distinctly suboptimal outcome. I’d prefer not to have to look up book titles in a biblio as well – it may be a bit more long-winded but the title is much more useful than the date (and FWIW more likely to be unique).

‘Ah that’s interesting – I might look into that – now, without losing my thread, must make a mental note of the source – oh, where’s the bibliography? Dum de dum, here we are – oh hang on, was it 1974 or 1976?’


richard 07.28.09 at 2:13 pm

…a corollary to my previous comment: as a general question, do you think people should be able to interact with/receive your text however they want, or do you think it is part of the author’s rights/responsibility to prescribe how their text should be used?
Obviously there are outlying cases – art texts, poetry – where one or the other of these options may be part of the point of the presentation, but what about academic texts in general: do you abhor people listening because they should be reading, or selecting short sections, etc?


David Foster Wallace (via Salient) 07.28.09 at 2:15 pm

It’s like reading a David Foster Wallace novel!

In Infinite Jest, the endnotes are very intentional, and they’re in there for certain structural reasons… you don’t, you don’t need to hear about it. It’s sort of embarassing to read this book [A Supposedly Fun Thing], you can almost chart when the essays are written ’cause the first couple don’t have any but the footnotes get very, very addictive, I mean, it’s almost like having a second voice in your head.

[There are 200 pages of footnotes in Infinite Jest…] Yes but the reader doesn’t experience them that way, because the endnote tags are in the text. Well. There are quite a few. Some of them are very short. Some of them are only one line long. It is a way… see this is… I’m just going to look pretentious, talking about this.

There is a way it seems to me that reality is fractured right now, at least the reality that I live in. The difficulty about writing one of those — writing about that reality is that text is very linear, and it’s very unified. You, I, anyway, are constantly on the lookout for ways to fracture the text that aren’t totally disoriented. You can, you know, you can take the lines and jumble them up, and that’s nicely fractured, but nobody, nobody’s going to read it, right?

So you’ve got — there’s got to be some interplay between how difficult you make it for the reader and how seductive it is for the reader, so the reader’s going to do it. The endnotes were for me a useful compromise, although there were a lot more when I delivered the manuscript, and one of the things that the editor did for me was had me pare the endnotes down to really the absolutely essential.

* Interview, 3/27/1997, http://www.charlierose.com/view/interview/5639


Ceri B. 07.28.09 at 2:38 pm

Richard@25 pretty well covers my view, I think. I tend to feel that anything that requires a substantial explanation ought to be in the body of the text anyway, though – for me it’s a sign that the flow of the material isn’t what I the writer thought it was going to be, and my responsibility to adjust. And yes, in any event, each citation should stand on its own with enough information to get from it to the original.


Eric 07.28.09 at 2:56 pm

Endnotes are preferred by publishers because they are the least repellent to an undergraduate, cross-disciplinary, or casual readership. They are unobtrusive to the flow of the text and allow the reader to ignore the notes altogether, if desired. They also take up the least amount of space in the printed form, which should translate to an otherwise lower list price than a book with a greater degree of reference apparatus.


Jim 07.28.09 at 3:38 pm

I agree with richard about endnotes for e-text. If you’re thinking about the future, reading a book on a Kindle or other e-book device makes endnotes very easy to use.


Sean Carroll 07.28.09 at 3:54 pm

Something I don’t think has yet been brought up: if you want your popular book to approach actual “popularity,” you need to have zero or nearly-zero footnotes. People are turned off by them, as surely as by equations. Endnotes are annoying, but will do much less harm to your bottom line.


Tim Wilkinson 07.28.09 at 3:57 pm

Re internet extras – I think these should indeed be extras. Internet access is inconvenient and for some unaffordable or otherwise impracticable – and anyway, a book should be self-contained.

It may not be quite to the point inasmuch as we’re are talking about references, which to be useful would normally need to be tracked down, but the gist of the position is that I want to be able to sit on the beach with my book (and possibly a pencil stub and some post-its) and read everything the author wanted to tell me.

I suppose the question remains of what would constitute dispensable extras – perhaps a ‘writing of’ feature? Deleted chapters? Actually the latter I suppose might not be far off.


Billikin 07.28.09 at 4:03 pm

For popular texts I think that endnotes are better for references than footnotes, but it is not a big deal. Speaking just for myself, I do not like explanatory footnotes. If the footnote contains only a sentence or two, using parentheses in the main text is more readable. If the footnote contains more, it can get really annoying. I have seen pages with only a few lines of main text, and a humungous footnote in small print that I can’t decide whether to read or not. I think that the solution for explanatory material is none of the above.

These days many, if not most general readers are used to hypertext and pop-up windows. The closest analog in popular texts that I have seen is the textbox. Enclose extended explanatory material within visible boundaries, and keep the text the same size as regular text. People can skip over the textbox, and if they read the material they do not have to strain their eyes. Another approach is to have technical or explanatory sections, and tell readers that they can skip to page XXX.


alkali 07.28.09 at 4:44 pm

This is much better than the all-footnotes system used, for example, in legal writing.

Not that this matters much, but that system is used almost exclusively in academic legal writing. In the real world, lawyers mostly use inline citation.


Keith 07.28.09 at 4:48 pm

Personally, I love explanatory footnotes, the longer the better, and as a layman in this subject, would find a citations/further reading section at the end of each chapter useful.


matthew kuzma 07.28.09 at 4:48 pm

In the interest of making your book accessible to the renaissance thinkers some centuries after the inevitable apocalypse, or to those on the other side of today’s digital divide, you shouldn’t rely on an internet-based hypertext version to handle all your references. If you have a lot of detailed explanation that is worthwhile enough to put in the book but not in the main narrative, I really like the idea of an end-of-chapter further reading section. But really, it comes down to how relevant and important are those explanations? I think the spectrum of importance goes:

Work the idea into the main narrative -> footnotes -> end-of-chapter -> end-of-book -> cut them.

So how important are these explanations to you and potential readers?


theo 07.28.09 at 4:59 pm

What Tom S. said.

I don’t care what you do, but so long as you’re using endnotes, the header of all endnote pages MUST contain the page numbers referencing the notes, not just the chapter number or title or the word “Endnotes.”

The sadly frequent practice of omitting page numbers makes me want to shoot someone in the face.


Nicolas 07.28.09 at 5:28 pm

Remember Barthes’s method in Camera Lucida.


bfeinberg 07.28.09 at 6:19 pm

nobody mentioned Anthony Grafton’s book on this topic???


Craig 07.28.09 at 6:25 pm

Complete footnotes are the bread and butter of scholarship, be they references or discursive. If you intend to write your book for serious scholars, then you have no option but to use footnotes. If you’re audience is not serious scholars, then scholarly standards can most certainly slide.

In-text textual citations are inelegant and mess up the flow of the prose. The only problem with legal citations is that they use silly mechanisms like “Supra note 156 at 22,” which forces the reader to go back twenty-nine pages to find note 156. This is why Chicago is superior: “Last Name, Short Title, page” for subsequent references and a bare full citation on first mention.

The only exception to in-text citation – which shouldn’t be a problem for you – is in commentaries on texts. In this case it is appropriate to use in-text citations. For instance (TII, §4) to refer to something John Locke’s Second Treatise of Government section four.


roac 07.28.09 at 6:31 pm

Is it really the case that Infinite Jest rather than Pale Fire is now the type specimen for the category “Novel with footnotes”? Makes me feel old.


richard 07.28.09 at 6:38 pm

“Last Name, Short Title, page” for subsequent references and a bare full citation on first mention.
That’s great as long as there’s a bibliography at the end. Where there’s no bib I generally find myself grinding my teeth and noticing how many citations there are to only a few sources.


roac 07.28.09 at 6:39 pm

“Novel with endnotes,” I meant to say in 43.


Bruce Sharp 07.28.09 at 7:51 pm

For printed works, I’m a fan of the method ajay describes in #9. For HTML, there’s a really elegant Javascript/CSS version at the blog Brand Spanking New. I stole their code for a slightly simplified version, described at Mekong.net.

One of the things I like about the BSN script is that it allows you to insert the source right next to the relevant text. For me, that makes it easier to avoid mixing up references.


Kenny Easwaran 07.29.09 at 12:56 am

roac – the novel with footnotes is Jonathan Strange and Mr. Norrell


Ray Davis 07.29.09 at 3:10 am

So what’s such a bad idea about “indulgent”? A happy writer makes a happy reader, I say.

Me, I like end notes, the more digressive the better. I save them up and read them all together at the end of a chapter. It’s like discussing the same topic with two different people (sometimes surprisingly different people) in the same day. The compare and contrast makes the trip more interesting.


ArC 07.29.09 at 3:16 am

This may be too geeky, but along with footnotes (or URL-ized footnotes), how about adding QR coding the URLs (2D barcodes) on the page?


Tracy W 07.29.09 at 8:48 am

FWIW: In printed works I like footnotes. Especially if you add in some that are not just references but also cover interesting points that are secondary to the main text.


Eszter Hargittai 07.29.09 at 11:58 am

I apologize for not reading all the comments above and likely repeating what’s already been said, but since multiple similar opinions may matter here, I thought I’d add my vote: footnotes with additional info and then a references section in the back with pointers to other work. In mainstream books I’ve seen this done by chapter, that seems like a reasonable approach. I hate endnotes for substantive info and I hate footnotes for references when mixed with substantive commentary, because interrupting reading to look at a footnote when it’s simply a reference can be disruptive and thus rather annoying.


Tim Wilkinson 07.29.09 at 1:13 pm

Eszter @51: I hate footnotes for references when mixed with substantive commentary

Obvously I’m not challenging this account of personal experience and preference, but isn’t it usually clear when a footnote is a reference? It’s generally attached to some mention of an author or a work, or to a quote.

I’ve never had any trouble of the kind described at all, but then maybe I’ve adapted to deal with it – in philosophy, refs are almost never in the text, and in any case most footnotes are references (sometimes with additional remarks). That’s my impression anyway – I haven’t tried to check it. Having said that, one of the most readable (if misguided and misused) philosophy books – Anarchy State and Utopia – has a beguiling combination of asterisked, obelisked and numbered notes, asides in parenthesis and even italicised digressions spanning several pages.

On another topic, the point made in previous comments about the popular aversion to annotations is certainly worth considering, though I wouldn’t be surprised if it were a conventional myth, perhaps born of snobbery, of the kind familiar enough in periodical publishing.

As the widespread inability even to make endnotes easy to find suggests, the orthodox opinions of publishers on how to organise books isn’t necessarily to be taken very seriously. Just how many people, I wonder, have decided against purchasing, reading or recommending a book solely because it had footnotes? And what evidence do publishers have about the matter? I can’t believe it’s that much of a problem.


roac 07.29.09 at 1:42 pm

the novel with footnotes is Jonathan Strange and Mr. Norrell

Indeed. And wonderful footnotes they are, too. (Yet if you look at the reviews on Amazon, for example, you will find plenty of people who bitch about them.)


Danny 07.30.09 at 4:27 pm

I’m with Tom Wilkinson on publishers’ reluctance to publish books with footnotes. I asked a fairly prominent editor about just this question – is there evidence for the idea that the general public doesn’t buy books with footnotes? His response was basically, “They just don’t like books with footnotes.” So, no, it doesn’t seem as if publishers actually have any evidence for the idea that the general reading public runs away screaming at the sight of some little text at the bottom of the page.

My suspicion is that readers who are scared by footnotes wouldn’t be that interested in the sort of book that would have footnotes in the first place. What’s really inexcusable are books whose audience is obviously limited to scholars that have endnotes.


Eszter Hargittai 08.01.09 at 1:09 pm

Obvously I’m not challenging this account of personal experience and preference, but isn’t it usually clear when a footnote is a reference? It’s generally attached to some mention of an author or a work, or to a quote.

In my case it hasn’t been clear. If it was clear then I wouldn’t find it such an inconvenience. (For one thing, it is completely plausible that one might have additional comments re a section of the paper that also needs citation.)


Tim Wilkinson 08.01.09 at 1:36 pm

yeah, sorry, that came across as faintly critical – and it wasn’t particularly well-thought, as I recognised just after clicking ‘submit’.

For mine own part, with footnotes I’m pretty sure I tend to glance down (or maybe even use peripheral vision) and check that it’s just a reference without interrupting the flow much – though that only works when there aren’t more than one (or maybe two) footnotes on the page.

It certainly wouldn’t work with endnotes of course – so as between the two types of note, it counts in favour of the foot variety I suppose.


Rafal Urbaniak 08.03.09 at 3:21 pm

Depends on what word-processing environment you’re used to. If you’re okay with learning some code and ready to give up wysiwyg edition, I’d recommend moving to LaTeX. It comes with its own bibliography management (you can use things like JabRef to manage your references, I think importing whatever database you might have should be quite painless). Then, having APA-style quotations [like: (Quine 1945), (Quine 1963: 23), (see Quine 1945)] with a full reference list at the end of the book, and having a separate sequence of explanatory footnotes is quite easy.

Comments on this entry are closed.