8 ms·

Usability of Footnotes

56 points by edent 6y ago

MengerSponge 6y ago
Tufte's footnotes are my favorite. They're especially brilliant when you have typeset documents on 8.5x11" paper, because you wind up with huge margins if you're using a humane line length for your main text. https://www.overleaf.com/articles/implementing-the-teachings-of-edward-tufte/hvtnngcxqnwm https://www.overleaf.com/articles/implementing-the-teachings...
- User23 6y ago
  Tufte is underappreciated for sure. He's often aped, but rarely do I see his ideas actually applied as he teaches them. The whole point is a theory and practice of communication. Before he gave up and decided to just focus on sculpture he was almost to the point of actually developing a solid theory of representing 5-6 dimensional information representation in 2space. Of course I understand why he gave up, because 3space is way more interesting when you've solved the 2space problem to your own satisfaction.
quadrifoliate 6y ago
From the article: > But I want to talk about the cognitive issues at play. When footnotes can contain anything from a citation to essential background information, the user is forced to break from the flow8 of reading – often in the middle of a sentence. This is a problem in both paper books and electronic books. I think this is...the idea of footnotes? That you don't break the flow and it's safe to skip over them, but if you're coming back for a leisurely read later, you can look them up and see what you missed. [1] When you go over the text a second time, you can follow the references to the footnotes and see the interesting crumbs that the author laid out, which often lead to a train of thought separate from the main flow of words. ------------ [1] At least, this is how I have used footnotes for a while when reading, and it works quite well. YMMV, since people read and comprehend differently.
- edent 6y ago
  I think my confusion is about how to read the footnotes which are more than just citations. If I read them "inline" then I don't understand why they're not part of the main text. If I wait until the end of the chapter, I've lost context. I'm prepared to believe that I haven't been taught to read in the right way. But there seems to be some disagreement on what the right way is.
- ringshall 6y ago
  Same idea with parentheses, at least I understand them.
hardwaregeek 6y ago
I find footnotes annoying even in print books. When I'm reading a book, I have to figure out if the book is the type that uses footnotes as citations or the type that uses it as actual content. Or worse, the type that does both. If it's the type that uses them as citations, then meh, I can skip over them. If it's the type that uses them as content, well ugh now I need to read this footnote and then get back to the original text, which is a context switch. Even worse if the footnotes are actually endnotes because then I need to physically flip the pages. Even worse if the endnotes are gigantic and have subnotes (cough Infinite Jest). If it's the type that does both then I have to look at every damn footnote only to be disappointed by a "Pfaffenberger, 2016" 90% of the time. I don't have a good answer to these problems. Tufte style is better because I don't have to move my eyes that much. But there's only so much one can put in a note in the margin. And it still distracts.
- User23 6y ago
  I like endnotes best. For a book where they are actually useful I either unbind or buy two copies so I can lay out the endnotes side by side with the book. Think of it like having two monitors where one is the working screen and the other is the reference screen.
  - JadeNB 6y ago
    Wouldn't it be more useful not to have to do this? Keeping them footnotes saves buying another copy, or damaging yours.
    - User23 6y ago
      Think of it as the difference between having two monitors and having one monitor where the bottom 1/8th to 1/2 randomly gets hijacked. Two monitors are also more expensive. My method has its trade-offs as you note, but I find having an uninterrupted flow through the main text while having access to the notes on an as desired basis works best for my enjoyment and comprehension. Without my method endnotes are so frustrating I almost never use them, and about half the time when I think ok this time it's worth it to context switch, it turns out not to be. This all comes down to Tufte's idea of letting the optic nerve do the work. A glance requires almost no conscious brainpower, this is why thumbing through a book you've read works so well. Thus I can glance over at the footnotes pages and decide if I care to read the note without even slightly interrupting my flow reading the main text.
just-ok 6y ago
Footnotes are great when done right. If you're using footnotes to cite things, you're doing it wrong: inline citations exist. If that's not enough for you, use a different syntax (say, ^[1] instead of ^1). If you're using it to define terms, you're doing it wrong: use a margin note. If that's not enough, you should probably typeset a nice box that makes definitions easy to identify. Footnotes (imo, of course) should only be used as a "here's more information for the curious" and should similarly only appear at the actual foot of a page (I despise endnotes). My experience is biased towards print and print-like (PDF) use of footnotes, but I don't think it's much of a problem to expect users to click every footnote when provided with a convenient pop-up UI like in the OP, provided that you've clearly defined a convention or pattern of their usage to your readers. I do wish websites showed footnotes on hover, though... jumping within the page is nonsense and often mucks with URL sharing.
- barryvan 6y ago
  Footnotes _are_ great when done right, but I have to say that inline citations are incredibly frustrating as a reader, because they're harder to skip over. A footnote marker is small enough that your eye can continue over it, whereas a citation like (Just-OK 2020) is much harder to read. My personal preference is to use footnotes as you suggest -- as a "here's more information for the curious" option -- whereas citations should be linked to a "references" section -- either at the end of the chapter or the book or the overall article.
  - abnercoimbre 6y ago
    > [...] a "references" section -- either at the end of the chapter or the book or the overall article. So footnotes retain the numbering system (1, 2, 3, etc.) throughout the chapter text, but how would you mark sentences for whom a reference is available?
    - barryvan 6y ago
      I'd probably use numbering for the references, and use the array of symbols* available for footnotes -- given that most pages shouldn't have more than a handful of footnotes, and you can restart your symbol order on each page. That's more difficult on longer-form web content, admittedly. * See https://en.wikipedia.org/wiki/Note_(typography)#Numbering_and_symbols https://en.wikipedia.org/wiki/Note_(typography)#Numbering_an... for a sampling. These can be doubled if necessary -- so , ††, etc.
sixstringtheory 6y ago
Before even clicking through to the article I was thinking about Infinite Jest. > As an author, I want to add more details which don’t fit in the main body of the text. Because I hate deleting text and my editor told me to take it out. I joked about IJ doing that exact thing, that the editor was all “no, you can’t just list nonsensical film titles for four straight pages” and DFW said “watch this.” Of course it’s us poor readers that pay the price having to use two bookmarks (or three, as some footnotes have footnotes). And there was plenty of this: > the user is forced to break from the flow of reading I’ve never had to reread so many passages. If you take it all in stride it’s quite fun, really. I found it humorous and chuckled each time I had to skip to the back, like I was privy to an inside joke. The article does raise good points, and I’ve never used an e-reader. I bet it helps a ton, I imagine something like the pop up doc/diff/def viewers like in a modern source code editor. I imagine it would’ve taken away some of the fun from IJ, but of course that’s not a typical case.
- codetrotter 6y ago
  I tried to read Infinite Jest once but eventually gave up because I found it kind of exhausting to read and I couldn’t really follow along. After reading your comment I decided to have another look at it. I knew last time that it was a weird book, and that it was enjoyed by a lot of people. But even looking at the first page of it again I just don’t feel compelled to read it. Maybe some day.
  - 1MachineElf 6y ago
    There's a decent audiobook too, if that helps.
    - thedanbob 6y ago
      As someone who has done a little work on the side recording audiobooks, I hope they paid that guy well. That sounds like a nightmare.
      - sixstringtheory 6y ago
maaark 6y ago
I can't believe he wrote a whole article on the usability of footnotes, using footnotes, on a site without links to return from the footnote back to where it was used.
- dmurray 6y ago
  The browser back button works, at least for me.
- edent 6y ago
  I'm being deliberately ironic to help underscore how usability issues affect footnotes.
dredmorbius 6y ago
...There have been many attempts to get a <footnote> element into HTML... HTML is now well into its 2nd quarter century, grew from academic origins, and yet still lacks a <footnote> or <note> element and standard presentation(s).¹ I've not followed developments at all closely, references here are based on current DDG searches, bit so far as I'm aware, the situation's not improved and has arguable regressed. Suggestions to use <cite> fail as not all footnotes are citations. Using <aside> overloads the element, creates ambiguities as to usage, and leaves us with nonstandard footnote mechanisms. There was a 2003 proposal for a <referral> tag which seems ... excessively Byzantine.² David MacDonald's <note> element proposal seems to cover most of the features (and objections) I'd anticipate.³ I'd consider footnotes (bottom of current page), endnotes (end of document or section/chaper/article), sidenotes (margin adjacent to text), inline notes (parenthetical communts, such aas this), or pop-up dialogues to be for the most part implementation details of an underlying notes element. An author might specify a default presentation, but the reader or device / media properties could override this in all cases to suit needs, limitations, or preference. ________________________________ Notes: 1. HTML3 had the <FN> element, long since abandoned: https://www.w3.org/MarkUp/html3/footnotes.html https://www.w3.org/MarkUp/html3/footnotes.html 2. https://lists.w3.org/Archives/Public/www-style/2003May/att-0050/proposal.draft.html https://lists.w3.org/Archives/Public/www-style/2003May/att-0... 3. https://www.davidmacd.com/blog/html51-footnotes.html https://www.davidmacd.com/blog/html51-footnotes.html See also http://spec-ops.github.io/html-note/index.html http://spec-ops.github.io/html-note/index.html
- apetresc 6y ago
  > Suggestions to use <cite> fail as not all footnotes are citations. Using <aside> overloads the element, creates ambiguities as to usage, and leaves us with nonstandard footnote mechanisms. I agree with all of your conclusions except this one. You appear to be contradicting yourself. How does <aside> overload the element? It does the opposite, by semantically separating <cite>s from mere textual notes and letting the client present the two however the user wants. I agree that within each category, it should be up to the presentation layer whether they are popups or footnotes or endnotes or just hidden altogether, but the distinction between citations and notes is an important one that can't be easily inferred. So what's the problem?
nsmog767 6y ago
One of the best uses of footnotes is in Softwar, the book on Ellison & Oracle. There, footnotes are an entertaining editorial compromise: the author refused to give up final approval, but he gave Ellison the right to respond within the book. Ellison will either expand on - or contradict - the author's points. Easily the most appropriate and defensible use of footnotes I've ever encountered. Link to book: https://www.amazon.com/dp/B00AK78QVI/ https://www.amazon.com/dp/B00AK78QVI/
- impendia 6y ago
  I skimmed through part of it and found this gem in a footnote: "LE writes: Gardens last for hundreds of years, companies don't. That's because people love and take care of gardens."
- simonebrunozzi 6y ago
  Interesting. What's the main takeaway of the book?
rimliu 6y ago
The worst case I had to deal with was when reading War and Peace by Leo Tolstoy. I can read Russian, that's not a problem. The problem is that about 20% is in French which I cannot read, and translation was given as footnotes. So annoying.
r34 6y ago
The worst thing ever is end-book-notes (or end-chapter ones) in a paper book. I have the habit of "precise reading", e.g. I usually read books from first to last page without skipping information. The "end-notes rich" books are the worst reading experience for me (bottom page foot-notes are OK). When I write text I quite often use footnotes - they are special for the organisation of the text, adding kind of additional dimensions to the main narration. I don't (or very rarely) use ebooks, so hard to tell.
bschne 6y ago
I personally like the approach that Gwern takes (e.g. [1]), which works fairly well on screens where you have a lot more horizontal space than needed for the main body of text (it's beneficial to keep the text somewhat narrow anyway, since lines that are too long become harder to scan/read). It seems to be somewhat similar to what Edward Tufte does in his books, which is also putting the footnotes in the margins close to where they are brought up, and often even interrupting the current line for an additional note or image at the point of reference. 1. https://www.gwern.net/GPT-3 https://www.gwern.net/GPT-3
- cxr 6y ago
  > I personally like the approach that Gwern takes (e.g. [1]), which works fairly well on screens where you have a lot more horizontal space I do almost all my reading and general browsing on half-width windows (side note: I highly recommend this practice), but sometimes I break out a tab into its own window and maximize it for pages that don't play nice at half-width. Gwern's footnotes break if you ever transition from a full-width window and then drop down to half-width. (Whereas if you start at half-width and keep it that way, they work as intended.)
- gwern 6y ago
  More info: https://www.gwern.net/Sidenotes https://www.gwern.net/Sidenotes
rocqua 6y ago
I have often used footnotes in my own reading. For things I want in the text for completeness, but do not consider essential to the story. However, thinking about it, they are completely nonfunctional for me when reading. Unless I am skimming a text, I will immediately look up any foot-note I come across. I suppose that, if it is a quick explanation of something I already know, or don't care about, it is easier to skip the footnote in full. Whereas an in-line explanation is harder to skip because it is part of the text.
jeremiahlee 6y ago
Nice to see Coil icon light up. The author just earned $0.02 from me because of their Web Monetization tag.
WorldMaker 6y ago
> How does the user know whether the footnote is a citation or is explanatory? Used to be that was one of the distinctions between footnotes and end notes. This is an affordance where that lost distinction in ebooks is most felt, I think. An equally old practice, however, is to use different marks for footnotes versus end notes, and that's something I've felt could be readopted in a digital context. In particular the asterisk, dagger, double dagger, very rare triple dagger symbology was used for foot notes only (as there are only a few such symbols, versus numerals, and they can't be "document global"). In a medium with hyperlinks, symbol uniqueness is less of a problem so asterisks for all footnotes and numbers for end notes is possible. Similarly, we aren't as limited typographically with what metal we have available in ebooks, we could enlist a much wider variety of symbols (emoji notes, perhaps?), or even pair symbols and numbers. Perhaps the biggest thing holding us back right now is that publishers want to use the same design tools for both print and ebook (for good reason), but ideas like increasing symbol variety or pairing symbols and numbers are both reasonable options to make it to print as well if they need to be kept close in similarity for ease of production.
zeveb 6y ago
Footnotes are great! There's something wonderful about the affordance of a digressive paragraph at the foot of the page. I do not need to break the flow of reading, just flick my eyes down, read and return to the body text. What are terrible are endnotes, which force one to page or click back and forth. That action breaks flow. I wonder if the issue is with how footnotes are displayed electronically? Perhaps a Sufficiently Smart E-Reader could display a footnote as long as its footnote marker, sentence, line or paragraph (would need research to see which is best, I think) is on-screen.

56 points by edent 6y ago

Tufte's footnotes are my favorite. They're especially brilliant when you have typeset documents on 8.5x11" paper, because you wind up with huge margins if you're using a humane line length for your main text. https://www.overleaf.com/articles/implementing-the-teachings-of-edward-tufte/hvtnngcxqnwm https://www.overleaf.com/articles/implementing-the-teachings...

User23 6y ago
Tufte is underappreciated for sure. He's often aped, but rarely do I see his ideas actually applied as he teaches them. The whole point is a theory and practice of communication. Before he gave up and decided to just focus on sculpture he was almost to the point of actually developing a solid theory of representing 5-6 dimensional information representation in 2space. Of course I understand why he gave up, because 3space is way more interesting when you've solved the 2space problem to your own satisfaction.

quadrifoliate 6y ago

From the article: > But I want to talk about the cognitive issues at play. When footnotes can contain anything from a citation to essential background information, the user is forced to break from the flow8 of reading – often in the middle of a sentence. This is a problem in both paper books and electronic books. I think this is...the idea of footnotes? That you don't break the flow and it's safe to skip over them, but if you're coming back for a leisurely read later, you can look them up and see what you missed. [1] When you go over the text a second time, you can follow the references to the footnotes and see the interesting crumbs that the author laid out, which often lead to a train of thought separate from the main flow of words. ------------ [1] At least, this is how I have used footnotes for a while when reading, and it works quite well. YMMV, since people read and comprehend differently.

edent 6y ago
I think my confusion is about how to read the footnotes which are more than just citations. If I read them "inline" then I don't understand why they're not part of the main text. If I wait until the end of the chapter, I've lost context. I'm prepared to believe that I haven't been taught to read in the right way. But there seems to be some disagreement on what the right way is.
ringshall 6y ago
Same idea with parentheses, at least I understand them.

hardwaregeek 6y ago

I find footnotes annoying even in print books. When I'm reading a book, I have to figure out if the book is the type that uses footnotes as citations or the type that uses it as actual content. Or worse, the type that does both. If it's the type that uses them as citations, then meh, I can skip over them. If it's the type that uses them as content, well ugh now I need to read this footnote and then get back to the original text, which is a context switch. Even worse if the footnotes are actually endnotes because then I need to physically flip the pages. Even worse if the endnotes are gigantic and have subnotes (cough Infinite Jest). If it's the type that does both then I have to look at every damn footnote only to be disappointed by a "Pfaffenberger, 2016" 90% of the time. I don't have a good answer to these problems. Tufte style is better because I don't have to move my eyes that much. But there's only so much one can put in a note in the margin. And it still distracts.

User23 6y ago
I like endnotes best. For a book where they are actually useful I either unbind or buy two copies so I can lay out the endnotes side by side with the book. Think of it like having two monitors where one is the working screen and the other is the reference screen.
- JadeNB 6y ago
  Wouldn't it be more useful not to have to do this? Keeping them footnotes saves buying another copy, or damaging yours.
  - User23 6y ago
    Think of it as the difference between having two monitors and having one monitor where the bottom 1/8th to 1/2 randomly gets hijacked. Two monitors are also more expensive. My method has its trade-offs as you note, but I find having an uninterrupted flow through the main text while having access to the notes on an as desired basis works best for my enjoyment and comprehension. Without my method endnotes are so frustrating I almost never use them, and about half the time when I think ok this time it's worth it to context switch, it turns out not to be. This all comes down to Tufte's idea of letting the optic nerve do the work. A glance requires almost no conscious brainpower, this is why thumbing through a book you've read works so well. Thus I can glance over at the footnotes pages and decide if I care to read the note without even slightly interrupting my flow reading the main text.

just-ok 6y ago

Footnotes are great when done right. If you're using footnotes to cite things, you're doing it wrong: inline citations exist. If that's not enough for you, use a different syntax (say, ^[1] instead of ^1). If you're using it to define terms, you're doing it wrong: use a margin note. If that's not enough, you should probably typeset a nice box that makes definitions easy to identify. Footnotes (imo, of course) should only be used as a "here's more information for the curious" and should similarly only appear at the actual foot of a page (I despise endnotes). My experience is biased towards print and print-like (PDF) use of footnotes, but I don't think it's much of a problem to expect users to click every footnote when provided with a convenient pop-up UI like in the OP, provided that you've clearly defined a convention or pattern of their usage to your readers. I do wish websites showed footnotes on hover, though... jumping within the page is nonsense and often mucks with URL sharing.

barryvan 6y ago
Footnotes _are_ great when done right, but I have to say that inline citations are incredibly frustrating as a reader, because they're harder to skip over. A footnote marker is small enough that your eye can continue over it, whereas a citation like (Just-OK 2020) is much harder to read. My personal preference is to use footnotes as you suggest -- as a "here's more information for the curious" option -- whereas citations should be linked to a "references" section -- either at the end of the chapter or the book or the overall article.
- abnercoimbre 6y ago
  > [...] a "references" section -- either at the end of the chapter or the book or the overall article. So footnotes retain the numbering system (1, 2, 3, etc.) throughout the chapter text, but how would you mark sentences for whom a reference is available?
  - barryvan 6y ago
    I'd probably use numbering for the references, and use the array of symbols* available for footnotes -- given that most pages shouldn't have more than a handful of footnotes, and you can restart your symbol order on each page. That's more difficult on longer-form web content, admittedly. * See https://en.wikipedia.org/wiki/Note_(typography)#Numbering_and_symbols https://en.wikipedia.org/wiki/Note_(typography)#Numbering_an... for a sampling. These can be doubled if necessary -- so , ††, etc.

sixstringtheory 6y ago

Before even clicking through to the article I was thinking about Infinite Jest. > As an author, I want to add more details which don’t fit in the main body of the text. Because I hate deleting text and my editor told me to take it out. I joked about IJ doing that exact thing, that the editor was all “no, you can’t just list nonsensical film titles for four straight pages” and DFW said “watch this.” Of course it’s us poor readers that pay the price having to use two bookmarks (or three, as some footnotes have footnotes). And there was plenty of this: > the user is forced to break from the flow of reading I’ve never had to reread so many passages. If you take it all in stride it’s quite fun, really. I found it humorous and chuckled each time I had to skip to the back, like I was privy to an inside joke. The article does raise good points, and I’ve never used an e-reader. I bet it helps a ton, I imagine something like the pop up doc/diff/def viewers like in a modern source code editor. I imagine it would’ve taken away some of the fun from IJ, but of course that’s not a typical case.

codetrotter 6y ago
I tried to read Infinite Jest once but eventually gave up because I found it kind of exhausting to read and I couldn’t really follow along. After reading your comment I decided to have another look at it. I knew last time that it was a weird book, and that it was enjoyed by a lot of people. But even looking at the first page of it again I just don’t feel compelled to read it. Maybe some day.
- 1MachineElf 6y ago
  There's a decent audiobook too, if that helps.
  - thedanbob 6y ago
    As someone who has done a little work on the side recording audiobooks, I hope they paid that guy well. That sounds like a nightmare.
    - sixstringtheory 6y ago

maaark 6y ago

I can't believe he wrote a whole article on the usability of footnotes, using footnotes, on a site without links to return from the footnote back to where it was used.

dmurray 6y ago
The browser back button works, at least for me.
edent 6y ago
I'm being deliberately ironic to help underscore how usability issues affect footnotes.

dredmorbius 6y ago

...There have been many attempts to get a <footnote> element into HTML... HTML is now well into its 2nd quarter century, grew from academic origins, and yet still lacks a <footnote> or <note> element and standard presentation(s).¹ I've not followed developments at all closely, references here are based on current DDG searches, bit so far as I'm aware, the situation's not improved and has arguable regressed. Suggestions to use <cite> fail as not all footnotes are citations. Using <aside> overloads the element, creates ambiguities as to usage, and leaves us with nonstandard footnote mechanisms. There was a 2003 proposal for a <referral> tag which seems ... excessively Byzantine.² David MacDonald's <note> element proposal seems to cover most of the features (and objections) I'd anticipate.³ I'd consider footnotes (bottom of current page), endnotes (end of document or section/chaper/article), sidenotes (margin adjacent to text), inline notes (parenthetical communts, such aas this), or pop-up dialogues to be for the most part implementation details of an underlying notes element. An author might specify a default presentation, but the reader or device / media properties could override this in all cases to suit needs, limitations, or preference. ________________________________ Notes: 1. HTML3 had the <FN> element, long since abandoned: https://www.w3.org/MarkUp/html3/footnotes.html https://www.w3.org/MarkUp/html3/footnotes.html 2. https://lists.w3.org/Archives/Public/www-style/2003May/att-0050/proposal.draft.html https://lists.w3.org/Archives/Public/www-style/2003May/att-0... 3. https://www.davidmacd.com/blog/html51-footnotes.html https://www.davidmacd.com/blog/html51-footnotes.html See also http://spec-ops.github.io/html-note/index.html http://spec-ops.github.io/html-note/index.html

apetresc 6y ago
> Suggestions to use <cite> fail as not all footnotes are citations. Using <aside> overloads the element, creates ambiguities as to usage, and leaves us with nonstandard footnote mechanisms. I agree with all of your conclusions except this one. You appear to be contradicting yourself. How does <aside> overload the element? It does the opposite, by semantically separating <cite>s from mere textual notes and letting the client present the two however the user wants. I agree that within each category, it should be up to the presentation layer whether they are popups or footnotes or endnotes or just hidden altogether, but the distinction between citations and notes is an important one that can't be easily inferred. So what's the problem?

nsmog767 6y ago

One of the best uses of footnotes is in Softwar, the book on Ellison & Oracle. There, footnotes are an entertaining editorial compromise: the author refused to give up final approval, but he gave Ellison the right to respond within the book. Ellison will either expand on - or contradict - the author's points. Easily the most appropriate and defensible use of footnotes I've ever encountered. Link to book: https://www.amazon.com/dp/B00AK78QVI/ https://www.amazon.com/dp/B00AK78QVI/

impendia 6y ago
I skimmed through part of it and found this gem in a footnote: "LE writes: Gardens last for hundreds of years, companies don't. That's because people love and take care of gardens."
simonebrunozzi 6y ago
Interesting. What's the main takeaway of the book?

rimliu 6y ago

The worst case I had to deal with was when reading War and Peace by Leo Tolstoy. I can read Russian, that's not a problem. The problem is that about 20% is in French which I cannot read, and translation was given as footnotes. So annoying.

r34 6y ago

The worst thing ever is end-book-notes (or end-chapter ones) in a paper book. I have the habit of "precise reading", e.g. I usually read books from first to last page without skipping information. The "end-notes rich" books are the worst reading experience for me (bottom page foot-notes are OK). When I write text I quite often use footnotes - they are special for the organisation of the text, adding kind of additional dimensions to the main narration. I don't (or very rarely) use ebooks, so hard to tell.

bschne 6y ago

I personally like the approach that Gwern takes (e.g. [1]), which works fairly well on screens where you have a lot more horizontal space than needed for the main body of text (it's beneficial to keep the text somewhat narrow anyway, since lines that are too long become harder to scan/read). It seems to be somewhat similar to what Edward Tufte does in his books, which is also putting the footnotes in the margins close to where they are brought up, and often even interrupting the current line for an additional note or image at the point of reference. 1. https://www.gwern.net/GPT-3 https://www.gwern.net/GPT-3

cxr 6y ago
> I personally like the approach that Gwern takes (e.g. [1]), which works fairly well on screens where you have a lot more horizontal space I do almost all my reading and general browsing on half-width windows (side note: I highly recommend this practice), but sometimes I break out a tab into its own window and maximize it for pages that don't play nice at half-width. Gwern's footnotes break if you ever transition from a full-width window and then drop down to half-width. (Whereas if you start at half-width and keep it that way, they work as intended.)
gwern 6y ago
More info: https://www.gwern.net/Sidenotes https://www.gwern.net/Sidenotes

rocqua 6y ago

I have often used footnotes in my own reading. For things I want in the text for completeness, but do not consider essential to the story. However, thinking about it, they are completely nonfunctional for me when reading. Unless I am skimming a text, I will immediately look up any foot-note I come across. I suppose that, if it is a quick explanation of something I already know, or don't care about, it is easier to skip the footnote in full. Whereas an in-line explanation is harder to skip because it is part of the text.

jeremiahlee 6y ago

Nice to see Coil icon light up. The author just earned $0.02 from me because of their Web Monetization tag.

WorldMaker 6y ago

> How does the user know whether the footnote is a citation or is explanatory? Used to be that was one of the distinctions between footnotes and end notes. This is an affordance where that lost distinction in ebooks is most felt, I think. An equally old practice, however, is to use different marks for footnotes versus end notes, and that's something I've felt could be readopted in a digital context. In particular the asterisk, dagger, double dagger, very rare triple dagger symbology was used for foot notes only (as there are only a few such symbols, versus numerals, and they can't be "document global"). In a medium with hyperlinks, symbol uniqueness is less of a problem so asterisks for all footnotes and numbers for end notes is possible. Similarly, we aren't as limited typographically with what metal we have available in ebooks, we could enlist a much wider variety of symbols (emoji notes, perhaps?), or even pair symbols and numbers. Perhaps the biggest thing holding us back right now is that publishers want to use the same design tools for both print and ebook (for good reason), but ideas like increasing symbol variety or pairing symbols and numbers are both reasonable options to make it to print as well if they need to be kept close in similarity for ease of production.

zeveb 6y ago

Footnotes are great! There's something wonderful about the affordance of a digressive paragraph at the foot of the page. I do not need to break the flow of reading, just flick my eyes down, read and return to the body text. What are terrible are endnotes, which force one to page or click back and forth. That action breaks flow. I wonder if the issue is with how footnotes are displayed electronically? Perhaps a Sufficiently Smart E-Reader could display a footnote as long as its footnote marker, sentence, line or paragraph (would need research to see which is best, I think) is on-screen.