21

I am a mathematician whose work often connects with physics, so I often have to read papers in physics journals. I find that the two-column format popular with physics journals annoying, and based on the following questions on this stackexchange I am not the only one that feels that way:

Difficulty reading scientific papers in two columns

How best to present long equations in two-column papers?

Do two-column format journals publish one column for special cases?

I contend that the two-column format is impractical for long mathematical equations and is harder to read, since your eyes have to move in a pattern different from most other printed texts. But I suppose that there must be some advantage to the two-column format, given it is so standard in some branches of academia. But I can't think of a single reason why it would be better than a single column printed page! Enlighten me, please.

6
  • 4
    I don't think there's an advantage. It's just ... tradition. For similar reasons, the citation styles in certain fields omit the paper name, which I find annoying but manageable.
    – Allure
    Sep 5, 2018 at 3:00
  • 1
    There's a good discussion in this answer.
    – Anyon
    Sep 5, 2018 at 3:02
  • 5
    A lot of these standards come from the dark ages before personal computers. So don’t be surprised if the answer has something to do with the limitations of punch cards and maximizing the number of articles that can be distributed using a single carrier pigeon.
    – Thomas
    Sep 5, 2018 at 3:15
  • 1
    @Thomas your comment would be more convincing, except that most of the formatting of computer produced documents has been unremittingly horrible (Of course the default options in some well known document-creation apps bear a lot of the blame for this).
    – alephzero
    Sep 5, 2018 at 8:40
  • 1
    I transitioned from an area with majorly single-column publications to an area with majorly double-column ones. I got used to it quite fast and like two columns more now, when reading papers. I'd also say the typical "expected" width of an equation has also something in common with the decision if the layout would be one-column. Sep 5, 2018 at 18:14

5 Answers 5

34

There's a widespread belief that shorter lines (fewer characters per line) are easier to read, because the eye doesn't have to move as far horizontally from the end of one line back to the start of the next. The left edge of the text (assuming left-to-right languages) may already be in your peripheral vision. Thus it's easier to visually find the correct line to read next, and avoid accidentally rereading the same line or skipping lines.

This is the justification for the extremely wide margins that LaTeX uses by default in one-column styles, for instance.

I don't know offhand if there is research supporting this belief, but maybe someone can fill me in if they know.

Two columns makes it easier to have short lines, without resorting to small paper size, large font sizes, or huge margins. Thus you still get a high density of text per page, and it keeps page counts down (and the associated costs).

7
  • 8
    Some references to studies on line length and readability here: en.wikipedia.org/wiki/Line_length. The newspaper industry has known about this for centuries - hence the large number of narrow columns per page in traditional broadsheet and tabloid page layouts.
    – alephzero
    Sep 5, 2018 at 8:37
  • Reinforcing this, online-only physics journals such as the Journal of High Energy Physics (JHEP) or Quantum, which don't incur printing costs, are often single-column with larger margins.
    – knzhou
    Sep 5, 2018 at 9:04
  • 5
    There actually is a lot of research on this. My copy of "the elements of typographic style" (version 3.2) covers it as: anything from 45 to 75 characters is considered satisfactory length for a single column work ... for a multiple-column work a better average is 40-50 characters (pg 26)
    – Racheet
    Sep 5, 2018 at 12:22
  • 4
    Personally I think this belief is misguided. No doubt shorter lines are easier to read with other things being equal, but other things are not equal. The contortions used to make shorter lines possible - frequent breaking of words across lines (journals/LaTeX) or stretching inter-word spaces to ridiculous proportions (newspapers) - for me impair readability to a greater extent than the line length helps it. Oct 5, 2022 at 9:40
  • 2
    For math, shorter lines forces us often to break formulas on 2 lines, which makes the reading harder in my oppinion.
    – Nick S
    Mar 14, 2023 at 15:23
3

I think it actually saves a lot of space.

Think about floats-like elements that need to be placed in between the text. Those include section titles, footnotes maybe, tables, equations, and, more importantly, figures.
When you place each such item, you need to leave blank space on either side since the text can only be above and below. This creates so much wasted space, which I suppose was(is) a big issue for printed journals.

And since normally journals allow for larger figures, tables, equations, etc. to span the entire width of the page - provided they are positioned at the top or the bottom of the page - I don't think there are serious drawbacks with two-column layouts.

In mathematics, I suppose you have larger equations and fewer figures so I can see your point, but at least in the areas I am familiar with, a typical 10-14 paged journal article will have about 5-7 figures and maybe 30 equations (with at most one or two of those will need to be larger than the column-width) so the space saved is considerable. As a matter of fact, when I use the IEEE Transactions template on LaTeX, when I switch from two-column to single column layout, the manuscript nearly doubles in size (I believe margins and fontsize stay the same, though I might be wrong).

4
  • Who cares about saving space? Print is dead.
    – barneypitt
    Nov 19, 2023 at 7:49
  • @barneypitt matter of fact, given the huge amount of literature that appears every day, it's no wonder that a 10 pages paper looks less scary than a 20 pages one -- and people are more likely to consider taking the time to read it.
    – sonarventu
    Feb 9 at 16:16
  • @sonarventu I can't speak for anyone else, but when I see 20 pages of readable paper using a decent font size and a modern font (why do half the papers use ugly eye-tiring serif fonts straight out of the 1970's?!! -- should be called sick page syndrome) I find that vastly less "scary" than 10 cramped pages.
    – barneypitt
    Feb 14 at 7:39
  • @barneypitt I also can't speak for others, but I usually find the opposite to be readable. It's also much much faster to read. Large fonts and larger movements make my eyes tired, although there are exceptions. Eg. I have IEEE's single page layout, but I find NeurIPS' style very nice.
    – kyriakosSt
    Feb 15 at 10:45
1

I think that reason could be; when reading academic paper, you are often re-reading sentences. It's not like reading literature, that you are just proceeding sentence-by-sentence without ever going back.

In order to grasp written text, sometimes you need to re-read, and having shorter lines makes it easier to go back, i.e. easier to visually capture several lines of text and eventually to advance faster.

1
  • Somewhat off-topic, but I re-read sentences in literature all the time. If it's a master of sentence-craft like Martin Amis or Kazuo Ishiguro I probably read 1/3 of the sentences again!
    – barneypitt
    Feb 14 at 7:44
0

As mentioned elsewhere, "[t]here's a widespread belief that shorter lines...are easier to read." Regarding,

I contend that the two-column format is impractical for long mathematical equations

Push long equations into one-column figures.

8
  • 3
    Your solution seems like it would greatly disrupt the flow of any paper with more than a few such equations. Sep 5, 2018 at 9:30
  • Related (on figures).
    – user68958
    Sep 5, 2018 at 9:34
  • 1
    @TobiasKildetoft This is a workaround, not a solution. It will work in some cases and not in others.
    – user2768
    Sep 5, 2018 at 11:22
  • 2
    And many journals stick to the principle that a figure is a figure, and an equation is an equation and you mustn't jumble the two up. There are other workarounds for full-width equations, but they're not great either
    – Chris H
    Sep 5, 2018 at 12:42
  • 1
    @user2768 there's a big difference between "contain" and "consist (mainly) of". The former is commonplace, but the latter is very rare.
    – Chris H
    Sep 5, 2018 at 12:52
0

None.

It is true that shorter lines are easier to read, but this only highlights the other main problem with the standard academic paper formatting, namely that the fonts are ridiculously small.

Increasing the fonts sizes (they should be twice as large to achieve the readability of any modern document) reduces the line length accordingly.

Another way to look at this is that the typical academic paper takes four pages of a modern document, and for no other reason than tradition, crams them onto one page, 2 by 2. (Okay. A reason could be to save paper, but it's a silly reason, given that journals are pretty much only read in electronic format.)

Academics don't (often) publish longer theses in this anachronistic format, and haven't for many years. Any arguments which try to make a case that there is somehow something "special" about academic content that somehow requires this antediluvian formatting, have also to explain why modern formatting is used for longer works.

It is high time for journals to enter the 21st century and accept papers only in a modern single-column large font format, using modern sans serif fonts.

1
  • 1
    The modern age goes even further: leave the format preferences to the end user and publish in a format that is digital first and printing second. Why fix a paper format in the first place?
    – Ambicion
    Mar 14, 2023 at 22:29

You must log in to answer this question.

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