Why two columns at all
The readable line is somewhere between about 45 and 75 characters. Much shorter and the eye jumps back too often; much longer and it loses the start of the next line on the way back. On a single-column page at a normal journal type size, a full-width line runs well past that, which is why single-column journals either set large type or carry wide margins.
Two columns put the line where it belongs and use the page. That is the whole argument, and it is why the convention has survived the move from print to screen.
What breaks it
A two-column page has a hard constraint a single-column page does not: a thing that is wider than a column has nowhere to go. There are only three answers, and a layout has to be explicit about which it uses:
- Set it across both columns, at the top or bottom of the page, with the text flowing around neither side. This is the usual answer for a wide table or a large figure.
- Shrink it to fit one column, which works for a figure and almost never works for a table with six columns of numbers.
- Give it a page of its own, which is right for a full-page figure and wasteful for anything smaller.
Tables are the hard case
A figure can usually be made smaller. A table cannot: its width is decided by how many columns it has and how long the entries are, and past a certain point the only way to fit it into one column is to make the type too small to read.
A table that spans both columns also cannot be split across a page the way a column of text can - or rather, it can, but only if its header row is repeated at the top of the continuation, or the reader is left looking at a page of numbers with no idea what they are.
This is worth deciding per table rather than per journal. A three-column table of summary statistics belongs in a column. A results table with twelve columns belongs across the page, and if it runs long, it belongs in an appendix or a supplementary file.
Equations in a narrow measure
A displayed equation in a two-column layout has about 80mm to work in. Many equations in a submitted manuscript were written in a full page width and do not fit.
The answers are to break the equation across lines at an operator, to set that one equation across both columns, or to define intermediate terms above it so the displayed line is shorter. All three are editorial decisions about the mathematics, which is why an author should see them on the proof.
When one column is the better choice
Two columns are not always right. A single column suits articles that are mostly prose and few figures - humanities and much of law - and it suits journals whose articles carry long block quotations, extensive footnotes, or verse, none of which survive a narrow measure.
It is also the easier layout to read on a phone, which matters more every year. A journal that sets two columns in print and reflows to one online is making the right decision twice rather than compromising once.
Last reviewed
