This question is a duplicate of this one I asked on graphicdesign.stackexchange.com, where commenters suggested to ask it here instead.

I am facing the following typography/formatting/spelling/punctuation problem:

In a scientific article I am writing, there naturally occur instances, where there are math symbols and their meaning scattered in prose text, e.g. "The input voltage u is given, as well as the inductivity L and the moment of inertia M, where [\n] i refers to the current through the motor." ("[\n]" symbolizes a potential newline character inserted by my teypesetting system.)

On some ocassions, my typesetting system sometimes inserts a line break before the math symbol, putting the symbol at the start of a new line.

Note, that I am not talking about the start of a sentence, but just the start of a line.

My supervisor tells me, this is bad practice, and that lines must not start with a mathematical symbol (1).

I know how to avoid this in my typesetting system (LaTeX), but I have never heard that rule before, so I am curious about learning more about it.


  1. Can you guide me to a reference work to confirm or abnegate (1)?
  2. If not a reference work, can you give an article where this is being written about?

If this is language-dependent: I'd like to know in particular about German, British English and American English.


I am not well-versed in typography literature, and my online search has not brought any meaningful insights for me.

In an answer attempt at graphicsdesign.stackexchange.com, the AMSMath reference guide Part 1, Chapter 1, 12.5 Wording (page 81) has been quoted:

Try to avoid reference numbers, variables, equation numbers, and mathematical expressions as the first word in a sentence.

While this is just a suggestion, it is not strictly an answer to my question about symbols after line breaks, but might hint at the nonexistence of the rule in question.

  • 11
    This is just anecdotical evidence, so a comment rather than an answer: I've never heard anybody claim before that there shouldn't be a math symbol at the beginning of a line. Most math papers certainly contain many lines that start with mathematical symbols. Aug 1, 2023 at 10:46
  • 4
    From the academic point of view - write the paper, submit it, and unless you find otherwise in instructions provided to authors, let the editorial staff worry about typesetting to the journal's style sheet. Aug 1, 2023 at 12:30
  • 1
    @ScottSeidman Your opinion generally seems like good advice; However, in this current scenario, I am the "editorial staff" who has been tasked to format others' texts into printed material.
    – marc
    Aug 1, 2023 at 13:35
  • 3
    If you are using LaTeX it will compute the line breaks itself. To avoid that line break, use this: where~$i$ refers to. If these are always single-letter symbols, that should cause little problem for LaTeX to compute a good way to do it.
    – GEdgar
    Aug 1, 2023 at 16:42
  • 2
    @GEdgar Thank you for the heads up, but as is written in the question, I know how to achieve this technically. The question I posed was about the justification (not in a typographical sense).
    – marc
    Aug 1, 2023 at 16:47

2 Answers 2


Rules, such as they are, in typography exist only by convention and by the acceptance or rejection by someone (e.g., a journal editor) of work that does, or does not conform to the rule. The so-called Oxford comma, or serial comma provides a good example; some journals and publishers follow the Oxford comma rule (i.e., convention), others do not.

More useful, I think, is to ask why it is frequently recommended that one avoid having symbols at start of a sentence, or start of a line. If, instead of searching the web for the narrow topic of "symbols at the beginning or and of a line", you instead search for different approaches that have been taken to choosing the position for a line-break, you will find a considerable amount of thoughtful material.

I searched using Google, with the words:

typography choice of position for line breaks

This article from the World Wide Web Consortium not only provides useful advice, it also provides useful insights and a good point from which to explore the issue further. It does not directly address the question of symbols at the beginning or end of lines, but I think you will see that the question about symbol positioning fits within a much larger framework of thought about how to make text easy to read.

  • Thank you for your insight. I will have a thorough look on this article, but you also confirmed the conventional character of such rules. In the meantime, I have had a talk with the mentioned supervisor, who backpedaled a bit and acknowledged your position as well. I probably am going to follow his conventions, now knowing that it is well within the rules of scientific writing do do otherwise, if I desire so.
    – marc
    Aug 1, 2023 at 13:39
  • +1 for the frame change - the point of typographical conventions is readability. Aug 1, 2023 at 20:58

As you note, the AMS is entirely agnostic about lines starting with symbols, and gives little advice about linebreaking around symbols, in general (on page 103 of the AMS Style Guide, the editor is advised to "Avoid breaking math in text between fences,"—this is expanded upon on page 115; indeed, section 13.16, starting on page 114, gives a lot of advice on breaking equations). That being said, the AMS Style Guide gives many, many examples of mathematics being broken across lines (both in inline and in displayed mathematics), hence it seems that the AMS has no problem with mathematics (and, therefore, symbols) starting a line.

The AMS Style Guide extensively cites Ellen Swanson's Mathematics into Type, which gives much the same advice. The table on page 47 is, perhaps, particularly relevant. Again, there are guidelines for breaking inline mathematical expressions and equations across lines, which necessitates starting some lines with symbols.

Generally speaking, the goal should be to avoid breaking lines in the middle of a "word" (which, as CrimsonDark notes, is a complicated issue. If a symbol is the first character of a word, then a line might start with a symbol (e.g. it is better to break "the bag weighs 47 pounds" before, rather than after, "47", since "47 pounds" is, more or less, a single syntactic unit, i.e. a number with units). On the other hand, a lot of people find single-letter variables at the start of a line to be "ugly" (I would include myself in that group—in many cases, these kinds of breaks also break up semantic units in an ugly way). Thus, to a person who finds single-letter variables at the start of a line ugly, the text

The input voltage u is given,
as well as the inductivity L
and the moment of inertia M,
where i refers to the current
through the motor."

is preferable to

The input voltage u is given,
as well as the inductivity
L and the moment of inertia
M, where i refers to the current
through the motor."

Note that, in either case, if you are using LaTeX, you can use ~ as a nonbreaking space, i.e. inertia~$M$, will prevent a linebreak between "inertial" and "M". And, to be clear, this last bit is entirely a matter of taste and personal preference—I can find no style guide which seems to have much of an opinion, one way or the other.

All of the above having been said, this seems like something to not get overly precious about. At the end of the day, most of these "rules" are just guidelines, which only really matter if an editor or supervisor is insisting on one version or another. I see no problem with just using basic TeX / LaTeX, and letting the typesetting engine decide what to do. If anyone complains, just say "Hey, if it's good enough for Knuth and Lamport, it's good enough for me." ;)

  • 1
    I largely agree but I don't think there is any problem at all with single letter variables at the beginning of a line per se. For example, "If such and such holds, then [line break] $X$ is a compact space" is perfectly fine. As you say, the problem is with instances like "the compact space [line break] $X$" where you break up a syntactic unit that belongs together. Aug 7, 2023 at 16:36
  • 2
    @AdamPřenosil Indeed. I tried to be careful to say that this is a matter of preference. Personally, I really dislike starting a line with a single letter variable. But this is completely a matter of taste. And rules for keeping syntactic units together should definitely take priority over more arbitrary rules like "don't start a line with a single letter variable. Aug 7, 2023 at 16:40

You must log in to answer this question.

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