# Do you care about (micro) typography?

My original question was based on the assumption that there is a universal scientific agreement on the typography of subscripts. However, I have been wrong.

The discussion if there was a universal typographic standard was closed as off-topic. So this question (about enforcing this standard) is probably off-topic too.

Assuming there was a universal typographic standard (but in physics, there isn't) that descriptive subscript should be typeset in upright font:

If the Physics.SE's Q&A's contain $G_m$ instead of $G_\mathrm{m}$, do you care? (It seems, you don't.) Would you edit a posting to correct that? (It seems you won't.) Would this change if it was a little bit easier to input?

• I don't think it's necessary and would reject such an edit as "doesn't make post easier to read". The ease of the inputs would not change (one could always use {\rm e}) – Kyle Kanos Feb 14 '17 at 17:49
• This is an excellent question to ask, even though I disagree with the proposal. Thanks for bringing it up. – David Z Feb 14 '17 at 18:36
• Answer in comment: I care, but I'm busy. I can't fix everything, so I prioritize. – rob Feb 14 '17 at 23:02

There is a point at which I stop caring.

Now, I think I care more about typography than most of our regular users (though not the most: we have a few users who are very detail oriented that way), but making one letter subscripts upright is on the wrong side of that line for me: I probably wouldn't even notice if someone else did this unless they called my attention to it.

I would probably consider an edit that did only this as too trivial, but I wouldn't reject a more comprehensive edit of the typography because this kind of adjustment was included.

• ^ pretty much just that. – Emilio Pisanty Feb 14 '17 at 18:27

In my time in physics I've noticed a convention where single-character subscripts are generally set in italic even if they represent an abbreviation rather than a variable. I grant that there's no good reason to actively choose this convention, other than ease of typing, but I don't think it's particularly harmful either. So on that basis, I dispute your premise that the subscripts you point out should be upright. (And I say this as one of the members here who is most picky about typography.)

Here's why I don't think this convention is a problem. Variable names in physics are pretty much always single characters. And, of course, they are always typeset in italic, as is standard. This means that when I see a run of consecutive italic characters, I immediately know that those characters are meant to be be interpreted individually, not as a unit. \begin{align} xyz &\implies \text{three things: x, y, z} \\ \mathrm{xyz} &\implies \text{one thing: xyz} \\ \text{(for comparison) }\mathrm{x,y,z} &\implies \text{three things: }\mathrm{x,y,z} \\ \end{align} In fact, I contend that this distinction of whether I should consider consecutive characters as a single unit or individually is the most useful piece of information conveyed by the choice of font shape, not whether the character represents something variable or something fixed. The latter is a piece of information I should already know based on the definition of the symbol (and if I don't know the definition of the symbol, I've got bigger problems). Now, obviously, the distinction between one or multiple things represented is completely irrelevant when dealing with a single character. There is no difference, no ambiguity, in the number of things represented by either $x$ (one) or $\mathrm{x}$ (one). Accordingly, when writing a single character, it's not really that useful to put it in an upright font. And, since we get used to this line of reasoning when looking at normal-size characters, we intuitively carry it over to subscripts and superscripts.

Anyway, if someone proposed an edit that merely changes the font shape of single-character subscripts as you suggest and does nothing else, I would consider that too minor, assuming all the subscripts were consistent throughout that one post. However, if a post used single-character subscripts inconsistently, i.e. sometimes italic and sometimes upright within the same post, I think an edit that fixed that inconsistency would just barely squeak over the line that makes it acceptable to me. (I'm not going to hold people to conventions but I will hold them to expectations of consistency within their own post, even down to fine typographical details.) Of course, as dmckee says in his answer, it would be much better if an edit that does this fixes other issues with the post as well.

• I just looked through the 50 papers I have open (thank goodness for browser tabs!) and not a single one uses upright fonts for subscripts. Every single subscript is italic. So... I agree, I don't think that there is a hard and fast rule that subscripts should be upright, even when they stand for something specific and are not a tensor index. – tpg2114 Feb 14 '17 at 20:19
• @tpg2114 I did not say all subscripts have to be upright. Just the descriptive mnemonic ones. See the PS I added to my original question, referencing IUPAP, IUPAC, NIST, ... standards. – mhchem Feb 14 '17 at 23:10
• @mhchem Right, but that's why I also added the last clause "even when they stand for something specific..." which means they are mnemonic ones. I was only pointing out that I have never seen one make the distinction rigorously in publications, and my current sample of papers reconfirms that. – tpg2114 Feb 15 '17 at 0:56