When I write

${\cal{A}} = \textbf{y}_{i}, \forall i = 1,\cdots, n^{2}$, 

I get:

enter image description here

However, if I remove the curly braces around \cal{A} and instead write

$\cal{A} = \textbf{y}_{i}, \forall i = 1,\cdots, n^{2}$, 

then I get

enter image description here

I'm just really curious what's causing this interesting issue.

  • 3
    Welcome to TeX.SE! \cal is not intended for use in math mode. Instead of it rather use \mathcal and your problem will disappear.
    – Zarko
    Mar 13, 2021 at 3:56
  • 1
    @Zarko - Please see the answer I just posted. \cal is a PlainTeX font switching command and, as such, is very much meant to be used in math mode. However, precisely because it's a PlainTeX holdover, it shouldn't be used in a LaTeX document written after 1994.
    – Mico
    Mar 13, 2021 at 5:38

1 Answer 1


You wrote,

I'm just really curious what's causing this interesting issue

Short answer: To use \cal in any LaTeX document written after 1994 [!] is either an accident waiting to happen -- as is the case with your sample code -- or won't work at all -- say, if the memoir document class is used.

Longer answer: \cal is a plain-TeX font switching command that is implemented in some, but not all, LaTeX2e document classes, as a legacy command to ease the changeover from LaTeX209 to LaTeX2e. (That changeover occurred in 1994.) Because \cal is a switch rather than a command that takes an argument, its scope is through to the end of the current TeX group. Hence, by writing \cal{A}, one creates the false impression that only A is affected by whatever \cal does; that's not the case, as you've (re)discovered. One must encase \cal A in curly braces -- i.e., write {\cal A} -- in order to avoid creating a typographic disaster. What you really ought to be doing, though, is to write \mathcal{A}. (In this regard, \cal is not different from the other legacy/PlainTeX font switching commands, such as \bf, \it, \sl, and \sc.)

Additional comment: Your sample code contains one further inaccuracy and one typographically questionable choice.

  • To write \textbf{y} in math mode will sooner or later come back to haunt you, especially if the document's text and the math fonts don't look alike. You should be writing \textbf{y} in text mode and \mathbf{y} in math mode.

  • To write 1,\cdots, n^{2} is somewhat unconventional from a typographic point of view, to put it charitably. Better to write 1,\ldots, n^{2} or, better still, 1,\dots, n^{2} -- and let LaTeX decide whether to generate a typographic ellipsis that's placed on the baseline or on the mathline. (The mathline is where the mathematical - (minus) symbol is placed.) \cdots should be employed, say, to elide a sequence of binary operations, as in \prod_{i=1}^N a_i = a_1 a_2 \cdots a_N. In contrast, \ldots should be used to abbreviate a list of comma-separated items.

To summarize, then, do consider replacing

${\cal{A}} = \textbf{y}_{i}, \forall i = 1,\cdots, n^{2}$,


$\mathcal{A} = \mathbf{y}_{i} \quad \forall i = 1,\dots, n^{2}$,
  • 1
    +1 But allow me a small correction: \dots decides where to put the ellipsis only with amsmath. The kernel definition simply discrimitates between text and math mode but always puts them on the baseline.
    – campa
    Mar 13, 2021 at 11:01

You must log in to answer this question.

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