Lately I have often seen excessive subscripts in posts. Please, use fewer subscripts!

A post with $$j_1$$ and $$j_2$$ but no other subscripted $$j$$'s would be easier to read with $$j$$ and $$k$$ instead.

Especially for $$j$$'s being subscripted or superscripted, $$a_{j_1}$$ and $$a^{j_2}$$ are awkward, where $$a_j$$ and $$a^k$$ are pleasant enough.

Here and here subscripts could be reduced. Here superscripts could be reduced. Here I didn't understand a post with all the subscripts until the edit showing an example without them; I suspect your eye will go first to the example without subscripts and then avoid the subscripted formula.

So eliminating subscripts may get you more answers, and in any case it will reduce headaches for some readers.

• This reminds me of a story about Eilenberg, who had a thing about subscripts: pdfs.semanticscholar.org/0f9e/… (see the article by Freyd). – Todd Trimble Oct 31 at 20:59
• I find it useful to keep subscripts in some cases where it makes clear some form of parallelism, e.g. if you want to talk about normal subgroups $H_1 \subset G_1, H_2 \subset G_2$, and their normalizers $Z_1, Z_2$, etc. So there are some cases where subscripts aren't necessary, but can provide more clarity than the alternative. – user44191 Oct 31 at 23:05
• Are you saying that the amount of subscripts is suboptimal, and one should substitute subscripts for something else? Perhaps a suburban Subaru submarine submachine? It's a good thing you're not formally complaining about superscripts, since that would be super hard to suppress and supposedly superior to the suboptimal case of the subscript. – Asaf Karagila Nov 1 at 9:53
• One easy hint for excessive indices: the notation $\exp$ for the exponential function. Write $\exp(a_2 y^2)$, not $e^{a_2 y^2}$ – Gerald Edgar Nov 1 at 15:15