You maybe saw this picture going around your social media a couple days ago. I did, but I’m connected to a lot of mathematics people who were naturally interested. Everyone who did see it was speculating about what the story behind it was. Thanks to the CBC, now we know.

So it’s the most obvious if least excitingly gossip-worthy explanation: this Middletown, Connecticut deli is close to the Wesleyan mathematics department’s office and at least one mathematician was too engrossed talking about the subject to actually place an order. We’ve all been stuck behind people like that. It’s enough to make you wonder whether the Cole slaw there is actually *that* good. (Don’t know, I haven’t been there, not sure I can dispatch my agent in Groton to check just for this.) The sign’s basically a loving joke, which is a relief. Could be any group of people who won’t stop talking about a thing they enjoy, really. And who have a good joking relationship with the deli-owner.

The CBC’s interview gets into whether mathematicians have a sense of humor. I certainly think we do. I think the habit of forming proofs builds a habit of making a wild assumption and seeing where that gets you, often to a contradiction. And it’s hard not to see that the same skills that will let you go from, say, “suppose every angle can be trisected” to a nonsensical conclusion will also let you suppose something whimsical and get to a silly result.

Dr Anna Haensch, who made the sign kind-of famous-ish, gave as an example of a quick little mathematician’s joke going to the board and declaring “let L be a group”. I should say that’s not a riotously funny mathematician’s joke, not the say (like) talking about things purple and commutative are. It’s just a little passing quip, like if you showed a map of New Jersey and labelled the big city just across the Hudson River as “East Hoboken” or something.

But L would be a slightly disjoint name for a group. Not *wrong,* just odd, unless the context of the problem gave us good reason for the name. Names of stuff are labels, and so are arbitrary and may be anything. But we use them to carry information. If we know something is a group then we know something about the way it behaves. So if in a dense mass of symbols we see that something is given one of the standard names for groups — G, H, maybe G or H with a subscript or a ‘ or * on top of it — we know that, however lost we might be, we know *this* thing is a group and we know it should have *these* properties.

It’s a bit of doing what science fiction fans term “incluing”. That’s giving someone the necessary backstory without drawing attention to the fact we’re doing it. To avoid G or H would be like avoiding “Jane [or John] Doe” as the name for a specific but unidentified person. You can do it, but it seems off.