Multidimensional signaling

Link post

What do you in­fer about a per­son who has ugly cloth­ing? Prob­a­bly that they have poor taste (in clothes, or sub­cul­tures). But it could also be that they are too poor to im­prove their wardrobe. Or can’t be both­ered.

What about some­one with poor grades? The ob­vi­ous in­fer­ence is that they aren’t so ca­pa­ble at the sub­ject, but it may again be that they can’t be both­ered, or that they have more ur­gent things to do with their time.

And some­one who makes clever jokes? Prob­a­bly that they are smart and nat­u­rally funny, but if they had more time or effort to spend on this, it prob­a­bly helped.

For all kinds of traits that peo­ple might try to sig­nal with their be­hav­ior, some­one can send a bet­ter sig­nal if they have more money or time or self-con­trol. Even when the main sig­nal be­ing sent is not usu­ally thought to be about any of those things.

The rea­son this in­ter­ests me: if sig­nals of­ten di­vide the pop­u­la­tion into ‘bet­ter or richer’ vs. ‘worse or poorer’, I won­der if this would cause us to imag­ine that be­ing rich is as­so­ci­ated with be­ing bet­ter, even if the two were en­tirely in­de­pen­dent. (And similarly for wealth in other gen­eral-use re­sources, like self-con­trol and time).

In a sim­ple case, sup­pose there are just peo­ple with pretty clothes (who are both rich and have good taste) and peo­ple with ugly clothes (who ei­ther have bad taste, or lack re­sources or will). Then do ob­servers come to think of ‘rich good taste’ type and a ‘poor bad taste’ type? Or do they pay more at­ten­tion to the ac­tual struc­ture of the space, and know for in­stance that this doesn’t mean learn­ing that some­one re­ally has bad taste ac­tu­ally means they are prob­a­bly poorer.

New Doc 2017-10-15 (1)

Note that I’m not merely sug­gest­ing that a per­son with more wealth can send sig­nals to look like they are bet­ter—that much is clear. I’m sug­gest­ing that at a pop­u­la­tion level, if the wealthier peo­ple can’t be dis­t­in­guished from the bet­ter peo­ple on some axis, then ob­servers may come to think that the two are as­so­ci­ated in gen­eral, even if they are not at all.

If so, this would be im­por­tant, be­cause it would ap­ply in a huge range of cases of sig­nal­ing. So that the prop­er­ties of poverty and weak-willed­ness and such would ap­pear to us to be much worse than they re­ally were.