On "The Elephant in the brain: Hidden Motives in Everyday Life" by Robin Hanson and Kevin Simler
"According to the reciprocal-exchange theory, conversations should be free to bounce around willy-nilly, as speakers take turns sharing new, unrelated information with each other. [...] Either listener might ask follow-up questions, of course. But as soon as their curiosity had been satisfied, they might be expected to turn around and share some new information of their own, regardless of whether it pertained to the previous discussion.
But this is not what human conversation looks like. Instead, we find that speakers are tightly constrained by the criterion of relevance. In general, whatever we say needs to relate to the topic or task at hand. Conversations can meander, of course, but the ideal is to meander gracefully. Speakers who change the topic too frequently or too abruptly are considered rude, even if[…]”
“And so it is with conversation. Participants evaluate each other not just as trading partners, but also as potential allies. Speakers are eager to impress listeners by saying new and useful things, but the facts themselves can be secondary. Instead, it’s more important for speakers to demonstrate that they have abilities that are attractive in an ally. In other words, speakers are eager to show off their backpacks.”
“If we return to the backpack analogy, we can see why relevance is so important. If you’re interested primarily in trading, you might ask, “What do you have in your backpack that could be useful to me?” And if your partner produces a tool that you’ve never seen, you’ll be grateful to have it (and you’ll try to return the favor). But anyone can produce a curiosity or two. The real test is whether your ally can consistently produce tools that are both new to you and relevant to the situations you face. [...] His backpack, you infer, must be chock-full of useful stuff. And while you could—and will—continue to engage him[…]”
the "criterion of relevance" as a constraint to signalling in conversations is not a proxy (signal) for the "size of the backpack" but instead shows a very concrete and rare skill: the ability to transfer and use knowledge one has to someone's benefit. I don't care about the number of tools someone has if he lacks the ability to use the appropriate tool in a given situation. The usefulness of someone's backpack is the product of the size of the backpack and the ability of the person to hand me and use the needed tool in a given situation. To produce relevant content is not just a signal, but a skill. I think Hanson and Simler took the signaling hypothesis too far in this case.