[roywiggins]

The problem with "signalling" theories is that they can explain nearly anything. Why did you do X? Why, to show other people that you are the sort of person who does X. Even if you didn't know it at the time.

It's hard to see how you would go about falsifying it. It's way too convenient as an explanation for whatever you want. It's less a theory and more a very particular lens to view the world through. Nearly anything can be slotted into it.

[Judgmentality]

Any theory that can't be invalidated is not worth seriously debating. It may be true or it may not - what difference does it make if it's impossible to know?

Person A: "Social signalling explains all of human behavior!"

Person B: "But how do we know that's true?"

Person A: "Who cares? It explains everything."

Person B: "So what can we do with this knowledge?"

Person A: "Explain things we already observed to be true."

Person B: "So what have we gained from this information?"

Person A: "A theory that explains everything."

Person B: "But that theory doesn't actually teach us anything new?"

Person A: "Yep! We already know everything! Wooo!"

[ravi-delia]

There are very real situations where signaling actually explains something. Mostly they involve people doing something that seems irrational or hypocritical on the surface; the hidden variable is of course social approval. In other situations, however, 'People do X to signal they do X' really is just a complicated way of saying 'People like to do X'. Once you have the action, it really doesn't make any sense to bother figuring out why.

[username90]

It doesn't explain porn. Most people watch porn in isolation and never talk about it so it can't be based on signaling. Hence it doesn't explain everything.

[LeifCarrotson]

It's valuable because it explains some actions that may be foreign to the observer. It may not be particularly common in most people's lives, and it may not occur to some people at all, but it's especially true in some politics or business relationships that are incomprehensible if you don't have it as a tool.

[Judgmentality]

I really don't see how a theory that can't be invalidated is useful for anything, other than a rhetorical device. Apart from a thought experiment along the lines of The Flying Spaghetti Monster, what good is a theory that can never be tested? It's a tautology.

[michaelt]

Well, you can also use it to transmute your feelings of inadequacy into feelings of cleverness and insight.

It's not that life's passing me by as I punch a keyboard at a big multinational, no siree, it's those those people posting pictures of themselves having fun who are the real fools.

[ZephyrBlu]

I find it hard to believe you have never seen social signaling 'in the wild'.

A lot of behaviour that otherwise seems nonsensical can be explained by social signaling.

Why do rich people often spend tons of money on things they don't need or use? To impress other rich people.

[Dylan16807]

You are describing a theory of "some things are signalling", which is pretty different from "everything is signalling"

[ZephyrBlu]

Yes I know. The person I was replying to seemed to not believe in signaling in general though.

[Judgmentality]

I find it hard to believe you have never seen the flying spaghetti monster 'in the wild'.

A lot of behaviour that otherwise seems nonsensical can be explained by the flying spaghetti monster.

Why do rich people often spend tons of money on things they don't need or use? To impress the flying spaghetti monster.

[feanaro]

This is not equivalent since it is reliant on the existence of the FSM, which is not usually taken to be proven, unlike (hopefully) the existence of other people.

[stefan_]

But it doesn't explain anything. It's ability to explain is entirely unproven.

The social sciences for a long time were just ancient white men writing ever larger treatises "explaining" everything. None of it had any explanatory power and fell apart at the slightest mention of an actual experiment.

[gweinberg]

Sure, but it's not to hard to think of examples of things which signalling theories would say shouldn't happen. For example, people should not be impressed by signals that are known to be easy to fake or that are in no way impressive in the first place. This may seem trivial, but it does have implications. For example, if you claim "I did X over the weekend" and I reply with "I call bullshit", it implies not only that I doubt your claim, but also that at least to some extent I care whether it is true, and also that you care whether I believe you or not (or at least that I think you care).

[Judgmentality]

> For example, if you claim "I did X over the weekend" and I reply with "I call bullshit", it implies not only that I doubt your claim, but also that at least to some extent I care whether it is true, and also that you care whether I believe you or not (or at least that I think you care).

This doesn't refute anything, in fact it reinforces the theory! Why did I make a claim about what you did over the weekend at all other than to signal to you about it? And of course I care what you think, because that's the entire point of social signalling.

I could say your entire response was a signal itself, and you're not consciously aware of how signalling is driving your behavior. At this point the theory becomes useless.

A good theory has hypotheses that can be (in)validated.

[roywiggins]

The point is that you might as well just not do the thing and say you did it, because nobody will ever verify it. So the doing of it can't have been motivated entirely by signalling; it would have been easier to do something else and just lie.

[fourier456]

Before we had a theory of planetary motion, epicycles were used. While obviously we stopped adding circles everytime a planetary prediction was off, the fact that some waves can approximate any reasonable function *is* interesting in it’s own right. I take a similar stance with signaling.

[nathias]

Every theory can be invalidated by analysis or empirical verification, the latter is not the only criterion of truth even for the most empirical epistemologies.

[nojs]

I think the difference is in how extensively you broadcast it. I like gardening, but I don’t care if nobody knows that I do. I think this is evidence that I’m not doing it for signalling reasons. So in that sense it’s falsifiable.

[lordgrenville]

I think maybe the right term here is not "unfalsifiable" but David Deutsch's "explaining too much". https://en.wikipedia.org/wiki/Explanatory_power#Overview

[SamBam]

But if you do post about it (perhaps you're hoping to find like-minded gardeners in your community?) and I say you're just doing it to signal, how would you falsify that?

[nojster]

Isn’t sharing it publicly already a counter argument to “I don’t care if nobody knows about it”? Thus you wouldn’t have to falsify.

We could think of a scenario where his neighbor shares information about his (the authors) garden. In this case, if the author would seek measures to delete these posts, it might be interpreted as “signaling that he does not care and even doesn’t want others to know”.

I agree with you that signaling always shows intention (ie wanting something). You can “want not to want something”, though (:

[SamBam]

I think if you define "signaling" as "wanting someone to know about it" then of course any social media communication is "signaling," but then is any communication at all. Oviously when I speak I want to be head, so it's not "I don't care if nobody knows about it."

But I think that's far too broad a definition of "signaling." The original article says "Building social capital through signaling."

When you post about your gardening to see if there are any other gardeners you can connect with, I don't think it's necessarily to "build your social capital," yet you can prove that it isn't.

[kubanczyk]

The theory has one big problem, but you are misdiagnosing it.

> Why did you do X? Why, to show other people

Try this approach instead:

Would X ("a trip to Athens") happen if there was no perspective to ever mention X in any future communication between humans? (Yes/No)

In this take the theory can become verifiable with some work.

Alas, the assumption is that you know you'll be cut off from transferring* bits of information from your brain to other brains. As humans are very social, this is almost useless. In actual world, even if you're going solo to Mars one-way, you probably will communicate back to Earth and more colonists may come join you (think => social status). If I am sailing around the world alone, I can still expect to return and write memoirs. So the only remaining things are the most shameful i-am-never-telling-that-anyone personal secrets. And that's quite a narrow use.

So, saying that Ivy League wouldn't happen in <some out-of-this-world scenario where humans do not socially interact> is very impractical.

[*] I'm saying the theory doesn't judge whether the signal (the information that flows) is to be trusted or untrusted. I think most commenters here wrongly conflate "signalling" with "slightly lying".