50 Comments
User's avatar
Steven's avatar

I think you may have fundamentally missed something about the implied social framework. A certain degree of ambiguity in ranking people within groups is deliberate and actively desired to support group cohesion. Unfortunately, my source on this is memory from reading hardcopy, so it would take some time for me to track down the exact reference, but IIRC the typical social group is most stable when "The one at the Top" and "The one at the bottom" are clearly identified (and sometimes "The #2" if continuity of the Top is needed or it's necessary to identify who is qualified to challenge for the Top), but ambiguity is otherwise preferred for those who are neither the top or bottom so that they may treat each other as "equals". This social structure provides a clear floor and ceiling to social status within the group while also minimizing the incentives for petty sabotage or arms races within the middle ranks.

This is a significant reason why indirect signaling is used: it maintains plausible ambiguity between those who are broadly similar on the quality being signaled, so that they may more easily maintain status as presumed "equals" rather than seeing their relative superiority/inferiority frequently fluctuate over minor changes in the quality signaled.

Expand full comment
Robin Hanson's avatar

We already post many other exact numbers available about people. Why are these numbers I propose worse than the others?

Expand full comment
DavesNotHere's avatar

High income numbers make you a target for burglary, kidnapping, or home invasion. At that point, you might want to signal that you consume a lot of private security services.

Expand full comment
Robin Hanson's avatar

I'm happy to have that stat included in your public stat profile.

Expand full comment
DavesNotHere's avatar

I think most people with the relevant level,of wealth would prefer the possibility of keeping that data private. I am not suggesting an addition to your list of stats, but pointing out that there are countervailing forces encouraging people not to flaunt their wealth.

Perhaps this leads to the further question, why are the wealthy so indifferent to the level and effectiveness of crime prevention services provided by the government? A shallow answer would be that this is a public good, and they are happy to consume it instead in a private form where the relationship between their expenditures and their consumption is more obvious and direct. Somehow, I’m not very satisfied with that answer, but it seems conventional enough.

Expand full comment
Steven's avatar

I honestly have no idea what other exact numbers you are suggesting are commonly available about people, so you're going to need to be more specific if you want me to provide a compare and contrast.

Expand full comment
Robin Hanson's avatar

For example, here on X you can see a follower count.

Expand full comment
Steven's avatar

Okay. Now, from that signal, what are you actually able to meaningfully conclude about anyone?

A signal that doesn't drive a decision is effectively just noise. What social decisions do you suggest would be driven by that number? Most people aren't X users and even most X users aren't significantly socially invested into it, so it's certainly not remotely as universally socially relevant as health or income. Does the exact number of followers really matter or is it more merely the magnitude of it by which comparisons are made? Are there particular thresholds that divide X users into effectively different social classes?

Expand full comment
Astiem's avatar

The signals you took are quite irrelevant but you could take some others signals. For example, you could see the total number of comments a certain used wrote, in this way you'd be able to see who are the most engaged in the platform. I don't think that knowing in which forum they commented on would be a little violation of privacy, but maybe, it could be a little weird but you could divide that comments on the emojis attached to it. Someone who posts his comments with an average of happy emojis, as the emojis says, it's happy or satisfied, while someone who has an average of angry emojis could be unsatisfied. The latter it could be quite ridiculous since we don't know can't know what they're commenting about, so every reaction could be justifiable, but this could serve an instance that new signals could also be created in order to identify social classes basing on the topic.

Expand full comment
Isha Yiras Hashem's avatar

You're missing the most important signal - the signal that you're socially aware

Expand full comment
Robin Hanson's avatar

How does posting stats prevent that?

Expand full comment
Isaac's avatar

The signal that you didn’t mean to signal at all.

Expand full comment
Boring Radical Centrism's avatar

Do countries or other communities where such things are public knowledge actually have less signalling?

Expand full comment
Robin Hanson's avatar

Good question.

Expand full comment
polscistoic's avatar

There are signals and signalling behavior. There are also meta-signals and meta-signalling behavior.

Meta-signalling: Signalling that you know how to signal something.

For example: “I like beautiful things” - which just happens to be expensive. To signal this without risking being seen as gauche, you should ideally by unaware that these are tightly correlated in your esthetic taste.

Knowing how to convincingly signal the above requires knowledge of what is the upper-caste “way of looking at things” (both literally and figuratively). To acquire such knowledge (to acquire any kind of knowledge) is costly. And that is important: The distinguishing feature of a signal, as opposed to a mere “message”, is that signals are not cost-free.

In addition to signals and messages (notice that the difference is a continuum, not a dichotomy), the concept of self-binding is useful. Convincing signalling behavior often borders on self-binding (it may be labelled pseudo self-binding). If you can convey to others that you are simply unable to conceive of “beauty” in any other way than you do, you are self-bound to a certain presentation of self. Meaning that you cannot be suspected of only being a very clever parvenu.

Such self-binding is difficult to achieve, because if you consciously try to achieve it, you are already way off the mark (as old zen masters would tell you in a different context). The most convincing way to signal good taste (which just so happens to be expensive) is to be brought up in a way that imprints a certain taste on you. So here it is primarily parents, not you yourself, who have borne the costs to imprint a particular type of signalling behavior (including meta-signalling behavior) in you.

What is here said using “beauty” as an example can be said related to most things that people normatively and/or esthetically value.

Most human interaction, including “culture” as such, can be grasped with variations of these concepts.

Expand full comment
Vasily Kuznetsov's avatar

While I support making signaling cheap and efficient, I also value a lot the freedom to choose what I want to signal and what I don't want to signal. The full transparency solutions remove this freedom and I find this cost excessive.

Expand full comment
James McDermott's avatar

But a person with a certificate of health is not as attractive as a person who wins at sports. Remember that just as behaviour is often about signalling, so too *reading signals* is not really about accessing the underlying information. This is a "fitness maximiser versus adaptation executioner" error.

Expand full comment
Robin Hanson's avatar

Posting stats doesn't prevent people from playing sports.

Expand full comment
Steve Phelps's avatar

The reason that we pay attention to costly signals is not necessarily because we consciously and rationally infer that they are reliable, but because costly signalling is an equilibrium under the coevolution between genes and culture (see "the handicap principle"). As a result there are strong biases in the kinds of signals we are likely to pay attention to. We may find ourselves attracted to articulate and good looking people despite the fact that we know that their Wikipedia page says they don't have degree and are suffering from terminal cancer.

https://en.m.wikipedia.org/wiki/Handicap_principle

Expand full comment
Catherine Caldwell-Harris's avatar

I appreciate Robin's idea, but note the distinctions between honest signalling, fake signalling, costly signalling,and hard-to-fake signalling. Lots of people wear expensive clothes, use big vocabulary, as part of trying to convince people of wealth, intelligence and so on.

Expand full comment
Caperu_Wesperizzon's avatar

Ockham’s razor: we like deceiving others, and maybe deceiving ourselves about our ability to come out ahead in the collectively losing game of defection.

> A related idea is to require paternity tests of newborns. Even though for most new dads the cost of a test is much less than the value of info they’d get from it, they don’t ask for such tests because of the bad signal that asking would send the mom.

Well, why do so many people take vows they won’t honor in the first place? Signalling, defection and having your cake and eating it too all the way down.

Expand full comment
Daniel Melgar's avatar

This would be awful for legacy admits and recruited athletes who actually graduate from Ivy League schools. To discover their fraud (and that of those schools) would have serious consequences. Not to mention all those members of congress and government who could now be exposed as frauds. No longer would a congressman or senator be able to impeach someone who didn’t graduate from college or who was intelligently discussing subjects outside their own field or profession.

I love it!

Expand full comment
Dominic Ignatius's avatar

Why is truth more valuable than status?

Expand full comment
Robin Hanson's avatar

Who said it was?

Expand full comment
Dominic Ignatius's avatar

It feels like a common thesis in your posts. Here, it looks like you would prefer making objective numbers visible ("truth") over the current regime ("status").

I don't entirely disagree, but I'm trying to figure out why truth is better than signaling for status beyond taking it as an axiom.

Expand full comment
DavesNotHere's avatar

Or put another way, why discourage people from expending their resources to gain status, unless you have a cheaper way for them to get it? In which case, it sounds like a big entrepreneurial opportunity. Or unless you think status is worthless, in which case just put it in line behind junk food and gambling.

Is the assumption that there is only a weak correlation (or none?) between the deceptive signal and the status gain it may achieve? Is there some better way to allocate status, or to predict the outcome and discourage the losers from wasting their resources?

Is a degree from an Ivy League school really a deceptive signal of smarts? I thought the problem was that actually attending them doesn’t actually improve your smarts much, it just filters out the applicants who are lacking.

The problem with signaling (for me) is that it keeps people in denial about the real reasons they do things, and so may mislead them into supporting things that they would actually prefer to avoid. It encourages overconfidence where caution is called for, and cover-ups where we need corrections.

Expand full comment
DavesNotHere's avatar

I guess that is my answer to the question in the OP - people don’t want direct signals because they want to,avoid admitting (even to themselves) what is motivating them. Seeking status is so low status!

Expand full comment
qbolec's avatar

What if the ambiguity of the signal is the feature? An expensive car can be bought because I am affluent or because I've took a loan to only appear wealthy. This is obviously good for pretenders. But it is also somewhat good for really rich in the context of peer groups - I think you wrote earlier that it is crucial that status within in a peer group is not legible as that's the pre-requistie for treating everyone as peers.

Expand full comment
Kris Buote's avatar

"But if you are willing to consider such rules using a cost-benefit framework, these seem promising."

What is the benefit? So we don't indulge in "wasteful" signaling? It's not clear to me that it's wasteful. Also, if you used "direct comparison methods" (e.g. public IQ records) it eliminates necessary competition. Veiled signaling allows less-smart people to pretend they're more smart and to compete in those realms. What you're suggesting, as I read it, is an absolutist hierarchy of "smartness" which doesn't allow healthy competition. People will simply point at the list and say "look, I'm smarter, you're invalid".

All in all, I would be curious to hear you elaborate on why making all personal information public is a good thing. That is the crux of this argument and it's not clear to me.

Expand full comment
Jack's avatar

The ways we signal wealth today are certainly very wasteful. One has to buy an expensive impractical car, or a yacht. What if that money could be saved?

A challenge with more direct means of signaling (say, a trusted third party vouching for your wealth) is that there has to be some measure of plausible deniability. No wealthy person wants to be caught trying to impress people. That's what poor people do, it's too needy. No, I just happen to love the freedom of my yacht, and the way my Ferrari drives, and the fine craftsmanship of my Rolex watch. Well yes it's true that those things are expensive, but really that's beside the point.

Expand full comment
Jack's avatar

Speaking of signaling, here's a clever dating strategy I heard about. A guy would raid the wastebaskets near ATMs to get the balance receipts people had thrown out. The ones with especially large balances he would keep, and carry them around with him. Whenever he met an attractive woman and struck up a conversation, he would nonchalantly take one from his pocket (as if it was the only convenient piece of paper on hand), write his name and number on the back, and give it to her.

Expand full comment
Apple Pie's avatar

I think this exaggerates the degree to which people send signals. Personally, I try harder to *avoid* sending signals than I try to send them. People usually get to know me pretty well by listening to me talk for a few minutes; I'm not that hard to read. In fact, most people aren't that hard to read. My sense is that the kind of person who spends a lot of thought actively signalling is basically a Narcissist.

The irony is that, if I'm right and signalling is mostly confined to people who are rather Narcissistic, then it actually leads to your conclusion more directly than if you are right that a large fraction of human behaviors are signalling. Taking options and tools away from Narcissists seems like a pretty smart idea to me!

Expand full comment