Crawl Across the Ocean

Tuesday, July 20, 2010

60. Signalling

Note: This post is the sixtieth in a series about government and commercial ethics. Click here for the full listing of the series. The first post in the series has more detail on the book 'Systems of Survival' by Jane Jacobs which inspired this series.


"Buying bread from a man in Brussels
He was six foot four and full of muscles
I said, 'Do you speak-a my language?'
He just smiled and gave me a vegemite sandwich"

In a world where everyone behaves the same way (e.g. the world of most economic models), being able to tell one person from another isn't all that useful. But as we saw a few posts back, and as common sense also indicates, people vary, with some people being more prone to cooperative behaviour than others.

If we imagine ourselves wandering around a world filled with a mix of those who are willing to cooperate with us, and those who will pretend to cooperate just to take advantage of us, then the importance of knowing who you can trust becomes obvious.

One way is to learn by experience. If you can remember who you've dealt with in the past, then you can shun those who've defected against you and only deal with people you've had successful dealings with before. But this still leaves a problem of what to do with people you are meeting for the first time, or people who you are dealing with in a new situation, or even people who defected on you before, but claim to have turned over a new leaf.

In these cases, we tend to look for signs that indicate we can trust someone. Maybe a common language or culture, or skin colour or alma mater, or a certain manner of dress or hairstyle, or a certain level of courteousness, or a certain credit score, or even a certain food choice (e.g. vegemite sandwich).

Thinking more generally, even someone's past behaviour could just be considered another type of signal to take into consideration.

Ideally, the signs that we look for will be hard to fake, since otherwise defectors might just try to pass themselves off as cooperators. Barring the use of memory altering technology, past experience with a person can be very hard to fake, which makes past experience with someone one of the best signals of their potential willingness to cooperate in the future.

Given a perfect ability to differentiate cooperators from defectors, even a group of cooperators as small as just 2 people could outperform a society full of defectors. But given a complete inability to differentiate cooperators from defectors, and lacking a way of keeping track of the results of prior dealings with people, then cooperators would become helpless against defectors - and the defectors will likely take advantage of the cooperators until eventually there are no cooperators left.

As we covered a while back, this distinction is what led David Gauthier to assume, in Morals By Agreement that people pursuing cooperation would be able to perfectly differentiate cooperators from defectors.

This is all pretty straightforward common sense, but as this series goes along and I talk more about how cooperative behaviour can evolve (or go extinct) in various circumstances, it will be useful to keep some basic points in mind, such as the importance of signalling cooperative intentions and remembering past interactions with people in sustaining cooperation.

* * *

A few other notes on the topic of signalling:

It's worth mentioning that signalling is only necessary when one party to a transaction has information that another party lacks (the information that will be signalled) so, from an economic perspective, situations requiring signalling are an example of one of the effects of asymmetric information , a topic which we discussed here a while back.

Of course, barring an ability to read minds or predict the future, you never really know for sure what the other person is going to do, so in that sense every transaction involves asymmetric information, which is a bit hard on economic theories which rely on the absence of asymmetric information as a key assumption.

* * *

Signalling problems can also be tied back to game theory via various 'signalling games' which investigate what messages might be sent and received between players under varying circumstances.

* * *

Some concepts related to the idea of signalling:

Cheap Talk - Signs that can be made with little effort and thus don't signify much. For example, if I ask you, 'Are you going to screw me over?' And you say, 'No', your answer would qualify as cheap talk, since it doesn't cost you much to make that statement. Or if cooperators tried to identify one another by wearing a yellow shirt, than anyone could wear a yellow shirt and pretend to be a cooperator. This notion is generally captured in the expression, 'actions speak louder than words' - since actions typically have a higher cost than words do.

Common Knowledge - Something that everybody knows, and everybody knows that everybody knows, and everybody knows that everybody knows and everybody knows, and so on. The Stanford Encyclopedia of Philosophy provides an example:

"A proposition A is mutual knowledge among a set of agents if each agent knows that A. Mutual knowledge by itself implies nothing about what, if any, knowledge anyone attributes to anyone else. Suppose each student arrives for a class meeting knowing that the instructor will be late. That the instructor will be late is mutual knowledge, but each student might think only she knows the instructor will be late. However, if one of the students says openly 'Peter told me he will be late again,' then each student knows that each student knows that the instructor will be late, each student knows that each student knows that each student knows that the instructor will be late, and so on, ad infinitum. The announcement made the mutually known fact common knowledge among the students."

Labels: , , , , , , ,


Post a Comment

<< Home