Frank's Passions Within Reason

Passions within reasonSince reading Robert Frank’s The Darwin Economy, I have been working through his back catalogue. The original and innovative Passions Within Reason: The Strategic Role of the Emotions is the best of Frank’s books I have read so far.

Frank’s major proposal is that the emotions act as a commitment device. When a dispassionate calculation shows that it is better to cheat, the emotion of guilt can act as a constraint to the “rational” course of action. When a dispassionate calculation shows that it would be better not to incur the costs of punishing someone for cheating you, the emotion of anger acts as a constraint, and prompts you to punish them regardless of the personal cost to you. These emotional responses are particularly important where someone has a high discount rate and give a greater weight to present costs and benefits than those that come in the future.

These emotional traits are able to persist in an evolutionary setting as the “irrational” behaviour allows people to make credible commitments, which then allow them to gain the benefits of increased cooperation or protection from cheating. For example, if you can signal that you are trustworthy and your guilt will prevent you from cheating someone, you are more likely to be able to enter beneficial trades and cooperative arrangements. If someone knows you will get angry and attack them if they cheat you, even though it is not a rational thing to do in that particular instance, they are less likely to cheat you in the first place.

Frank deals well with two of the obvious attacks on this argument: why did such a convoluted way of undertaking these actions evolve; and what might prevent someone faking the emotion to prevent being cheated or so that they can enter into a trade where they can cheat?

On the first, Frank recognises that evolution does not come up with the perfect solution, but evolves responses to the environment over time. There is path dependence in the range of traits we have. In this case, the mix of high discount rates and emotional constraints to those discount rates were the solution that was found, and not the ability to rationally assess that the short-term costs are worth it for the sake of reputation. If humans are cognitively limited, the other option may not be feasible.

As an example of this path dependence (although possibly not an accurate example), consider the time before the formation of complex societies where there are repeated opportunities for mutually beneficial trade. Then, a high discount rate might have been generally appropriate, with those who cheated being favoured as they incurred no long-term costs when they did. However, when greater complexity of society arose, those with a constraint to their high discount rates did better as they were able to refrain from short-term cheating behaviour and develop reputations for fairness. Rather then tempering those high discount rates, which might still be important in some circumstances, the evolved response was emotions to constrain them in other circumstances. This story might not be true, but it creates a picture whereby opposing tendencies cancel each other out rather than a person evolving the perfect foresight necessary to consider the repeated interactions and how to act strategically.

For the second argument, the prevention of cheating, Frank relies on signalling theory. Frank argues that emotions are not easy to fake, which makes them reliable signals of someone’s intentions. If they could be easily copied, they would cease to be used as signals and the benefits to copying the emotion would disappear. It is the fact that they are hard to fake that leads to them being used as signals in the first place.

There are parts of this argument that trouble me, as the costs to faking emotions seem low. The costs exist, as work by Centorrino and colleagues demonstrates, but there are successful fakers of emotion. Many psychopaths can appear trustworthy despite their lack of trustworthiness. Possibly cheaters do not proliferate because the benefits to faking emotion are low, but if that is the case, why do we put such faith in these emotions as signals? Possibly cheaters are simply found out too often, so despite one successful cheat, their reputation ruins future opportunity. In that case, the signalling role is not as important as establishing reputation.

Regardless, Frank shows that it is possible for mixes of cheaters and honest people to emerge, with the frequency dependent selection ensuring that neither type disappears. If everyone is honest, there is no point signalling honesty or monitoring for cheating behaviour, so cheaters can thrive. If everyone cheats, the few honest people will spend great effort to seek each other out, and reap the benefits from specialisation and trade.

The book is not without its flaws, but given the strength of the fundamental ideas and the way Frank argues them, it feels like nitpicking to mention them. The only one I will note for the moment is that Frank does not give justice to rational behaviour, such as doing the right thing because people will find out. As experimental evidence shows, while humans can act in what seem to be non-self interested ways, people adapt to the probability of being caught cheating and the available rewards. Emotions are a commitment device, but in some cases, rational calculation can do the job. People with higher IQ are more trustworthy, yet I would argue they are not more emotional.

However, that is being pedantic about an excellent book. It is entertaining and interesting throughout. Twenty four years after release, it is still a must read.

One comment

Comments welcome

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s