When reputation is at stake, punishment becomes more responsible

Jun 20, 2012

(Phys.org) -- The evolution of cooperative behaviour in people is often explained by the fact that it provides the opportunity to punish undesirable behaviour. However, such punishment is costly and the benefits for the person being punished therefore unclear. If the costs of the punishment exceed the benefits of the cooperation promoted by it, punishment would, from an evolutionary biological point of view, be pointless. Moreover, several behaviour experiments suggest that sanction opportunities are abused and that even cooperative behaviour is sometimes punished by fellow subjects.

Using the , Christian Hilbe and Arne Traulsen of the Max Planck Institute for in Plön have now developed a model that shows the dependence between , the evolution of and punishment. According to this, reputation could be the key for the successful evolution of responsible sanctions.

Responsible punishment does not initially seem to have any advantage for the person being punished and therefore does not fulfil the requirement for the of such behaviour. To resolve this problem, Hilbe and Traulsen developed a two-stage mathematical model.  In the first stage, the players can be either cooperative or uncooperative. Based on this, they must decide in the second stage whether or not they will punish others for previous behaviour. This shows that and justified sanctions are only carried through if the interactions can be observed by others.

“The decision by someone to punish others then affects not only the short-term relative advantages of the players, but also their reputation”, says Arne Traulsen.

A subject’s own reputation seems then to have a very high value, as individuals punish unfair behaviour, even when they must reckon with resistance. “We are prepared to pay a high price to maintain our reputation. The suspicion that someone is watching us is enough to increase our willingness to cooperate”, explains Christian Hilbe.

Punishment is therefore primarily worthwhile if it is used responsibly. Sanctions not only have the purpose of punishing uncooperative behaviour, but also work as a signal to outsiders. Only through responsible sanctions can the willingness to cooperate in the population increase. The tendency shown prior to this, to punish unfair behaviour, would thus be an advantage in the long term and balance out the costs incurred. The ability of humans to collect information through others, pass it on and so build up a reputation consequently seems one of the fundamental reasons for the particularly pronounced willingness to cooperate among people.

Explore further: The rich have more political clout in states, but stricter lobbying rules can narrow gap

More information: Christan Hilbe und Arne Traulsen, Emergence of responsible sanctions without second order free riders,antisocial punishment or spite, Scientific Reports 2, Article Number 458, doi:10.1038/srep00458

Related Stories

Punishment of egoistic behavior is not rewarded

Nov 14, 2011

The heated debate surrounding the German "state Trojan" software for the online monitoring of telecommunication between citizens shows that the concealed observation of our private decisions provokes public ...

Statistcal Physics Offers Insight Into Moral Behavior

May 06, 2010

(PhysOrg.com) -- It seems a little strange for statistical physicists to consider questions of morality in behavior. However, that is is just what a study at ETH in Zurich, Switzeralnd, is doing. Led by Dirk Helbing, the ...

Recommended for you

User comments : 11

Adjust slider to filter visible comments by rank

Display comments: newest first

Birger
5 / 5 (3) Jun 20, 2012
My favourite aricle about "altruistic punishment" in the jornal "Science" showed a famous film scene with this text:
"Make my day. Dirty Harry succintly informs a norm violator that he will take pleasure in administering altruistic punishment"
dogbert
2.5 / 5 (2) Jun 20, 2012
To benefit a group, bad behavior must not only be made detrimental to the person behaving badly, such consequences must be known to all and must be consistently applied.

You may call knowledge of consequences reputation, but it is the certain knowledge of adverse consequences which can be expected to moderate behavior.
antialias_physorg
2.3 / 5 (3) Jun 20, 2012
but it is the certain knowledge of adverse consequences which can be expected to moderate behavior.

Only in people with a selfish outlook (which arguably are the majority)
Moderation can also come from foresight. Willingness to work for the group (and being assured of reaping the benefits the group accrues). this is basically why all types of communist systems have tried to educate the people as to the benefits of common action and being aware as ones' contributer to the group rather than being a parasite of it. (Not saying that it did work particularly well - we all know how communism turnes out if some aren't willing to participate or retain the "give me more than my share"-attitude of capitalist societies)
TkClick
3.3 / 5 (3) Jun 20, 2012
even cooperative behaviour is sometimes punished by fellow subject
We can ask, for example, why the less successful scientists tend to punish these more successful ones? It's indicia of sectarian society, where people become entangled with their common belief and/or professional interests. In particular, in science the reputation is very important, because the contribution of scientific work cannot be judged in real time. Which is why the scientists prefer the publishing in reputed journals. This indeed introduces a bias into quality control of scientific work, which should be judged from longer perspective and it shouldn't depend on the form or quality of journal, where it was presented first. Whereas in another areas of human activity the quality of work is checkable a way sooner.
antialias_physorg
3 / 5 (2) Jun 20, 2012
We can ask, for example, why the less successful scientists tend to punish these more successful ones?

Do they? In my experience 'success' in science is determined by the quality of your work. And the work is the important part - not the success. If scientists were after money or fame they would be working for companies - not doing research.

In particular, in science the reputation is very important
Not at all. That's why working ins science is so neat. There are hardly any big egos floating around. And the few that there are have an unambiguous body of work to back it up with.

Which is why the scientists prefer the publishing in reputed journals

You want to do that because then you know it will be read. It's great to know that you could do work that was accepted to a big journal. It's neat to know that others think your work is good enough to be cited.
antialias_physorg
1.7 / 5 (3) Jun 20, 2012
This indeed introduces a bias into quality control of scientific work,

Peer review is done on anonymized papers. How can such a process be biased towards people instead of biased towards quality? That's the whole point: to take the bias (if any existed) OUT of the publishing process. The entire peer review process is set up so that the quality of work is checked BEFORE being published - as opposed to any other kind of publication process (e.g. in general journalism where the fact checking is usually confined to the author and there is no double checking going on at all)
TkClick
3 / 5 (2) Jun 20, 2012
Anonymous peer-review is OK for judging of general problem in large groups. When the subject or methods of solution become specialized, then the risk of competitive bias increases fast. From this reason for example string theorists are presenting most of their work in ArXiv preprints only: not only the string theory enables to pile math articles fast and their publishing in printed form would become too expensive, but in this small community everyone knows everyone. Therefore the string theory community faces the lack of qualified feedback, being too specialized. And the anonymous peer-review has a hidden drawback: just because the judges are anonymous, they're more vulnerable to hidden personal animosities at the moment, when they recognize the authors. After all, in similar way like the various critics at the anonymous internet environment. All things have two sides: the anonymity not only removes the cognitive bias - it means the lack of public control of the review process too.
TkClick
2.5 / 5 (2) Jun 20, 2012
This introduces a bias into quality control of scientific work
I didn't talk about peer-review right now, as the problem of reputation in science is more deeper and general. When the cooker prepares a bad meal, then the quality of his work can be judged immediately. Such a cooker is not dependent of bad reputation at all. But when some scientist writes the wrong or just plain useless article, then the lack of impact will manifest itself just after years. But this scientist wants his salary and another grants right now - how to solve this paradox? Well, the scientist will publish such an article into journal with high impact, like the Science and Nature and he will be judged by the impact of this journal. But from the same reason just these journals suffer with high degree of false dismissal of quality works. Nature Journal itself leads the record in the sloppy rejections of important findings, often leading into Nobel prices later
antialias_physorg
2.3 / 5 (3) Jun 20, 2012
When the subject or methods of solution become specialized, then the risk of competitive bias increases fast.

That may appear at first glance to be the case, but if you have ever worked in science you know it just isn't so. Publications are VERY specialized. There are no two people working on EXACTLY the same problem. Publications also are small steps buidling on previous work.
So whenever you get stuff to peer review it's no threat to your current work at all.

Additionally you're not the only reviewer. The reviews I have been involved in had these articles sent out to at least 5 different people (and no reviewer knows who the others are). So even if you figure out who the paper is from and if you have a personal grudge then it won't do any good to slam the paper even if you could (which you can't because you have to DEFEND your criticisms with valid scientific arguments. You can't just say "sloppy paper")
antialias_physorg
2.3 / 5 (3) Jun 20, 2012
Animosities would show up quickly in a peer review process (your criticism would not match that of the other reviewers).

But when some scientist writes the wrong or just plain useless article

If some scientist does then it doesn't get past peer review. That's what peer review is for.

But this scientist wants his salary and another grants right now - how to solve this paradox?

Science projects aren't funded based on the paper you write. They are already funded beginning to end (whether you write a paper in the interim or not). There's no connection between journal papers and grants.

Well, the scientist will publish such an article into journal with high impact, like the Science and Nature and he will be judged by the impact of this journal.
You make it sound as if someone just decides to do that independent of the quality of their work. Not so. Peer review, remember?

You should really stop making conspiracy theories out of something you know nothing about.
patnclaire
3 / 5 (2) Jun 20, 2012
I have been reading peer reviewed journals for fifty years. Peer Review in theory sounds great. Peer Review in practice can go awry. These same psychology journals have published articles over the years about anonymous comments having a harsh tone; Snarky is the current term. Anonymous commentors can write and say hurtful words but hide in the darkness of anonymity. Could this be one reason that the American Constitution Bill of Rights specifies being confronted by your accusers?
All reviewers know who the author is. Suppose the name is with held to protect the innocent, then there is the attribute of reputation. If the author has presented at any conference then the identity might be guessed by a reviewer. Slim chance but still possible.
Peer Review in software engineering does work, and works quite well. My own opinion of peer reviews can best be summed up by the quote that democracy is the worst form of government except for all the others.