Posts Tagged ‘game theory’

How Twitter Can Spark Revolutions

Penetrate the hoo-ha and self-aggrandising bluster about attributions of what, or who, caused the Arab Spring, and there might actually be some truth to the notion that social media, and Twitter in particular, were pivotal to sparking the revolutions that have punctuated the Middle East and North Africa.

This is a bold claim, but one I think might be justified by looking at the problem of starting a revolution through the lens of game theory, and the Stag Hunt in particular.

The problem with kicking off a revolution in a nation run by a paranoid autocrat is you simply don’t know whom to trust. Sure, you can probably assume that most of the population would be happy to see the tyrant turfed from power – after all, the nature of such regimes is that few benefit at the expense of many, and it’s the many who want to see the revolution get underway. But even if only one in ten individuals is pro-regime, or likely to dob you in to the regime’s thugs to save their own neck, the risk of voicing your anti-government attitudes is high, especially given the potential cost if you grumble to the wrong person.

This reluctance to fly one’s flag is one of the great hurdles to overcome before the people can band together and use that potent resource of overwhelming numbers to bring down even a powerful minority autocracy.

This dilemma can nicely represented by the Stag Hunt:

In this context, hunting Stag is declaring one’s revolutionary intentions and agitating for a better government that can benefit all. Hunting Hare is keeping one’s head down and suffering the regime. Should two revolutionaries get together, they can cooperate to help bring down the government and implement a better system with less corruption and waste etc. However, should a revolutionary meet up with a non-revolutionary, they risk being exposed, and punishment could ensue, while the non-revolutionary just continues plodding along under the oppressive regime.

The trick with the Stag Hunt is there are two equilibria: Stag-Stag (SS); and Hare-Hare (HH). It’s the mixed strategy of Stag-Hare (SH) that is unstable, and any system over time (excluding factors like non-random pairing for the time being) would expect to gravitate towards one of the two equilibria. The thing is, the basin of attraction of HH is significantly larger than that for SS, so most systems (75 per cent, to be precise) descend into the lower-payoff equilibrium of HH (read Skyrms, 2004 for more).

So, most societies where declaring one’s anti-government sentiments in public can be unwise, you’d expect people to more or less uniformly keep their heads down and play Hare all the way. And that’s what you saw in places like Libya. People were doing it tough, but that’s preferable than agitating towards something better and getting dragged away in the middle of night.

There are several ways of nudging the equilibrium from HH to SS, but there’s a considerable gulf of SH to overcome in the process. Spend too long in that gulf, and things will rapidly drop back to HH.

One way of helping bridge that gulf is through sharing reliable information: if agents don’t know whether their neighbour is going to play S or H, they’re more likely to play the risk-averse strategy and play H themselves. As will their neighbour. Yet if they have information that makes it more likely their neighbour will play S, then they can more safely play S themselves.

This is precisely what I believe Twitter did: it facilitated the communication to the broad population that each individual wasn’t alone in their anti-government sentiments and, as such, if they broadcast their anti-government intentions, they’re likely to be received by another anti-government individual.

Now, Twitter can’t solve the whole problem of starting a revolution, and the Stag Hunt doesn’t represent every nuance of the revolutionary dilemma, but if there’s something to this analysis, then perhaps Twitter, and other social media, did in fact play a role in sparking off, if not finishing, the Arab Spring revolutions.

Ends and Means

I call it “pulling a Cameron,” in reference not to the present British Prime Minister, but to the broadcaster Deborah Cameron who handles the morning slot on Sydney’s ABC Radio 702.

A common refrain a few minutes in to her maddeningly predictable morning routine of following the happenings of the first several pages of the Sydney Morning Herald is to enquire of some expert or other: “what are we doing to prevent X from ever happening again?”

And by “X” I mean whatever undesirable event has appeared on the front pages, whether that’s a case of callous bullying in our schools, a death by accident or some other unsavoury turn of events.

One recent example was the tragic death of a young university student at a rural college after being thrown from an ex-racing horse that was being used to train horse riding skills. The horse was deemed safe for students yet it possessed a fierce distemper that flared on that day, throwing the student

The question posed by Cameron, seemingly predictable and justifiable in the circumstances, was along the lines of: “what are we doing to prevent more deaths of students during riding training?”

The presumption is that the outcome is unacceptable so, clearly, our current process that led to this outcome must be deficient.

Yet that’s a presumption that is unfortunately as fallacious as it is common.

For we chase outcomes on an ad hoc basis at the risk of employing processes that undermine our very intentions to produce better outcomes. In short: we focus myopically on each good or bad outcome at the danger of forgetting that it’s good processes that are of primary importance.

And even a good process – nay, the optimal process – can result in bad outcomes from time to time.


Moral Dynamics

One of the privileges of being a philosopher is you can create new terms, define them how you please, and damn be to any conventions that would have the term used otherwise. So, I’ve created a few new terms – well, at lease one is new.

Here they are, in conceptual order:

Moral Diversity: the phenomenon I’m interested in understanding and explaining, namely the existence of persistent and intractable differences of opinion over what is considered good and bad, and the norms that promote good behaviour.

A key element of Moral Diversity is moral disagreement which, if it truly is intractable, poses a problem for any realist, objectivist or generally monist approach to ethics. I would suggest that Moral Diversity is a very real phenomenon, although I acknowledge that I could be proven wrong.

Moral Ecology: the notion that moral norms are not hard-and-fast rules but strategies employed to foster in-group cooperation and out-group competition, and these norms emerge in response to the environmental conditions around them, which includes the strategies employed by other individuals and communities.

As there is no one set of norms that best promotes in-group cooperation and out-group competition in all environments, and there is no one set that forms a stable equilibrium within one environment, and because new norms will inevitably emerge and compete with existing norms, there will always be a pluralism of norms that interact in a dynamic way.

I argue that Moral Ecology is the best way to understand morality as a natural phenomenon and to explain the existence of Moral Diversity.

Moral Dynamics: the study of the moral norms within a particular environment, ostensibly with the intention of finding the optimal set of norms that will form the most stable equilibrium and which yields the aggregate outcome closes to the Pareto optimal level, while resisting invasion by new norms and behaviours, particularly ones that are inclined towards defection.

If Moral Ecology is the correct way to understand morality as a natural phenomenon, the Moral Dynamics is a new approach to studying morality, not to find the single best hard-and-fast set of rules that works in every situation, but to find the different and dynamic sets of norms that work in different environments.

OK, rip in to them.

Why Cooperate?

There’s every possibility that I’ve missed something utterly obvious, but I’ve been reading up on the fickle nature of cooperation for my thesis, and I’ve found what appears to be a gaping hole in the literature.

There are countless studies that explore the challenges of encouraging cooperation – primarily via the use of the Prisoner’s Dilemma as a tool. There are also plenty of studies that look at cooperation from a social perspective, at the motivations that individual agents employ and the psychological benefits they receive from group activity.

But I’m yet to find a paper that clearly defines what cooperation really is and, more importantly, it’s benefits.

It seems the existing studies skip the question: why cooperate?, and barrel forth into questions about the difficulty of encouraging cooperation, apparently assuming that cooperation is a universally desirable outcome. Maybe it is, but I want to know precisely what are the benefits of cooperation that makes it so attractive.

So far I’ve come up with six reasons why cooperation is deemed desirable:

1) Force multiplication

Cooperation allows two or more individuals to combine forces to perform a task that would be impossible by one individual alone. A simple example would be lifting a heavy weight (we’ve all asked friends and family to pop around to shift that couch).

In fact, the task could be defined as anything that is beyond the capacities of an individual, whether that capacity is physical (strength), intellectual or spatiotemporal (manipulating two distant objects simultaneously). This doesn’t necessarily imply an improvement to efficiency, just that a previously impossible task is rendered possible.

2) Division of labour

Cooperation enables a large task to be broken up into smaller sub-tasks which are easier to perform than the whole task taken as one. This not only makes one large complex task into smaller simple tasks, but it allows a large serial task to be parallelised, thus improving efficiency.

3) Sepcialisation

Cooperation allows individuals to devote a greater proportion of their finite resources towards improving performance at a particular task. Combined with division of labour this not only improves efficiency but can also enable tasks that were previously impossible without specialisation.

Specialisation also allows individuals to capitalise on their intrinsic strengths and mitigate their intrinsic weaknesses by cooperating with a sympathetic individual.

4) Coordination

Division of labour and specialisation in turn allow greater coordination through devoting finite resources specifically towards directing effort in a way that is more efficient than if it is undertaken on an ad hoc basis. The benefit of coordination would only emerge if the energy expended coordinating is more than made up for by increased efficiency or productivity in the task at hand (a problem for corporations riddled with middle managers today…).

Coordination can also prevent conflicts of interest and potentially costly clashes (see the donkeys in the pic).

5) Trade

Cooperation also allows one individual to ‘trade’ a surplus for a surplus produced by another individual. This trade can be literal, such as trading goods, or it can be figurative, such as trading labour.

6) Risk mitigation

Cooperation enables a task to be unshackled from being dependent on any one individual, such that if that individual is in some way prevented from performing that task, the entire endeavour doesn’t collapse around their ears.

I’m sure there must be more benefits to cooperation. I reckon economics must have studied cooperation extensively, but I’m not as familiar with economic texts, so don’t really know where to dig to find the answers. Most of my research has been in evolutionary biology, behavioural ecology, game theory and ethics, and it seems the benefits of cooperation are largely taken for granted in these fields.

If you think I’ve missed anything, or you have some tips on where I can read up on research on cooperation, please do let me know.

Political Philosophy and EVE Online

Even if you don’t dabble in massively multiplayer games, EVE Online is worth a look just for the revelations that emerge from creating a loosely regulated world and opening it up for nerds to play with.

This interview with The Mittani, who is CEO (or guild leader) of Goon Fleet, the largest corporation in EVE, is solid gold. Enlightening even. It touches on politics, human nature, psychology and even has traces of game theory lurking just under the surface

One of the greatest advertisements of all time.

First, a bit of context for those unfamiliar with the game. EVE is a space-based massively multiplayer online (MMO) game where thousands of players flit between hundreds of solar systems, each with unique planets, moons, asteroid fields and space stations, and they mine, trade and fight. Often they fight each other.

EVE has one of the most active player-versus-player (PvP) communities of any MMO. A large tract of space in EVE is called nullsec, meaning it is effectively lawless. While a player will become an outlaw and be hunted by computer-controlled authorities for attacking another player in high security (highsec) space, in lowsec there’s no automated retribution. It’s true frontier stuff.

Players can also band together to create corporations, mining the rich resources in nullsec, using those resources to build ships, equipment and even space stations. These corporations effectively gain sovereignty over that sector, and they protect it from invasion by other corporations.


Chaos, Levels of Explanation and Interdisciplinarity

I’ve been thinking a lot about interdisciplinary research (IDR) of late. (One day I’ll spend a lot of time thinking about finishing my thesis, but hey.)

It seems that one of the most fundamental questions to ask is: why do we have separate disciplines at all?

Seems obvious, but often the unanswered obvious questions are the most interesting. Delving into them can reveal something illuminating about our assumptions about how things are, and even reveal some false intuitions.

The simple answer might be that there’s no one discipline that can tackle every question we might want to ask. Okay, why?

Well, probably because such a discipline would be unmanageably complex. Far easier to carve up nature – and the questions we want to ask about her – into bite size pieces.

But why carve it where we do?


What WikiLeaks is About (Hint: It’s Not About Exposing Individual Wrongdoing)

It seems my previous post may have been made in haste. After reading this enlightening post on Zunguzungu about Julian Assange’s philosophy, it’s now clear that his intention isn’t just to provide a medium by which whistleblowers can expose individual cases of wrongdoing, he’s instead attempting to alter the communication and secrecy landscape entirely, thus eroding the very possibility of what he calls ‘conspiracy’.

Where details are known as to the inner workings of authoritarian regimes, we see conspiratorial interactions among the political elite not merely for preferment or favor within the regime but as the primary planning methodology behind maintaining or strengthening authoritarian power.

Authoritarian regimes give rise to forces which oppose them by pushing against the individual and collective will to freedom, truth and self realization. Plans which assist authoritarian rule, once discovered, induce resistance. Hence these plans are concealed by successful authoritarian powers. This is enough to define their behavior as conspiratorial.

His idea is to basically remove the capability of ‘conspirators’ to be able to communicate effectively and in secret, thus removing their ability to conspire. The leaks also turn the organisation on itself and fosters paranoia, thus further reducing its ability to ‘conspire’.

The more secretive or unjust an organization is, the more leaks induce fear and paranoia in its leadership and planning coterie. This must result in minimization of efficient internal communications mechanisms (an increase in cognitive “secrecy tax”) and consequent system-wide cognitive decline resulting in decreased ability to hold onto power as the environment demands adaption. Hence in a world where leaking is easy, secretive or unjust systems are nonlinearly hit relative to open, just systems. Since unjust systems, by their nature induce opponents, and in many places barely have the upper hand, mass leaking leaves them exquisitely vulnerable to those who seek to replace them with more open forms of governance.

This is a kind of assymetric information war that could only be fought in the internet age. And, in a way, it was probably inevitable – remember Stuart Brand’s “information wants to be free”, meaning not that information ought to be made freely available, but that the cost of distribution is so low that it becomes harder and harder to keep it from moving around.

I find Assange’s essay compelling, but I still have concerns. They centre around trust:

How do we know we can trust Assange and WikiLeaks?

For the time being, I think we can. But the very lack of transparency in its own workings – by Assange’s own theory – make it ripe for descending into a conspiracy of its own.

Yet opening up WikiLeaks would likely cause it to stop functioning, as the leakers would suddenly be exposed, and the organisation itself would be vulnerable to being taken down, whether through its leaders being arrested or its servers disabled.

This suggests a paradox: in order for openness to thrive, there must be some restrictions on some information.

And I don’t believe there is any way to resolve this paradox satisfactorily.

That’s not to say this paradox is unique to this context; Hobbes faced the very same paradox when considering how to prevent people from descending into a war of all against all. His solution was an all-powerful sovereign state that has the power to force people to cooperate without defecting – to take a Prisoner’s Dilemma analogy.

But… how do the citizens know the sovereign won’t defect on them? I.e., how do they know the state won’t abuse its power and engage in some kind of conspiracy to further its own ends to the detriment of the people?

There is, to date, no perfect solution to this problem. The separation of powers, a free press and as much transparency as possible is the best we’ve come up with until now. And the very existence of WikiLeaks suggests that even this hasn’t sufficient to date.

Another concern is whether full transparency is always a good thing. I see it as being a sliding scale: more transparency means more bad acts are caught, but also more benign acts might also be inappropriately exposed (say, revealing a political leader is gay or atheist, causing a perfectly competent leader to be ousted for reasons irrelevant to their capability to perform their job).

Some things should remain private. Although too much privacy and secrecy means less benign acts are exposed, but also makes it easier for bad acts to occur unseen.

There’s no perfect solution to this problem. It’s a trade-off.

So, check and balances. Checks and balances. It’s the best we can do. And while WikiLeaks is putting itself forward – commendably, in my opinion – as a check on authoritarian and ‘conspiratorial’ power, it needs to be aware of the potential trap it’s digging for itself.

Seems we’re at a crossroads in how information is used, and the current state is unstable. It’ll have to settle into equilibrium at some point, and it’ll be interesting to see how it does.

Evolution and Moral Ecology, Mini PhD Version

I’ve posted a new static page with an outline of my PhD thesis on evolution and moral ecology. If you’re interested in my overarching theory, it’s worth reading. Hopefully it’ll put a lot of the other missives I write in context. Although I don’t doubt it’ll also raise a lot of questions and objections. Happy to hear them. Any criticism that can steer me in a better direction will improve my thesis. I call it PhD 2.0.

Morality and the Obsession with Harm and Fairness

Where you can find contemporary moral philosophers talking about the content of morality (instead of their preferred pastime of quibbling over metaethics), you’ll often find them talking about issues concerning harm and fairness. But is this all there is to morality? What of moral prohibitions concerning food, or cleansing rituals, or burial practices? Can you just translate such norms into norms about harm and fairness? Or is the domain of morality larger than many philosophers might readily suggest?

This was one of the questions broached by ANU’s Ben Fraser in a seminar at Sydney University yesterday. Fraser’s paper was about the limits of the moral domain, specifically defending Richard Joyce’s account of morality from criticisms mounted by Stephen Stich. I won’t cover everything said by Fraser (you can read his entire paper here), but I am particularly interested in what it is that we’re really talking about when we’re talking about morality.

And I tend to believe that defining morality in terms of harm/fairness exclusively is a bit narrow – but understandable. Even so, we shouldn’t limit ourselves to issues of harm/fairness if we want to understand the full scope of morality and moral phenomena.


Liberalism and Value Pluralism

Does a commitment to normative value pluralism logically entail a commitment to liberalism? Isaiah Berlin is a known proponent of both pluralism and liberalism, and at times he’s appeared to suggest there is a logical connection between the two – although at other times he suggests the connection is only a psychological one; that it’s choice that makes human beings human, that we’re made pluralistic, and liberalism is the best system to enable these two psychological forces to coexist.

Isaiah Berlin, from Steve Pyke's collection.

I had the opportunity last night to attend a debate between two leading Isaiah Berlin scholars on this notion of the link between liberalism and value pluralism, Beata Polanowska-Sygulska from Jagiellonian University in Poland, and George Crowder from Flinders University.

I have to profess a vast ignorance when it comes to Berlin’s work – he doesn’t feature prominently (or at all) in the philosophical texts I’ve been wading through, although his views about pluralism and liberalism appear to be remarkably close to my own. I shall have to read his work more thoroughly before I complete the normative chapter of my own thesis on evolution and moral pluralism.

From the outset, I can pick one stark point of difference between mine and Berlin’s views: he believes in a plurality of incommensurable objective values, such as equality and liberty, whereas I don’t believe any objective values exist. However, that’s not a show stopper, as I think you can extricate the objectivity from Berlin’s values without too much trouble (and call it a fictionalism, if you will) and the crux of his argument will remain largely the same – the only difference is in some distant metaphysical justification for his pluralism.

On the connection between pluralism and liberalism, I have a slightly different take from those advanced last night. Mine is, unsurprisingly, informed by evolution and moral psychology. It goes a little something like this:

Humans have been grappling with the problems of social coordination for hundreds of millennia. These problems are effectively: how do you get large numbers of unrelated individuals to live and work together in a way that advances their interests (biological and psychological) without suffering the ill effects of socially disruptive behaviour, like cheating or ‘defection’ (in game theory terms), and without  succumbing to invasion by outsiders.

Evolutionary forces have worked such that those individuals who were able to solve these problems more effectively were able to leave a greater number of offspring for future generations. Thus, those individuals who evolved the psychological mechanisms that promote prosocial behaviour and censure socially disruptive behaviour, were lent a selective advantage. These psychological mechanisms include the moral emotions, problem solving heuristics and moral reasoning.

However, there is no one solution to the problems of social coordination that works best in every environment, particularly as the environment is also made up of the other ‘strategies’ employed by others, thus is dynamic rather than static. As such, evolution has not settled upon one set of psychological mechanisms or predispositions, for to do so would have been unstable and left that population prone to ‘invasion’ by other strategies – invasion from within, through mutation, or from without by other individuals with different psychological makeups.

The upshot – and Edward Westermark acknowledges this as early as 1906 – is that human psychology varies (genetically as well as a result of environmental influences), and this variation yields a broad spectrum of moral outlooks and values. Westermark’s passage is as follows:

The emotional constitution of man does not present the same uniformity as the human intellect. Certain cognitions inspire fear in nearly every breast; but there are brave men and cowards in the world, independently of the accuracy with which they realise impending danger. Some cases of suffering can hardly fail to awaken compassion in the most pitiless heart; but the sympathetic dispositions of men vary greatly, both in regard to the being which whose sufferings they are ready to sympathise, and with reference to the intensity of the emotion. The same holds good for moral the emotions. The existing diversity of opinion as to the rights of different classes of men, and of the lower animals, which springs from emotional differences, may no doubt be modified by clearer insight into certain facts, but no perfect agreement can be expected as long as the conditions under which the emotional dispositions are formed remain unchanged.

As such, value pluralism is an empirical fact about human psychology. But it’s more than just psychological; just because evolution has primed us with this variation, it doesn’t mean it’s good. There could be one moral value that actually serves our interests better than others. However, I don’t believe this is the case. As I mentioned above, there is no one solution to the problems of social coordination. And if you take solving the problems of social coordination to be important, then you will also find pluralism to be important, as it allows a range of solutions to arise, one of which might be the best in any particular situation.

On to the strengths of political liberalism: I don’t actually think it’s promoting autonomy that is the fundamental justification for liberalism, instead it’s that promoting autonomy allows the pluralism of values (or ‘strategies’) to work in tension with each other, thus preventing any one strategy from dominating and causing the society to become unstable (or to be in disequilibrium). Autonomy is a second-order value, but one that enables first-order values to be promoted most effectively.

Thus liberalism is the most effective political framework (that we know of to date) that allows the various strategies for social coordination to balance each other out, and best enables a society to meet the challenges of social coordination. It’s not without cost; as Berlin states, there will be situations where values conflict and you’ll inevitably get dilemmas with no perfect solution. But that’s the price you pay, and it’s a smaller price than adopting a monist approach and, say, placing egalitarianism or order, above all other values.

I suspect that Berlin would have disagreed with several points in my account of liberalism, but I think they’re largely detail. On the whole, I think my account is very similar in action to Berlin’s, although I’ll have to read a great deal more of his work to know for sure.

1 2 3