Can Math Help Social Science?

A very different kind of event took place last week in the room at the top of the Mathematics building called ‘Mallvinden”. The words resonating in the room were not “Banach Space” and ‘Lebesgue Measure”, but “society” and “justice”. This was Society and Social Dilemmas: Game Theoretic and other Mathematical Modelling,a workshop organized by Olle Häggström, Mathematical Statistics and Kristian Lindgren, Physical Resource Theory at Chalmers. (Logically speaking, the event should have been subtitled “Computational and Mathematical Modelling”, for there was a lot more computational aspects to the modelling aspects on show than purely mathematical.)

The event began with corruption and human institutions and ended with climate change, with biology, psychology and linguistics in between!

The kickoff talk was by Bo Rothstein, a very well known personality in Sweden for his apparently unlimited access to the comment and viewpoint pages of the leading Swedish dailies such as Dagens Nyheter and Goteborgs Posten. Rothstein’s questions are very interesting ones. First, how do institutions develop that lead to a non-corrupt society, for example, Sweden today? The question becomes even more interesting when one discovers that Sweden was just as corrupt as other places in Europe not so long ago in history. The answer is probably some very complex combination of historical, cultural and religious factors. What could mathematical or computational modelling have to contribute to understanding this question? I really don’t know. Rothstein’s second question was: when corruption is the norm, it takes very different forms – why? I must confess to being puzzled by this question – given the very different cultural, historical and other settings in places like Nigeria, Naples or India, one would naturally expect corruption there to exhibit very different features as a rule. Why should it be surprising? One would be surprised if exactly the same type of corruption arose in these very different kinds of places, and that would need explanation! Rothstein showed us emails he exchanged with a leading game theorist asking why corruption takes very different forms, and the answer was essentially, “good question!”.

The final talk by Olof Johansson Stenman was entitled “Behavioural Economics Insights Applied to Climate Change”. What are the insights from behavioural economics? That people don’t always make decisions fully rationally, that they do not make them entirely motivated by material self interest, that they often cooperate only conditionally i.e. provided the others do too, and that they respond to rewards and punishments. I confess wondering if one should churn out dozens of experimental studies and sizable journal papers to arrive at these insights, but perhaps they serve a purpose in a discipline where ritual sacrifices are offered routinely at the altar of self interest. As for the application to climate change, this translated essentially into the prescription that one should punish the defectors from the collective agreement. All very well said, but who is going to punish the greatest culprit – the EE? Sanctions were mentioned.Well, the EE is great at threatening and imposing sanctions on others, but who is going to impose sanctions on the EE? A proposal by Stiglitz was mentioned about imposing trade sanctions on the EE. Once again, sounds good on paper, but who is going to impose the trade sanctions? The EU? Or China? They would shoot themselves in their own legs? In the end, the most depressing thing about combating climate change is the EE – who won’t even sign a very weak and non-binding agreement like Kyoto, what hope for a serious binding and drastic treaty that is required? There is some hope though, because with the exit of the Coward government in Australia, the coalition of the unwilling has been reduced to essentially one.

The highlight of the event for me was the talk by Olof Leimar which showed that while we humans run hopeless contortions around social dilemmas, animals appear to solve it effortlessly, demonstrated beautifully with the example of the cleaner fish.

Although, as hinted above, I am sceptical about what Math and CS can contribute to some grand questions from social science, I do hope that this event will initiate more collaboration between CS, Math and other disciplines in the social sciences. Life buried in Math and CS internals can be monastic and at some point at least, a bit uninteresting, whereas such interactions could make us, to use a phrase that was hammered at me growing up in India, “healthy, wealthy and wise”!

The Foreigner’s Curse

It was Camus who observed in his Notebooks that foreigners in a country inevitably end up being critical about its ways. With my recent postings, I ruffled more feathers, and strengthened my already strong reputation for being one of these. So if not an apology, at least a (further) clarification.

First, one should not generalise by ethnicities, nationalities, race … etc. Of course I agree. Indeed I am bugged most by this personally when people start typecasting me with traits they claim are Indian. My response is an agitated: “Hey! Whether or not x may be an Indian trait, you have to judge me by what I say or do!”. So obviously I shouldn’t be accusing anyone of going this or that based on the fact that they’re Swedes! Individuals are different, and they are responsible for their own thoughts and actions, not those of others.

Second, I am often incensed at comments made about India (for example in recent meetings in the new fever of Indo-Swedish samarbete) by people who have no clue. Obviously, it takes a while to understand the norms and mores of a country, and to pass judgements without doing so is the height of ignorant condescension. So, by the same token, I shouldn’t go around spouting wisdom on Swedish ways without having understood them correctly.

That said, I do believe some stereotypes are true! Or, at least have a grain of truth about them. And, second, it sometimes requires a foreigner’s eyes to spot something that is invisible to the locals.

In my offending blogs I made two related claims: first that Swedes have a tendency to be defer-rant to authority, that of the Government say, or to that conferred by a position or status. Second, that Swedes are infected by “coursophilia” – to attach too high an importance to formal courses and points as a way of judging competence.

Now, following (very briefly and superficially), David Sloan Wilson’s adage to view everything in the light of Evolution for Everyone, there may be good reasons fo these memes to have set in. After all, Swedish Governments, for eaxmple, have long been models of unsurpassed integrity and honesty, and the Swedish education system has similarly worked marvels in creating the present state of society. So there may be very good reasons for the deference to authority and coursophilia.

But, there is also the danger that it can be misused and exploited …

Det Sovande Folket

That’s the title of Fredrik Reinfeldt’s book from 1993 as a “ung moderate” claiming that Swedes were a brainwashed people. Now, you won’t find me agreeing with Reinfeldt on too many things, but a discussion yesterday on pedagogy makes me wonder if he has a point …

I went to the discussion hoping to make what I thought was a rather uncontroversial point: that a university like Chalmers, which aspires to class itself as among the best in the world, should adopt a flexible and multi-faceted approach in deciding on “pedagogic competence” among its staff, new and existing. In particular, pedagogic courses could be one of several criteria, but they should not necessarily be mandatory.

How wrong and naive I was! The discussion started with claims about how Chalmers has higher requirements for pedagogic excellence than other places, which turns out to consist of insisting on the requirement of pedagogic courses, and the discussion never seemed to get away from this point.

Rhetorical arguments were presented of the following form: suppose I was looking to appoint someone to teach algorithms at Chalmers – would I accept someone who had 20 years working at Ericsson as a programmer who had no formal qualifications to teach algorithms? Well, let’s consider this a bit more carefully. Suppose A has worked 10 years at Ericsson in a strong research group working with networking algorithms. A has been active on the international scene, presenting cutting edge research at premier algorithms conferences. A has taught courses as a guest lecturer at Chalmers and KTH in algorithms, been in demand for giving summer courses at international schools on algorithms … However he doesn’t have something that B has: namely a pedagogical certificate pinned to the wall asserting competence in teaching algorithms. Moreover, B has no research record at the international level, has never taught at Chalmers or KTH , Lund or Uppsala … and never even heard of international summer schools in algorithms.

At the trial lecture, B proceeds in a very “pedagogic” fashion: he has neat powerpoint slides one labelled “Introduction”, the final one labelled “Summary and Conclusions” and in between a slide for “Greedy Algorithms”, “Dynamic programming” and “Divide and Conquer” with textbook examples of toy problems. A on the other hand, begins with a real problem that Ericksson had to solve in its mobile networks and starts by challenging the audience to offer solutions. Then he proceeds to illustrate how the problem could be approached with different strategies, their respective strengths and weaknesses, false turns, errors … and finally the solution (all with non Microsoft crap, to top it all!).

Who should one hire at Chalmers? No brainer you say? You’re right: by Chalmers high standards of pedagogic excellence, we would disqualify A for not having the competence to teach and hire B!

Time to make yet another of my bold sweeping invandrare claims about Swedes. First a clarification of my motivations. Like most other invandrare, I recognise that compared to most other places – foremost India, but also the EE for example – Sweden is, in general, a very enlightened and progressive place where reason, integrity and harmony hold sway. Thus it is all the more frustrating for an invandrare that when they are so reasonable about most things, how could they possibly be so stupid about others? For example, “coursophilia“: to be considered competent in something, you must have certificates and points from courses – pedagogic courses! – otherwise you’re incompetent to discuss the subject. Thus:

Noam Chomsky shouldn’t write about politics

Jared Diamond shouldn’t write about history and societies

Olle Häggström shouldn’t write about climate change …

Because none of them have taken 101 courses on the subject, nor courses on how to write … hence are incompetent on those respective topics, no matter how much independent reading and research they may have done, no matter what their experience and track record is, and no matter how insightful and accurate their arguments may be. (Evidence for my claim: I’ve explicitly been asked about the claims listed above, and other invandrare strikingly had similar experiences. Counterexamples: three people from my own Dept. who made outstanding presentations against the pedagogy course requirements.)

There was also an interesting bootstrapping argument: Chalmers should insist on pedagogic course requirements because other universities  – in Sweden! – are doing so. Thus, people at D&IT shouldn’t end up at a disadvantage when they apply to other places in Sweden and are required to show their pedagogy course points … Meanwhile these universities of course are probably using the same argument in reverse to enforec their pedagogic course requirements. A beautiful self-reinforcing – mutually recursive as CS people would say – circle.

So it seems to be a long and hard battle ahead to get Chalmers to think outside of this framework of courses and points in judging pedagogical competence. One thought that occurred to me – fleetingly – was to compile a list of the world’s leading universities that are universally acknowledged to achieve the highest standards in research and teaching excellence – MIT, Stanford, Oxford , Cambridge … – and document how they evaluate competence, in particular that they do not have pedagogic courses as mandatory requirements. But then Chalmers will merely take that as definitive proof that we have higher standards of excellence!

How we Mathematicians and Computer Scientists Contributed to the Financialization Mess

Two nice accounts of how we mathematicians (probabilists in particular) and computer scientists contributed to the financialization mess:
End-of-the-World Trade
THE SUBPRIME CRISIS
The probabilists dreamt up the fancy models (all wrong!) and we computer scientists crunched the numbers. The computer scientists managed to hit two birds with the same stone: by the amount of air conditioning needed to cool their number crunchers, they hit the climate too!

“Free Market” = Shock, Torture, Terror …

Naomi Klein, The Shock Doctrine, Penguin 2007 (Swedish edition Ordfront 2007)

“Free Market” = Democracy, Freedom, Peace, Prosperity …

is supposed to be the equation of the zeitgeist. But, as Naomi Klein demonstrates in this book, her latest and by far the best, the truth is just the opposite, namely the title of this blog!

Klein demonstrates this by an enlightening survey of the economic history of the last 50 years which has seen the world increasingly gripped by the tentacles of “free market” fundamentalism.

The Shock Doctrine begins with a literal story of shock treatment by the Canadian doctor Ewen Cameron at McGill University, research sponsored by – who else? – the CIA! Cameron’s theory was that once “complete depatterning” had been achieved, the patients could be reprogrammed as desired. Klein intends to use this as a metaphor for how the “free market” ideology has proceeded time and again. This story is certainly revealing in showing that torture is as American as apple pie, so Abu Ghraib and Guantanamo are worthy successors to a long tradition. However, for Klein’s main story, this shock treatment of her own is in danger of claiming her as the first victim, for it opens the door to “conspiracy theory” attacks. Indeed this is what two recent reviews – one by the Nobel prize winning economist turned public intellectual, Joseph Stoiglitz and the other in LRB – have picked on, while being otherwise appreciative of the book’s merits.

From Chile and Argentina to Iraq and Palestine, the “War on Terror” and back home in Katrina ravaged New Orleans, Klein paints a rich description of how catastrophes, natural and man made – have been the shock used to force unsavoury policies down the throats of unwilling victims.

A lot of this is done under the rhetoric of “free market” fundamentalism. The LRB reviewer takes Klein to task for not distinguishing between the “free market” ideas themselves and corporate misbehaviour. What he doesn’t seem to get is that the louder the proclamations are made about “free markets”, the easier it becomes to see that the use of the term has very little to do with the fantasises set out in Econ 101 textbooks.

One common pattern is the repeated appearance of Chicago school economists at the key points in these acts. “Conspiracy Theory” or just a coincidence? Are economists just disinterested scientists seeking the truth? Well, as Stiglitz writes, the economic policies being pushed “were never based on solid empirical and theoretical foundations, and even as many of these policies were being pushed, academic economists were explaining the limitations of markets”. Reading Klein’s book makes it clear that while Marx it may have been who uttered the famous line that “philosophers have only studied the world, the point however is to change it”, the ones who have taken it seriously and successfully are the Chicago Boys of Milton Friedman.

Doing Science by Talking Loudly and Guessing Outrageously

Francis Crick: Discoveror of the Genetic Code By Matt Ridley 213 pp. Atlas Books/HarperCollins, 2006.

“I have never seen Francis in a modest mood”, so begins the first line in James Watson’s account of the momentous discovery of The Double Helix for which he shared the Nobel prize with Francis Crick in 1962. Indeed the entire scientific life of Francis Crick is a tale of immodesty: first the structure of the DNA, then the genetic code and finally neuroscience and consciousness.

Yet, the beginning could hardly have been more different. Known in school as a moderately clever but distracted kid, Crick soon acquired a reputation for someone who just talked a lot – very loudly at that – and never managed to finish anything satisfactorily. His real scientific career started very late, when he was over 30 and the discovery of DNA structure came when he was 37.

The decisive influence that started and then continued to shape his scientific career seems to have been the presence of the other. All his major discoveries were made in partnerships with a partner who could serve as a sounding board for ideas. First was George Kriesel, an eccentric Austrian logician who specialised, among other things, in making random proposals to women on the Riviera, see Kreiselinana . From Kreisel, he learned how to organise his thoughts precisely and logically. Next was Watson with whom he developed an instant rapport. One of the best administrative decisions his bosses at Cambridge made was to put the two together into a room where they could talk loudly and endlessly. Then later it was Sydney Brenner with whom Crick cracked the genetic code and finally in the last 18 years, an unfulfilled quest for understanding consciousness with Christof Koch.

Crick’s approach to doing science is a great way to debunk common public parodies. The scientist is depicted as careful and stodgy, collecting fact after fact, and refusing to speculate wildly. Exactly the opposite of Crick! Time after time, Crick tossed up idea and hypotheses out of nowhere, with not an ounce of evidence in support. The best example of this is probably the paper “On Protein Synthesis” from 1957 in which he made a set of bold assertions each depending on one another: the function of genes is to make proteins, there are 20 kinds of amino acids in proteins, and all occur in nearly all proteins, whatever the species or organisms … all guesses, and all, as later turned out outrageously, correct! The most remarkable part of the paper were the two general principles Crick set forth:

My own thinking is based on two general principles, which I shall call the Sequence Hypothesis and the Central Dogma. The direct evidence for them both is negligible, but I have found them to be of great help in coming to grips with … very complex problems.

The “Sequence Hypothesis” states that a sequence of bases determines a amino acid and nothing else is needed to tell a protein how to fold. This was complete heresy at the time, later a fundamental surprise of molecular biology, but just standard bioinformatics today. The “Central Dogma” in original form was:

Once “information” has passed into protein, it cannot get out again … the transfer of information from nucleic acid to nucleic acid or from nucleic acid to protein may be possible, but the transfer from protein to protein or from protein to nucleic acid is impossible.

Today, this is again,a central tenet in molecular biology, namely that information passes uni directionally, from DNA to RNA to proteins.

The common parody of the fact-addicted cautious scientist fits much better, Rosalind Franklin , the crystallographer and contemporary at King’s College who had important data, but refused to speculate on it, possibly denying her a proper share of the discovery of the double helix.

There is a beautiful vignette for computer scientists, during a visit by Francois Jacob, one half of the great French team with Jacques Monod. Monod and Jacob had done some experiments that seemed to decisively refute Crick’s assertions about protein synthesis: when new genes were introduced, proteins were synthesised far too quickly for there to be time to build up new ribosomes as Crick’s theory seemed to require. During Jacob’s talk, suddenly Brenner let out a yelp and started talking very fast, and in response, so did Crick. They had seen the solution! And the solution came from Brenner recalling a paper he had studied several years ago by the great Hungarian mathematician John von Neumann on the design of self-reproducing automata. There von Neumann had laid out the basic principle of computer architecture: that the “tape” which held information or instructions was separate from the “tape reader” that interpreted or executed those instructions. This was precisely what the ribosomes were: they merely read the instructions carried to them by messenger RNA and executed them to synthesise the proteins!

On ecan’t but speculate that Crick would have loved to stay a bit longer as the field of Systems Biology takes shape and starts addressing some of the problems in neuroscience he was obssessed with at the end.  In particular, his speculations on the claustrum and his highly connected structure sound straight out of recent network biology!

Finally, science funders have lessons to learn from the career of Francis Crick, especially the early days with Watson. Crick and Watson were commonly regarded as ignorant fools with no background in the area, meddling in things far beyond their ken, hoping futilely to compete with luminaries like Linus Pauling. They were constantly straying from their assigned tasks for which their positions were funded, and never did their reports etc on time. Thank God for science!

Matt Ridley has proved yet again, that he is much better at writing wonderful popular science books than at running banks.