“Its not science until I can mathematize it …”

A rushed day last Thursday, between taking the cat to the Vet, completing a paper on modulators in complex systems and dashing up to London for a just-in time-arrival at the Royal Opera House for what turned out to be a mixed performance of Madama Butterfly: The lead soprano missing a note at the start of Un Bel Di Vedremo but retrieved with a well sung but over-acted suicide in Act III. . I also had to attend a seminar in London the next day, so stayed up in one of those dire hotels in the west of London that charge more for less and on which I have previously commented. No internet access, a battery that did not last the train journey home and a traumatic weekend [please do not mention 10 seconds or Law 5.7 (e)], all resulted in my absence from the blogosphere for several days. For this I apologise, although this week will also be difficult as I am getting ready to present our work for the last two years at a symposium in Singapore next week.

To the subject of today’s blog; I stayed up in London as I had an EPSRC day seminar looking at the state of complexity science research in the UK all day Friday.

The seminar had been organised by their complexity science programme. The same group also organised one of the more enterprising examples I know of a government agency in action (and I use that word advisedly)): its sandpit on emergence late last year (of which I was Director), allocated over a million pounds of research money to inter disciplinary research without filling in forms until after the decisions had been made.

This seminar was a little different. It had a work product to achieve, namely a list to a pre-determined format so we were somewhat shoe horned into a process, but such is life, these things have to be done and meeting old friends and new people is always worth some investment of time. Sitting there for a day as a participant, with no special responsibilities was enjoyable and gave me time for reflection. These academic events are interesting, a mix of intellectual stimulation and much frustration at the bureaucracy of research funding and its consequences for the behaviour of otherwise very intelligent individuals. As I moved between engagement and disengagement with the process I also jotted down some reflections which I share here. Not all of these relate to the seminar itself, but arise from other events and reading.

I claim no coherence for these, they are as I say jottings and I hope by the end to be able to make a connection to the suicide of Madama Butterfly in front of her blindfolded son; the son holding a US flag, surrounded by falling Lotus blossoms. I have no idea how I am going to do that, but complexity is of course about starting journeys without knowledge of endpoint, so I am sure I can.

So here goes, in the form of a list assembled in no particular order:

1. Scientists are not immune to fashion hopping. The seminar was about complexity, but some but forward the idea that Britain’s strength in systems thinking was in effect an existing practice in complexity science. Complexity is a more fashionable label, so rather as people in Information Management started to adopt the KM label in the late 90’s we see the same thing here. Now I think that early work in systems thinking is a strong base for complexity, but to confuse the two is an error. One of the many reasons that KM is gradually fading from the scene, is that it just became another name for information management; we do not want the same thing to happen with complexity
2. The measurement system for academics is producing pervert systems. I have lost count of stories I have heard of people getting chairs, not due the brilliance of their research, but due to their ability to attract grants and publish articles in the right journals. It is a classic example of Goodhart’s Law, in which a measurement system (which had some validity) is turned into a targeting system at which point the value of a measure is lost. The process of awarding grants, promotions etc is now tied into a rule based and eminently game-able system.
The system also seems to reward specialisation, rather than integration. Interestingly several people at the meeting commented that some of the most valuable grants had been small ones to allow people to hold meetings and meeting people from other institutions and disciplines. However to get money allocated to genuine trans-disciplinary projects seemed to be considered impossible. In effect we have a basic mistake, the belief that a collection of specialists is as good a generalist. In this connection we did list as a strength the remnants (as it is going) of a liberal education tradition in Britain. I do think that it is of critical importance, during periods of change, that we train and promote people who are able to make connections between the previously unconnected. The more people focus on one aspect of their discipline the less this will be possible.New toys are dangerous. The fact that neuro-science can explain many things that we did not know before, does not mean that neuro-science will explain everything. I can see the old free will debates and mind-brain identity theory starting to come out again in the scientific community. We need to see those tools in context, accepting that some are radical (not everything builds on the old, despite that being a requirement in most research proposals) but all take place within a complex and changing context. To be an expert in one discipline, does not necessitate your discipline being able to explain everything.
3. Most people working in this area, myself included, make a key distinction between a system which is complicated and one which is complex. A complicated system many have many parts, but the relationships between them are know and over all through computation we can predict the behavior of a system. On the other hand in a complex system all coherence is emergent and the same thing never happens again the same way twice. Now I must admit that I thought this was more or less accepted, however it was challenged, interestingly by many of those who saw systems thinking as complexity. It was from this group that I picked up the phrase that forms the heading of the blog: Its not science until I can mathematize it

Now that last statement deserves some commentary in its own right. Most of the groups had identified the high potential of complexity to apply to social systems (and remember this was the Engineering and Physical Sciences Research council). I have already commented on some interesting potential uses of technology such as robot swarms, and many others could be listed. I also do not deny the fundamental nature of mathematics to the physical sciences, although I think it is over played at times. However for many things we do not have appropriate mathematics, and in some cases we may never have. The absence of an ability to mathematize should not prevent science from being carried out. I was reminded here of Hari Selden’s psycho-history, in which Asimov explores the consequences of being able to predict the future of mass human interactions. What Asimov saw in the 1940’s finds modern manifestation in prediction markets (much over-hyped by the way, but that is for another day) and other manifestations of group interaction. Asimov however is writing in a day, and a scientific context in which to understand the mathematics of something, was to understand it absolutely and thus prediction would be possible. Complexity science is only just starting at the time that Asimov writes, but to my mind (and my mathematics peters out at University when I moved into Physics and Philosophy) deals with systems where prediction is by its very nature is impossible. Yes, mathematics allows us to improve simulation, but simulation should not be confused with prediction any more than correlation should be confused with causation. Yes, techniques such as power laws are starting to give new insight into extreme events. But all that said, science is going to have to deal with things that cannot be mathematized. I have seen some horrendous attempts to do this in the social sciences by the way. One recent article I had to review on knowledge management seeks to prove a series of nonsense statements about human systems by virtue of some highly complex equations built on sand. My general experience at the moment is that engineering, biology and chemistry seem to be closer to my understanding (and application) of complexity than physics and mathematics. In fact many people in physics and mathematics seem to deny complexity.

I will pick up on various themes here over the next month or so, but for the moment they are random thoughts. Having arrived at the end, how to I make a connection to Madama Butterfly. Well there are some themes. Butterfly is deserted by her American “husband” but remains loyal in the face of all the evidence, until finally she meets his new American wife who offers to take the child. Butterfly’s greatest arias are of hope and optimism, but at the end she accepts the logic of her culture and uses her father’s knife to commit suicide having first blindfolded her child and left him to play. The opera speaks at many levels, but the one theme I think worth emphasising here is the danger of being trapped into a way of thinking or social structure that leaves you at the end of the day trapped into either/or choices. The essence of scientific enquiry is (to my mind) not to compromise on rigour in prove, but to be prepared to see things anew. Major progress in science has nearly always been achieved in opposition to the scientific establishment of the day. Now we can understand why this happens, it is a form of entrained pattern of thinking. The more expert we are, the more our patterns entrain around that expertise. To break those patterns require radical and often painful encounters with reality. The more that we can bring the physical sciences into interaction with human systems the more those encounters will be beneficial for both parties. We should reject both modernism and post-modernism in the respect, life is just so much richer than that sterile debate.