Monday, March 20, 2006

Insecurity rules

Dave Lull has pasted some fascinating excerpts from "Fooled by Randomness" in my posting about the not-so-grumpy bookman. (Thank you for persisting, Dave, seems like blogger has been having some technical problems.) The first excerpt is about the author, Taleb, and his friend's comparison of two philosophers: "We surely closed our minds by following Descartes’ model of formal thinking rather than Montaigne’s brand of vague and informal (but critical) judgment. Half a millennium later the severely introspecting and insecure Montaigne stands tall as a role model for the modern thinker." The second excerpt contains some thinking about science. "Science is a fundamentally skeptical enterprise. How? By some fallacy of aggregation (i.e., the sum is not the parts), empirical-experimental science is not the sum of scientists but the upper bound of competing results; scientists are in a ruthless contest, frequently at each other’s throat. Each individual is disciplined by a few annoying peers going after the robustness of his results, not by his own intrinsic devotion to truths, a system quite similar to the assumed role of competition in a capitalist system." Yet science must evolve in this direction if we are to make advances -- in some key areas, at least. My posting about the future of science alluded to this necessity, in the context of biology. Biology's future lies in "big science" --- analysis of nature's unimaginably complex systems and networks and their perturbation by chemicals, currents and other environmental factors that cause a cell to move or a protein to form, for example. Our relatively new knowledge of genome sequences is the tip of this particular iceberg. So vast is the challenge that more and more leading scientists are becoming keenly aware of the value of this uncertainty and introspection that Dave has highlighted. The science of bioinformatics has developed this way, with its roots in the open-source free spirit of the Web, but a problem for this field has been persuading the 'at each other's throat' biologists to provide their data (genomes and so on) for the analyses. And who can blame the biologists, who are forced into the "ruthless contest" by the career structure in which they exist? (That's where the Taleb extracts come into it.) But things are changing. Systems biology is a nascent discipline but growing. And an example of the collaborative, as opposed to the competitive, approach to science is demonstrated by the Alliance for Cell Signaling , whose goal is to understand not only how cells interpret signals, but how they interpret them in context -- that means looking at signalling networks and how they are affected by minute changes in circumstance. Many world-leading, 'in competition' laboratories have signed up to the alliance, and share out the work, as they know the challenge is too vast for one person or lab to get anywhere alone. Disclaimer: if the above posting reads a bit incoherently, forgive me. While writing it I have also made dinner for my daughters, read an essay on Pride and Prejudice by one of them, listened to piano practice, helped the other one create a "links" page for her Fruits Basket website, and fixed up by phone for a friend to come and play after school later in the week. Networking, schmetworking -- such is the multitasking life. Space to think is at a premium. Thanks again to Dave Lull for providing these highly pertinent extracts.