I was a tad disingenuous in an earlier post: in the back of my mind, I have always been planning something specific to do with the reverted wilderness acreage we’re buying in the country.
Victorian amateur scientists have always fascinated me. I imagine fondly that someday in the next few months you will find me ensconced at a portable table out in “the back”, wearing my sun hat and glasses, with my WANned iBook and cheap USB microscope, live-blogging pictures of my very own algae, rotifers, seed capsules and suchlike. Better by far, in my technophilic opinion, than a mouldering leather-bound personal journal filed with watercolors of toadstools and calligraphic noodling.
See, I often pine for the days when not just landed gentry but regular folks had microscopes and telescopes and fossil-collecting handbooks and terraria and bred doves and lilies and otherwise learned something first-hand about the real world in their own gardens and town auditoria. The social norm of public scientific inquiry faded long ago, of course, but now I practically despair over it. For example, home-schooling parents are probably the biggest purchasers of microscopes and science training stuff for their kids, but the demographics (and general anti-intellectualism) of the majority of home-school parents don’t encourage me that biological learning is being thoroughly elucidated in these efforts. Most “nature stuff” people do these days pays attention only to the sort of big dramatic cheetah-kills-antelope stuff they’re exposed to on TV: whale-watching, hiking, hunting, birding and the like. They tramp miles through equally interesting but ignored life to go and see the animals, and then tramp back home and sit back down in front of the TV, their boots covered in fascinating stuff on the mat by the door.
Some small part of the reason people don’t “do science” is the cost of equipment and supplies. Yes, a nice gas chromatograph is still rather pricey, and a useful telescope will set you back a few grand. But I spent $30 on my 200x plastic USB microscope (it’s a discontinued toy), and I have this computer just sortof sitting around warming my lap up all the time anyway. So I’m not entirely certain that it’s reason enough ever. Except maybe nuclear physics, and maybe radio astronomy.
Some other part of the reason is supposed to be the difficulty of getting your head around today’s super-specialized scientific knowledge. People (kids) are not trained in science, therefore not qualified to do it. They need somebody to train them in the methods, and show them what they’re supposed to be looking for, and what it means in context. This indicates to many people that science teachers are required, and parents therefore off the hook. But take it from me: I taught botany to wannabe science teachers for three years; you would be frightened or very very sad if you really understood how bad they were at thinking or understanding, let alone teaching about science.
But I think the biggest reason hobbyists don’t do science is that they just don’t know they can. All you really need to do is think and understand the process to be qualified to do it.
By what will be seen to be a very direct path, buying muck and dreaming of sitting in the shade with a microscope and putting it all right here on the Web has reminded me of one of the other projects I’m gearing up for.
A huge and very important chunk of complex systems research consists, in a reduced sense, of thinking about how systems are put together of agents following simple rules. Writing little stories, in other words: “What would happen if people in a market simply traded according to random rules?” and “What would happen if proteins were composed of two types of subunit (hydrophilic and hydrophobic) on a chain constrained to a planar lattice, and you let them wiggle around and ‘fold’? What would you see if you did that? Does it suffice to explain some of what really happens in protein folding?” Of course, before they’re published these what-if questions are prettied up and presented as if the researcher knew all along that they were doing a rational experiment, but because you’re a diligent and faithful reader to have worked your way along this far already, I’m letting you know the Big Secret of Professional Science: we really mostly just try stuff and see what happens.
The science part of complex systems happens in at least three stages. Two of these are: (1) analysis and reframing of stuff that really exists in terms that let you talk about it reasonably using concepts that easily become simple models, and (3) in interpreting the computer simulations you build according to those models to see what they tell you about the real world. The bit in the middle, the (2) that differentiates a lot of complex systems research, is what I refer to as building analogous systems — artificial worlds in which your model of the real world is literally true. So for example, the previous notion about “people in a market trading using random strategies” is in a sense a prospective model of real-world market traders using bounded rationality other weird non-rational stuff we see all the time. The analogous system you can build is the actual running computer program in which little agents representing people trade some tokens representing real market goods and currency according to rules you code as “random” according to your interpretation of the term. The resulting program is not the model: your model is your analysis of the real world, summarized as “perhaps it’s like this” (or hidden in “what if it were like this?”)
The third part, mainly observational but informed by your original modeling effort, basically lies in collecting data in the analogous world and seeing how that may explain or apply to the real one. For example, in collecting a million different protein-folding results in a simulation based on your two-component model of proteins, and then seeing how the statistical distribution of the results might match that seen in nature.
I’m wordy because I’m excited and writing-to-think. All I’m trying to say is this, really: Much of complex systems research is just:
- Look at what’s around you and frame a model that summarizes what you think you see
- Write and run a little computer program (an “analogous system”) in which the model is literally true
- See if the behavior of the analogous system gibes with what you observe.
That’s it.
Point: Complex systems research is easy.
See, the interesting thing about complex systems research—simultaneously the thing that makes the systems interesting, and the field—is that even the analogous systems we build are capable of unexpected and often nigh inexplicable emergent behavior. That’s the point: the model is not tractable by traditional math approaches, so for example a traditional economist would simplify away the stuff that’s emergent because the equations are too hard to solve. But you — you cunning complexologist you — build a simulation based on the model and work around the hard math bit. Yes, maybe even the computer implementation is wild and does weird stuff, but it’s much faster than the real world and so you try it 100,000 times and see what happens.
I harbor secret desires. Many of those I will choose not reveal here, but among the others are: I would like people who are not credentialed union card-holding ivory tower scientists to be able to undertake scientific exploration and investigation personally, collect and manage the observations that will arise, and publish the results in valid peer-reviewed scientific journals (that they can afford).
I think something like the Open Source approach to software development would work, and for exactly the same reasons. I will write about that here in a bit.
In the meantime: almost anybody who knows what it is (and can write code) has written a Game of Life program. Almost everybody who knows what it is (and can write code) has written a Mandelbrot set generator program. The same goes for genetic algorithms, Markov text generators, and innumerable other canonical “chaos and complexity” simulations and algorithms which have been popularized through the years. Yet, recreational or not, these simple programs are exactly the sort of thing that makes complex systems research go.
I’ll bet that at least a dozen of the thousands of people who wrote their own Game of Life (at least those who played with the parameters) encountered phenomena that would have warranted publication in a peer-reviewed journal. And at the same time, I bet that most of the thousands of other people (if only they had been exposed to the work in the context of a community of like-minded collaborators and background information) might have moved on beyond screen-saver diddling and addressed real and serious unanswered scientific questions.
But as amateurs, these folks worked alone and were thus hemmed in by a limited social capital and intellectual context. Their results are forever relegated to recreational status in the “umbra” of science, never published and thus doomed to oblivion. No matter how many interesting “what would happen if…?” and “what does it mean that…?” questions they asked, the answers were for the most part unattainable or unshared.
That’s sad. It’s just as if they lived in the country, went out occasionally and poked around a bit, caught a few butterflies nobody had ever seen before, and not knowing what they had let them go, got bored, and went back in to watch TV.
Working alone, these folks (which I would number in the thousands) remain hobbyists re-creating simple toys. Working together, I think they might become a potent distributed scientific workforce, as powerful and effective as more traditional labs and warranted scientists.
By my argument, you need three tools to do valid complex systems work yourself: One is what you are sitting in front of right now. Another is the mess of meat perched up there at the top of your neck. And the third? Access to other people working on the same thing.
And that’s one thing you can do with muck and the Web that you can’t do with just muck: begin to disintermediate—or enhance and expand—the traditional scientific establishment.