Thursday, June 26, 2008

Presented ....

... for your consideration: Yellow Science. (Thanks to Dave Lull for alerting me to the broken link.)

Someone should alert the WSJ copydesk that it's Sir Arthur Eddington, not Sir Thomas. The Nature of the Physical World is still worth reading, I think, if only because it is so well written.

Compare The Irrelevance of "Probability". (Hat tip, Dave Lull.) Taleb seems to be arguing prudentially and I largely agree. (I live a pretty "green" life, actually, and have always been passionately respectful of the environment. But I also know that "Nature" is an abstraction and does not take care of anything. Forests, to cite just one example, need to be managed - actively cared for - not just left to take care of themselves. If you don't believe this, stop taking care of your garden and see what happens.)

Update: The End of Theory: The Data Deluge Makes the Scientific Method Obsolete. (Hat tip, Dave Lull.)

Consider physics: Newtonian models were crude approximations of the truth (wrong at the atomic level, but still useful). A hundred years ago, statistically based quantum mechanics offered a better picture — but quantum mechanics is yet another model, and as such it, too, is flawed, no doubt a caricature of a more complex underlying reality. The reason physics has drifted into theoretical speculation about n-dimensional grand unified models over the past few decades (the "beautiful story" phase of a discipline starved of data) is that we don't know how to run the experiments that would falsify the hypotheses — the energies are too high, the accelerators too expensive, and so on.

Now biology is heading in the same direction. The models we were taught in school about "dominant" and "recessive" genes steering a strictly Mendelian process have turned out to be an even greater simplification of reality than Newton's laws. The discovery of gene-protein interactions and other aspects of epigenetics has challenged the view of DNA as destiny and even introduced evidence that environment can influence inheritable traits, something once considered a genetic impossibility.


  1. Mountains of data for pure science research certainly can change the approach. But I think a lot of working engineers (like me) would tell you that given enough data and the time to analyze it they could design a better bridge, build a better levee, etc. But getting enough good data is often the hard and expensive part - if it can be done at all. Talk of the wonders of data reminds me of the joke about how to be a millionaire - "first, get a million dollars, then......" Such is my admittedly limited perspective.

  2. Thanks, Jim. I appreciate it that you comment on things like this, which are well outside my area of expertise. Actually, I have no area of expertise.

  3. Hi Frank,

    You're starting to sound like Socrates, saying that you have no area of expertise. In Philly, there must be an oracle somewhere, a Philadelphic Oracle to ask, "Who is the wisest in the city?"

    You know Socrates response in Plato's Apology:

    Well, Chaerephon, as you know, was very impetuous in all his doings, and he went to Delphi and boldly asked the oracle to tell him whether--as I was saying, I must beg you not to interrupt--he asked the oracle to tell him whether there was anyone wiser than I was, and the Pythian prophetess answered that there was no man wiser. Chaerephon is dead himself, but his brother, who is in court, will confirm the truth of this story.

    Why do I mention this? Because I am going to explain to you why I have such an evil name. When I heard the answer, I said to myself, What can the god mean? and what is the interpretation of this riddle? for I know that I have no wisdom, small or great. What can he mean when he says that I am the wisest of men? And yet he is a god and cannot lie; that would be against his nature. After a long consideration, I at last thought of a method of trying the question. I reflected that if I could only find a man wiser than myself, then I might go to the god with a refutation in my hand. I should say to him, "Here is a man who is wiser than I am; but you said that I was the wisest."

    Anyway . . . my thought on these data collections and improbable determinations of probabilities, led me to wonder a bit at what would take the place of probability models. Possibly: possibility models.

    If we know something cannot be true, then it can be disregarded. That's the easy part, what we've been consciously up to, up until now.

    However, if we don't know the falseness of a supposed possibility and, furthermore, cannot find it out; and we don't even know the probability of a proposed possibility, and cannot find that out either, what remains is possibility, what comes from the power of speculation, what physicists have, indeed, been up to.

    One step further, though, and we get that in order to solve problems, we may then select from these unfalsifiable and undiminishable possibilities, albeit even necessarily conflicting possibilities, and find that certain ones are more practical for a purpose at hand. For instance, in psychology, sometimes we are better off applying behavior modification, reinforcing the behavior that we want to accentuate, something we do when we get paid, when our recent work gets reinforced and leads to us working more. Yet, this Behavior Modification, this possibility model of how our psychologies work, does not work as well when trying to explain Freudian slips that reveal to others and often to ourselves, what our unconscious minds have been up to.

    What science would set as its task, then, would be to construct possibility models, that could then be used by engineers et al, to broach practical scientific advancements and solved problems, or even simply (possibly) explain some things.


  4. jeff mauvais1:29 AM

    Forests need to be managed? Forests have covered parts of the earth for 350 million years. The earliest fossil evidence for a recognizably modern man (albeit a man who probably didn't possess the technology to practice forest management) is 200,000 years old. Apparently, forests managed themselves just fine, through predictable patterns of ecological succession, for at least 349,800,000 years. Forests are managed by humans to maximize their economic value, but certainly would not disappear without us.

    This is not a screed against active management of some forests, which I support wholeheartedly. In fact, forests are a buffer against rising atmospheric CO2 concentrations (see your recent link to Freeman Dyson's NYRB review) and need to be protected and managed to maximize their role as CO2 sinks. But that's only because mankind has increased CO2 levels while drastically reducing forested areas.

    My original point is that biological systems, whether ecosystems or human beings, tend to manage themselves quite well unless perturbed by outside forces. And this occurs without the intervention of some abstract force called Nature.

  5. Hi Rus,
    Socrates is secure so long as I am the competition.

  6. Hi Jeff,
    The condition of forests prior ot the arrival of man - as with all prehistory - large a matter for speculation. But obviously they thrived well enough to survive.
    But there is no reason to presume that the arrival of man wasn't of benefit to them I'm reasonably well off today, but a favorable change of circumstances could make me much better off tomorrow. At any rate, we must deal with the world as it is today not as it was before we arrived.
    I have spent time in Pennsylvania's oldest so-called virgin forest. It dates back to about the time Penn arrived. In other words, the Indians had probably burned its predecessor, as was their wont, and which is why Penn's Woods was precisely that. That old forest in the northwest of the state is a gloomy place. The canopy shuts out most sunlight, so there is little underbrush and, hence, little wildlife, which depend on the underbrush for food and shelter. The wildlife you see is just passing through - like the humans who visit there (damned few of them, I might add). That forest is just a collection of very old, mostly sick and dying trees, the arboreal equivalent of an old folks' home.
    Visit the forests managed by the state's Bureau of Forestry and you see quite a different picture. The state has been in the forefront of forest management since the days of Gov. Gifford Pinchot, who was himself a forester. As something I just posted indicates I like a little wildness even in my garden. I'm not arguing for manicured woods. But, as the American Indians demonstrated, a symbiotic relationship between man and forest can be achieved as is good for both. It is the leave-it-to-nature school of thought that we can thank for those out-of-control fires out west.

  7. Anonymous2:40 PM

    Chris Anderson's article makes some good points, but it is by no means the whole story, and it confounds issues.

    For example, he rightly points to the fact that genetics is more complicated than some people thought it would be at the time of Mendel, or at the time Watson and Crick worked out the structure of DNA, etc. But this is not because data has got "bigger", it is just that the more we have learnt about the biology, the more complex it has seemed.

    Chris Anderson is right to say that "big biology" and the masses of data being accumulated are a challenge to analyse, annotate and make sense of. But it is not right to assume from that, that "biology has got too big to understand". There is still a lot you can learn from a fossil fragment, for example! And plenty of cell and molecular biologists are learning many new things about the systems they study, using single cells, single molecules, etc. Just read the scientific literature to get the picture.

  8. Thanks, Maxine. That is something I could never have known otherwise.Once again, the value of the blogging conversation: I link, you enlighten.