Is Biology a science, or a Science?

There is an interesting debate surfacing in the pages of Nature about the future of biology.  In February, Evelyn Fox Keller published an essay entitled "A clash of two cultures" that started off asking the following questions:

Physicists come from a tradition oflooking for all-encompassing laws, but is this the best approach to use when probing complex biological systems?

Biologists often pay little attention to debates in the philosophy of science. But one question that has concerned philosophers is rapidly coming to have direct relevance to researchers in the life sciences: are there laws of biology? That is, does biology have laws of its own that are universally applicable? Or are the physical sciences the exclusive domain of such laws?

Keller concludes her essay by asserting, "Even though we cannot expect to find any laws governing the search for generalities in biology, some rough, pragmatic guidelines could be very useful indeed."

By "law", Keller means something akin to Conservation of Momentum, or Conservation of Energy.  Rather than simply a quantitatively predictive model, these are deep principles that govern the way the world works.  As a brief example, Keller focusses on the debate over the existence of scale free networks in biology.  Oddly, despite the fact that there are loads of papers on such scaling laws, Keller asserts those laws are rare: "First, power laws, although common, are not as ubiquitous as was thought; second, and far more importantly, the presence of such distributions tells us nothing about the mechanisms that give rise to them."

Professor Keller makes no mention of other kinds of apparent laws in biology, such as MacArthur and Wilson's species-area relationship presented first in The Theory of Island Biogeography, which deserves being called a law due to its appearance in a great deal of experimental data.  But I am most confused by the notion that physical laws might tell us about the mechanisms that give rise to them.  There is no "why" in conservation of momentum, nor in conservation of energy.  The deepest "explanation" of those laws is in the spatial and temporal symmetry of the universe -- translational symmetry gives you conservation of momentum and temporal symmetry gives you conservation of energy.  These are examples of Noether's Theorem, which says for every continuous symmetry there exists a conservation law.  This is one of the deepest results in all of physics, but there is no mechanism, no why, to be found anywhere in the mathematical statement of the theorem or in its physical consequences.

But even before Emily Noether, before James Clerk-Maxwell, before Newton and Leibniz -- before there was any modern mathematical systematization of physics -- there were a great many experimentalists accumulating data that looked a great deal like distributions.  Such as "air resistance tends to scale with volume"; or "the acceleration due to gravity is independent of mass"; or "planets sweep out equal areas in equal times during orbits around the sun".  But with considerable effort, and after several hundred years, we have well proven physical laws that describe all these observations.

Rather than there being no fundamental laws of biology, as Keller suggests, it seems far more likely to me that we are still collecting sufficient data to spot those laws and write them down.  And the biggest barrier to deriving any such laws is the quality of the data.  I've been hearing for 15 years how quantitative proteomics via mass spectroscopy is at hand, but the first clear demonstrations of truly quantitative, label-free mass spec were published only just last month in Nature Biotechnology (here is the News and Views piece by Bergeron and Hallett).  Similarly, only recently were guidelines for producing (and reproducing) quality mRNA data via gene chips decided upon.

Last week in Nature, Brian Enquist and Scott directly confront Keller's pessimism and come to quite the opposite conclusion:

In the opening of his seminal 1917 book On Growth and Form, D'Arcy Thompson quoted the eighteenth-century philosopher Immanuel Kant, who lamented that the field of chemistry had not yet embraced a mechanistic and mathematical expression of chemical phenomena. As a result, according to Kant, chemistry at that time was just a science, rather than a Science with a capital S. Despite Kant's view, however, as Thompson emphasizes, a great quantitative revolution proceeded to transform chemistry into a capital-S Science every bit as rigorous as physics. Thompson goes on to argue that biology is poised for just such a quantitative revolution.

Today, Thompson's thesis is being borne out; biology is becoming an increasingly rigorous quantitative Science that is finding more generality with each publication cycle. Most would agree that mathematical theories of quantitative genetics (including the modern synthesis), populations dynamics, organism interactions, epidemiology, ecosystem processes and growth and metabolism have together revolutionized biology, transforming it into a capital-S Science. This quantitative revolution would have been greatly muted, though, had investigators not been compelled, by Thompson's explicit advice, to identify general patterns and laws, to describe these quantitatively and to search for underlying mechanisms.

In this light, Keller's thesis that biology is a series of exceptional cases is a great leap backwards.

I find Keller's thesis all the more confusing given her biography of Biology, "Making Sense of Life".  In that excellent book, Keller recounts the history of modern biology with the theme, "Explaining Biological Development with Models, Metaphors, and Machines".  There is a clear trajectory in her history from natural language models of biological function towards more quantitative descriptions, all the while with participants flirting with writing down mathematical laws of biology.

In her Nature essay, Keller seems to suggest that the present profusion of data in biology suggests a complexity beyond description by general laws, and that this putative state of affairs is both acceptable to biologists and something to be more broadly expected and accepted:

In the past, biologists have been little concerned about whether their findings might achieve the status of a law. And even when findings seem to be so general as to warrant thinking of them as a law, the discovery of limits to their generality has not been seen as a problem. Think, for example, of Mendel's laws, the central dogma or even the 'law' of natural selection. Exceptions to these presumed laws are no cause for alarm; nor do they send biologists back to the drawing board in search of better, exception-free laws. They are simply reminders of how complex biology is in reality.

Again, I would differ and observe that experiments in physics didn't start to reveal any laws until people started working with simple systems.  Indeed, every physics experiment is highly engineered to be as simple as possible, so that only one thing is being measured, and only that one thing can vary.  The practice of engineering simple biological systems for the purpose of understanding how biology works is one of the driving forces behind Synthetic Biology.  As I recount in my forthcoming book, this effort is really only just getting under way, as the necessary tools have previously been either too expensive for broad use or have only just become available in any form.

Biology is definitely a Science, capitol "S" and all.  It will just take some time to write down the laws.