Why biology is different
Last night, I read an old paper by Ernst Mayr on the position of biology amongst the sciences (Quarterly Review of Biology, 71(1): 97-106, 1996). In it, he posits that biology is rather radically different from other physical sciences; as a result, he says biology ought to be considered an autonomous branch of science, and not a provincial branch of physics or chemistry.
This, along with a more strident defense of biology (Science, 133: 1745-1748, 1961), surprised me more than a little bit. Since I do biophysics, I am not properly a biologist or a physicist, per se, but I do research that straddles the traditional boundaries of both fields. So I tend to look for good questions to be answered, and have not worried myself about whether my work is more "physics" or more "biology."
However, in his review, Mayr argues that there are questions dinstinctly biological in nature, and that these questions (along with some conceptual and methodological differences), make biology unique amongst the sciences. I thought I would paraphrase a few of the most interesting here:
Conceptual differences in biologyMethodological differences in biology
- The importance of historical narrative as an explanatory device.
- The prevalence of indeterminacy owing to the high frequency of stochastic processes, unknown factors, the presence of constraints, the interaction of multiple causes.
- The importance of quality (structure, form, function) in the properties and actions of objects, and a correlated reduction in the importance of purely quantitative differences.
- Presence of an historical constituent in the inherited program; hence legitimacy of "why" questions; capacity for the storage of historical information.
Other autonomous aspects of biology
- The importance of observation in addition to experiment.
- The frequency of independent multiple solutions to the same problem.
- All biological phenomena have two sets of causations, those controlled by the historically accumulated information of the genetic program (evolutionary or ultimate causations), and those controlled by the properties of the interacting system (proximate causations). The study of the historical components of each system is as legitimate a concern of biological science as the study of proximate causations.
- The outcome of biological processes is usually affected simultaneously by multiple causations, owing to the complexity of the systems interacting with complex biotic and physical environments
- Many properties of systems cannot be explained by a study of their isolated components.
It's interesting that many older biologists saw the successes of biochemistry and biophysics as a danger to the field of biology proper. Many early critics complained that such interdisciplinary research gave much to the other field, while returning relatively little to biology. In this light, the point about multiple causations (evolutionary and proximate) is still quite relevent, because the evolutionary causes are often unaddressed in many of the recent "hot" papers in biophysics.
Comments
The contribution of 'biological physics' (i.e. the biology of PRL) does not offer much to biology more based upon its style of presentation rather than any other factor. It is a classic example of a field divided along literature lines with little cross citation.
Posted by: Josh | June 25, 2006 2:07 PM
That's an interesting point that I hadn't thought about — the way research contributes to a field is in part by embedding itself in the "network of cross-citations." It seems that one way to contribute is by putting one's work in the context of others' work.
And I think you're right, the biological physics articles in PRL tend to concern themselves more with the previous publications in physics than with the recent work in biology.
The style of presentation is profoundly important too. I think concepts like "consensus" and "discussion" are far more important in biology than they are in physics. This is probably due to the nature of physics: it's much easier to create an isolated system to measure the impact of one independent variable on one dependent variable, thereby making it much easier to directly prove causation. This is often not possible in biology, where multiple interacting factors can give rise to unexpected phenomena (thus, our differentiation between in vitro and in vivo work). So in biology, we depend on discussion and consensus to a much higher degree than what is considered useful in the physical sciences.
Posted by: Courtney Hodges | June 27, 2006 8:58 AM