Biology's First Law The Tendency for Diversity & Complexity to Increase in Evolutionary Systems

The subtitle of this book points to an observation that most biologists will anecdotally agree with. Looking at the long sweep of evolutionary history, there is indeed a clear overall tendency for life forms to become more diverse and complex. Daniel W. McShea and Robert N. Brandon, the one a biologist with a secondary appointment in philosophy, the other a philosopher with a secondary appointment in biology, here declare it the Zero-Force Evolutionary Law or ZFEL. But is this a law of nature? And does it really differ from stochastic processes or even entropy?

The authors dive right in with some definitions, and to discuss this book it will be helpful to repeat these. If you have an evolutionary system, and you assume variation and heredity, then diversity and complexity will on average increase, unless other forces, e.g. natural selection, constrain or override this tendency. The authors give two definitions and point out they do not consider the ZFEL a new discovery. Rather it is a recognition of, and unifying principle for a range of findings across biological disciplines. When not subject to selection, differences tend to accumulate, both within organisms (e.g. random mutations or gene duplications leading to pseudogenes) and between them (the increasing diversity and complexity of life over evolutionary history).

McShea and Brandon quickly move to clarify what they mean with diversity and complexity. Especially the latter has a colloquial usage indicating sophistication, functionality, etc. They are not interested in this and define “pure” complexity: the number of part types making up a structure, or the degree of differentiation between parts – nothing more. And this creates a hierarchical relationship between the two: diversity at one level makes for complexity the next level up. For example, a great diversity of cells makes for a complex organism.

An important and very interesting consequence of their line of thinking is that it invites a so-called gestalt shift: a change in what is considered background and what foreground. In their view, change is the constant, with natural selection acting more to reign in the creative maelstrom generated by the ZFEL. In their final chapter, they consider how this provides a satisfying answer to the challenge often raised by creationists: how do complex structures such as eyes arise? Biologists often respond by suggesting a series of intermediate steps, each a little bit better than the last one, each selected for, which risks descending into just-so stories. But if the ZFEL creates diversity and complexity in abundance, natural selection acts more by whittling down and constraining that which is useful, meaning intermediate steps are not strictly necessary (which is not to say they do not occur, see e.g. Evolution's Witness).

Biology’s First Law is a brief book and throws out an interesting if provocative idea. Although the topic is rather technical and abstract, the authors do a great job keeping the reader on board, repeating important caveats where necessary, and being explicit in what they mean. Even so, some of the examples and argumentation feel, well, a bit clunky.

My first thought a few pages in was: “isn’t this just entropy?” The second law of thermodynamics states that there is a natural tendency for systems to become increasingly disordered and chaotic with time, while maintaining order takes a constant input of energy and effort. This is why stuff breaks and falls apart. The authors are adamant that it is not, and, to their credit, marshall convincing arguments (though see Cosmic Evolution for an example that does use entropy as an explanation). What does not help their argument, though, is that the two analogies they repeatedly use in the book (a picket fence breaking down and the wind scattering leaves in a garden) are prime examples of entropy in action.

Second, in some places the writing starts to reek of ad-hoc explanations and special pleading. They suggest that if complexity and diversity do not increase, you just need to invoke a mechanism (genetic drift, an absorbing barrier, stabilising selection) to explain why not. And any such inhibiting mechanism would not challenge the ZFEL, say the authors. Take it away and the tendency for the ZFEL to increase diversity and complexity will resume. Basically, they take the ZFEL to always be true, which does not leave much room for falsification.

Third, given the starting points of several of their hypothetical examples, the outcomes are self-evident: a picket fence with identical pickets, a population of initially identical organisms, a wormlike animal with identical segments. These systems are at the extreme end of morphospace (the space of possible configurations) and can only develop in one direction. What about real-life systems, in all their messy glory? Wagner’s computational work and identification of astronomically large spaces of possibility are interesting in this context (see Arrival of the Fittest).

Tying in with that, despite the author’s purported humility, calling their idea a law and drawing parallels to Newton’s first law is anything but. Taking the ZFEL to be true is putting the cart in front of the horse – possibly a purposeful provocation to get people thinking. For now, it is a hypothesis. To their credit, they outline what research needs to be done to falsify their idea and they provide some interesting evidence. Some of this is not supportive and they evade this by the epistemological escape hatch I mentioned above: just find an ad-hoc constraint to invoke.

Usually I review new books, but since this is a 2010 book we can ask: what has become of this idea? I have not done an exhaustive search, but there are two noteworthy follow-up papers I immediately found. Puzzles for ZFEL, McShea and Brandon’s Zero Force Evolutionary Law was published in 2012 in Biology & Philosophy and raises several points of criticism, different from mine and more thoroughly developed. A Quantitative Formulation of Biology’s First Law was published only last month in and is a modelling exercise for discrete and continuous traits.

Biology’s First Law is a very interesting, thought-provoking book and I think the authors are on to something, but this idea needs developing. Some of the examples and arguments here are self-evident or not very strong. And I’m not yet convinced that we need a name for this, calling it “stochastic processes” or “randomness” fits the bill just fine in my view.