Drift as Biology’s First Law

Synopsis: Robert Brandon agrees with the “evolutionary theory as a ‘theory of forces’ ” framework established by Elliott Sober but disagrees with Sober’s choice of Hardy-Weinberg as a zero-force law. Instead, Brandon believes genetic drift is to evolution as inertia is to Newtonian mechanics, i.e., drift is the “first law of biology.”

Last time I discussed Elliott Sober’s conception of evolutionary theory as a “theory of forces” in which evolutionary ideas like selection, drift, mutation and migration can be construed as “forces” pushing around allele frequencies. This post will not necessarily criticize Sober’s view (but I do have papers lined up to read that do so), but will expand on it a bit further with Robert N Brandon’s chapter in Philosophy of Biology, entitled “Biology’s First Law.”

Again, like the Sober chapter, Brandon’s chapter is freely available at Google Books, if you are interested.

According to Sober, a theory of forces requires a zero-force state or law which describes the state of an object when there is zero net force acting upon it. Sober believes biology’s zero-force law to be Hardy-Weinberg because it describes the state of allele frequencies when there is no selection, drift, mutation or migration – allele frequencies reach equilibrium and no longer change.

Robert Brandon disagrees with Sober’s choice of the Hardy-Weinberg law as evolution’s zero-force state. Instead, Brandon chooses genetic drift as biology’s zero-force law, even deeming it “biology’s first law.” This may initially seem counter-intuitive, but I believe Brandon makes many good points. I find his argument convincing even if evolution is not a theory of forces.

He agrees with Sober in that a force is a vector quantity – a force has magnitude AND direction. This readily applies to mutation and migration (according to Brandon). Brandon views selection a bit differently which will be explained below – I’ll make sure to note when that is – but he believes selection is indeed a force.

Drift, however, has problems when construed as a force.

While drift has magnitude and direction, its direction is unpredictable. Given alleles A1 and A2, drift will eventually drive one to fixation… but which one? Brandon acknowledges that drift’s direction can be defined as a decrease in heterozygosity in a population but again, which homozygous combination (A1/A1 or A2/A2) will reach fixation? He likens this problem to saying that a 20N force acts on an object – this is either nonsensical or incomplete as a force requires requires direction.

Brandon’s key point is that genetic drift is not a force, but the default position. “It is part and parcel of a constitutive process of any evolutionary system – namely, the sampling process” (87). He points out that sampling processes – such as gamete formation (Mendel’s segregation) – are inherent to the biological system. I believe this is akin to Sober’s conception of Mendelism as the background upon which the forces work (Sober 36).

However, can’t mutation also be explained in such a way – that mutations are inherent to the biologically system? The difference between an active force and the background makes sense to me (in an Einsteinian context (Sober 36)), but I am not sure how the two are different in a biological sense. Interestingly, Brandon notes that there has been much debate on differentiating selection from drift as both are sampling processes. He believes the two can be differentiated but those are in other papers that I have yet to even look for.

So the question must be asked – what is wrong with Hardy-Weinberg as the zero-force law?

Brandon describes Hardy-Weinberg (HW) as: If there are no evolutionary forces acting upon gene or genotype frequencies, then those frequencies will remain the same. However, H-W assumes large (or infinite) population size in order to rule out the effect of drift, but real populations are of finite size. Because drift is always acting upon a population (to different degrees), frequencies will change. Changing gene frequencies due to drift, not stasis (as described by Hardy-Weinberg), is the default state in biology..

This is intuitive. Brandon believes that while his article is the first articulation of drift as the First Law of Biology, molecular biologists and geneticists have been applying the principle for decades. When detecting selection, biologists compare the number of synonymous base substitutions against the number of non-synonymous substitutions. Synonymous substitutions, or neutral/silent substitutions, are allowed to drift and are theoretically invisible to selection (ignoring codon bias), while non-synonymous substitutions are immediately seized upon by selection. Essentially, the number of neutral substitutions (drift) serves as the null model – or inertial state.

Furthermore, DNA sequences that have not changed over time are under the effect of purifying selection, i.e., selection can keep allele frequencies from changing! This is where Brandon views selection’s force qualities differently (I told you I’d note it!). Normally selection is conceived as a specific kind of selection: directional. However, there are other “kinds” of selection like stabilizing and disruptive selection. If the default state is for gene frequencies to remain the same, stabilizing selection as a force is slightly problematic because it also keeps gene frequencies stable. However, when change is viewed as the default state, the zero-force state, the null hypothesis – purifying and directional selection can both be construed as forces. As stated before, stasis is not the zero-force state in evolutionary biology – change is.

Brandon’s articulation of the Principle of Drift is as follows (Brandon 89):

A population at equilibrium will tend to drift from that equilibrium unless acted upon by an evolutionary force. (A population at rest will tend to start moving unless acted upon by an external force.)

A population on evolutionary trajectory t caused by some net evolutionary force F will tend to depart from the extrapolated path predicted based on F alone (in either magnitude or direction or both) even if no other evolutionary force intervenes, unless F continues to act. (A population in motion will tend to stay in motion, but change its trajectory unless continually acted on by an external force.)

The parenthetical statements are to make the principle’s analogies to physics explicit. Clearly, however, the principle of drift is conceptually opposed to the idea of inertia (change as default vs no change) but they both parallel each other functionally as descriptions of the default state, or null hypothesis, of a system.

Brandon then corrects some misconceptions. The Principle of Drift does not take a side on the selection/drift debate nor in the debate on whether evolution is fundamentally deterministic or stochastic. He also notes that he has not discounted Hardy-Weinberg for its utility, only on its standing as a zero-force state. Brandon further says H-W is only applicable to diploid populations (as Sober noted through Kimura (Sober 37)) and thus is not general enough to be considered a law. According to Brandon, the principle of drift, as well as the principle of natural selection, are generalizable to any evolutionary system and can be conceived of as biology’s two laws.

Even if evolution cannot be portrayed as a theory of forces, I think the principle of drift is still effective, especially since it is clearly used by working biologists all the time. I find this conception of evolution particularly powerful and useful.

Are there any problems with drift as biology’s first law? As this was published just this year, I don’t believe any counter-arguments have had the time to be constructed and published.

Next time I will discuss arguments against evolutionary theory as a theory of forces. Not sure if I should be blogging about this stuff when I have classes, TAing, a history paper, a senior seminar, GRE, and grad school applications to work on…


Brandon, Robert N. “The Principle of Drift: Biology’s First Law.” Philosophy of Biology: an Anthology. Ed. Alexander Rosenberg and Robert Arp. Chichester (U.K.): Wiley-Blackwell, 2010. 84-94.

Sober, Elliott. “Evolutionary Theory as a Theory of Forces.” The Nature of Selection: Evolutionary Theory in Philosophical Focus. Cambridge, MA: MIT, 1984. 13-59.

6 thoughts on “Drift as Biology’s First Law

  1. Very nice. While the laws of physics may directly govern fundamental particles, which are distinctly a-biotic, it would certainly aid biologists in directing the field to a more substantially mathematical subject if more of a background in advanced physics was stressed.

    I’m definitely all for implementing the language of Biological Laws, though I think that the correct wording would make it possible to call them simply universal laws, though having to qualify the statement with such things as “organisms consisting of one or more cells which are reproduced via the molecular copying of nucleic acid residues…etc.” sort of dampens the significance.

    But seriously, don’t skip-out on the GRE or those applications!


    • Indeed. It seems weird to link Newtonian forces with evolutionary processes just because they affect such different phenomena.

      And yeah, I don’t know if drift and selection are truly universal. Can we imagine a form of life where neither are the case? Not sure.


  2. Isn’t that a characteristic of evolution that’s derived from the tendency for entropy to increase (the second law of thermodynamics)? Systems (gene pools) become disordered over time (genetic drift/diversification), to the extent that their situation within an organism allows them to (that is, excluding quickly fatal mutations) until something outside the system constrains that disordering – that is, another “force” acts on the system. I can’t quite make the argument myself, because I don’t yet know enough about physics or biology, but there’s probably a case to be made for a direct analogy there. Why is the proposed “law” more fundamental than the law of entropy?

    I also disagree with the assertion that the proposed law is “opposed” to the law of inertia. Isn’t entropy (on a molecular level) the result of Brownian motion, which is the manifestation of the law of inertia and the first law of thermodynamics? Aren’t genetic events essentially molecular events, and isn’t genetic drift due directly to the molecular processes surrounding genetic replication and transmission? Aren’t, then, “increasing entropy” and “genetic drift” or “genetic spread” equivalent terms, if the basic units of your system are genetic? Isn’t the primacy of genetic drift, the increase of entropy within a genetic system, thus consistent with the law of inertia, and in fact directly derivative of it?

    It’s not as simple as “change vs. no change”, because change is constant in all systems. What might be different is its immediate appearance to us, but it doesn’t make sense that the laws driving genetic and molecular events should be “opposed” to each other, if the current scientific dogma that all things that science studies are constrained by the laws of physics holds true.

    Those questions asked, the concept of dynamic equilibrium (that is, a state of little or no overall change as well as constant small-scale changes) isn’t really new. As the Hardy-Weinberg theory (from what I can tell after having read on it for 5 minutes) explicit excludes genetic drift, it’s only directly applicable to situations where one expects no changes in alleles themselves and no new alleles or genes to pop up. In that formulation of the problem, the states we’re talking about (finite populations sizes, all that real-world stuff) make the states that produce drift non-zero-force states. I think this is a nice extension of the approach to deal with this problem, but otherwise it’s hard for me to get excited about such an apparently logical step.

    Also, good to find another undergraduate blogger.


    • I don’t know much about entropy and how it affects life on the planet so I can’t comment on what you say to any great detail.

      Drift is usually seen as reducing genetic diversity and so I don’t know if the drift/entropy analogy works. An argument for mutation/entropy may be better, but I believe it’s still a stretch. I also know that entropy on Earth is a little wonky due to the planet constantly receiving a surplus of energy, i.e., Earth is not a closed system. I don’t know that affects life specifically though.

      As a side note, in the book, Unifying Biology, Smocovitis mentions that Fisher developed the Fundamental Theorem of Natural Selection as an analogy to the Second Law of Thermodynamics. I don’t know if this analogy means they were directly related or if he viewed it how I view drift as the first law: they both serve a similar purpose. In the case of selection and entropy, Fisher may have seen both as the universal laws that control biology and physics respectively.

      Thank you for the comment though!


  3. Speaking from a physicist’s viewpoint:

    This seems somewhat like saying, “If I leave a rock on the ground, it will eventually roll away.” The issue is not with the sudden motion of the rock. If that’s how nature works, that’s how it works. The issue is with the incredible vagueness of the statement. Roll in what direction? With what momentum? As a non-scientific idea it’s fine, but you’ll never calculate anything with it the way that physicists do with forces, and the analogy to trajectories may get you in trouble in the long run.

    Perhaps you’re looking more for something like the way a gaussian wavepacket spreads out in position space over time?


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s