Ongoing debates: directionality and progress

Over the last two hundred years or so in which evolutionary ideas have been seriously proposed we can find a full spectrum of opinions on evolution and progress. As philosopher Michael Ruse has observed “evolutionists, including Darwinian evolutionists, are badly split on the question of whether or not the path of evolution is progressive, from simple to complex, from blob to the human.”

Writing in the Origin of Species Darwin had spoken of a “Tree of Life” formed by the branching of species, with twigs and leaves reaching skyward. As we saw earlier, Gould has championed an alternate image of a bush, as he believes that the tree image incorrectly implies there is a degree of upward directionality within evolution. With Darwin’s model it is easy to suppose that the uppermost branches and twigs of the tree are superior to those further down. By contrast, Gould wishes to emphasise the non-directionality, suggesting that we view life as “a copiously branching bush,” with some branches near the top, and some at the bottom, all being “continually pruned by the grim reaper of extinction, [and] not a ladder of predictable progress.”

Gould has had to defend his commitment to non-directionality against the observation that more complex organisms have, in fact, been preceded by less complex ones. In his classic text On Evolution, John Maynard Smith offered a rather ambiguous response to this dilemma: “It is in some sense true,” he writes, “that evolution has led from the simple to the complex ... I do not think that biology has at present anything very profound to say about this.” Maynard Smith’s conclusion is justified if we consider natural selection in abstract algorithmic terms. At this level of description we can expect to see an increase in fitness, but this does not automatically imply an increase in complexity since a complex organism is not necessarily more fit.

Perhaps the most optimistic advocate for progress and directionality in evolution was Jesuit palaeontologist Pierre Teilhard de Chardin. Writing in The Phenomenon of Man, first published after his death in 1955, he proposed that a trend was visible in evolutionary history which he referred to as the ‘law of increasing complexity consciousness.’He proposed that this law, which has brought about human consciousness, will continue influencing evolution into the future, culminating in what he called the ‘omega point.’ As we might expect, Gould was severely critical of Teilhard, describing any directionality as a “noxious, culturally embedded, untestable, nonoperational, intractable idea that must be replaced...” Ruse has noted that Gould’s position may have been in part due his abhorrence at the possibility that science could be used to justify the claim that one species or race is further evolved and so superior to another. Ruse is an even handed philosopher who is not at all sympathetic to the ideas of Teilhard, in fact he has characterised evolution as “going nowhere ... and rather slowly at that” but also allows that “we might suppose that evolution, even Darwinian evolution, is a lot more directed than someone like Gould allows.”

How might Conway Morris’s work on convergence play into the question of directionality? Once again fitness landscapes can help us visualise the situation. While at first glance evolutionary convergence appears to introduce a directionality that Gould would find repugnant, this is not the case. Convergence implies that life will inevitably draw from a fixed set of fitness strategies, but it does not tell us which of these strategies is superior. If we picture the situation on a fitness landscape (viewed at the largest scale) we can imagine a peak for bacteria-style fitness (i.e. highly robust, massive growth rate, etc.) and a different peak for human-style fitness (i.e. fragile body, the ability to respond intelligently to threats and opportunities). Neither is more ‘natural’ than the other, nor is one superior to the other.

One more parameter needs to be added to our model to allow it to represent the full array of options; population density. This is tightly related to the time parameter since the longer evolution runs, the greater the number of organisms. The effect of this parameter is more easily visualised if we invert the landscape such that the lower points now correspond to greater fitness to that local environment. Organisms can be represented by ball bearings dropped at random locations at a specific rate onto the landscape. In the inverted landscape, natural selection acts just like gravity in physical landscapes, drawing the ball bearings into valleys and basins.

The parameters that influence the character of evolution are: the shape of the fitness landscape (flat or highly contoured), the source of the fitness landscape (arbitrary, planetary, or universal) and the population density (i.e. the time allowed for evolution). The question of directionality in Earth’s biological history depends very much on the kind of evolutionary scenario we are actually in. The noteworthy options are as follows.

Weak convergence

On a weakly convergent (i.e. mostly flat) landscape the density factor does not significantly affect the result. No part of the landscape is particularly noteworthy, and we always find the entire landscape populated at random. In this case, there is no overall directionality.

Strong convergence, high population density

However, for cases where convergence is significant, or if biology is held to be universal, landscape density is critical. If 100 balls are dropped into a landscape that has 9 major basins, it is likely that all of the basins will be occupied. In this scenario, there is also no directionality per se, because the landscape is likely to be ‘flooded.’ In real terms, this could mean that, yes, human-like organisms and bacteria-like organisms were likely from the very beginning, but now that they have evolved, evolution has reached a kind of stasis. This is essentially the view described by Conway Morris: “Darwinian evolution remains central as the agency,” he says, “but the nodes of occupation are effectively predetermined from the big bang.”

Strong convergence, low population density

On a highly convergent but sparsely populated landscape, organisms may not yet be at their optimum fitness. Crucially, there may be entire basins that have not yet been occupied at all, but which will be in time; i.e. if only 9 balls have been dropped into a landscape with ten major basins, one of them will be unoccupied. This last scenario could be considered a directionality of sorts, since evolution has ‘somewhere significant to go.’ This conception of directionality is very different from Teilhard’s omega point, and does not appear to be the kind that Gould would find repulsive.

 Printer-friendly | Feedback | Contributed by: Adrian Wyard