philosophy science

methinks no telos in evolution

My ‘techno-skepto’ mate, Damian, has posted an updated version of his funky cool little evolving tool.  Head over and have a play with the numbers (offspring & mutation rate) and the target phrase.

25 replies on “methinks no telos in evolution”

Ha! Quite right (and thanks for the link). There’s really no telos in evolution. That’s the main reason why the ‘weasel’ calculator is not a good evolution simulator; it’s got a goal in mind and it selects accordingly. If it were a real evolution simulation it would be the environment doing the selection. But I think it does a great job of showing how selection produces very different results from the randomness implied by ‘monkeys with typewriters’ or ‘a tornado in a junkyard’.

Cheers Damian,
I’ll take this opportunity to ask a question I’ve been thinking about recently.

We can imagine a massive asteroid blizzard colliding with us, a) blocking out almost all sunlight and b) almost totally corrupting the air, and thus wiping out all life on earth, except any simple organisms that could survive with the hugely reduced quality of air & sunlight, etc.

My question is (and you know I don’t have theological/biblical or other wise ‘christian’ problems with evolution): is it inaccurate to propose that there would have been ‘set-back’ after ‘set-back’ where biological life would/could have been either reduced to it’s basic forms (if not wiped out a few times to re-organise/restart?)? I guess that would assume a large number of catastrophic events large enough to do that… But still, it seems that we must say that in addition to having no ‘telos’, the mechanism of natural selection has no preference for complexity over simplicity?

(sorry for rambling!)

Oh oh oh can I answer? (Sorry for butting into a conversation again, you’ve randomly picked on my pet subjects in the last few days :))

It seems simple life (bacteria and the like) are the dominant form of life on earth right now by almost any imaginable measure – number, biomass, impact, adaptability, environments colonised, capabilities, diversity etc. In that sense I would argue nothing substantial would change from a non-human viewpoint.

Stephen Jay Gould describes increasing complexity like the statisticians drunken walk. Here a drunk leaves a pub and stumbles on down the road randomly moving left/right/ahead. His movement to the left is restricted by a wall but he can move as far right as he wants. Statistically over time he will eventually move further and further to the right despite there being no goal to do so.

Gould argues that complexity of life is similar – where too-simple for life is the left wall and an infinite possibility of complexity stretches out to the right. In that sense complexity is a statistical probability rather than anything that is “preferred”.

Cheers Ian,
The ‘drunken walk’ could be a very helpful metaphor. If I may adapt it to my imaginative uses ( ;p ), I’d want to draw attention to the possibility of the man (or woman, you sexist – kidding) to simply fall down or indeed to do a 180 and head back the way he came?

As for the ‘wall’ and probability, etc. wouldn’t some/many/even-most? drunks use the wall to balance against!?? But moving on, consider this further adaptation:

I’d want to see ‘left/right/ahead’ as the particular path toward this or that kind of complexity; the man cannot and does not ‘aim’ at any kind of complexity (and – again – may well get himself turned ’round and walk back toward simplicity if he finds himself leaning that way!). As for a ‘wall’, I wonder if this is simply the ground? If he is no longer walking (living & passing on genes), he is passed out (evolutionary dead end).

The reason I (provisionally) like this better is that I don’t yet see how complexity was in any sense ‘eventual’ in the sense of being necessary. But then again, you only spoke of ‘probability’, not necessity.

It seems to me that life (after self-organising – something I’m still baffled by) didn’t need (or have the odds in its favour?) to get complex. Heck, in the whole biological history of earth (within the larger story of the cooling of the universe), there’s a window that life had to appear and diversify within. It could have been that even with the window it didn’t happen or happen as well (as complex & as diverse) as it did.

rambling again!

The actual forward motion down the road represents time in the analogy so a 180 or staying the same place doesn’t quite work. Incidentally the standard drunkards walk assumes either a 50/50 left/right or 33/33/33 left/same/right distribution but it works just as well (just slower) provided there are non-zero probabilities left and right.

One way the drunkards walk falls down in this analogy is that it deals with a single position. In terms of the complexity of life it is better to think of the path as the far extreme of complexity with all points towards the wall filled in – i.e. it represents the range of complexity. The same basic principles apply though.

So yeah, in my view an increase in the complexity of life was not necessary at all, but it is certainly not against expectations. It was also by no means required that life went in the particular direction it did.

Just for fun you can set this up in Excel:

In A1 Put “X” (title for column containing the location of the drunk)
In B1 Put “dX” (title for change in X)
In A2 put “0” (starting point at the wall)
Leave B2 blank
In A3 put “=MAX(A2+B3,0)” (this adds the change to the previous value unless it would create a value lower than zero representing the wall)
In B3 put “=RANDBETWEEN(-1,1)” which creates a random integer of -1, 0 or 1
Copy A3 and B3 down as many rows as you feel like.
Select column A and then insert a chart of these values (I suggest a vertical bar chart but most options work.)

This will give you a particular random walk. Press F9 (or change any cell) to recalculate the random numbers and watch the chart change.

I like that Gould analogy Ian.

Here are my ill-informed thoughts on how complexity might fit in.

First, it would be useful to define ‘complexity’. Perhaps a good starting definition would be “a system comprised of many parts with many interactions”. That way we could say that an alligator is complex because it’s made of lots of cells that interact with each other. Or that a colony of penguins is complex because there are many individuals who are, in turn, made of many cells and so on. Or that the planet is complex, etc. A rain cloud is also complex but nowhere near as complex as a single bacterium in its interactions.

Ian, you might have a better definition to hand.

As Ian pointed out, it’s actually the simpler lifeforms that dominate our planet (we’re big and so we tend to only notice bigger things). I think that Dale brings up an interesting idea that perhaps this is because of the several major extinction events. But it would seem to me that there is a massive niche that will always be there for the small and less complex things. In fact, if all the bacteria were to ‘move on’ and evolve into bigger and more complex creatures all of us biggies would die because we’re entirely dependent on them. And presumably a bacteria that became more complex would also be as dependent. A catch 22 that would indicate that one of the most successful evolutionary strategies would be to be the little guy.

But the environment we find ourselves in has many niches and we see good examples of this when there is a major extinction event these niches are quickly filled (thinking here of large dinosaur herbivores who were rapidly replaced by large mammalian herbivores who’d sprung up from shrew-like mammals).

It would seem to me that there is also a niche available in any ecosystem to be the ‘smartest’. And I think that entails being able to make good predictions about the future which would seem to require a brain of some kind that can run reasonably accurate simulations. Mammals and birds seem to be pretty good at this (not so sure about reptiles?) and humans are the kings of this niche. At one stage we had at least one serious contender for this niche but either we killed them off, bred them out or they got unlucky somehow and died off by other means. We came pretty close going extinct ourselves perhaps 70,000 years ago.

I think that with complexity comes a kind of fragility that means that more complex creatures can more easily become extinct during an upheaval. I think that, like the drunken man analogy, some creatures will find themselves in niches where it was their complexity that led to their selection but I also think that, perhaps like the drunken man wandering too far out onto the road and in front of a bus, the environment we find ourselves in happens to include things like meteroites which may not be survivable by all but the most simple organisms.

Perhaps humans have found a new niche on this planet as creatures that have reached the top of the forward-thinking-but-fragile niche. Right now if there were to be an extinction event the size of the Permian-Triassic event I doubt we’d survive it. But give us a couple of hundred more years (and a complete abandonment of spending 90% of our resources on wars and cosmetics) we might just get ourselves into a position of being able to at least dodge something like this. (And perhaps in two hundred years time we’ll find that these extinction events are pipsqueaks compared to some major cosmological solar-system-destroying event lurking just around the corner!)

Like I said at the start, I’m pretty ill-informed about the mechanisms of evolution and what kinds of trends in the development of complexity we should expect so take all that with a grain of salt.

loving the thinking and the questions! thanks gentlemen.

Damian, the idea of ‘niches’ makes me imagine a randomly swinging (and sometimes not swinging) nerf gun that randomly fires (and sometimes doesn’t fire) other little nerf guns in any and all directions (and sometimes in no direction). The ‘niche’ would be like a plateau nearby (or distant?) for the ball to (hopefully?) land on and develop into a full, able-to-fire gun itself. Depending on the size of the plateau (and distance away from the originating nerf gun), there would be less or more probability for the little nerf guns-buds to make it there.

a crude, imperfect analogy, of course – but might be an evolutionary step toward a better analogy!

I think complexity is best described in terms of the amount of information required to communicate what something is. Imagine a game of life board that is running over time. From simplest to most complex, the board could be described as

1. the whole board ends up dead (a simple system)
2. the board settles on distinct patches of alive and dead that might oscillate but otherwise stay the same (a system in equilibrium)
3. some parts of the board continuously change among a background of stability (a complex system)
4. the entire board is continuously changing chaotically (a chaotic)

What is interesting is that the amount of information required to communicate it increases from 1-3 and then actually drops again for 4. Complexity theory essentially points to the highest information content right on “the edge of chaos”.

You’ll note this leads to two subtly different meanings for the term complexity. One is the other end of a scale from simple to complex, the other describes a subset of system behaviour along the scale from static or chaotic. It is very easy to subtly misuse these two meanings. Complexity theory is not (per se) about the most complex things being the most interesting, it is about the right amount of complexity being interesting.

Gould actually uses “size” as a metric for complexity in the first sense in his analogy and it is a reasonable proxy but not perfect – some organisms are simply bigger versions of smaller organisms and require no real added info to describe.

The “niche” concept always makes me slightly uneasy because it implies limits to possibilities when I think these are virtually unlimited and ones position in the possibility field would not hold any significance viewed absent its current inhabitant. In that sense I think niche’s are apparent only in hindsight.

From an evolution point of view there must be a continuous potential connection (plateau) between all possibilities. I am not sure evolution has the capacity to make big jumps. Perhaps a closer analogy would be water flowing over a bumpy plane, and settling in lower areas representing continuous exploration. The puddles that remain are niches of a sort (although even that analogy is pretty crude).

As you could probably guess, I could ramble on about this stuff all day :)

yeah it’s interesting and hard playing with metaphors and models. I like different ones for different reasons.

I agree that niches and species would both be only apparent in hindsight in a purely evolutionary (and thus non-telic) account.

I think the idea of the ‘thickness’ of complexity may be helpful. Life would have started at its simplest and ‘thinnest’, and then – without leaving the level of simplicity – would have ‘thickened’ to include more and more complexity.

I guess what I’m wondering is how ‘eventual’ or ‘likely’ were organisms as complex as humans? We all agree, I think, that there were no prior goal(s) of a) any particular plant or animal species (say, Cedrus atlantica or homo erectus) or b) any particular ‘thickness’ of complexity. Surely it would have been quite possible for life to have remained relatively ‘thin on the ground’ – if that makes sense?

I’m not sure I follow what you mean by “thickness of complexity” unless you are referring to quantity/diversity of different species at a particular level of complexity?

I would suspect that, in a world where multi-cellular life is possible, animals the complexity of humans was virtually inevitable (in a statistical sense) given a long enough time span. However animals specifically like humans would be obviously far less likely.

by ‘thickness’ I mean the range (perhaps I should have used that word!?). When the least and most complex organism aren’t that different from one another, that would reflect a ‘thin’ biosphere – in terms of complexity. But when the difference is like from amoeba to anthropoid, that is a ‘thick’ biosphere.

And yeah, the question of ‘given enough time’ is exactly what I’m asking. It seems to me that animals with human-like complexity would not have been inevitable. Should we not be able to say that life could possibly have only advanced to the amoeba level of complexity? Or indeed never have arisen at all? This seems well possible to me?

Two other thoughts:

-I note that there would be (perhaps?) a temptation to assume that because we ARE here it must be by default an ordinary and inevitable thing for life & complexity to arise??

-Ian & Damian, from what you know or have heard, was the first biological life on earth plant, animal or neither?

Ahhh I see what you meant – yeah thickness is good way to look at it.

In terms of possibility over time I am quite sure is possible life of some form would never have arisen. I am not as sure this is “probable” however.

I think there is a misplaced tendency to assume human-like life is inevitable. I think the high chance of complex life of some form arising from simpler life however is a reasonable idea.

I think the boundary between life and non-life is fuzzy but I would imagine the first “life” was simply a molecule capable of connecting other substances together to make copies of itself. This molecule would have rapidly emptied whatever environment it found itself in and then a copying error may have produced a version that did something better or used resources the other didn’t and so on and so forth.

Yeah I guess I’m just wanting to know how we judge the reasonable-ness of the idea or the probability of it.

And whilst I don’t want to go there here/now, and not bent on widening gaps for God, abiogenesis remains the biggest gap, does it not? One can imagine stuff connecting with other stuff, but going from non-replicating to replicating seems a giant leap – even at that basic level of (relative) simplicity. How many times would/could the inorganic molecules have ‘missed’ making it to replicability?

I think the key to what limited judging we can do falls back on the sheer magnitude of time and space available for it to happen, and the diversity of things that could potentially count as it happening. I like the analogy to poker hands – the odds of drawing any specific 5 card hand off the top is about 0.00004%. However the odds of getting say any hand a straight or better is much better, at 0.76%.

And I agree that abiogenesis is the biggest gap in our knowledge, but I think that’s mostly a function of the time gap between when it happened and where we are now. I am not as sure it is conceptually much bigger than any of the other gaps and I’m not convinced it required any more improbable leap than any other step.

Interesting stuff Ian. Indeed, probabilities in our universe are always problematic because we’ve a sample size of 1, and we don’t know how much of it we’re even able to see!

And re abiogenesis, don’t take me for a young-earther, but I think it is a different (and larger/greater?) kind of gap in that natural selection (by mutation) cannot be the mechanism that brings about the first replicator

I am not sure that is really the case – the steps from atoms to molecules to combined molecules to combined molecules that do things to combined molecules that replicate are all conceptually similar size leaps. I agree that our knowledge of the step from molecules that do things to molecules that replicate is probably the biggest gap, but that is purely a function of perspective imo.

Natural selection actually operates on non-replicating systems in that the molecules that survive are those that survive and on average those more suited to an environment will tend to be present in higher numbers. The missing element is information storage which allows replication of innovation. When that kicks in you get a much more patterned behaviour as opposed to a more static system.

Catalysts, structure forming proteins, etc.

(this might be a double post, the last comment seemed to disappear. Delete this one if so).

ok, sure, in a very basic sense, there are (non-replicating) molecules that ‘do things’, and in one sense these can have varying degrees of suitedness to a given environment. Indeed, even a rock can have varying degrees of suitedness to an environment. But replicability would to me seem to exponentially multiply the degrees of suitedness in that the ‘selectiveness’ of natural selection really comes into its own with replicability.

It is more the “stickiness” of change that comes in with replicability – the selection component is more environmental. Because change can stick you get substantially more rapid “progress”.

It’s essentially the entire step :)

It is also worth noting that natural selection is not necessarily the only possible mechanism for rapid and pseudo-directed “progress”, it is just the only one we know of and the only one that appears to be operating.

Comments are closed.