Thread: "Evolution follows a fractal pattern"

Would such a statement be considered a "hypothesis" or an "obvious fact"?

I guess it could be a hypothesis (given a way to test it).

But I'm not really sure what it means. Which aspect(s) of evolution would be fractal? Rate of change, number or distribution of species? Or something else? And fractal in what sense? "Fractalness" is a geometric property; how would you apply that to evolution? To me, it sounds a bit like saying that evolution is isoceles or spherical....

Originally Posted by Strange

Which aspect(s) of evolution would be fractal? Rate of change, number or distribution of species? Or something else?

Perhaps all aspects of natural evolution could be fractal. Everything else in nature seems to be. Whether they they would all be part of the same pattern or each aspect has its own individual progression, I don't know.

I was thinking of the relationship between so-called "micro" and "macro" evolution, and even though the scales are different, they're both similar to each other.

Single-celled and multicellular organisms, plants and animals, fish & mammals, trees & grasses, etc. You could say the differences between each pair is as the differences between the following pair.

I'm not sure if I'm really explaining my thoughts well enough.

As Strange has suggested and you seem to have confirmed your use of the word fractal in this context is ill-defined. It is in danger of coming across as a buzz -word that sounds scientific, but is too wishy-washy to have significant meaning. Could you tighten it up a little?

Maybe I'll be able to explain my thinking a little better with an analogy. A very simplified analogy.

For example, imagine the Cantor Set:


Suppose the top bar was "Life" and the two below it were the Domains, and underneath them were the Kingdoms, etc.

Would the amount of difference between the Domains be proportionate to the value of the differences between each of the Kingdoms? Likewise for the differences between the Phylums relative to the Kingdoms containing them?

Say the 2nd-from-bottom line in the picture represented dogs, cats, horses, and so on. Below the dog sections you'd have all the individual breeds of dog, below the cats you'd have lions & tigers, etc. and the same with the horses, being horses, donkeys and zebras, etc.

The species on the bottom line can (just about) cross-breed with their neighbor as the "distance" (as displayed in the picture) is rather small, but the distance between each breed of dog and each breed of equine is proportionate to the distance between dogs as a whole, and horses as a whole.

Could the relative evolutionary distances between each lifeform be proportionately similar to the evolutionary distance between groups on the tier above it? Fish are to mammals as animals are to plants, that kind of thing.

(Wow, it took me ages to think of how to write all that down in words and I'm still not sure I've explained my thought process clearly.)

Ah, I see what you mean now. (Kind of)

I think there is still a challenge of quantitatively defining distance, which would be required before you could test for fractalness (fractalicity?).

But it is possible that looking at the genomes would allow this. I believe there is already a concept of distance used there.

Originally Posted by Daecon

Would the amount of difference between the Domains be proportionate to the value of the differences between each of the Kingdoms? Likewise for the differences between the Phylums relative to the Kingdoms containing them?

Those groups are pretty much entirely arbitrary - more or less based on the whim of biologists (though certainly not irrational). Taxa are, as far as I understand it, not real objects.

Bifurcating lines drawn in a tree-shape may look fractal-like in the sense that if you zoom in to a specific area of a tree it will tend to look roughly like any other area or any "zoomed out" area. Problems with this pictorial analogy arise when you consider that all lineages have end points, that is, the splitting lines terminate. So the fractal pattern is not really fractal. Second, taxa do not have neat quantifiable differences between their genomes. For example, some E. coil "species" may have up to 60% difference genetically, while other species within a genus may differ by less than 1%.

Originally Posted by Zwirko

Originally Posted by Daecon

Would the amount of difference between the Domains be proportionate to the value of the differences between each of the Kingdoms? Likewise for the differences between the Phylums relative to the Kingdoms containing them?

Those groups are pretty much entirely arbitrary - more or less based on the whim of biologists (though certainly not irrational). Taxa are, as far as I understand it, not real objects.

Bifurcating lines drawn in a tree-shape may look fractal-like in the sense that if you zoom in to a specific area of a tree it will tend to look roughly like any other area or any "zoomed out" area. Problems with this pictorial analogy arise when you consider that all lineages have end points, that is, the splitting lines terminate. So the fractal pattern is not really fractal. Second, taxa do not have neat quantifiable differences between their genomes. For example, some E. coil "species" may have up to 60% difference genetically, while other species within a genus may differ by less than 1%.

Ah, I didn't know that. I just assumed that the principles of fractal geometry would apply to the way life changes over time.

Fractals and evolution are both examples of emergent complexity but life itself isn't necessarily fractionally though some aspects of it are.

You could describe the branching of the tree of life as a fractal but it's too vague, even the definition of species becomes a bit fuzzy when you look at ring species.