Convergent Evolution - Reasonable conclusion? or convenient excuse?

And consider that this example has been picked out precisely because of its remarkable feature.

Why did Carl Linnaeus, who died 80 years before Darwin published the Origin of Species and believed that all life existed in immutable forms laid down by God, place all animal life into a nested hierarchy then? Why does that system, laid down 250 years ago, broadly persist to this day? Why haven't the tens of thousands of taxonomists working since that time noticed that this nested hierarchy doesn't match reality as you assert?The answer, of course, is that as a matter of fact life does form nested hierarchies. And the hierarchies it forms look very similar whether you look at genetics, morphology or fossil data.

Why? What survival benefit would we gain from hearing ultrasound? How about a dog? An elephant? Generally speaking smaller things make higher pitch noises, and the bigger you are the less interested you are in what small things are doing. Know the high pitch whine of various insects wings? Irritating, isn't it? But where's the benefit? Stick the threshold of hearing into the ultrasound and all you get is a load more insects whose wing-beats are an irritating whine intruding in your world but whose existence is entirely irrelevant to you.So I'm not seeing a big difference benefit to higher hearing thresholds.Now the cost. Hearing "hair" cells are arranged in the inner ear in order so you get the cells that respond to the lowest pitch first, and so on deeper into the ear until you reach the cells which respond to the highest pitch (IIUC the cells are similar, but the shape of the cochlea focus different sounds to different points but don't quote me on that). Increasing the auditory range would either mean altering the "step" on these cells, reducing the fidelity of the auditory input so you had less ability to distinguish sounds across the most useful part of the range or having a deeper cochlea with more cells. That means more cells, a bigger cochlea, more nerves to receive this input and more brain required to interpret the input. That's a cost in terms of energy to build and maintain.So, you see, a higher frequency hearing would have few benefits and non-zero cost so like so many other things, we see a trade off between costs and benefits. Something evolution works to optimise.

No, you've got it backwards. This is how it goes:Science: the overwhelming consilience of evidence points to the evolutionary explanation.
You: But Bats and Dolphins have similar proteins that's so improbable evolution can't have happened.
Us: Show us the probabilitySee how this is you being challenged to back your argument? We accept evolution because of the evidence for it. If you wish to challenge that position with a probabilistic argument you have to have some actual maths to back you up otherwise all you're doing is arguing from personal incredulity, and that's not terribly convincing at all.

How is that evidence against common design? Do designers always use the same elements?

I ran both of your examples through PARS and it gave these trees:For your nested version, it finds a single tree of tree length 12:

+--------Taxon 8
           +--------6
           |        |        +Taxon 7
           |        +--------5
           |                 |        +----------------Taxon 6
  +--------3                 +--------4
  |        |                          +Taxon 5
  |        |
  |        |        +--------Taxon 4
  |        +--------2
  |                 +--------Taxon 3
  |
  1--------Taxon 2
  |
  +--------Taxon 1

Unfortunately, PARS requires an outgroup to work best and treats the first taxon as the outgroup if it doesn't get one so this is a bit off, but note that even so it has grouped most of the final pairings together and identified taxons 7&8 and 5&6 as more closely related to each other than to any other taxon.Now if we look at the random sequences, two most parsimonious trees are found, both with tree length 28:

+---------Taxon 5
            |
  +---------3            +-------------------------Taxon 6
  |         |            |
  |         +------------4             +-----------------Taxon 8
  |                      |             |
  |                      +-------------5-----------------Taxon 7
  |                                    |
  |                                    +-----------------Taxon 3
  |
  |           +-----------Taxon 4
  1-----------2
  |           +-----Taxon 2
  |
  +-----------Taxon 1

and

+---------Taxon 5
            |
            |                                 +-------------Taxon 8
  +---------3              +------------------5
  |         |              |                  +-------------Taxon 7
  |         |              |
  |         +--------------4----------------------------Taxon 6
  |                        |
  |                        +----------------Taxon 3
  |
  |           +-----------Taxon 4
  1-----------2
  |           +-----Taxon 2
  |
  +-----------Taxon 1

To be honest, I'm pretty surprised at how nested that came out.(edit) Actually, when I think about it, I'm not. If you think about it there are only 7 characters, so the maximum possible tree length is going to 49*. So that a random set produces an intermediate value kind of makes sense. There are a bunch of long arms in that "tree".* - Actually that's an overestimate because there are only 6 states for each character but I can't think what that makes it off the top of my head.Edited by Mr Jack, : No reason given.Edited by Mr Jack, : No reason given.

Hi Percy, It can work by treating one of the taxons as the outgroup (taxon 1, here). The analysis will then be less good at determining the rooting of the tree - as, in fact, we see - since it doesn't know which states are primitive (unchanged from the ancestral species) and which are derived (changed from the ancestral species). The tree length is the number of individual changes required to form the tree, this is, each time we swap a single letter to another to make the tree the tree length increases by one. The most parsimonious tree(s) for a given set of data is the one with the lowest tree length - in other words, the one that applies the same mutation the least number of times. If you start with the assumption of a nested tree, you can fit data to it. What will be different is the nature of that tree, and the length of that tree, I'd guess. You'd also expect to have more different most parsimonious trees, I think. Er, I got it with my course materials, I assume Wounded King's link will take you to where you can get it. I needed to tweak a couple of settings to produce the correct formation of the output, but I didn't post-process it at all or make any attempt to alter the trees (although I did cut those trees out of a more detailed output).

Hi RAZD, I'm curious too The first reason I can see is that there's no outgroup, although, as Wounded as already shown, even that doesn't help much. The outgroup is important because it helps the program determine which states are later mutations and which states were there originally. The second reason is that Percy's example happens to have multiple mutations at the same site, specifically the first mutation on the left branch is then overwritten by the second mutation on the leftmost branch (AAAAAAA->AAAAABA->AAAAAGA). Because this was the only mutation in that branch, the information needed to group the later pairings together has been completely lost. The same has happened on the righthand branch (AAAEAAA->AAAGAAA).In real data, there'd be many more datapoints, and more states; usually allowing multiple changes in the same character to the be resolved.

Note that the tree Wounded King is referring to is the one in which he added the outgroup, whereas I was referring to the one without.