MrDarwin

Any ideas that a system is "optimally designed" assumes that we can identify those that are "suboptimal", and that for any system, it could have been configured in any one of several different ways, some of them better than others.

So how do we tell the difference? How would we go about identifying a "suboptimal" design? What criteria might we use? Is it even possible to identify suboptimal designs? (Because if it is not, then it is likewise impossible to identify optimal designs.)

Anybody? Anybody? Vanderzyden?

Hastur

I'm not sure I follow, MD.

It may not be possible to define the optimal design for a compact automobile, for instance, but it's pretty easy to point out that having the fuel tank right next to the back bumper isn't a good idea.

MrDarwin

And just why isn't it a good idea? How do you know? You are apparently judging, somehow, that one design is, or would be, better than another. So how do you arrive at that conclusion?

Hastur

Ahh.

Now I see. What you're after is the basis for saying that one design is "good" and another "bad".

Sorry, can't help you. I'm not a philosphy major.

It would seem though, that even without knowing what a perfect design is, one can compare two given designs to see which one accomplishes its' mission more effectively.

Tom Sawyer

I think you have to ask if there is a way that it COULD be designed better, even if it works fine the way it is designed.

Using the car analogy, there is nothing inherently wrong with a blueprint design that puts the fuel tank by the back bumper. The car works perfectly well when the fuel tank is there. It's only when you factor in environmental influences, like another car rear-ending it and causing the fuel tank to explode very easily due to where it's placed that you can see that this is a suboptimal design for fuel tank placement. Putting the fuel tank in a different place gives less of a blowing up factor, which is important in any car, so the a car design with a fuel tank in a different place would be more optimal.

In terms of biology, there are many instances where you could look at something and say "hey, I could have done that better if I were in charge." There are many, many places throughout these forums and elsewhere where these can be found. Take the fact that each cell has a complete copy of the body's DNA when it only needs or uses a small portion of it to work. If you were the manager of a factory that built cells and your workers were putting the whole DNA into each unit they produced, instead of just what was needed for each individual cell, that would significantly lower the productivity and profit margin for the factory and you'd tell your workers to stop putting the whole DNA in and only use what they need. The fact that nobody's told the factory to do that suggests that there's no manager and the place just runs on its own.

There's lots more examples like this. When looking at something, just ask yourself "Could I have designed it better?" If the answer is yes, you have a suboptimal design.

[Edited for dyslexia]

[ October 18, 2002: Message edited by: peteyh ]

[ October 18, 2002: Message edited by: peteyh ]</p>

davidH

According to natural selection the fact that each cell has a complete copy of the DNA of the organism shouldn't be a fact.

Way back in the evolutionary timeline this "excess" information should have been lost since it was causing a lot of unnecessary energy usuage.
So the question that has to be asked is - why is the excess DNA present?

What possible advantage did it give to an organism over one that slowly lost the excess DNA through mutation?

If you can think of no advantage then evolution is called into question. If however there is an explanation - maybe it applies for the reason why cells having excess DNA is an optimal design.

Tom Sawyer

Not at all. Evolutionary advancements tend to use what's already there. If you have single-celled organisms that give a complete copy of their DNA to each half when they split in two to reproduce, then if an error happens in that split and the two children don't subdivide properly, but are stuck together - sort of like the difference between conjoined twins and regular twins - then each half would have a full copy of the DNA. If that organism splits in two to reproduce, it's children would both have two cells and so on and so on for a couple of billion years until you have a bunch of multicelled organisms.

Although it would be better not to give a full copy of the DNA to each cell, that just so happens to be the way that it happened and it works fine that way, so there was no need for that aspect of development to be selected out. It works the way it is and remains as a neutral trait in the design of the organism while other factors select against nature and evolve all the species and just carry this historic design, which was useful in one-celled organisms, but not anymore, along for the ride.