What does natural selection operate upon?

*

Natural selection is about incremental modification of already-existing forms or traits.

There is no problem about this happening - I mean that traits such as body size and structure or behaviours and abilities are modified by natural selection - obviously they sometimes are modified by selection (whether natural or artificial).

But there are difficulties which an explanation using natural selection must acknowledge.

One is this: How do we know what are the forms which natural selection modifies?

*

From the open-ended complexity of an organism - the limitless ways in which we might analyze its structure, behaviour and everything about it - how do we decide what are its 'traits' in order that we can potentially explain how they have evolved?

What is the objective evidence that a hand, or a wing, or a particular patten of food seeking behaviour is a trait? - something unitary, cohesive, which can be explained?

*

It seems 'obvious' to us that a hand is a trait - but that obviousness may be denied, or it may be an artifact of our own evolution. To an amoeba a hand is not a trait. If somebody doubts that a hand is a trait, how can they be convinced?

This is the deepest metaphysical problem of using natural selection as an explanation - the problem of explaining how we know what it is that has to be explained. 

*

This problem has only one answer: that the knowledge of what has to be explained, if it is to be valid, must come from outside of the discourse of natural selection: thus natural selection cannot be an ultimate explanation.

*

Natural selection operates within the discourse that tells us what needs to be explained - it just does - so the discourse that tells us what needs to be explained is more ultimate than natural selection.

Consequently, natural selection must be (and I mean must be - this is logically entailed) just one way of explaining what it is that needs to be explained.

*

Consequently, in relation to discussing the status of natural selection (the philosophical meta-subject which is evaluating the validity of natural selection as a discourse, rather than the specific instances of proposed natural selection - which are just normal science) the discourse that must be made explicit is that which describes what is a species, what is an organism, what is a trait.

How seldom this is addressed by those who argue for the ultimate and universal validity of natural selection...

*

(Obviously that meta-discourse which tells us what natural selection validly does, cannot itself use the explanation of natural selection! Natural selection cannot be used to validate natural selection; when the validity of natural selection is the matter being evaluated!)

*

From this we can infer that natural selection is based on assumptions concerning the nature of reality - that there are species (or some equivalent of organisms related by descent), that there are organisms as basic units of explanation, and that organisms may be analyzed into traits each of which can evolve by natural selection (with some significant degree of autonomy), and that we can recognize all these - that we humans can look-upon the natural world and validly perceive that there are such entities as species, organisms and traits, and we recognize what they are individually and specifically.

A pretty big set of necessary assumptions...

A pretty big class of problems which are routinely ignored.

*