In my last post, I pointed out that there were some things that are unknowable to science.
But since science is the best way to know anything (that can't be directly experienced), it might be useful for me to talk about why this is the case.
In cladistics, phylogenetic diagrams are displayed with taxa along one side in row, with their connecting branches (clado == branch) with their relationships revealed deeper in the diagram:
(A, B, C, D, and E can all be species or more inclusive groups.)
You'll notice, however, that no cladogram ever shows B evolving directly from A:
This diagram is completely invalid if A is supposed to be a specific organism (as opposed to a group not represented by a single organism). There is no way to show A is the direct ancestor of B or C for fossils.
If B and C appear both slightly derived from A, this really gives no insight as to which of the three is the true ancestor (assuming, of course, the true ancestor is present--it may not even be under consideration. All three may have had an ancestor which is unknown in the fossil record known at the time this cladogram was "generated").
Features are derived and lost all the time in organisms--recursions happen frequently in evolution, and often, even which organism is oldest offers little insight into the matter, since the entire extent of any given population is not something that is preserved in the fossil record (from beginning to end--or good enough to qualify for knowing the temporal occupation of a given species).
More to the point, a cladogram is a hypothesis--a testable statement of organisms' relationships. Producing a cladogram that suggests a true ancestor has no value because it is unfalsifiable. (Maybe it's true, but how can you evaluate if it is or not?)
The problem can be illustrated pretty well (I think) by this analogy:
Say I have two boxes. And I have two different pieces of paper which go into each into a box. So I cart the lot to the darkroom and--blindfolded, yet--put one piece of paper into a box, and the other into the other.
Outside the darkroom, I'm confronted with not knowing which piece of paper went into which box first.
And there really is no way to determine which box got its paper earlier thant the other. The event happened. And it happened without an observer being present. And there is just no way to know which got what earlier than the other. And that is what unknowable means.
Now, is this conundrum of who speciates what really a problem? Not really. Not being able to talk about ancestry at this level of detail isn't a problem, because inductively it is known that parents have children, and children also become parents, and this is true for populations like species as well.