I've just been having a look at some of the Wikipedia pages about certain bony fish groups, particularly Sarcopterygii and Rhipidistia. These need some serious fixing. There's a lot of stuff out there about 'ancestral groups'. That is to say, describing a group as the ancestor of another group.
"Tetrapods — four legged[sic] vertebrates were the terapodomorphs'[sic] descendants."
"Rhipidistia is now understood to be an ancestr for the whole of Tetrapoda."
The notion that a group is 'ancestral' is a bit misleading, especially if we accept that groups (i.e. clades) are actually the descendents of a single common ancestor. That's not really the problem, though.
Let's look at each of the quotations below the fold.
The first one claims that tetrapods descended from tetrapodomorphs. The actual meaning of 'Tetrapodomorpha' is the tetrapod total group (Ahlberg, 1991). By definition, this group includes all tetrapods, and any and all fossil taxa that are more closely related to tetrapods than to any other extant group (in this case, probably lungfishes: the Dipnomorpha). Tetrapods are a subset of Tetrapodomorpha, not descendents of them.
The second quotation is similar. Ahlberg (1991) also re-defined 'Rhipidistia' cladistically as the Dipnomorpha + Tetrapodomorpha. In this sense, Rhipidistia is monophyletic. However, the Rhipidistia was a pre-cladistic grouping meant to encompass porolepiforms and 'osteolepiforms'. Porolepiforms (probably a real clade) and the 'osteolepiforms' (not a real clade) represent an assemblage of lobe-finned fishes that would look quite similar to an 'untrained observer'. This similarity is mostly just shared primitiveness. That is, it does not unite them to the exclusion of other taxa (namely lungfishes in the case of porolepiforms; and tetrapods in teh case of 'osteolepiforms').
Figure 1. Some 'rhipidistian fishes'. Top: Holoptychius. Bottom: Eusthenopteron along with an illustration of its pelvic and pectoral fin endoskeletons.
What is significant about 'rhipidistians' in the classical sense (i.e. before Ahlberg's 1991 paper) is that they lack synapomorphies or homologies. They have to be defined on the basis of a set of characters and character absences, implicit and explicit, that is hand picked to exclude other groups. They are 'fishes', meaning they have fins (not digits) with lepidotricia, bony dermal rays. But these are also found in the early tetrapods Ichthyostega and Acanthostega. However, these latter taxa lack an intracranial joint, a division of the front part of the braincase from the back part that contains the ear capsules. Coelacanths also have this division, but they are not rhipidistians. However, coelacanths lack the dermal skull bone characters, such as a maxilla, that are found in 'rhipdidistians' such as Eustheonopteron and Holoptychius showing in Figure 1., above.
As you can see, it quickly becomes obvious how the defining characters are arbitrary, in some sense. They are picked to justify a group of taxa that share some overall similarity. It does not reflect an attempt to discover the hierarchical relationships among characters. This latter process is the discovery of homology: the characters that unite monophyletic groups. It is in this way that real evolutionary relationships are discovered; not through the nomination of "ancestral groups".
Ahlberg, P.E. 1991. A re-examination of sarcopterygian interrelationships, with special reference to the Porolepiformes. Zoological Journal of the Linnean Society 103: 241-287.