Well, this is an open question in evolutionary biology :) One important thing to keep in mind is that species are hypotheses that can be tested, and the interp

Well, this is an open question in evolutionary biology :) One important thing to keep in mind is that species are hypotheses that can be tested, and the interpretations can vary from person to person and field to field. For example, many molecular geneticists will often discover new cryptic species on the basis of their DNA sequences, that a morphologist might not consider to be a true species. For fossils, of course, there almost always only morphological evidence for species. The relative roles of anagenesis (change within a species) and cladogenesis (splitting event that creates two new species) in evolution is still being investigated, especially by paleontologists. One of many problems is that the fossil record is extremely incomplete, and so it's difficult to determine whether anagenesis or cladogenesis dominates. For me (as a non-paleontologist) my simplifying assumption about evolution is that anagenetic change doesn't occur, but again this is still an area of active research.