‘Species’ are often treated as existing only at a single specific time slice. In fact, they are timely limited entities with a beginning, a time of persistence and an end, the consequences of which are discussed here. Here we try to understand how reliably we can precisely reconstruct the events that took place close to a species split. Therefore, we discuss different scenarios of a hypothetical population undergoing split and/or character evolution. Subsequently, we compare how a Hennigian ideal reconstruction would look like in comparison to a possible actual reconstruction. We also evaluate the character evolution at such splits. Last, we compare how the different reconstructions resolve appearance of new species and new characters through time. We summarize the major consequences of these observations on certain problems, notably the sister species vs. ancestor problem, “chronospecies”, fossil species, character evolution, and mapping on stratigraphic charts. We strongly argue for basing scientific investigations on proper philosophical backgrounds (epistemology), but epistemology still needs to be practically applicable. With this contribution, we aim at providing additional philosophical bases for certain aspects of evolutionary reconstructions, while still keeping the approaches practically applicable.