Occassionally an author will suggest that recovery of a specimen is representative of a new species, and this is generally true. The problem with this type of nomenclatural act is that when it is required on the part of the author to “downgrade” a known taxon to do it. One can find a variety of reasons to dispense with reasonable certainty as to the identity of the new material and how it can apply to the old. When the old gets a name, and the new one seems like it should be named, conventionally one refers the new material to the old.
Trickier still is when the original material is very much incomplete, scrappier, or otherwise limited in how much information can be garnered from it. Generally, this is a matter of perspective.
One of the most recent dinosaurian culprits of this is Suchomimus tenerensis [1]
Posted by Jaime A. Headden 