Not circular at all. There are all sorts of techniques that have been used throughout the ages to separate creatures into groups, from physical similaries to DNA measurements.   So it is quite possible to identify groupings. But you asked for a test to differentiate man from other animals which carries the implication - perhaps unintended - that man is in some way more distinct from the rest of animals than a tiger is from the rest of animals, and that's what I doubted. If you were not making such a claim, I apologise. I'd be grateful if you can confirm that one way or the other.