Took a class on the rise of complex civilizations once...it was really interesting stuff. I wish I could remember more from it right now.
Anyway, I think even people aren't a naturally civilization-building species. Perfectly modern humans lived in hunter-gatherer bands for 100,000 years at least before starting to settle down in the past 10,000 years. We took up civilization eventually, but not quickly. We'd have once fit the "no groups with more than 30 members" pretty well. So I don't think your species need any sort of unusual characteristics to get the no-civ tag, even with normal humans you could just say they haven't invented civilization yet. Still, certain characteristics could make civilization even less likely....my bet is that asociality is the best one. Humans have always lived in groups or bands, it was only a matter of extending them. But some species are strictly solitary. I don't think aggression alone would do it, because that tends to get channeled into dominance hierarchies.
This made me think of some other things too. Neolithic humans we know, and elves are pretty neolithic as is. But what about stone age dwarves....interesting! I'd never thought of it. Secondly, a number of the civilization related things we do aren't "natural" exactly, but are side effects of having a big and versatile brain. So talking and social interaction come naturally to us, while large societies and building do not. What if you had an intelligent species that was naturally solitary but had an instinctive knack for building (beavermen?) or something that naturally built cities and huge societies (antmen?).