Ah ha, so the degree of abstraction may serve as a bound for thought.
Just to paraphrase.
And one avenue for extrapolation: so, if true this would argue for lower animals than humans do not exhibit the capacity for abstract thought. Those who feel language is not necessary for this thought, then, may argue that lower animals do have such capacity.
And I also think that our current understanding of encoding in brains (wholly incomplete) allows for a necessary definition of thought as to only include conscious processes. Otherwise any neural activity suggesting, say, fear might be construed as an animal being scared, which I'm operationally separating here in the following way: fear is a testable, induceable state that includes, among other things, amygdalic stimulation. Being scared is the state of recognizing and acknowledging fear.
Do animals exhibit consciousness? Can they be scared or are their reactions of fear simply that -- reactions?
The beautiful implications of this is that particular neural encoding might be sufficient to define "thought." Which would make it a viable candidate to search for in neural activity of lower animals.