I think there may be other opinions on the matter, but I dislike a because of the line "an abstract model." While the passage is about one particular abstract model, a seems to generalize the authors suspicion of this one model to all abstract models in general.