In my view, both BJT's and FET's are voltage controlled current sources more than anything else, but they have major differences in their voltage-to-current transfer characteristics. The drain current of an FET depends quite linearly on its gate-source voltage. Likewise, the collector current of a BJT depends on its base-emitter voltage, but this relationship is logarithmic. Both relationships are equally real and both should be treated as cause-effect relationships.
For a BJT, the key to understanding its operation is this: When one varies the base-emitter voltage, the charge in the transistor's base region changes proportionally. The amount of movable charge carriers in the base region is increased if you increase the forward voltage bias on the base-emitter junction. If, on the other hand, you decrease the forward bias voltage on the BE junction, the number of movable charge carriers in the base region becomes less. The more movable charge carriers in the base region, the more charges can flow between the collector and emitter. This flow of charges is what we call the collector current.
I can't explain this stuff very clearly. In case any of you wants to learn more about this viewpoint -- that BJT's are controlled by the charge in the base region, which, in turn, is controlled by the base-emitter voltage -- the best thing you can do is try and look for better explanations.
Some may find this useful:
SCIENCE HOBBYIST: how transistor works, an alternate viewpoint