Yes, that was the point. Words are labels, non-descriptive signs that hold no information regarding the signified. They only have meanings within particular communities. However, communities sometimes collectively label things inconsistently, mixing up the similar and the different. This causes discrepancies between the strucure of language and the structure of the (percieved) world. On the other hand, sometimes individuals and smaller communities make up their own labels, and have no problems as long as they apply them consistently. Babies and academics are especially prone to that sort of linguistic activity. If we agree to call a Cutoff knob a "Voodoo knob" instead, this will impair our communication with the outsiders, but not our operation of synthesizers.
I take all that as evidence that you can make "correct" propositions using language "incorrectly", and vice versa, i.e. correct use of words is insubstantial regarding the problem of knowledge.
As a kind of related side note, if we start dealing with descriptive signs, e.g. "the Morning Star" and "the Evening Star", things only get worse. Then there are metaphors, such as "a squirrel is a rat in a fancy outfit", and this is hell
Agreed. One needs a Theory of Truth first, and there are many, and a Theory of Justification, of which there also many, and, if we look at Gettier's paper in particular, a Theory of Entailment, and there are literally infinitely many of those.
Kitties have no knowledge. Only behaviour patterns learned through positive and negative reinforcement. And toxoplasma