Sorry, I wasn't sure where to post this. If it doesn't belong here, please move it to the appropriate forum! Anyway... I'm under the impression that most linguistcs believe our linguistic cognitive processes are made up of a set of parameters, which can be "switched" on and off. Examples might inclue the pro-drop parameter which would determine whether the subject of a sentence must be overly pronounced. However, imagine a scenario in which no languages required an explicitly stated subject. Would the pro-drop parameter really seem like a parameter at all, if no existing language demanded it? The concept of a pro-drop parameter would seem unncessary, as there are an infinite number of parameters that _could_ be expressed, but are not--but we must define the parameters expressed in _actual_ languages. Now suppose we find a remote tribe language which would be incapable of being described without the addition of a new parameter. And then another language, requiring a new parameter. Ad infinitum. Would linguists really say that all of these parameters are part of the human brain, but just not expressed in the majority of most languages? The alternative would be to say that no such parameter could be found, but that's an even uglier scenario, because it would leave us with the same problem i expressed earlier. We can never be certain that a parameter won't be "switched" the same way on every known language. Neither of these options seem appealing. I'm sure there's an answer, but I am unaware of it. How would you explain the discrepancy?