Interesting question, thanks! A word used in a deviant way would only have meaning if those who listen or read understand it. For example, I sometimes get confused in casual conversation and come out with a spoonerism -- a mixed up word. Usually, though, my wife understands what I mean anyway, by interpolating from a shared context. A new word, or a new meaning of a word, might gradually come to be accepted usage more widely. Slang words, in particular, tend to get picked up rapidly in this way. Let us say that the 'correct' meaning and usage of a word is determined by the dictionaries. But a modern dictionary is itself only a reflection of broad usage by speakers and writers. We have to go back quite a way in time to find a dictionary that sets out to adjudicate meanings, rather than simply record or describe them. So, the 'official' meaning of word comes about because of common usage. If enough people use a slang word, it ends up in the dictionary. Dictionaries tend to be a conservative force in...

I really dislike doing this, but variations on your questions have been asked before, and some good answers put up. Please see: http://www.askphilosophers.org/question/3587 Other questions and answers pertain to your broader question, which is not about works of fiction, but about 'speakers and writers' more generally. But I'll add, re your nicely absurd J. K. Rowling interpretation, that even this would have SOME bearing on how we read the Harry Potter books. We might suspect that Rowling was clinically insane, and scour the books for further evidence; we might suspect that she was a prankster, and again scour the books looking for meta-fictional jokes; we might suspect she was writing a time-travelling cyborg novel and looking to promote it, and again we might then return to Harry Potter using it to help us imagine what the new novel would be like. The point is, the author clearly has some connection to the book(s). Either, then, we ENTIRELY discount the author when...

Like any real experiment, a thoughtexperiment (or analogy, case study or example) in order to be validevidence for some position, has to be conceived of as beingrepeatable. So, my thought experiment should be compelling onits own terms, and not because of some special context that makes itcompelling. Only then will the thought experiment (or whatever) havevalidity beyond that context. 'All things being equal' is thus akinto the notion of controlling variables.

An extremely sophisticated question, and one I'm not sure I could address in its entirety. However, I will start and perhaps one of my colleagues can take things much further than I can. I take it that the core of your question is this: why do we assume that problems are 'one' while answers are 'many'? Is this in fact the case? Although as a matter of fact, there will be hundreds (maybe thousands) of papers addressing a particular problem, is there not also a working or regulative assumption that all of these are contributions to an answer (that is, a single answer)? If it turns out that one answer cannot be arrived at then so be it, but I do not believe that is the starting assumption. Also, if there is an irreducible variety of answers then this suggests that the problem itself has been misstated, or perhaps there is more than one problem. Which leads nicely to the second part of my answer. Is it the case that philosophers tend to assume that a problem is the same problem, intersubjectively shared...