In the Cartesian argument (for Skepticism) from dreaming we are to assume that if we were dreaming, then were we to see hands in that dream, those would not be hands; but of course, they would be dream-hands in a dream-world, and so why should dream-reference to them fail? If we think of someone dreaming about hands, then clearly those are not real hands; but, were this a dream (not a dream-within-a-dream, which is what our "dreaming" would refer to), then what is meant by "real hands" within that dream would be dream-hands. You might wonder if that would be the case, had we fallen asleep having already learnt the meaning of "real hands" in the real world; but presumably we learnt the meaning of "real hands" in this world, and were this a dream then that would be a dream-world. Might we have learnt the meanings of our words in some higher realm? But, as soon as we clarify such worries, say in some Moorean way, by describing what is meant by "external thing," we tie the meanings of our words to this world: worrying about that problem resolves that problem!
In 2003 I was published in the British Journal for the Philosophy of Science, but I've DONE little since then (I am currently writing a book). Blogging since 2007, originally as enigMan (a "Meaning"-full name), my main involvement was via the Philosophers' Carnival because I started a PhD in Philosophy in 2007. (The preliminary work for my book having got boring, in 2014 I started taking photos of my village, sharing them with similar amateurs and others on google+ and now on MeWe (as google+ is no more) and also short videos on YouTube. For what I think about the Antichrist Zuckerberg, see my posts :