We had a lot of kids round on Halloween for some mass trick-or-treating. Even when we split them up, the groups were too big. As we were strolling between houses, JessieL asked me 'what's this linguisty jobbie Apoa's been telling me about?'
She was referring to the utterance 'Everybody doesn't like chocolate'. My students find it hard to believe that for some speakers it can mean 'not everybody likes chocolate' as well as 'nobody likes chocolate' (I've got a feeling I agreed with them once - maybe I've been contaminated by linguists). I promised I'd ask a few more folk.
At breakfast, we found 3 out of 4 people accepted both senses. Apoa was the odd one out and just found it weird. JessieL was incredibly articulate as she explained her opinion to Apoa:
'Technically, like if you're talking proper and everything, it just means that every single person doesn't like chocolate. But I can see that some people might say it sometimes and mean that not everybody likes it.'
It's a good illustration of the problems of investigating competence when all of your evidence has to be based on performance data (at least that's what I tell the students).
B-)
Nice example of how it can be difficult to 'get' both senses of a scope ambiguity. I find the non-natural reading of the example you give quite clumsy, but it still works okay.
Generally, in scope ambiguities, the most natural reading is where the quantifier scope is in the same order as the quantifiers in the sentence, with left-most quantifiers having the wider scope.
Thanks, Frankie. I think my intuitions are like yours but, as ever, context can change things. It sounds a lot better, I think, if you can interpret the speaker as focusing on the wrongness of 'everybody' and attrobuting that assumption to someone else.