I argued that this is one of the sources of the difficulty of the discussion about secondary senses: the unclarity about how a certain expression is supposed to be used is one of the things that make it difficult to give examples. I suggested that when giving examples, we also need to clarify—to ourselves first and foremost—the use of the expressions we examine or give as examples; it is not enough to just give the expressions.
I would like to connect this to the discussion about metaphors. Giving examples in a discussion about metaphors is also difficult, and for similar reasons: When we give examples of metaphorical expressions, it is not always sufficiently clear how these expressions are meant to be used. If I am right, then everything depends on that—everything depends on clarity about use.
As in the case of secondary uses of terms, it may be the case that an expression that was meant to be metaphorical is not really metaphorical at all: for it might have a literal use. That might be the case, even if it is hard to think of a literal use for the expression.
Take ‘Juliet is the sun.’ More than any other example, perhaps, there is consensus that this is a metaphor. Nevertheless, even this expression can have a literal use—for instance, when naming celestial objects in an attempt to memorize them. So Mars is Harold, and Neptune is Jack, the Moon is Suzy, and Juliet is the sun. In any case, my claim is, again, that there really is no substitute for clarifying the use of the expression we examine or wish to use as examples.