On page 69-70 of her book Anscombe writes, It is clear that [in negation] one must convey what situation one is saying does not exist, and this will be conveyed precisely by the picture depicting that situation. No other picture could be involved: you could not for example make a picture of the situation's not existing. We must be careful not to confuse what is not the case with what is the case instead of it; if you tried to make a picture of a situation's not existing you would only make a picture of what did exist instead of it. The only exception to this is when we have the convention that not shewing something shews that the thing does not exist: as when a map shews that no large river passes through Birmingham by not shewing a river passing through Birmingham. Today at least one convention for a pictorial "not" is very commonplace: a circle with a diagonal line through it placed over the original picture. Of course, this may be (and often is) taken to mean "Do not ____", but there's nothing sacrosanct about this indicator being used as a command rather than descriptively. It seems to me that both W and Anscombe go through a lot of convolutions to avoid relying on anything like Frege's assertion sign. Pictures are different from props in that they depict only--and don't also say that what is depicted is the case the way asserting a prop does. But it doesn't seem problematic to have a convention according to which adding a mark to any picture would mean that it is intended to be "non-fiction." I get the sense that W may here be overly fond of avoiding reliance on "saying" in favor of some sort of intrinsic property/element that can "show" what would otherwise have to be said. Walto