I propose that the best way to understand the meaning of smiling is that the signaller who smiles stands for itself and, further, that smiling is a self-expression about the signaller. Green (2007) understands face expressions as self-expressions without describing the structure which allows a face expression to be a self-expression. I employ Millikan’s notion of selfsign, but only her basic idea. The core of the concept of a selfsign is that a thing stands for itself as a proper part of a complex sign family. Here smiling is the sign family. I suggest smiling is a sign where as part of the sign the signaller stands for itself as a selfsign. My analysis can be extended to facial expressions and certain gestures including several forms of non-human primate expressions. What do we gain by taking smiling to be a case of self-signing? There was a lot of debate about what is the cognitive significance of essential indexicals and their opacity or unsustainability (Millikan 1990, Cappelen and Dever 2013, Jaszczolt 2013, Magidor 2015). What could sustain cognitive significance? I propose that the analysis of selfsigns explains why smiling has proto-cognitive significance. My idea is that a smile conveys something about the signaller without the possibility one error: whom the property expressed concerns. This enables us to flesh out a core characterisation of cognitive significance and an understanding of what enables it. The self-signing structure is one way to achieve cognitive significance in the form of proto-cognitive significance.