Self-depiction errr… self-assertion (TLP 3.332)

Wittgenstein, in TLP 3.332, says something about sentences which sounds awkward, but is perfectly clear when applied to pictures. So the insight in the lemma may well depend on his picture theory of language.

3.332 No proposition can make a statement about itself, because a propositional sign cannot be contained in itself (that is the whole of the ‘theory of types’).

In regard to pictures it is common to think that pictures cannot depict themselves, and that this is due to the fact that depiction rests not on thoroughly regulated conventionality (elaborate syntax plus semantics, a vocabulary), but on something that is traditionally understood as similarity.

Which Naturalist Conception of Depiction

There is ample debate on whether similarity can be held to do the job (Goodman thinks it cannot), or whether the similarity involved must be understood as similarity between two things which must then be compared before we can say that a picture depicts something. So the most ingenious approaches have been suggested: Greg Currie argued that depiction mobilises natural powers of recognition; Ernst Gombrich that we see the surface of the picture as the thing it depicts; Richard Wollheim that depiction depends on a peculiar act of perception: seeing-in; Ken Walton that seeing something depicted engages us in a game of make-believedly seeing the thing depicted, and so on. But these subtleties need not defer us from the insight shared by all these theoreticians that the way pictures depict is profoundly unlike the way a sentence describes.

Pictures and Language

The critical difference between language and depiction seems to be this that looking at a picture we can perceive what it depicts whereas merely seeing a sentence does not yet generate any insights about its meaning—on top, we need knowledge of syntactic rules, vocabulary, and something to answer what Hegel would call “Arbeit des Begriffs”: allowing the words to have an impact on each other, and to produce a meaning for the sentence by the sheer combination of the words (plus the vocabulary meanings, and, the later Wittgenstein would add: their use).

Why Pictures Cannot Depict Themselves

Pictures cannot depict themselves because if they would they would simply produce another copy of the same picture, and this would not count as depicting themselves. Why language can describe itself is due to its conventional nature which allows for indexicals (which refer to the context of the utterance), negation (pictures cannot deny that something is the case), questioning (pictures cannot question whether something is the case), and so on. Pictures must mean through naturalist means of “similarity” (and see above for the qualifications).

In all, we are puzzled about the Wittgenstein passage because he is saying something about language which seems to conflict our most plausible intuitions.

References

— Gombrich, Ernst. 1963. “Meditations on a Hobby Horse; or, the Roots of Artistic Form.” In Meditations on a Hobby Horse and Other Essays on the Theory of Art, edited by Ernst Gombrich, 1–11. London: Phaidon Press.
— Goodman, Nelson. 1985. Languages of Art. Indianapolis: Hackett Publishing Company.
— Currie, Gregory. 1998. “The Aesthetics of Photography.” In Image and Mind, 72–74. Cambridge, New York: Cambridge University Press.
— van Gerwen, Rob. 1999. “Representaties waarnemen.” Feit & fictie IV:67–80.
— Walton, Kendall L. 1990. Mimesis as Make-Believe. On the Foundations of the Representational Arts. Cambridge, Massachussetts: Harvard University Press.
— Wollheim, Richard. 1980. “Seeing-As, Seeing-In, and Pictorial Representation.” In Art and its Objects. Second edition, 205–226. Cambridge, New York: Cambridge University Press.
— Wollheim, Richard. 1993. “Pictures and Language.” In The Mind and its Depths, 185–192. Cambridge (Mass.), London (England): Harvard University Press.