‘The logic of the concept of colour is just much more complicated’

III.96–106 were most probably composed after Wittgenstein moved from London to Cambridge in early April. Whether set down in a single sitting or several, they were certainly written before 11 April, the next date in the manuscript and published text. As in the remarks already discussed, Wittgenstein examines or re-examines a variety of colour concepts. He does not tackle the problem of explaining why there is no transparent white or reddish green, which Runge mentions in his letter to Goethe (cited in III.94) but continues instead to ponder how we think and talk about colour. Perhaps thinking the impossibility of transparent white is satisfactorily explained in Part II, the question I am taking to have drawn him back to the topic of colour, he now casts his net still wider and takes up a diverse set of colour-related problems, all pretty much from scratch. He proceeds, as before, in an exploratory and tentative fashion and from time to time abruptly shifts directions, occasionally with advance notice but mostly without warning. The object of the exercise is, it is fair to say, to clarify the logic of colour concepts, to alert philosophers to pitfalls they need to avoid and to isolate philosophical questions about the nature of colour deserving attention, the possibility of certain colours and the impossibility of others above all.

After noting at III.95 that specifying the colour of a particular area of a table may be more difficult than specifying the colour of objects in a room, Wittgenstein remarks on the difference between appearance and reality, a topic of much philosophical discussion. At III.96 he notes that what ‘seem so’ to a person (even to everyone) is no guarantee that ‘it is so’. And twice so, he would say, for colour. The fact that a table seems brown to everyone does not ‘therefore’ mean it is brown. One that appears yellow in sunlight may, he notes surely correctly, appear brown in shade and actually be neither. There remains a residual question, however. What does it mean to say a table seen by everyone as a particular colour is ‘not really’ that colour? This does not seem right. A table seen by all as brown may, if not always, sometimes, one wants to say, be properly regarded as brown.