Frank Jackson is a well-known philosopher from Australia. Perception, first published in 1977, is an argument for a Representative theory of visual perception similar to John Locke’s. Jackson sums up the book in the last paragraph:

The first four chapters present my case for a Sense-datum theory of perception. Chapter 5 gives the reason for holding that all sense-data are mental. This forces a choice between Idealism and Representationalism. In chapter 6, I argue that there is no good reason for not choosing Representationalism. And, finally, in [the last] chapter, I have, first, attempted to justify my taking the perception of things as basic throughout; and, secondly, I have tried to make more precise … the particular kind of Representationalism that should be chosen.

The Sense-Data theory of perception has had a long history in philosophy. Its principal tenet is that we never perceive physical objects directly. Instead, what we immediately perceive are mental objects called “sense-data”. Thus, when I see the wall in front of me, I immediately perceive a mind-dependent blue expanse, a sense-datum (or set of sense-data), not the wall itself. In Jackson’s words, a visual sense-datum is “something seen, but not in virtue of [seeing] anything else”. Physical objects, on the other hand, are seen, but always by virtue of seeing something else, namely, sense-data. Furthermore, there is a kind of causal relationship between physical objects and the sense-data that “belong to” those objects. That’s why sense-data are said to “represent” the external world.

I’ve read that Jackson no longer endorses the theory he presents here, but Perception still provides a good account of one version of the Sense-Data theory. Professor Jackson makes many interesting distinctions and responds to a number of criticisms. One criticism he anticipates is that he is spending too much time analyzing the language of perception (Chapter 2, for example, is called “Three Uses of ‘Looks'”). He responds that his analysis of language isn’t meant to show whether sense-data actually exist. Instead, it’s meant to help us decide whether believing in the existence of sense-data (accepting the theory) is a reasonable thing to do and how the theory should be stated.

I read Perception because I’ve been thinking about the topic a lot lately and think the Sense-Data theory is probably the best one (even though it’s probably considered a bit old-fashioned now; for one thing, philosophers talk about “qualia” these days instead of sense-data). It’s common for philosophers to think of the mind as software running on the hardware of the brain. But if we’re going to think of our minds/brains as computers running programs, it makes sense to think of the input to these computers as data. Just as a computer program processes data and not the real stuff in the world, we process data as well. In our case, colors, sounds, smells, and so on, are different kinds of data. It just so happens that the data we process isn’t made up of numbers or ASCII characters (or electronic on/off settings); our perceptual data is made up of red or blue expanses, soft or loud noises and pleasant or unpleasant aromas. We don’t perceive the world “directly”, because that can’t be done. We perceive sense-data that represent the world; in similar fashion, computers process electronic values that represent the world too.