By Carolyn Ursabia
In Sense-data, a chapter in his book Some Main Problems of Philosophy, George Edward Moore builds a case for a theory of perception called sense-data theory. There are many modes of gaining knowledge about the world, including but not limited to perception, memory, testimony of others and inference. Perception seems more fundamental than all the other modes of inquiry; the other modes of inquiry rely on it. A theory of perception is a metaphysical account of the nature of perception; that is, it is a proposed answer to the question of what is happening when we perceive something with our senses. Sense-data theory is one such theory of perception. I will argue that Moore’s case for sense-data theory is weak on the grounds that it assumes too much.
According to Moore, what happens when we all saw the same envelope is the following:
o The subject directly apprehends sense-data.
o The subject indirectly apprehends the object.
To illustrate what he means by this, he uses the sense of sight and viewing an envelope as his example. He assumes that the result will be transferrable to all of the other senses. In his example, the object is the envelope. Sense-data are the colour, size, and shape of the envelope. Recall that we are only considering sight. The subject is the person perceiving the object. Moore explains the first part—that the subject directly apprehends the sense-data—through a distinction he makes between sense-data and apprehending sense-data. The former are the sensible properties of the object, where the latter is what goes on inside the subject. He gives two reasons for believing this distinction. The first is that whereas the experience of seeing the whitish colour of the envelope ceases—namely, apprehending the sense-datum—when we look away from it, the whitish colour—namely, the sense-data—conceivably continues to exist when we look away. The second reason he believes that the two are distinct is because he believes the whitish colour of the envelope—the sense-datum—is really somewhere on the surface of the envelope, whereas the experience of seeing the whitish colour—apprehending the sense-datum—appears to happen somewhere within the subject’s body. Moore calls the experience the subject has of seeing the whitish colour the direct apprehension of sense-data.
To explain the second part—that the subject indirectly apprehends the object—Moore first establishes that the object cannot be identical with the sense-data. He does this with an appeal to what we will call perspectival variation; that is, how sense-data for the object will vary for every subject because each perceives the object from a different perspective. For example, each subject will view a different colour, shape and size of the envelope, based on their location relative to it. Some will see lighter or darker shades of white, and various quadrilateral shapes and sizes for the envelope. Our perspective is dependent on our location relative to the object. Moore argues that since each subject observes different sense-data for the same object, and the sense-data from multiple perspectives cannot all be identical with the object, then the object cannot be identical with sense-data; the two are distinct. When we see the envelope, we directly apprehend a shade of whitish colour. If we were to look at it again from a different perspective, we would perceive it as having a different shade of white, but we know that it is still the same object. The object does not change: it has a true colour, shape and size that does not vary. It exists independently of our minds, and when we apprehend sense-data, even when our perspective changes, we know that it is still the same object. In summary, Moore argued that what happens when we perceive something with our senses is that the subject directly apprehends the sense-data; sense-data are distinct from the object; the object exists independently of our minds; and that we come to know the object indirectly through its sense-data.
The main reason that I find Moore’s argument to be weak is because he does not provide a satisfactory description of the relationship between the object and sense-data. He argues once again from perspectival variation that we can never know the real size, shape or colour of an object. If all the colours as viewed from different perspectives are part of the real object, then he says that all of the colours would occupy the same surface, and this “is difficult to suppose.” He concludes, therefore, that the colour, shape and size of an object is never “a part” of the actual object. He then argues that it is possible that the sense-data that we directly apprehend could be qualitatively identical with the features of the real object, but we cannot know if it is numerically identical.
“This seems to be the state of things with regard to these sense-data—the colour, the size and the shape. They seem, in a sense, to have had very little to do with the real envelope, if there was a real envelope. It seems very probable that none of the colours seen was really part of the envelope: and that none of the sizes and shapes seen were the size or the shape of the real envelope.”
To show the relationship between the object and the sense-data, he instead appeals to the space that the object holds in time. What each subject sees is “a part” of the real space that the object occupies.
“…even if the colour presented by your senses is not a part of the real envelope, and even if the shape and size presented by your senses are not the shape and size of the real envelope, yet at least there is presented by your senses a part of the space occupied by the real envelope. And against this supposition I confess I cannot find any argument, which seems to me very strong.”
Even if this is so, this still fails to account for the relationship between sense-data and the object. He holds that all of the colours, sizes and shapes could not be identical with the object; that we cannot know the real colour, size and shape of the object; but that we can directly apprehend the location in space and time of the parts of the object. However, knowing the location in space and time of the object does not explain anything about the other sense-data. We still do not know where they are located, and how they interact with the object. He seems to end on the note that it is self-evident; a truism about the nature of objects and their sense-data. Sense-data are distinct from the object; the former is not a part of the latter. The subject directly apprehends sense-data, which could happen to be qualitatively identical with the “real” properties of the object, but we cannot know if they are numerically identical with them. We do not where the sense-data reside relative to the object. What then is the purpose in the object having a real colour, shape and size, if we can never know it since all we can be acquainted with are sense-data?
Without a satisfactory account of the relationship between sense-data and the object, even though he has established that they are distinct, it is not clear that the object is a necessary part of perception. Moore himself says that, “We must know, when we directly apprehend certain sense-data, that there exists also something other than these sense-data, that there exists also something other than these sense-data—something which we do not directly apprehend. And there seems no sort of reason why we should not know at least this, once we have dismissed the prejudice that we cannot know of the existence of anything except what we directly apprehend.” But must we? It is necessarily true? No, I say it is not. To make his case for direct apprehension of sense-data as something that distinct from indirect apprehension of the object, Moore explained away any need for the object. It ceased to have reasonable relationship with what we directly apprehend. If we are able to perceive all of the features of the object directly through sense-data, there no longer is a need for the object. We should instead be looking for responses to his argument from perspectival variation—namely, that a location in space and time cannot have all of the perspectives, else find another theory.
Crane, Tim and French, Craig, "The Problem of Perception", The Stanford Encyclopedia of Philosophy (Spring 2016 Edition), Edward N. Zalta (ed.), URL =
Huemer, Michael, "Sense-Data", The Stanford Encyclopedia of Philosophy (Spring 2011 Edition), Edward N. Zalta (ed.), URL =
Moore, G. E. 1953. Some main problems of philosophy. London: Allen & Unwin.