Berkeley and Extension
In Berkeley’s “Three Dialogues between Hylas and Philonous,” Philonous, speaking for Berkeley, argues that extension and all other properties of objects are mind-dependent. Hylas, his counterpart, takes up the position that objects have two types of qualities: primary qualities, which are mind-independent (i.e. they are inherent to the object, and the object would have them if even if no one could perceive them), and secondary qualities, which are mind-dependent (i.e. they would not exist without the mind). He says that extension (i.e. taking up space in the physical world) is a primary quality. This conception is similar to Locke’s resemblance thesis, which states that our ideas of primary qualities resemble properties with mind-independent existence, but our ideas of secondary qualities do not resemble anything with mind-independent existence. The two theories are not identical, however, because Hylas does not explain Locke’s definition of an idea: an idea is either any physical sensation or any mental reflection upon that sensation. Locke agrees with Berkeley that all ideas are mind-dependent, but they differ about universals (i.e. ideas which many people can share). Locke believes in them, and Berkeley does not. This is an important distinction which Berkeley ignores in the p423-424 argument.
To explain that extension is perception-dependent, Philonous introduces a thought experiment centered on a mite (i.e. a very, very small animal with sensory organs). For this animal to survive, it must be able to perceive objects in proportion to the size of its own body. Because it is much smaller than a human, an object like a marble, which is tiny to us, would be gargantuan to the mite. Furthermore, because the mite’s eyes are made of different material than a human’s, the world would also look different in other ways (color, texture, etc.). To Berkeley, an idea is nothing but a sensory representation of a particular object from some particular point of view with respect to that object. Thus, the human’s and mite’s ideas are similar because they are both sensory perceptions, but they differ because they come from different points of view.
Philonous then argues that Hylas’s opinions entail a contradiction. He reminds Hylas that it is possible for water to feel warm to one hand and cold to the other, and an object cannot have contradictory properties, so the difference in heat must exist in the mind alone. For this reason, “secondary” qualities are mind-dependent. In the same way, an object can appear to have “little, smooth, and round” properties to one perceiver and contradictory “large, uneven, and angular” properties to another. A human could even have both these ideas if he examines the object with a bare left eye and a microscope on his right eye. Furthermore, an object appears small from far away and large from close up. By the definition of an “inherent” property (see “primary” quality explanation above), one cannot change the property without changing the object itself. Because the perceivers’ evaluations of extension are contradictory, and because an object cannot have contradictory properties, the difference in extension must exist in the mind alone. Therefore, the mind is the measure of extension, and this property is mind-dependent. Each perceiver has a different point of view; each person’s idea of an object’s extension is unique, and there is no “correct” or “real” extension. The difference between primary and secondary qualities was in mind-dependence, and since both are mind-dependent, they are identical.
Berkeley’s argument refutes Hylas, but it does not refute Locke. Since Locke believes in universals, he can say that general ideas which all people have, not specific ideas, define an object. Perceivers may all have different opinions about the specific, determinate nature of the extension of a marble (how much space it takes up), but all people can understand that it has some determinable extension (that it takes up space). Berkeley has proven that there are contradictions in determinate extension, but the determinable is what is actually inherent to an object. (In response, Berkeley, in p425-426, denies the existence of determinables on the grounds that it is impossible to abstract a determinable quality out of an object. Whenever we try to do so, we end up thinking about determinates, as well; since these two “different” traits always coexist, they actually are one and the same.)