Moore’s Open Question Argument Moore argued that a definitional analysis of ‘good’ in naturalistic terms is impossible. He advances the open question argument in defense of this conclusion. The open question argument points out that for any naturalistic property P, one can intelligibly doubt whether P is good. But if it were a conceptual truth that goodness is P, that is, if P captured the meaning of ‘good,’ then according to Moore it would not be open to us to doubt whether P is good. That is, whether P is good would not be an open question. If it were necessarily true that goodness is P, then to doubt that identification would be to commit a logical error. Since, for any naturalistic property whatsoever, it is an open question whether that property is good, goodness cannot be defined in naturalistic terms. Thus, Moore concludes, goodness is not a naturalistic property. For instance, Moore considers the view that goodness is that which we desire to desire. Because it is open to us to intelligibly doubt whether what we desire to desire is good, it cannot be the case, according to Moore, that goodness just is that which we desire to desire. Because we can make sense of doubting whether that which we desire to desire is good, it cannot be that ‘good’ means that which we desire to desire. It cannot be that the terms ‘good’ and ‘that which we desire to desire’ pick out identical properties. The open question argument can be thought of as purporting to establish two distinct conclusions, one semantic and the other ontological. The semantic claim about ‘good’, which Moore takes himself to have proven true, is that the meaning of ‘good’ cannot be captured in naturalistic terms. The ontological thesis is that goodness is not identical with a naturalistic property. Moore obscures the distinction between these two theses, and takes both to follow from the open question argument. In doing so, he makes two mistakes. The first mistake is that of supposing that if there were an a priori link between ‘good’ and some naturalistic property (a link, that is, that we can learn of independently of our experiences) then competent speakers would be unable to sensibly deny it. The second is to assume that identifying one property with another requires that this identification be a priori. Contemporary advancements in the philosophy of language show why the open question argument fails to establish Moore’s ontological conclusion. Like the identification of water with H2O, goodness may be identical with a naturalistic property even though it may be open to us to doubt that identification. The open question argument presupposes that if two terms, ‘A’ and ‘B’, pick out the same property, then ‘A’ and ‘B’ must be synonymous. But we can see why this presupposition is mistaken when we consider, for instance, the identification of water with H2O. ‘Water’ and ‘H2O,’ pick out the same property, but the terms are not synonyms, and the identification of water with H2O is not knowable a priori. Likewise, goodness may be identical with a naturalistic property, despite the fact that that this identification may not be knowable a priori, and although the predicate that expresses that naturalistic property may not be synonymous with ‘good.’ If goodness is a naturalistic property P, but ‘P’ is not synonymous with ‘goodness’, then it will remain an open question whether P is good. This open question, however, will not threaten the ontological claim that goodness is P.
Pages to are hidden for
"moore"Please download to view full document