Part Two: The Metaphysics of Normative Facts
In part two of The Nature of Normativity, Wedgwood delves into the messy metaphysics of normativity, and takes some pretty remarkable positions, which include, that “the intentional is normative”, that normative facts are causally efficacious, and that though normative facts and mental states are irreducible, his view is consistent with naturalism.
The Metaphysical Project
Wedgwood begins the section with an elucidation of the sort of metaphysical account he is searching for – a constitutive account. Universally quantified biconditionals do not succeed for such an account because the result is reductive in a way that doesn’t provide important information about the property of which we are attempting to account for. What we are searching for in a constitutive account is the real definition of a property that tells us something about the essence of that property, in such a manner that is not reductive. Those philosophers who do accommodate “essence” into their ontology allow for it in modal terms, holding that the essence of an object is the properties that it cannot exist without. Wedgwood doubts this characterization, citing work by Kit Fine, as it leads to the conclusion that the experiential qualities of water are part of the essence of water, which indeed seems strange, as it entails that part of the essence of water is our experience of it. Instead, the essence of an object is the “basic metaphysical principle that states the nature of that object” (141), and this definition must directly or indirectly explain, “why certain propositions about the object are metaphysically necessary or metaphysically impossible” (141). As such, these metaphysical definitions of essence constitute fundamental modal truths that are used to explain all other modal truths, as the latter will be logical consequences of the former.
This characterization of the metaphysical project at hand allows for a clear explication of what a reductive account of normativity would amount to – it would provide necessary and sufficient conditions for a normative property in non-normative terms, and this would have to follow from the essence of the normative property itself. While meeting the latter part of this account will be difficult, we still must be argued out of reductivism, as it seems to follow from supervenience (though he argues against this assumption later), and Wedgwood notes that arguments for the irreducibility of the normative are often flawed, as they rest on an equivocation of property and concept in G.E. Moore’s open question argument. Put roughly, that it is an open question whether a normative concept is identical to a non-normative concept doesn’t logically entail that those concepts are not realized in the same natural property, with the result that it is possible that that normative concept is reducible to that natural property. What is needed is an argument that the nature of normative concepts requires that what they refer to is different from what non-normative concepts refer to. Wedgwood intends to make good on this argument shortly.
The Normativity of the Intentional
Wedgwood lays the foundations for this argument in the next chapter, wherein he argues for the normativity of the intentional, the thesis that an account of mental states that are about something (intentional states) cannot be offered without reference to normative properties and relations. It is conceivable that non-normative terms can be used, assuming they are sufficiently complex psychological terms, for this role, but Wedgwood’s point is that these non-normative terms would be referring to essentially normative properties (in line with the property/concept distinction noted above), such that if one wanted to argue that this amounted to a reduction of normative properties to non-normative properties, one would have to reduce normative and mental properties to non-normative and non-mental properties in one fell swoop, as the one cannot be understood without the other.
The intentional cannot be understood without normative properties and relations because intentional states have a content that is composed out of concepts, and a mental relation towards or about that content (such as belief, desire, etc.). In continuation with his support for conceptual role semantics, Wedgwood argues that the nature of each concept at play in intentional states is given by a principle that defines when beliefs involving that concept are correct, and basic principles of rationality that specify the use of that concept in thought and reasoning as rational or irrational. For example, with the concept “yellow”, the principle that defines the concept in such a way that beliefs that use it can be true or false is the semantic value of the concept, the property that “yellow” stands for, and the basic principles of rationality for reasoning might include the principle that it is rational to make the judgment that an object exemplifies this concept when there is the visual experience of the property this concept refers to.
Now we need an account of what it is for a thinker to possess a concept and have a mental attitude towards that concept, or in other words, what it is for an agent to have an intentional state. Wedgwood’s answer is that to possess a concept is to have a disposition to use the concept in question in ways that are rational, as determined by the rules of rationality for that concept. These principles would specify a body of information or set of mental conditions as making it rational to have a further mental state, in virtue of that body of information or prior mental state. A visual experience of an object as having the property yellow makes it rational to form a judgment or belief that that object instantiates the concept yellow.
It is a matter of debate as to the extent of what dispositions are central to the possession of a concept; for some philosophers, a holistic model seems necessary, in that all dispositions, rational and irrational alike, fix the concept. Wedgwood rejects this holistic notion convincingly, pointing out that if this holistic view were correct, then correct uses of concepts could involve irrational dispositions from that concept all the same. While we can concede that we use concepts correctly even when there is a dormant irrational disposition that has been over powered in that case, the holistic view contends that the proper use of that concept in that instance is due in part to that irrational disposition, entailing that an irrational disposition aids in rational thinking, which seems absurd. So Wedgwood is correct to conclude that it is only our rational dispositions that “are involved in determining the identity of the concepts that feature in our thoughts” (169).
At first glance it might seem possible that these rational dispositions could be specified in non-normative terms, but this would be mistaken, as rational dispositions must include the specification of defeasibility conditions that make essential reference to normative properties and relations. A rational disposition is a function from stimulus to response conditions, and it is certainly true that conditions can block the function’s output, as in if one has a reason to doubt their visual experience (perhaps they ingested a psychedelic), then the response condition is impeded. What we must include in the specification of rational dispositions is the proviso “in the absence of any special reason to believe one’s experiences to be unreliable” (170). But this makes explicit reference to a normative relation of being a reason to believe something, and that that our dispositions can be blocked by mitigating circumstances seems to be an integral part of our rational dispositions reveals that there is no way to specify rational dispositions without such normative relations as conditions for belief. But as these rational dispositions serve as the basis for our possession of concepts and mental attitudes, and these concepts and mental attitudes are the substance of intentional states, it follows that intentional states cannot be accounted for without normative relations.
The Irreducibility of the Normative
In the following chapter, Wedgwood explores some of the metaphysical implications of his thesis that the intentional is normative. He begins modestly, noting that intentional mental properties are not natural properties if they are not more fundamental than normative relations, such that the normative cannot be reduced to the mental, as normative relations and properties would be on both sides of the reductive biconditional, if one were to be offered. It seems possible that the intentional is reducible to the normative on this account, but Wedgwood rightly rejects this notion as what is “essential to these normative properties is precisely their relation to intentional mental properties and relations” (175). In other words, normative relations are true of mental states such as planning or desiring, such that any account of normative relations must make reference to intentional states. As a result, “the two domains of properties and relations are essentially interdependent, without either of the two being reducible to the other” (175).
Wedgwood posits that some philosophers mistakenly take this conclusion to imply that all naturalistic reductions of the intentional must necessarily fail. However, Wedgwood doubts that this logically follows from the normativity of the intentional, as on his view, a naturalistic reduction would succeed if it reduced the intentional and the normative simultaneously – it was admitted before that non-normative terms could be found to refer to normative properties, such that a reduction of the normative is possible in tandem with a reduction of the intentional.
While the irreducibility of the normative and intentional does not immediately follow from the normativity of the intentional, if the normative and intentional must be reduced simultaneously, and there is good reason to conclude that the intention is irreducible, then we have reason to conclude that the intentional and the normative are irreducible. Given the plausibility of Wedgwood’s conclusions that the essential features of intentional states are the normative relations that allow for rational dispositions, and that the essential features of normative properties are in their relations to mental properties, the most promising form of reductivism will be a form of functionalism. But as we just noted, if functionalism fails to offer an adequate reduction of mental states, then it will fail to offer a simultaneous account of mental and normative properties.
Wedgwood rejects reductive functionalism due to a powerful argument from George Bealer. The argument is, in bare form, to count as a reduction at all, functionalism must start with a theory that captures the essential features of the mental states being reduced, and that a sequence of physical properties satisfies that theory. To avoid being forced into the identity theory, and thus allowing for a central tenet of functionalism, multiple realizability, the theory must require that these physical properties are distinct from the mental properties the original theory is about. “But if the theory is true of these non-mental physical properties as well, that seems to conflict with the original claim that the theory captured what was distinctive and special about the mental properties in the first place. In this way, reductive functionalism seems doomed to undermine itself” (183). Thus, as functionalism fails to offer an adequate reduction of mental states, it fails to jointly reduce the mental and the normative. The normative, like the mental, is irreducible.
The Causal Efficacy of Normative Facts
A further implication of the normativity of the intentional is that normative facts are causally efficacious, as for Wedgwood, this is the only way to account for proper rational reasoning. The notion of a disposition is a causal notion, in that an entity’s responding to the stimulus in the proper way is caused by its being in the relevant stimulus condition. It is the normative fact that we are in a certain mental state that makes it rational to form a further mental state that causes the formation of that new mental state, such that it must be said that normative facts are causally efficacious. Wedgwood argues that this is the approach that must be taken, as we cannot separate out the normativity of a reason for forming a new mental state and the causal power of a prior mental state to a new mental state without undermining the credibility of rational reasoning, which, in order to consistently and reliably meet its target, requires that the normative reason for the new mental state is the cause of the new mental state.
There are of course some objections to be handled against the notion that normative facts rationalize further mental states in a way that causes those further mental states, including the points that certainly not all rationalizers cause further mental states, and that even when all the proper rationalizers are present, they do not cause a further mental state (such as when understanding a certain mathematical equation should rationalize the formation of another belief, but for some reason doesn’t cause it). These objections lead Wedgwood to refine his view by noting that rational reasoning is composed of basic steps, which are steps in reasoning with no step in between, such that Wedgwood posits that mental states rationalize the end belief even if they only cause the mental state that is a basic step in reasoning to that end belief. So this is meant to explain how we can have a mental state that rationalizes a distant belief even though that distant belief is not caused by that mental state, because the basic steps to that distant belief “make it rational for you to form that belief or intention by means of a long and complicated process of reasoning” (189). In many cases the rationalized belief is only a basic step a way, such that the normative fact both rationalizes and causes that belief, but it does seem then that Wedgwood’s theory has the resources to account for how a mental state far removed from the current mental state is rationalized by the current mental state.
Wedgwood then turns to Gilbert Harman’s objection to the causal efficacy of moral and normative facts, which is roughly that if normative properties were causally efficacious then they would offer explanatory power as to why one event occurred rather than another, and as it seems that our best explanations for such scenarios appeal to the non-normative facts of the situation, normative facts are dispensable for causal explanations, and therefore there is no reason to conclude that they do have causal efficacy. Wedgwood models his response to an argument for the causal efficacy of mental properties by Stephen Yablo. That argument is roughly that though mental state 1 is realized in physical state 1, mental state 1 is better placed to explain why mental state 2 occurred than is physical state 1, because, “the physical state P1 contains not only the mental state M1, but numerous other elements as well that are quite unnecessary to causing the new mental state M2. In this way, M1 is more proportional to the effect of M2 than P1” (194). Likewise, though a moral property is realized by physical properties, this moral property does a better job of explaining why persons reacted to that action in the way that they did, as, for example, when a person says something offensive, the shocked reactions of the spectators are better explained by the offensiveness of the act than the physical description of what the person said and its context.
Nonetheless, Wedgwood concedes that there are important differences between the case of mental properties and normative properties, the first being that the explanatory power of mental properties is powerful because it is possible to doubt that physical state 1 caused mental state 2, at best, physical state 1 caused physical state 2 which realized mental state 2, which allows for the room to argue that mental state 1 is the best explanation for mental state 2. This route is not open to the proponent of the causal efficacy of normative properties, because it is impossible to deny that the shocked reaction of the audience was indeed caused by what was said. Moreover, the physical description of what was said explains further reaching facts about the situation, such as why a “boo!” is heard on a tape recording of the event.
To avoid this problem, Wedgwood utilizes Yablo’s distinction between “world-driven” and “effect-driven” causes, where the effect-driven cause contains as little as possible that is not causally necessary for the effect and the world-drive effect contains more “about how the effect about in the actual world” (195). In essence the effect-driven cause focuses on the supervenening properties that could have been realized in other ways, such as normative properties and mental properties, while the world-driven cause focuses on the physical properties that actually realized the supervening property in the actual world. So, this distinction allows for the relevance of normative explanations, but the question remains as to what value is gained from this distinction, other than keeping the possibility of causally efficacious normative facts alive.
Wedgwood does make the observation that many philosophers seem to assume that to vindicate the causal efficacy of normative properties we would need to find scenarios when the normative facts provided a better explanation than the non-normative facts. Wedgwood thinks that this assumption is untenable, and all that is appropriately required is that the normative facts provide a correct explanation, not the best explanation. Moreover, Wedgwood suspects that many philosophers hold this position on the grounds that causally efficacious facts do not fit into a naturalistic world view, such that showing that they can fit into such a picture would, ideally, open up the possibility of causally efficacious normative facts. His quick answer for how this view is consistent with naturalism is that normative facts are true of the mental realm, that ordinary causal relations obtain between normative and mental facts, and that “these causal relations between normative and mental facts are themselves realized in physical causal relations between the physical realizations of those facts” (197).