Davidson argues that there can be no law-like explanations of psychological events (beliefs, desire, and the like), since the concepts deployed by psychologists necessarily elude the sort of law-like generalizations possible in physical theories. Davidson thinks this is so because psychological concepts used to explain behavior are external impositions upon events; they are not natural features of the events in question. Davidson ﬁts this thesis into his larger theory about mentality – the theory of ’anomalous monism’ – which holds that psychological events are physical events and yet reside in an autonomous domain of explanation. The majority of the paper, however, is dedicated to defending the claim that psychological concepts are not amenable to law-like generalizations.
In the ﬁrst place, Davidson thinks it is impossible to provide necessary and suﬃcient conditions for persons acting on reasons in every case. If it were possible to predict behavior on the basis of beliefs and desires, two psychological concepts, we would need to be able to account for the totality of a person’s beliefs and desires (since they are all causally related to one another). But such an endeavor is impossible; so, explanations involving the psychological concepts of belief and desire shall always elude law like generalization.
Further, when we attempt to explain behavior using directly observable
events, such as punishment and reward, spoken commands, and deprivations, such as behaviorism uses (a once inﬂuential paradigm in psychology) the theory cannot predict behavior unless patterns of belief, desire, and emotion are inferred from the actions of the agent. So a purely external route to law-like generalization is barred.
Davidson then considers a theory which uses actions of agents to predict behavior, construing beliefs and desires as mere theoretical constructs in a theory of behavior. Such is Frank Ramsey’s theory of decision making, which computes the subjective probabilities and beliefs towards various propositions by weighting the decisions of persons when choosing between alternatives. The particulars of the theory are complex – and not particularly relevant. The point is that such a theory, despite being scientiﬁc in many respects, fails to predict behavior because it cannot account for the inﬂuence of a previous decision (between alternatives) on a subject’s current decision. So such a theory does not yield accurate predictions. If it cannot even yield accurate predictions, then it is certainly not law-like.
So it seems the necessary concepts of psychology – rationality, consistency, coherence, belief, desire, and all the rest – do not properly ﬁt into any theory of law-like generalizations about human behavior. Yet, such concepts are indispensible to psychology. So it seems that psychology can never predict behavior like physics can, despite the fact that psychological events necessarily are physical events. When we resort to psychological explanation of behavior, we use a unique set of concepts and operate in an autonomous realm of explanation – one separate and distinct from the one used to explain the physical events which undergird all behavior.
Primary Argument: Anomalous Monism and the independence of psychological explanation. (This is an outline of Davidson’s famous Anomolous Monism, though this article merely argues for premise 4. But it is helpful to see where the article’s argument fits into Davidson’s wider scheme).
1. Psychological events cause, and are caused by, physical events.
2. Events which enter into cause/eﬀect relations with one another can ﬁt into a deterministic set of laws when appropriately described.
3. Therefore, psychological and physical events can ﬁt into a deterministic
set of laws with one another, when appropriately described. [1,2]
4. There are no psychophysical laws.
S1. In order to explain psychological events, we necessarily rely upon
concepts like rationality, coherence, and consistency.
S2. For there to be a law-like account of psychological events, we
would need to dispense with such concepts, since they are external impositions upon the events.
S3. We cannot dispense with such concepts when we describe psychological events.
S4. Therefore, there are no psychophysical laws.
5. Therefore, psychological events necessarily ﬁt into a set of physical, deterministic laws (i.e., they are physical events). [3,4]
6. Psychological events, when described in the vocabulary of psychology,
cannot be reduced to physical explanations.
S1. For psychological events described in the vocabulary of psychol-
ogy to be reduced to physical explanations, there would have to be physical equivalents to terms such as ’rationality.’
S2. There are no physical equivalents of such terms.
S3. Therefore, psychological events described in psychological terms
cannot be reduced to physical explanations. [S1,S2]
7. So, psychological events are physical events which, when described in the vocabulary of psychology, cannot be reduced to physical explanations. [5,6]
(Or, psychological events reside in an autonomous level of explanation when described in psychological terms) [6,7]