§2. An Assessment of Anomalous Monism:
Anomalous monism can be characterised by way of a number of theses; identity, non-reduction, the nomological connection of events under their physical descriptions only, the anomalousness of the mental, and supervenience. It is the last two that are of primary concern in this paper; however, they cannot be understood in isolation from the former. I therefore examine each in turn before directing my attention in §3 to the last two theses. Let any seeming repetition here serve to emphasis their inter-relatedness.
2.1 Identity
This is the claim that each single mental event is numerically identical to a single physical event. To avoid making mental type to physical type identifications Davidson offers us a token-identity theory. This allows that the mind-brain identity theory is true, despite the fact it is not always true that similar types of mental events are correlated with similar types of physical events. This in turn, suggests Davidson, rules out the possibility that mental events are related to physical events in a lawlike way. Numerical identity does not however mean we are talking about two things, a mental and a physical event. Rather we have a single event with a mental description and a physical description. For Davidson an event is mental if and only if has a mental description where the mental verb is essential and irreplaceable or non-substitutable with a physical verb.
However, there seems to be a problem for Davidson’s identity thesis. His original work on events took the form of a defence of a claim about identity conditions. Given that events can be the cause of or effects of other events Davidson’s intuition was that their conditions of individuation should be couched in terms of causes and effects. He proposed that ‘events are identical if and only if they have exactly the same causes and effects’. On that understanding a token mental event would be identical to a token physical event if each had the same cause and effect. This claim says nothing about what events are but ensures only that they be understood as particulars if one also holds that causation is a relation between particulars. However, Davidson offers us little reason to suppose this is the case other than on the grounds it relates events and they are particulars. So it provides no independent grounds for thinking events are particulars. Furthermore, it says nothing about what kinds of particulars events might be. Indeed it is consistent with events being abstract or non-material particulars.
Notwithstanding these observations Davidson later abandons his intuition that the conditions of individuation of events should be couched in terms of causes and effects. He writes, ‘[Quine] says my suggested criterion for individuating events is radically unsatisfactory, and I agree… Quine has made it clear to me what was wrong with my original suggestion, and I hereby abandon it …Quine’s criterion is nearer and better’. What Quine’s criterion amounts to is that events are identical if and only if they occur in the same space at the same time. Quine was content to allow that his thesis turned events into material objects; he says ‘a physical object … is the material content of any portion of space-time, however small, large, irregular, or discontinuous. I have been wont to view events simply as physical objects in this sense’. Davidson, however, needs to resist this form of assimilation if he is to maintain his non-reductive thesis. He does so by calling for a distinction between events and objects reflected by our ordinary ways of speaking or sorting.
But how are we to understand his new allegiance to a spatiotemporal theory of event individuation? If we agree to the plausible assumption that anything which exists in space and time is physical, then it follows from his new theory of event identity alone that all mental events are physical events. A mental event, being an event, must have spatiotemporal location, and therefore must be a physical event. But if mental events are identical to physical events it seems to me that this should be the result of some argument, rather than simply the consequence of a theory about the nature of events.
Such an argument could run thus: Davidson could restrict the theory of spatiotemporal individuation to physical events. The argument for anomalous monism would then establish that all mental events that interact causally with physical events are physical events without relying on a spatiotemporal theory of events. He is then free to apply his spatiotemporal theory to mental events as well without begging the question. This approach is far from problematic for Davidson is then confronted with a choice. He can either keep his spatiotemporal theory as a general theory of event identity, thereby proving his materialism on the sly, or he can restrict his theory of event identity in the way suggested and use his argument for anomalous monism to establish that mental events are physical events. In which case he forgoes a general theory of events.
However, such problems are not pressing in either case for they do not challenge his claim about the relation of mental to physical events. We have a single event with a mental description and a physical description, where an event is mental if and only if it has a mental description where the mental verb is essential and irreplaceable or non-substitutable with a physical verb. This is explained by his non-reductive thesis below.
2.2 Non-reduction
Davidson claims we cannot characterise the mental properties of an event physically even though each single mental event is identical to a physical event. This is to be contrasted with a view held by some philosophers, particularly those of the logical positivist school, who have been committed to the view that all science could be reduced to physics. Since these philosophers also held that psychology either was, or could be made into a science, they believed that psychological properties are, at least theoretically, reducible to physical properties. So, for any psychological property, say, believing I am thirsty, there is some property, probably highly complex, belonging to basic physics, which is identical to it. To make this claim more plausible it was allowed that psychology might not be directly reducible to physics, but might instead be reducible to neurophysiology, which in turn would be reducible to physics.
What makes reductionist versions of the identity theory unacceptable to Davidson is that by linking mental properties to physical properties, they contravene the thesis of anomalousness of the mental. He insists that if there are bridging laws between the mental and the physical, then the attribution of mental states to people will no longer be answerable only to the normative and holistic considerations to which he thinks they ought. The possibility will then arise that the physical states of some person will pre-empt the attribution of mental states based solely on the evidence appropriate to such attribution. So for Davidson no matter how detailed or complicated the purely physical description of an action, some mental terms will always be needed for its adequate characterisation.
Support for Davidson non-reductive thesis is to be found in the argument for variable or multiple realisation. This is the claim that a certain type of mental event can be realised by a potentially infinite number of different brain events. For example, let p stand for a token physical event, and M for a type of mental event: p1 instantiates M1, but so may p2, p3,…pn. Multiple realisation therefore gives us reason to think that Davidson has a point, namely; that a single physical description fails to fully characterise a mental event, for that mental event could be identical to any number of different physical events.
If the foregoing argument holds it refutes psychophysical reduction, though this is far from the last word on the matter as I will show in 3.1. Suffice it to say, it was, thinks Davidson, a mistake to assume that it was possible to explain what someone thought or did on a particular occasion in abstraction from the whole network of that person’s beliefs, desires, intentions, hopes and other mental states. In other words, that a person’s belief and desires could be specified independently of one another, and decided solely by the inspection of individual tokens of brain behaviour.
2.3. Nomological connection of events under their physical descriptions only.
Davidson holds that a lawlike connection holds between physically described events and we cannot, from §2.2, reduce mental events to physical events, so we cannot include the mental descriptions within the law. He states his claim thus:
Suppose M, a mental event, caused P, a physical event; then under some description M and P instantiate a strict law. This law can only be physical … But if M falls under a physical law, it has a physical description, which is to say it is a physical event … So every mental event that is causally related to a physical event is a physical event.
Given that Davidson has made differing claims about the conditions of event individuation there are two ways we can examine this claim. In §2.1we saw that Davidson initially viewed events as bare particulars individuated by their causes and effects, later he abandoned this in favour of a spatiotemporal view of events individuated by their spatiotemporal locations. Underlying the former position is the idea that events, in themselves, are neither physical nor mental. What makes an event physical and what makes an event mental is whether or not it has a mental or physical description. As stated in §2.1 for Davidson ‘an event is mental if and only if it has a mental description’. A worrying aspect of this claim is that events seem to form a single, and ontologically neutral, class of entities. However, if we simply identified being a physical event with being describable in physical vocabulary then every event would trivially count as physical, since a physical description would be true of every event. Similarly of course, every event is describable in mental terms and hence every event would trivially count as mental. While alert to the problem, Davidson’s attempt to deal with it is less than convincing.
In ‘Mental Events’ he introduces the idea of descriptions that use mental or physical vocabulary essentially. Yet this is not at all helpful for he accepts that this will allow all events to be counted as mental. However, he justifies this by saying, ‘we can afford Spinozistic extravagance with the mental since accidental inclusions can only strengthen the hypothesis that all mental events are identical with physical events’. His interlocutor might reasonably protest at such extravagance claiming that it would allow that all events are describable as physical, thus strengthening the idealist hypothesis that all physical events are identical with mental events. However, it might be argued that since events themselves are considered only as bare particulars nothing very much hangs on whether all events qualify as being mental or physical through being trivially subsumed by mental or physical descriptions.
The view of events as bare particulars contrasts however with the view of events individuated by their spatiotemporal locations (outlined in §2.1). On Davidson’s new spatiotemporal conception events are essentially physical. Instead of bare events, on the one hand, and mental or physical descriptions on the other, each event is now seen as having one description, the physical description of its space-time location. Given the two positions how are we to understand the above passage? Such equivocation seems only to obscure the issue of which events a lawlike relation holds between.
Obfuscation of this sort has motivated more general criticisms against anomalous monism. Hornsby reasonably asks what assurance there is that the way that one form of discourse which articulates the world into different events will be the same as another? The events of psychology, beliefs, perceptions, and the like, may well cut across events as discerned by the neurophysiologist. Take for example the desire to raise my arm in a signalling motion causing me to raise my arm. While neurophysiology will describe what happens in terms of neurophysiological events, psychology will discern a desire and an action. But can we assume we can superimpose these two accounts and find that there are discrete events that each describes in its own way. Where exactly does the action begin and the desire end? Must the desire and action be constituted as neurophysiological events that are continuous, or could there be parts of this chain between the desire and the action?
Such questions put pressure on the view of events as bare particulars, for this conception of events provides no reason for not seeing the psychological description as being about one set of events, and the neurophysiological descriptions as being about another set of events. The ontological neutrality of events makes it difficult to insist that both descriptions must be of the same set of events. What is being challenged here is not so much Davidson’s belief that events, as particulars, can be given different descriptions, rather that there are no compelling grounds for thinking that the events picked out by, say, neurophysiology and psychology must be the same things under different descriptions. If they are not the same things under different description then the identity theory is false.
This challenge to Davidson’s identity thesis seems to require abandonment of one of the propositions outlined in §1.2. Hornsby suggestion is that we give up the principle of the nomological character of causality since it is this principle which requires that a mental event that causally interacts with some physical event is the very thing which fits some physical description under which it instantiates a physical law.
Yet I think Davidson can resist Hornsby’s suggestion. He can do so by showing she is wrong to say the descriptions offered by psychology and neurophysiology are not talking about the same event in different ways. Hornsby does in fact posit a sense which gives credence to the notion of physics and psychology talking about the same things in different ways but rejects this view as untenable. She argues that some events are complex events, composed of smaller events, e.g. the event of Shakespeare writing Hamlet was composed of the smaller events of his writing each act, sentence, and word. Likewise, the event of Lichtenstein producing a completed painting was composed of the smaller events of placing individual coloured dots on his canvass. Similarly for physical events, if we accept that all matter is composed of atoms, all events are composed of some basic unit. Hornsby calls this the ‘mereological’ conception of events. On this understanding, she says, we might well say psychology and physics appear to identify different events by regarding each event as making different collections out of some basic unit. She thereby claims that there must be some other event that is the composition or fusion of these other events. She insists, however, that such ‘putative events lack any conceivable value to us in giving explanations … [I]nasmuch as it is the nature of events to cause and be caused, we expect individual events to be members of kinds that pull their weight in illuminating accounts of why one thing followed another’. She thus concludes we should reject the mereological conception of events and with it the common stock of events which Davidson insists are of interest to, say, neurophysiology and psychology. I think she is wrong on this point.
While the mereological conception of events offered by Hornsby does not entail Davidson’s spatiotemporal theory of their individuation I see no reason for Davidson not insisting that it work with it. He can argue that the basic units out of which complex events are composed could be seen as spatiotemporal point-instants. This is consistent with his new allegiance to the Quinean position, viz. that, ‘a physical object … is the material content of any portion of space-time, however small, large, irregular, or discontinuous’ (italics mine). Thus Davidson could allow that there is an event composed of Shakespeare’s writing of Hamlet. In other words, he can endorse Hornsby’s mereological conception of events as a consequence of his spatiotemporal view of event individuation, and in so doing answer Hornsby’s criticism. Furthermore, in abandoning the theory that events are individuated by their causes and effects Davidson will disagree with Hornsby that ‘it is the nature of events to cause and be caused’ . On his spatiotemporal theory of event individuation events may be causes and effects, but it is not part of their nature that they are. Moreover, Davidson can insist that physics and psychology are talking about the same thing in different ways. If this argument holds he can now say that a mental event that causally interacts with some physical event is the very thing which fits some physical description under which it instantiates a physical law. On that basis he does not need to cede to Hornsby’s request that he abandon his second principle of the nomological character of causality.
2.4 The anomalousness of the mental
Davidson holds that we cannot formulate laws connecting physical events with mental events. In other words, the mental cannot be explained or predicted by strict deterministic physical laws. That we cannot formulate such laws is, for Davidson, not a contingent matter but rather it is the claim that we necessarily cannot formulate such laws. Much of course hangs on such a claim for if it cannot be substantiated we are, by Davidson’s lights, simply not free given he holds that freedom entails anomaly. Given the implications for the claim it requires close and careful examination.
Davidson maintains that laws are linguistic and thus hold between events only as described in certain ways. Further, he also holds that causality and identity are extensional relationships, either holding or not holding, between events no matter how described. This allows for the possibility that two events can stand in a causal relationship extensionally but that the subsequent law can only cover certain descriptions of those events. So understood he insists there are no psychophysical laws.
Psychophysical laws, if they obtained, would identify mental properties, such as believing that it is raining with physical properties. Such a law, where M denotes some mental state and P some physical state, would say the following.
(L) for any x, x is in M iff x is in P
The subject of the predicate M is the same as the subject of P, it is a relation of identity not of description, giving us a nomological connectedness.
However, Davidson claims the following: ‘nomological statements bring together predicates that we know a priori are made for each other – know, that is, independently of knowing whether the evidence supports a connection between them’. And, he insists, given that we know in this way we should know a priori when predicates are not made for each other, and such is the case with mental and physical predicates. The polemic in §2.4 give us reason to be sceptical of this claim. However, he offers as an explanation why he believes there cannot be psychophysical laws:
[I]t is a feature of physical reality that physical change can be explained by laws that connect it with other changes and conditions physically described. It is a feature of the mental that the attribution of mental phenomena must be responsible to the background of reasons, beliefs, and intentions of the individual. There cannot be tight connections if each is to retain allegiance to its proper source of evidence.
The problem with laws according to Davidson is that they would connect two sets of predicates, mental and physical, thereby serving as conduits through which principles distinctive of one set would be transmitted to the other set. If this were the case then the possibility will arise that the physical states of some person will pre-empt the attribution of mental states based solely on the evidence appropriate to such attribution, e.g. it would be regarded as valid to deny x believing that F because she is a physical statement . For Davidson this is unacceptable for he claims that in recognising that mental and physical predicates are not made for each other, we are recognising that ‘conditions of coherence, rationality, and consistency … have no echo in physical theory’. This certainly asserts the anomalousness of the mental but more needs to be said if we are to accept as an undeniable fact the anomalousness of the mental claimed in §1.1.
Let p be the proposition that some point is green and q the proposition that that point is extended. Let us suppose also bridging laws connect the belief that p and the belief that q with the neural states m and n. So, whenever someone was in the neural state m, she was also in the neural state n. If bridging laws hold then we should be able to infer that whenever she believes p she also believes that q. But this does not seem to be the case. It is plausible to assume that someone may have strange beliefs about colour such that our overall requirements of rationality, the normative principle, directs us to attribute to her the belief that p and not-q. So, while Davidson thinks events under their mental descriptions are predictable we cannot infer anything from the physical description of mental events. Thus, he concludes, there are no strict laws at all on the basis of which we can predict and explain mental phenomena. This amounts to arguing for ‘nomological slack’ relating the mental to the physical. Such ‘slack’ allows for a more relaxed position over what constitutes causal explanation such that the laws governing practical reasoning are of a totally different form to those governing the operation of causes.
Psychophysical laws should of course be distinguished from psychophysical generalisations. For example, in a fixed domain all the objects are either blue or red. They are also either soluble or insoluble. Now it could be true that all the blue objects are soluble. If so, we could formulate a true generalisation about this fixed domain of objects,
(G) if x is blue then x is soluble.
But no one would say that (G) is a law; that all blue objects are soluble. Rather (G) is a true generalisation for our fixed domain relating solubility to colour. To be a law it must support counterfactual claims about what that something would be like if described by the antecedent of the conditional. Clearly (G) falls short. Davidson does not deny that there may be true psychophysical generalisations, for him this is merely an empirical matter of no philosophical interest itself. His point is rather that even if there are such generalisations our a priori conception of the nature of the mental means that there cannot necessarily be laws.
This seems plausible, but his interlocutor might plausibly counter that the normative principles governing the mental certainly seem to fit the criterion of what it takes to be a psychological law. How might Davidson respond? He would have no difficulty ceding that they may be laws in some sense but would be adamant they are not laws whose existence enables us to make accurate predictions. There is nothing in Davidson’s thesis for the anomalousness of the mental that requires he countenance the possibility that any event is undetermined. Merely that there are events, when described in the vocabulary of thought and action, that resist capture into a closed deterministic system. As in the case of psychophysical laws the argument against psychological laws rests on Davidson’s holism and normativity of the mental. Davidson’s makes explicit his claim
[I]t is an error to compare a truism like ‘If a man wants to eat an acorn omelette, then he generally will if the opportunity exists and no other desire overrides’ with a law that says how fast a body will fall in a vacuum. It is an error, because in the latter case, but not the former, we can tell in advance whether the condition holds, and we know what allowance to make if it doesn’t.
If the truism about the omelette were to be considered a putative psychological law for the antecedent to hold someone must want to eat an acorn omelette. But knowing whether this condition holds is not straightforwardly something that can be predicted with precision thinks Davidson. The antecedent desire and the consequent action both feature in the interpretation of an agent and are hence related to one another as part of a holistic web. So knowing whether the condition holds is part of having some extensive view of the agent as a rational being. But the normative character of the mental make it impossible to specify all of someone’s beliefs and desires individually such that we could know whether the conditions hold. What is needed in the case of action if we are to predict on the basis of desires and beliefs, is a quantitative calculus that brings all relevant beliefs and desires into the picture. However Davidson believes there is no hope of refining the pattern of explanation on the basis of reason into such a calculus. There are, however, arguments to suggest Davidson too readily rejects the possibility of such a calculus. A refinement of the pattern of explanation is offered by Fodor.
Fodor argues for a scientific psychology with laws that contain failsafe clauses, such as ‘all other things being equal’, or ‘and there are no overriding desires’. So while a strict law of nature might tell us that F events always cause G events, or every F event has such-and-such objective chance of causing G event, a ceteris paribus law might tell us that F events cause G events ceteris paribus. With such failsafe laws psychology would be in no worse state, he claims, than those of the special sciences. Such special sciences would include all empirically explanatory schemes other than physics, e.g. geology, chemistry, and the like.
Fodor insists that if Davidson were to reject psychological laws because of the necessity of failsafe clauses then he must also reject as laws the generalisations of all the special sciences, concluding that the only real science is physics. But why should this matter? Nothing hangs on this for any of Davidson’s claims. What is worthy of consideration about Fodor’s claim is that the presence of failsafe clauses reflects the fact that the regularities found in say, geology, are contingently subject to ‘interferences’ which are too diverse themselves to be subsumed by any purely geological description. Similarly, he thinks there is also a mental reality that underlies any attempt to interpret it. As a result any complications introduced by holism and normativity can simply be ‘idealised’ away. Here idealisation is an epistemological device used by Fodor to remove ‘interference’ that might obscure the way things really are underneath. Thus the difficulty in giving a ‘quantitative calculus that brings all relevant beliefs and desires into the picture’ is merely an epistemological difficulty.
Plausible as Fodor’s argument seems it is not persuasive against Davidson. He need only argue that because the normativity and holism are constitutive of what people really believe and desire it means that there is no underlying mental reality whose laws can be studied in abstraction from normative and holistic perspectives of interpretation. Thus he can go on to argue that the failsafe clauses in psychological generalisations reflect a fact about the mental in a way in which their counterparts in, say, geological generalisations, do not reflect any deep fact about geology.
Support for Davidson against Fodor’s proposal comes from Schiffer. He says ‘[m]y trouble with psychological ceteris paribus laws is that I doubt there are any.’ He does however believe there are ceteris paribus sentences but questions whether they do in fact express propositions, and if they do, whether those propositions are laws. Schiffer claims that sentences like
(S1) If a person wants something, then, all other things being equal, she’ll take steps to get it
are deceptive. While it looks to be expressing a determinate proposition, because ‘all other things’ seems to be referring to some contextually determinate thing and ‘equal’ some determinate relation among them. But what, we might reasonably ask, are these ‘other things’ and what is it for them to be ‘equal’. If the ceteris paribus clause doesn’t make a contribution to the proposition expressed by (S1) then all we have is a sentence like
(S2) If a person wants something and …, then she’ll take steps to get it.
This is no good at all. It does not express a complete proposition, or contain anything that could be believed, or play an explanatory role. What the supporter of ceteris paribus laws needs to show is, i) what true proposition is expressed by ‘Ms cause Bs ceteris paribus’, and, ii) that the truth of the proposition expressed by ceteris paribus sentences determine anything worth calling psychological laws. In other words, we do not show there are ceteris paribus laws merely by showing that there are ceteris paribus sentences.
Despite Schiffer urgings that we reject the possibility of psychological ceteris paribus laws Davidson might be prepared to countenance them if they did not tell against the anomalism of the mental in the strictest sense. I will consider such a possibility in §4.3.
For the moment at least we can say that if the arguments against psychophysical, psychological and psychological ceteris paribus laws hold then a case for the anomalousness of the mental has been made. Whether this is sufficient to establish free will is another matter. But such speculation is premature; a serious problem awaits this thesis in §3.1.
2.5 Supervenience
We saw in §1.1 Davidson holds ‘that both the causal dependence, and the anomalousness, of mental events are undeniable facts.’ Even if we are prepared to countenance claims for the anomalousness of the mental he needs to tell us how the mental supervenes on the physical, especially given he wants a causal account of freedom. As with the topic of §2.4, then, much hangs on this thesis. However, given I raise a number of objections in §3.1 to Davidson’s supervenience claims I will largely confine myself in this section to exposition.
Davidson supplements his theory of anomalous monism, which as we have seen combines ontological monism with the anomalousness of the mental, with the claim that the mental is supervenient on the physical. He does so because without it the mental is causally irrelevant; causal relevance is not assured by ontological monism alone.
In ‘Mental Events’ he defined supervenience in two ways. He says, ‘[i]t might be taken to mean that there cannot be two events alike in all physical respects but differing in some mental respect, or that an object cannot alter in some mental respect without altering in some physical respect’. I take the first definition to mean that what mental event occurs depends on what physical event occurs, and not the other way around as this would lead to the accusation of espousing a form of idealism. Davidson seems to endorse this first definition when he later writes;
[T]he notion of supervenience, as I have used it, is best thought of as a relation between a predicate and a set of predicates in a language: a predicate p is supervenient on a set of predicates S if for every pair of objects such that p is true of one and not of the other there is a predicate of S that is true of one and not the other.
But why does he hold that the mental is supervenient on the physical? Following the passage quoted he says that: ‘supervenience as I have defined it here is clearly all I need for the argument in ‘Mental Events’’. What he is arguing for here is the identity of mental events with physical events. Failure to do so makes it difficult for him to show how the mental is dependent on the physical. Showing the dependence avoids the charge of epiphenomenalism. With the mental dependent on the physical he can then explain what it is to act for a reason. In other words, he requires that reasons, intentional mental states, be intervening distinguishable states of an internal behaviour-causing system, i.e. causally relevant. So, when one acts for a reason that reason causes the action for which it is a reason. This is only possible if he can identify mental events with physical events. This having been said, the law which a reason instantiates is a physical law instantiated under a physical description. However, in arguing for the identity of mental events with physical events Davidson wants to emphasis that such ontological reduction does not imply that mental properties are physical properties, nor that there are causal or bridging laws relating events classed by mental properties with events classed by physical properties. In so claiming he preserves the anomalousness of the mental.
It is puzzling to see why such emphasis is provided by supplementing anomalous monism, which on its own would not rule out two physically identical objects differing with respect to mental properties, with a claim about the dependence of mental properties on physical properties. Supervenience may, or may not, fall short of causal or bridging laws but it does represent a tightening of the relation between mental and physical properties as compared with anomalous monism. It cannot therefore be needed to emphasis the looseness of the connection. Therefore, his need to identify mental events with physical events by supplementing anomalous monism with supervenience seems problematic. Davidson’s views on supervenience have of course evolved since first given expression in ‘Mental Events’. However, I will consider these developments having first examined the problems with the thesis as articulated.
Blog #3 - Supervenience and the Anomalousness of the Mental