The Canberra Plan and the Diversification of Australasian Philosophy: 1990s

Abstract

The story of Philosophy in Australasia during the 1990s is one of diversification and maturation. The metaphysical naturalism of previous decades continues though it becomes more sophisticated and itself diversifies. The Canberra Plan became a self-conscious approach to philosophy though not without its critics. Perhaps in contrast to expectations fostered by preceding decades, one finds in Australasian philosophy a resurgence of Dualism, interest in arguments for God and anti-Utilitarian moral philosophy. In this chapter, I concentrate on the development of Functionalist philosophy of mind leaving the Dualism to one side, the resurgence in philosophy of religion, the emergence of Practical Ethics and Applied Philosophy in a widened context for doing philosophy and questions raised about the method and nature of philosophy by the moral philosophy of Raimond Gaita. The treatment is selective and cannot do justice to the breadth and variety of the discipline during the 1990s. I select to try to convey some of that breadth and variety. As always, philosophy in Australia is adventurous, vigorous and fundamental.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

I expect that if one were to seek a characterisation of philosophy in Australasia prior to the 1990s from someone in the know outside the region, one would expect the focus to be on ‘Australian materialism’. Were philosophy of religion to figure in the characterisation at all, I imagine that Australian philosophy would be thought of as predominantly atheistic or at least naturalistic, supposing that God is by definition a supernatural being. In moral philosophy, the impact of Peter Singer among others would suggest that utilitarianism would count strongly in the characterisation. Of course, these points would involve something of an abstraction from what was to be found on the ground: there always have been plenty of theistic and other kinds of religious philosophers in Australia; there have been anti-utilitarians and anti-materialists also. However, the international image of Australian philosophy would be along the lines sketched.

Something I would emphasise in this chapter is that the 1990s saw a diversification of philosophy in Australia, even in the presence of the rather challenging conditions of the Australian university scene during this decade (surely part of any adequate account, more than I can give here). And in particular, perhaps surprisingly, dualism, God and Platonic/Kantian ethics got something of a rehabilitation. Dualism will have to be ignored here unfortunately. For the main and most influential dualist argument during the 1990s came from the Australian philosopher, David Chalmers. However, he worked in the United States until 2004, and so I omit discussion of his work. Instead, I will focus on intentionality and mental causation in the philosophy of mind, arguments for God in philosophy of religion and on Plato (and Kant) in the moral philosophy of Raimond Gaita.

In the 1990s, Australian philosophy was still mainly quite sanguine about metaphysics—a continuing feature since the 1950s. Generally, its temperament was naturalistic—even if we include Chalmers’ dualism, the arguments for God and the ethics of Raimond Gaita. Philosophical research was predominantly pursued within the analytic tradition, involving work in philosophy of science, epistemology and philosophy of language and logic. However, we also see during this period the establishment of a society for the study of European philosophy, an energetic Women in Philosophy association and an emerging lively interest in comparative philosophy and the establishment of schools of philosophy at two Catholic universities. Further, in my view, work deepened and matured in the history of philosophy, a discipline (as it was done increasingly in the 1990s) that was not always central to Australian analytic philosophy. I will not be able to cover all this. As I said, I will focus on philosophy of mind, philosophy of religion and moral philosophy—and I will not do anything like justice to all the high-quality work done in these areas during the decade in Australasia. So, this chapter has a section on each of these areas in turn. I will use the discussion of metaethics to include some reflections on, for want of a better term, ‘philosophical method’ worth turning over in the light of some of work in the 1990s. I would, however, stress (something which I cannot substantiate here) that a fuller story will include the wide range of diversification to be found in the discipline and, of course, an accounting of the emergence of various institutions of philosophy outside the standard university department. For during this decade, much work was done of signal importance both for the inner workings of the discipline and its place in the broader society, in centres of various kinds, interdisciplinary entities and the like. But now, I turn to the mind, then God, then the good.

The Mind: Intentionality and Mental Causation

During the 1980s, arguably the most influential and distinctive contribution of Australasian philosophy to debates concerning the nature of the mind was made by Frank Jackson’s well-known argument about Mary. Having lived all her life in a monochrome environment and knowing all there is to know about physics, human physiology and particularly the materialist science of the senses, Mary learns something she did know not know before when she first steps into the polychromatic world the rest of us are familiar with. That is, the phenomenology of the world’s sensible qualities is a fact further to the facts detailed in the complete materialist science of perception and its physical causes. The so-called knowledge argument seems to show that the reductivist physicalist consensus about the mind stumbles across this fact that human mental life is replete with qualia. The implication would seem to be some kind of property dualism—presumably, an epiphenomenalism wherein this aspect of mind is causally impotent, the physical world being causally closed. So it seemed to some during the 1980s. If there wasn’t an adequate alternative, materialistically acceptable, way to capture what Jackson was pointing to in the Mary example, either materialism was in trouble or eliminativism had more going for it than most wanted to admit. Neither option seemed palatable: if epiphenomenalism is true, the stabbing pain doesn’t cause us to wince; if eliminativism is true, the gulf between science and commonsense seems to widen to a point even those who are not real friends of commonsense would not welcome.

On the other hand, Australasian philosophers are often thought to be bold and brash in the theses and arguments they defend. Michael Devitt and Kim Sterelny could be seen this way in their work in philosophy of mind and language during the 1980s Devitt and Sterelny (1987). Heavily influenced by Quine’s view of philosophy as continuous with science and his arguments against the analytic/synthetic distinction, they advocated putting metaphysics before either philosophy of language or epistemology (Devitt 1997). Theirs was a naturalistic program in the theory of content or intentionality. They argued for a ‘language of thought’ as the syntactically structured medium in which inner belief and desire states proximately caused behaviour against the background of the causal account of representational meaning. Devitt, in particular, defended an aspect of this representationalist theory of mind in which not only the ‘broad’ semantics of truth-conditions but also the ‘narrow’ semantics of states’ internal conceptual-role played a part.

In what follows, I will lay out the development of the views of Sterelny and Jackson on intentionality during the 1990s. One might also discuss David Chalmers’ work on consciousness in The Conscious Mind (1996) and the two-dimensional semantics integral to it, but as mentioned earlier, until 2004 he worked in the United States, and I therefore omit to discuss his work here.

For some time now, the unifying theme of analytic philosophy of mind has been functionalism. More or less, this is so in 1990s Australasian philosophy of mind. However, it has been taken in different directions. I will start with Sterelny.

Representationalism and the Language of Thought: Kim Sterelny

Sterelny’s is a form of functionalism where beliefs and desires are internal states of a subject which consist of relations, of the right kind (one for belief, one for desire), of the subject to internal, syntactically structured symbols which, in turn, represent ways the world is or could be by virtue of the right kind(s) of causal relations of the elementary parts of the structured symbols to the world. This is a version of functionalism in the sense that the relation, to inner representations, of which the state consists (i.e., that it is a belief state that p rather than wondering whether p or desiring that p) is a matter of how the state is involved in the causal network of perceptual inputs and relations to other contentful states which causally produce the behavioural outputs of the subject. Sterelny is often careful to avoid claiming for this account of the mind that it is ‘functionalist’ in other senses. For instance, he avoids the thought that what makes the belief that p, the belief that p, in general has anything to do with the evolutionary history of the species to which the subject belongs—though in other ways, being a self-confessed ‘born again Darwinist (Sterelny 2001, p. xiv), his large program is to defend a conception of the mind as a particular kind of representational system (an intentional agency system) which has evolved as an adaptation to solve problems of environmental complexity (Sterelny 2001, Chaps. 8–9). But such a broad role for biological function will not work as a general theory of mental content (Sterelny 1990).¹ Nor is Sterelny sanguine about the prospects of a theory of propositional attitudes in terms of functional role interpreted as inferential role. For inferential role theories miss the point that it is the reference of concepts that demands explanation (and inferential role is of sentence-like representations ‘made out of’ concepts). And, further, the frame problem(s) arises sharply at the stage where perceptual inputs end and belief takes over, if content is individuated by inferential role (Sterelny 1990, p. 136ff.).² If what we are put in mind of to believe under the pressure of perceptual inputs is inferentially individuated, then, because in principle any kind of knowledge we have could be relevant to determining what to make of what we ‘seem to see’, it is hard to see how one is to assign determinate content to perceptual beliefs at all—too much, of indeterminate domain, could matter in too many different kinds of circumstances of being appeared to, for the inferences we should draw from perceptual inputs (Sterelny 1990, p. 137).³ We would lose grip on individuating contents entirely.

Instead, Sterelny proposes a program for an updated concept empiricism as a naturalistic theory of content. He distinguishes between kinds of concept: base level concepts (relatively few), concepts completely defined in terms of other concepts and concepts whose content is fixed by a combination of causal relations to the world and descriptive conditions associated with the concept whether directly keyed to the base level concepts or less than directly keyed to the base level concepts.

This is an empiricist theory of content because the base level concepts are the products of perceptual input devices which are modular in nature and the meanings of these products are fixed by the environmental sensitivities for which evolution selected them. Their modularity yields innateness and so furnishes independence from learned concepts, avoiding holism and permitting them to act as basic contents. Relatedly, the modularity yields determinacy of content, solving the qua- or depth problem: why our face recognition module recognises faces and not faces or retinal pattern X or…: recognising retinal patterns has no evolutionary advantage while face recognition does in the circumstances under which the face recognition modules developed and were selected for. And it solves the misrepresentation problem because the concept of green developed to recognise green objects under normal visual conditions and had evolutionary advantage in them, but does not work in the dark or under abnormal light conditions such as the light cast by sodium lamps (Sterelny 1990, p. 138). The evolutionary role for which the module was selected furnishes those states with the environmental reference which are partial determinants of content. Causal relations to the immediate occasion have a role too: a particular episode of the activation of the facial recognition module feeds into the perception of Melanie’s face because it is Melanie who, then, caused the activation (Sterelny 1990, p. 137).

Central to these solutions which this suggestion proposes are the constancy mechanisms involved in the operation of the perceptual modules. Things don’t look like they have changed colour when the light changes because a cloud has passed over them, nor do they seem to change shape as they move across the glade. Marr’s theory of vision proposes such mechanisms (Sterelny 1990, Chap. 4). Sterelny offers Marr’s theory as a kind of model of the representational theory of mind he favours, without being overly dogmatic about it. The constancy mechanisms of the visual system have clear evolutionary advantages. In terms of Sterelny’s later ideas,⁴ the tracking of environmental features and objects (the representational content of the states into which we are put by our tracking operations) which the constancy mechanisms afford is robust over variations in the routes information from the world takes into the mind and the cues by which it is registered (Sterelny 2001, pp. 207–211, 262–266).

Non-base concepts mostly will not be definable completely in terms of other concepts. Only some will be. Instead, for most non-base concepts, their reference conditions will be provided by a combination of causal and descriptive conditions. S has the concept of tigers, tiger, partly because of his causal contact with tigers and partly because of his descriptive knowledge of tigers. For such natural kind concepts, Sterelny suggests an advantage of his new proposal over what he and Devitt had proposed in their (1987). He now suggests that some clarification is possible of the descriptive conditions for such concepts. A clue is taken from Marr’s account of the way information borne by light incident on the eye permits the construction of 3-D object-centred representations prior to recognition of the object itself: e.g., animals can be represented as a hierarchy of generalised cones with major cones representing the trunk of the body and others, the head and limbs. The 2-D projection of the light encodes this geometry (Sterelny 1990, p. 71). To be sure, the details of Marr’s theory may be incorrect. However, it makes the point that if the basic idea of the modularity thesis is right, there will be some coarse-grained representation of tigers that is part of the visual perception of them. So Sterelny suggests that the descriptive condition that helps tie down the content of natural kind concepts is bound up in it: the descriptive condition can be understood as given in terms of this Gestalt representation (whatever more exactly that is) yielded by the operation of the perceptual inputs.

Hence, some concepts can be acquired on the basis of causal links and modular representations, and others can be acquired on the basis of these concepts. But all concepts bottom out in the perceptual modular representations. However, this account is importantly different from traditional versions of concept empiricism. First, Sterelny’s stress on the importance of our perceptual systems’ constancy mechanisms grounds taking the content of the representations they produce not to be our experiences but rather objective features of the world, those features of the world which the constancy mechanisms are keyed to robustly under variations of environmental conditions and so on. Second, Sterelny invests no stock in the importance of definitions in terms of base concepts as traditional concept empiricism generally did. Very few concepts are genuinely definable in terms of other concepts. Rather, most non-base concepts get their semantic properties from causal relations to the world (Sterelny 1990, p. 139ff.).

Sterelny maintains that his proposal also illuminates concept acquisition. He discusses Patricia Churchland’s (1986) criticism of Fodor’s nativism about concepts, which she sees as a reductio of the whole idea of a language of thought. Sterelny makes the point that if the language of thought hypothesis is not committed to Fodorian nativism about the possession of concepts, nor is it so committed about the acquisition of them. For the reasons Fodor has for his nativism can be undercut with Sterelny’s theory of content. That is, Fodor had argued that to learn a concept is to acquire the ability to form a hypothesis in the language of thought which specifies the extension of the concept and so requires a rich conceptual stock to be in place. Nativism seems unavoidable to Fodor. In contrast, Sterelny’s theory of content requires that possession of the concept is the result of the right kind of causal contact with a specimen instance of the concept. To be sure, for many or most concepts, a descriptive element is required. But no definition is required which will serve to specify extension. If sententialism can avoid the need for definitions, it can avoid nativism. If it can avoid nativism for possession of a concept, it can avoid it for acquisition. Though some concepts are innate—those of the modular perceptual systems whose explanation is adaptationist (Sterelny 1990, §6.9, 2001, Chap. 8, §3)—most are not. They are acquired by the right kind of causal commerce with the world together with a modicum of non-definitional description based in the perceptual system (Sterelny 1990, p. 162).

Jackson et al.: Content and Mental Causation

One very natural way to introduce the functionalist idea about the mental is to characterise mental states such as beliefs and pain as internal states which occupy a certain kind of role in the causation of behaviour. This causal role connects a mental state to perceptual inputs and other mental states in the production of behaviour. Imagine Joe who is thirsty after a hard day’s work in the paddock. He would like to quench his thirst and wonders where he might get a beer. After talking to Jack, he goes to the fridge out the back, pulls out an ale and… and so on. Joan shows up and wonders out loud why Joe has gone to the fridge out the back. Given his other mental states (thirst, desire to slake it, etc.), we would explain his going to the fridge out the back to get the beer, by explaining that he believed there was a drink in the fridge out there. So, it seems natural to think, given his thirst, that his belief caused him to go to the fridge out the back to look for a beer. Thus, the functionalist thought would seem to be that, for instance, a subject S has the belief that p if and only if S has an inner state which occupies a certain kind of role (the belief role, to be specified) which involves it in mediating between inputs from the senses and output in behaviours, interacting with other mental states, such as the thirst and his desire to slake it, in doing so. Since the inner state in question is causal, it seems natural to say that the belief is that inner state, as belief is causal. And after all, to reclaim the idea that mental states are real, causally operative internal states of subjects is a very natural way to understand the motivation of functionalism in moving away from behaviourism.

This is pretty much how Sterelny understands the main point of functionalism. Contentful inner states cause what we do by virtue of their content. Jackson and his fellow authors (Jackson et al. 2004; Jackson 1996; cf. Jackson and Pettit 2004a, c; Jackson 1998c, 1998d) demur importantly from this more or less straight reading of the basic idea of functionalism. Sterelny describes the view of Jackson et aliorum as ‘Dennettish’,⁵ apparently meaning that it is somewhat less ‘realist’ about content than he is comfortable with. And yet, though Jackson notes that Dennett (in The Intentional Stance, 1989) defends a version of the ‘map-like structures’ account of mental representation (just as Jackson means to), this is glossed as ‘the less instrumentalistic reading’ of (some of) Dennett’s papers in that collection (Jackson 1996, p. 405, n36). So, Jackson sees his account of content and mental causation as broadly ‘realist’ (cf. Jackson 1996, p. 377f.), though others are not so sure.

Jackson’s view is carefully constructed. He distinguishes what might be called ‘minimal functionalism’ from the various more detailed versions of functionalism which take diverse positions on various questions, especially the more controversial ones, in philosophy of mind. Minimal functionalism is simply the thought that ‘to have beliefs and desires is to have internal states playing certain common-sense functional roles’ (Jackson and Pettit 2004c, p. 16). Or again, ‘… the folk conception of belief and desire is that of internal states playing inter alia certain kinds of causal-explanatory roles with respect to behaviour’ (Jackson and Pettit 2004a, p. 38); and ‘common-sense functionalism defines mental states holistically by their place in a network… for a subject to be in any particular mental state is (a) for there to be a set of states interconnected in a certain way, and (b) for the state the subject is in to be in the relevant place in that network’ (Braddon-Mitchell and Jackson 1996, p. 52). Minimal functionalism, then, leaves open many questions. For instance, it is an open question for functionalism so understood whether it is the functional role that is the state referred to in such formulations or the realising state which occupies this role (Jackson and Pettit 2004b, p. 99; Braddon-Mitchell and Jackson 1996, p. 100ff.). Again, it is a question further to minimal functionalism how exactly functionalism ought to account for the content of propositional attitudes—what to make of the ‘externalist’ arguments of Putnam, Burge and others; what to make of arguments for so-called narrow content; and so on. More particularly, there are various strategies functionalists could take towards the question of the place of mental states, especially propositional attitudes precisely as contentful, in causal explanations of behaviour (Jackson 1996, 1998d). All that minimal functionalism requires is that there is some internal state of the subject which is causally involved in the production of behaviour. It does not require any particular characterisation of the internal state—whether as beliefs and other intentional states or in terms of content or some other set of terms. Putting it another way, assigning content to a belief is a matter of describing the overall causal role of the belief rather than the nature of any particular element in that causal role. If content assignments assign sets of possible worlds to beliefs, then ‘the set of worlds that is the content of the belief is a way of describing the causal role of the head states: the head state produces behaviour which is such that were the way things actually are a member of the content set, the behaviour produced would tend to satisfy the subject’s desires in the sense of bringing about a world that is the way the subject would like things to be’ (Braddon-Mitchell and Jackson 1996, p. 182ff.). In minimal functionalism, we abstract both from the intrinsic nature of the internal state and from those of its properties which cause the bodily movements of behaviour.

Jackson’s motivations for characterising functionalism minimally are twofold: (a) to place functionalism strategically in order to defend it against eliminativism as a genuine empirical theory of mind with highly entrenched epistemic credibility (see esp. Jackson and Pettit 2004a, c; Braddon-Mitchell and Jackson 1996, Chaps. 3, 13) and (b) to defend the essentials of a view of ordinary understanding about the mind as in good order and more or less continuous with emerging scientific knowledge about the causation of behaviour in human beings and the possibilities of cognitive life in other kinds of creature or (possible) artefact. The second of these raises the question of the so-called Canberra Plan for philosophy which no account of Australasian philosophy in the 1990s can ignore. However, I will postpone discussing this until the section dealing with philosophical method below. For now, I resume the account of content.

In going beyond minimal functionalism, Jackson’s view of the content of mental states combines a number of moves which, in combination, seem to be in tension. For, on the one hand, he defends the following propositions:

• 1. The internal states which are the mental states such as belief that p or desire that q are the brain states which occupy the functional roles of mediating between inputs, other mental states and behaviour—the realiser states, rather than the second-order states of having a state which fills the functional role, the role states.

• 2. Causal explanations of behaviour which appeal to functional states involve this claim:

  • (a) It is the (internal) categorical basis of the functional state which causes the bodily movement involved in the behaviour (cf. Jackson 1996, p. 397).

• 3. Psychological explanations cannot proceed autonomously vis-à-vis neuro-physiological explanations. The truth of physicalism shows that there must be a relatively smooth way, even if its details are beyond us, of typing psychological states in terms of neuro-physiological-cum-physical types. Supervenience and causal closure of the physical require this.

• 4. The explanation of human behaviour in terms of the categorical basis of functional psychological properties of internal states will be given in the neuro-physiological details about the realiser states which are the mental states, specifically in the neuro-physiological explanations of bodily movements as caused by the brain states.

• 5. These detailed neuro-physiological explanations will not mention, nor need to mention, the content/intentionalistic characteristics of the states which are our mental states. The movements of the body have proximal causes in the neuro-physiological properties of the brain states.

• 6. Functionalists should accept the moral from ‘externalist’ arguments concerning content that the content of many mental states is ‘broad’: that is, they have content which is (at least in part) constituted by such facts about the mental states as what it is in the subject’s environment that causes them, the word usage of the subject’s linguistic community, the effects in the world of the subject’s action, how things are in the subject’s environment and so on.

Together, these propositions suggest that there is no role for mental content in adequate causal explanations of behaviour. Content is an aspect of the specification of functional roles cited in belief-desire explanations of behaviour. But beliefs and desires are the internal neuro-physiological states of brains, and the causes of the bodily movements involved in behaviour will be the internal, intrinsic neuro-physiological properties of these brain states. As the categorical basis of the functional properties involved in having beliefs and desires, these neuro-physiological properties will do the casual explaining. Content is therefore redundant in the explanation of behaviour. Functional properties, like dispositions, do not cause. Further, the intentional content of propositional attitudes is a matter of their relations to the environment, the linguistic community of the subject and so on. Being in this way relational and keyed to the external world, these intentionalistic properties cannot be causes involved in the production of bodily movement. They could be different without making a difference to the subject or her behaviour (as in Twin Earth and like thought experiments). To be sure, ascribing beliefs individuated by the criteria of broad content implies that the internal state possessed of the content causes the behaviour: Joe believes that there is beer in the fridge out the back if there is in Joe an internal state which causes behaviour that would satisfy Joe’s desire for beer if there is beer in the fridge out the back. Thus it seems a necessary truth that having this belief will, absenting other considerations, typically cause that sort of behaviour. Yet if Hume is right, it is no part of the nature of a genuinely causal connection that the nature of the cause is, even in part, to be such as to cause the effect it explains: the nature of a thing—the thing which is the cause—may be taken in abstraction from its relations to other things, such as the causal relation. Causes, understood in terms of their intrinsic natures, are contingently related to their effects. But the belief that there is a beer in the back fridge is, for want of a better term, internally related to behaviours of seeking beer in the back fridge. Its causal powers with respect to such behaviour are necessary. So, the causally explanatory properties of the internal state which causes the bodily movements involved in such behaviours must not be those of the content of the state, but must be others which can be taken in abstraction from content and which characterise the internal state intrinsically. Again, content has no causal role.

On the other hand, Jackson also defends the following propositions:

• 7. Commonsense functionalism is true, or for the most part is, and is well enough epitemically entrenched for there to be no serious threat to it from the eventual development of the neurosciences.

• 8. Commonsense functionalism states, among other things, that beliefs contribute causally to the production of behaviour, and do this by virtue of the beliefs they are, that is, by virtue of their content.

• 9. Causal explanations of behaviour which appeal to functional states involve also this claim, in addition to 2a above:

  • (a) The functional state picked out by the specification of the inputs and behavioural outputs distinctive of the functional state was involved in the production of the explanandum.

• 10. And by six above, what belief a belief is is, often, a matter of its broad content.

Taken together, these theses, 7–8, 9a (ignoring 2a) and 10, suggest that there is an autonomous kind of psychology which proceeds in broad semantic terms but independently of the micro-explanations of neuroscience. Given the broadness of mental content, its taxonomisation of mental states could, in principle, cut across those of neuroscience. However, three above makes it clear that Jackson does not want to go in this direction. How, then, does Jackson render coherent the apparently eliminativist tendency of his emphasis on neuroscience and physicalism with his confidence in the content-causal picture of commonsense functionalism together with broad content?

There are a number of ways to approach Jackson’s solution. One is as follows. A crucial move is his idea of the ‘here-and-now’ intuition. For, he argues, it is an intuition of commonsense that when Joe heads for the fridge out the back, it is something in him that has to have changed, after talking to Jack, to explain his behaviour. It isn’t enough that there be shifts in his relationship to things in the environment, as if he could remain exactly as he was before talking to Jack. He has to change. If the properties of the resulting change to his neuro-physiological state which cause his bodily movements are neurological properties, what causal role is there for the content, that there is beer in the fridge out the back, to take up?

Jackson suggests two strategies to save a causal role for mental content. The first is derived from a distinction between causal relevance and causal efficacy and an argument that there is a way for a property to be causally relevant and crucial in certain kinds of causal explanation, without being causally efficacious. Take some chemotherapeutic concoction. It is a medication. But describing it this way is to assign it a functional property in relation to human beings. Explaining a person’s warding off a malignant tumour in terms of her taking the chemotherapeutic substance as a medication, we cite a causal-relational property of the substance. However, chemotherapeutic concoctions can be used to control the pain of otherwise unmanageable arthritis. In its use in reducing tumours, we understand the causal pathways which explain its effects. There is a detailed aetiological story tellable about its efficacy. Simply describing the concoction as a medication does not yield such a detailed aetiological story. However, by specifying the medical indications for (there being a certain kind of tumour in certain tissues, etc.) and the effects of the concoction, one brings out that the concoction is causally relevant to the reduction or death of the tumour. However, the same explanation implies that there is a more detailed explanation of exactly how the concoction produces that effect on the tumourous cells. In contrast, we do not have such an aetiology of the efficacy of such a substance in managing the pain of arthritis. However, it is obvious that there must be one. It will be an explanation which explains exactly how, in detail, the chemical attributes of the concoction, perhaps, block the nerve pathways by which the pain is experienced. Of course, this explanation which I am presuming could be wrong—perhaps, for instance, the concoction’s chemical structures actually destroy problematic cells involved in the actual dysfunction of the painful joints. But there must be such an explanation of the causal efficacy of the chemotherapeutic concoction. However, even without this explanation of efficacy, the explanation in terms of the analgesic functional powers of the concoction is an explanation. Jackson and Pettit call such explanations ‘program explanations’ because, though at an abstract level, they ‘program for’ the micro-level explanation which brings out the causal efficacy of the chemical properties of the concoction (cf. Jackson and Pettit 2004e, f).

The important point about such programming explanations is that they aren’t excluded by the micro-level explanation. Indeed, argue Jackson and Pettit, the micro-level explanations which bring out the causal efficacy of the causally relevant item described in functional level, more abstract terms, include these functional properties. For instance, to say that such-and-such a chemical structure in the concoction causes certain changes in certain cells along the nerve pathways between some joint and the brain, preventing the transmission of certain electrical signals to the brain, is to include the analgesic effect of the chemotherapeutic concoction on the pain of the arthritis. The items referred to by the functional explanation are not rejected but are included in the micro-level explanation.

Similarly, the functional role played by the internal states of the brain which are belief states is not ‘disappeared’ by the discovery of a detailed aetiology of the production of bodily movements involved in behaviour. That detailed aetiological story of the production of bodily movement, told in terms of the brain state’s neuro-physiological properties, just as much includes the causal role played by the belief-cum-content in causing behaviour as does the detailed biochemical explanation of the efficacy of the chemotherapeutic concoction, chemically described, on certain cell structures (those of arthritic joints), which includes the concoction’s causal role as an analgesic for arthritis pain. Just as it is a mistake to think that the analgesic effect is an addition to the biochemical aetiological story, it is a mistake to think that the belief-cum-content is an addition to the neuro-physiological causal story. The posits and the properties of program explanations are not excluded by aetiological explanations; they are part of them. So, commonsense explanations of behaviour are safe, in principle, from progress in neuroscientific explanations of bodily movement. Though beliefs and their content themselves may not be causally efficacious, they are casually relevant, as is brought out by the program explanation at the more abstract, functional level.

Further, there are good reasons for taking program explanations as important for bringing out various relations of the causally efficacious states to factors in their environment and to behaviour. Consider, for instance, Lewis’ example of a metal whose microstructure is the categorical base of its optical opacity, its thermal conductivity and its electrical conductivity. Intrinsically, the metal’s microstructure is what it is regardless of whether it is put in a field of electric potential difference, between two regions of different temperature, or between a light source and a receiver. Yet, this same intrinsic microstructure provides the stuff of the detailed micro-causal explanation of any light absorption, any electrical current, or any heat transfer that the metal becomes involved in. Intrinsically, the metal is what it is. But electrical conductivity, opacity and thermal conductivity are different dispositions of the metal. Further, it is importantly different to explain the death of someone due to the metal in terms of its opacity, its electrical conductivity or its thermal conductivity. If it is the metal’s opacity that caused the person’s death, the person did not see something behind the metal which turned out to kill her. If it was the electrical conductivity that was relevant, her death was by electrocution. If it was the thermal conductivity, she was burnt to death. Even if the properties cited in them are not causally efficacious, program explanations pick out different causal powers as the relevant ones by specifying what happens in terms of the relevant inputs and outputs. So, program explanations are not rendered redundant by the explanations for which they program. The same point applies to explanations of behaviour in terms of beliefs and their contents, in relation to neuro-physiological explanations of bodily movements. As we saw in discussing Jackson’s idea of minimal functionalism, content ascription is a way of describing the relations of behaviour caused by internal states in terms of inputs and other internal states: the content of a belief is the way the world would have to be for the behaviour caused by the state to make the world to be the way the subject desires it to be (Jackson 1996, p. 401; Braddon-Mitchell and Jackson 1996, p. 182ff.). Thus, the belief that there is a beer in the back fridge is causally relevant to the explanation of the direction of Joe’s movement, for that movement is what will contribute to causing the slaking of Joe’s thirst by drinking beer. Had Joe drunk some lemonade from the back fridge, his belief would have to have been that there is lemonade there. Of course, there is a problem here to which I will return shortly—how to distinguish which of two beliefs that could explain behaviour actually does. For now, it suffices that content ascription relates inputs and outputs and other internal states abstracting from aetiological details and thereby picks out functional roles that are played in the production of behaviour within a subject. These functional properties are not excluded by detailed micro-explanations in neuro-physiological terms but are included as parts of them; nor are they rendered redundant by these detailed aetiologies for they (these broad-content relations of the internal states) serve to locate the internal story of the aetiologies against their environmental causal background and history and so on, which helps to capture important aspects of the causal goings-on in the world—what beliefs and desires are involved, what their content is and so on.

However, Jackson is concerned that this concession that content is causally relevant though not causally efficacious in a complete aetiology of bodily movements may not be enough to save commonsense functionalism. It may involve too much repudiation of commonsense’s ‘here-and-now’ intuition. This is the intuition that it takes some change in Joe to get him moving towards the fridge out the back. In contrast, broad-content ascriptions and the explanations going with them involve setting Joe into various relations with his environment, and though the explanations built on these relations to the environment program for detailed internal state aetiological stories in terms of neurophysiology, this notion of causal relevance may not suffice to capture the thought that it is precisely a change in Joe which literally causes his movement towards the fridge out the back. For the change in Joe could be the same and produce the same bodily movements (and raw behaviours) even though the environment to which he is related, and the semantic properties consequent upon these relations, be very different (e.g., if a Twin Earth was done on Joe). The point is clearer perhaps in terms of Joe and his doppelganger. Joe and his doppelganger will undergo the same changes internally and their bodily movements be the same though their broad semantic environments differ radically. Jackson, therefore, holds out for the possibility of a notion of content that is narrow and truth-conditional. This way, internal states with truth-conditional, representational properties are shared by a subject and her doppelganger. This intentional content is narrow because it is shared with the doppelganger. It is truth-conditional since content explains behaviour because the content is taken to be true or because her desire is for what is desired to become true. Whether this narrow content is causally efficacious or not depends on how close to the categorical base of belief it is. Jackson writes that, ‘if we can identify this narrow local fact with a relatively intrinsic feature, presumably a neurological property of some kind, or something of a more structural kind that supervenes on the neurology, then the narrow content that underlies broad content will be causally efficacious; if we cannot, then the narrow content will be causally relevant without being efficacious, but at least it will be local’ (Jackson 1996, p. 403).

There are two points here. The first is that this notion of narrow content substantially deflates the significance of broad content. For Jackson suggests that the combination of an internal state’s narrow truth-conditional content and its causal and certain other relations to the environment determine broad content. As Jackson notes, this is controversial and requires a theory of narrow, truth-conditional content to show that it is possible. Inferential role semantics is a possible model here. (See Braddon-Mitchell and Jackson 1996, Chap. 12, for Jackson’s reasons for this deflationary account of broad content).

For my purposes, however, it is a second point that is more relevant to Jackson’s thinking about the causal efficacy of content. This is that the local, narrow fact on which narrow content, if it is to be causally efficacious, supervenes is either one about a neurological property or a more structural fact supervening on the neurology. Obviously, this is a somewhat speculative matter. Be that as it may, Jackson’s support for the map metaphor of mental representation against the language of thought hypothesis seems important to note in this connection.

Jackson generally discusses the map metaphor for mental representation by introducing it to explore what implications it has for mental causation, should it turn out that brain states represent ways the world is or could be in a manner more like a map than like the sentences of a language (Jackson 1996, p. 403; Jackson 1998d, p. 113). In another place, he (and Braddon-Mitchell) argues for the advantages of the map metaphor over the language of thought approach though presents the map proposal as neutral on the neurophysiology of representation in the brain (Braddon-Mitchell and Jackson 1996, p. 188). I take it that there is some ground clearing necessary to defend the possibility of map-like mental representation as opposed to anything else. The question as to whether mental representation is map-like or language-like (or something else if there is mental representation as conceived in these debates) is an empirical one. In the ground-clearing connection, Jackson argues that maps, holograms and such are systematic and productive representational systems. Thus far, they fare no worse than language of thought views. They are systematic and productive because they are structured representational systems. But it is not clear that proponents of map-like representation are neutral about the neurophysiology of representation in the brain. Nor do Braddon-Mitchell and Jackson argue in their 1996 book in such a way as to retain neutrality about ‘the neurophysiology of representation in the brain’. For they argue for the advantages of the map metaphor over the sentences-in-a-language metaphor. Further, I would have thought that the map-like representation hypothesis fits very naturally with the idea of narrow, truth-conditional content supervening on neuro-physiological structure. That is, if mental representational content is (a) narrow, (b) truth-conditional and (c) causally efficacious, Jackson can quite naturally put the map-like representation hypothesis to work. For (a) maps are structural representations and as such are at least candidates for supervening on structural neurological properties. So, they should be systems of narrow representational content. (b) They can have content assigned them in the usual minimal functionalist way—that is, in terms of inputs and behavioural outputs—‘we match head states that are beliefs with possible states of the world by the rule that each state of the head gets assigned the possible state of the world which is such that if it were the way things actually are, the behaviour that head state causes would realise what the subject desires’ (Braddon-Mitchell and Jackson 1996, p. 181). As such, individual beliefs fall out of systems of belief in the light of input-behaviour specifications assigning content first to the belief system, holistically (Jackson 1998d, p. 117ff.). The model here is the metal which is opaque, thermally conductive and electrically conductive even though the intrinsic nature of the metal does not distinguish between these functional properties as such. In light of this, the possibility seems clear that mental representational content is truth-conditional given the right correspondence relation between internal states and possible states of the world. And (c) if the right kind of relation of supervenience is found to exist between such map-like representational systems and the structural neuro-physiological properties of the brain states, presumably the mental states can be not only causally relevant but also causally efficacious.

Is Jackson’s view of content and mental causation ‘Dennettish’? Kind of. Individual beliefs are, as it were, ontologically consequent upon content assignments to the system of belief in the light of input-behaviour connections and intentional systems theory (on which cf. Braddon-Mitchell and Jackson 1996, p. 145ff.). Is this incompatible with realism about mental causation? Apparently not. For just as the dispositions of opacity, electrical conductivity and thermal conductivity are real features of metal, different from each other though all based on the single molecular structure and so on of the metal, their different linkages between inputs and outputs are not just a matter of pretending that the metal is opaque, electrically conductive and so on. Similarly, distinguishing individual beliefs from each other in the light of different input–output linkages attributed to a holistic intentional system is not a matter of pretending that human beings believe that, for instance, there is a beer in the fridge out the back, merely for predictive purposes but not taking this seriously as a causal explanation of what the subject does. For on Jackson’s picture, there are individual contents incorporated into the intentional, representational, map-like system in Joe’s head, even if talk of individual beliefs is something of a hypostatisation. Is there an element of instrumentalism about mental states in this? To be sure. It is an instrumentalism about individual beliefs and desires, for they are not ‘basic’ in the way whole systems of intentional content are—there is an element of ‘reporting relativity’ involved in the characterisation of individual belief contents, and this characterisation is keyed to the description and prediction of behaviour in ways that some intentionalist realists will not be entirely comfortable about. Is this an objection to Jackson? I doubt it.

Philosophy of Religion

Scientific Theism: Peter Forrest

In his 1996 book, God without the Supernatural: A Defense of Scientific Theism, Peter Forrest seeks to defend a version of the thesis that God exists. His is a distinctive line of thought because it seeks to steer a course between the metaphysical naturalism that is pretty much orthodoxy in Australian analytic philosophy and the more traditional religious understanding that God is a supernatural being transcending nature in ways inaccessible from a scientific outlook on the world and from a perspective furnishable from precedents in our common experience of the world. He characterises metaphysical naturalism as the view that nothing exists except what is that for which well-confirmed science provides precedent. Supernaturalism, on the other hand, posits things and properties which neither science nor common experience provides precedent. The concept of God which he defends is one of a personal God, who creates the natural order and does so for the sake, at least in part, of ourselves in it—and for the sake of ourselves in it not just as so many elements of an aggregate whose good is to be maximised, but for us as individuals whose good is a target of divine concern. Forrest’s theism is, then, an anthropic theism. While he is prepared from within this theistic view to posit things or properties for which science provides no precedent (common ground with supernaturalism), he will not posit anything for which no precedent can be found in common experience (departing company from traditional supernatural religion).

The argument for this theism is, in the broad, a kind of abductive one. Forrest styles it ‘inference to the best explanation’ apologetics. Three obligations are incurred by adopting this approach. First, it has to be shown that anthropic theism is a better explanation over a wide(r) range of phenomena than naturalistic explanation; second, it has to be shown that the hypothesis that God exists is epistemically possible, that the probability is not too low that God exists prior to the consideration of the explanatory power of the hypothesis; and third, it has to be shown that the strongest objection to it is not so strong that, despite its explanatory power, we are not entitled to it. Once all these debts are discharged, it can be concluded that anthropic theism is a better hypothesis to believe than naturalistic atheism. In this way, anthropic theism is warranted and atheistic naturalism is not. However, the possibility has to be allowed for that an even better hypothesis than anthropic theism might exist. To the extent that it is, anthropic theism is not warranted to the exclusion of agnosticism, based on the possibility of a better hypothesis. Hence, both agnosticism and anthropic theism are warranted on intellectual grounds. On the other hand, since there is value in commitment on important matters such as the existence of God, Forrest offers other, non-theoretical, reasons for being an anthropic theist rather than an agnostic.

Such an argument is clearly ambitious within the scope of a single work. Explananda for which Forrest defends anthropic theism as a good explanation include the following: the fitness of the universe for life, the regularity of the laws of nature, our capacity for intellectual progress, the overridingness of moral obligations, beauty and the serendipity of mathematics. Then, each of these involves its own subsidiary topics. Further, Forrest’s strategy requires him to enter the lists of the philosophy of mind as he characterises the God for which he argues in terms of consciousness: God is ‘unrestricted consciousness’. Since consciousness here is about the appearances of things, all the usual conundra arise about the accommodation of qualia and so on in the physicalist-cum-functionalist picture of the mind. Then, even if all this is made good, Forrest has to fend off the objection to accepting the theistic hypothesis from the problem of evil. This is no small philosophical garden just on its own. So it is very clear that the argument of the work is very ambitious.

I think it has to be owned that in such an ambitious project, a trade-off is inevitable between the philosophical conscience on matters of detailed argument on narrowly focused topics and achieving the kind of broad vision of the sweep of the argument that this kind of project requires. To his great credit, Forrest’s discussions of topics consistently introduce thoughtful observations, insightful ideas, carefully trod paths between opposed views, creative suggestions and so on. Many problems are offered a variety of anti-supernaturalist theistic conceptions (for instance, there are five suggestions for what the anti-supernaturalist might think the afterlife is like: Forrest 1996a, Chap. 2, §4). Further, he brings to bear on familiar topics in philosophy of religion helpful insights from philosophy of science as well as general metaphysics and epistemology. In this it is arguable that compared to, say, Swinburne’s comparably ambitious project over his entire career, Forrest’s book genuinely gives the naturalist approach an open-hearted hearing. However, the trade-off will be inevitable for such an ambitious project as Forrest’s. Many will read the back-and-forth of arguments, the numerous theses, principles and speculations and feel that many important topics have been treated too cursorily or that particular theses have been dismissed too peremptorily or their resources for meeting challenges taken with inadequate attention and so on. For in each of the subordinate topics he broaches (e.g., the explananda mentioned above), the best explanation epistemology raises a complex set of questions and responsibilities concerning theories rival to his own. These sets of questions and responsibilities will be parallel to those for which he is responsible, in the large, if he is to defend anthropic theism at the most general level of the overarching issue of God’s existence or nonexistence. If, as it has always been a virtue of analytic philosophy to insist, ‘God is in the details’ (pun intended), it cannot be granted that Forrest convincingly achieves what he sets out to. Despite the wealth of ideas, the frequent subtlety of argument and the width of argumentative mastery, the ambition and the epistemology pull in opposite directions. There is just too much debate about each of the topics addressed in the book for anthropic theism to achieve a defence in the comprehensive way Forrest seeks here.

However, an equally critical problem is the force of the conclusion if it does eventually, given the details, work. For there are (at least) two quite central features of Forrest’s theism that deserve attention. One is the objective though non-objectual reading of God’s existence that he prefers. For there is no entity or object which is the God to which we conclude from this argument: better to say ‘there is a God’ than that ‘God exists’ (Forrest 1996a, p. 209f.). God is unrestricted consciousness, and consciousness is the (intransitive) appearing of things. Being unrestricted, God is the consciousness of all things—God is the appearing of all things. Although the fact that things appear is an objective fact, God is not an entity in any sense involving the ‘collecting’ of all consciousness (or appearing) in a unifying self. Nevertheless, it is an objective fact that God exists.

There are important questions this thesis raises, many of which Forrest addresses. However, many he does not address, such as: how is one to reinterpret more traditional claims such as that God is an agent? Or that God is a necessary being? These more traditional formulae need reformulation. Further, Forrest admits that the non-objectual view he advocates is ‘peculiar enough’ (p. 209) but, in a very brief discussion, suggests that the peculiarity is removed by observing that emotional states (such as devotion to God) take propositions as content rather than persons or other objects and that it is possible to have a relationship with such a God in near enough to ‘the popular sense’ without strictly being in a relation to God (logically). In the abstract, the logical point might be fair enough. In this logical context, the point of such a propositional (de dicto) rendering of the emotions of relationships as Forrest suggests seems to be to avoid rendering talk about God transparent to quantification. Forrest’s preference for ‘there is a God’ (i.e., (∃x) Gx) over ‘God exists’ (i.e., (∃x) x = g) suggests something like this, as only the latter strictly quantifies over God in the domain of quantification. Moreover, since one can worship non-existent objects, reading a context such as ‘s worships g’ de dicto closes it off from existential generalisation on ‘g’ (i.e., it blocks inferring from ‘W(s, g)’ to ‘(∃x) W(s, x)’) and so avoids treating God as a member of the domain of quantification, in the way a de re reading would permit (which would permit inferring from ‘R(s, g)’ to ‘(∃x) R(s, x)’). Alternatively, or perhaps complementarily, Forrest may be alluding to a non-objectual interpretation of quantification over God, and this may fit naturally with God conceived as the appearing of all things in some way.

Whatever the correct interpretation is, the non-objectual character of God raises some questions. First, does this mean that earlier speculations about God as ‘the unified consciousness to which all things appear’ (p. 177) are façons de parler? On Forrest’s view, it seems that one does not love God or stand in right relation to her in the grace of the sacraments or when convicted in guilt by her in her unfathomable goodness (these sound like relations to me). Rather, one is lovingly mindful that there is the forgiveness of God, that there is the divine grace available in the sacraments, or the like. To my religious ear, this sounds like a parody of religious devotion. To be sure, as Forrest says, when one fears death there isn’t a thing, death, as it were, ‘out there before one’ which one fears; rather, one fears (say) that one will be dead soon (p. 210). But that is because of ‘the sort of thing’ death is. We all know that. In contrast, in loving relationships, one’s love isn’t just possessed of loving thoughts that, say, one’s son is a good man and he makes one proud. It is a relationship with him. One is confronted by him, and (for want of better language for it—substance language doesn’t help here either) one’s love for him opens one out to him precisely in his otherness. In mature relationships with one’s son, when you say you love him and are proud of him, and you mean this, he knows you mean it. When you don’t mean it and are just being conventional or not rising to the challenge of honesty in the relationship, he can’t be conned. There is real encounter here that gets below the shields we put up in ordinary social interactions—real encounter in the knowing recognition by him of one’s pride and love and by oneself in seeing that he recognises it. One is in a relationship with this man. The issue doesn’t primarily concern logical form, but it is not without implications for logical form. Or rather, we can recast bits of natural language in different logical forms till the cows come home. If we are talking about ontological categories in something like Aristotle’s sense, it should be recalled that we can quantify over and render transparent to inferences of existential exportation not only ‘objects’, but events, processes, non-existent Meingongiana, first-order properties, states of affairs, things expressed by lambda or epsilon operators… whatever we like or not—in which case we represent apparent natural language quantifications over such things very differently without quantification. Our view of the metaphysics (or ontology) of the domain under investigation rules on such matters.

Arguably, Forrest’s de dicto reading of religious attitudes fails to carry conviction. For, at the very least, the appearings of a totality of things, uncollected by some sort of focal Other, lack the kind of separate, unified, self-consciousness that eludes one but that is capable of moments of truth for oneself, putting one on notice and the like. Appearings, and so consciousness, even though they escape materialist explanation, are nevertheless epiphenomenal and so not part of the causal story about what happens. But a wife could not have ‘such fury as a woman’s scorn’ were she not an other, like I am trying to evince. But such is God. It seems to me that religious talk about the nature of devotion and relationship with God is not doable entirely propositionally, de dicto, with the quantificational (or referential) opacity apparently intended. Forrest’s God might be epistemically possible, but it isn’t clear such a God is worth having. Or, should I say, it isn’t clear that all things appear and there is unrestricted appearing which is worth taking as the content of a propositional pro-attitude of devotion, taken de dicto? One might hope that more of the religious phenomena were saved.

All this said, however, Forrest’s work is of great value for its resolute relocation of the conduct of religious apologetics in an abductive epistemology which takes into account the broad canvass of issues and human experience as well as the power and possibilities of contemporary naturalism. Religious sensibility, even if in ‘good order’ as a type of form of life, cannot stay where it used to be given these two factors.

As a transition to the next section, I will outline the debate between Peter Forrest and Mark Wynn concerning the anti-supernaturalism of Forrest’s project.

I said above that there are (at least) two features of Forrest’s project which deserve attention. The second feature is that his overarching aim is to get a theism which takes us beyond naturalism though without supernaturalism. Wynn questions whether Forrest’s anti-supernaturalist theism is ultimately stable, given his accounts of both divine ineffability and the nature of emergent orders of things in the world (cf. Bishop 1999, p. 107).⁶ I will only treat the ineffability point. The anti-supernaturalism which Forrest defends is the idea that we should not believe in entities (a) for which there is no precedent either in science or in the familiar and (b) whose operation violates the laws of nature (Forrest 1996a, p. 2ff.). He justifies this by arguing that there is an Ockhamist presumption in favour of positing only entities of kinds we are familiar with or entities, even if unfamiliar, which are required posits of well-confirmed scientific theory. The latter part of this presumption is weak if a posit is precedented in the familiar; it is strong if the posit is not precedented in the familiar (Forrest 1996a, p. 4).⁷ The intellectual cost of positing such an unprecedented entity may be balanced if the explanatory gain is high. Anti-supernaturalism, then, must argue that for any highly explanatory supernaturalist explanation, there is a better, anti-supernaturalist, perhaps even naturalist, one (Forrest 2001, p. 101).

Wynn’s argument can be put quite briefly. In relation to divine ineffability, Forrest wants to maintain some version of the idea that we are unable to know the divine nature, even though we may know some things about God. His acceptance of this modest ineffability thesis arises from the concern that the requirement of precedence in either science or common experience could undermine divine transcendence. Yet, accepting our ignorance of the divine nature means that the divine nature is at least epistemically unprecedented. Forrest uses various gambits to clarify how his views work in relation to this thesis. What is clear is that the characterisation of God as unrestricted consciousness is speculative. This speculation is coherent with the anti-supernaturalism to no small extent (though one could urge that the disembodiment of the divine consciousness is unprecedented). An alternative is the supernaturalist’s suggestion that a more plausible hypothesis regarding our ignorance of the divine nature is that God is unprecedented in our world. Wynn’s main point, however, is that if, as it seems Forrest concedes, we cannot identify the precedents of the divine nature or, once they are identified, are unable to trace the extension from the familiar to the divine, the difference—i.e., the difference between the supernaturalist’s insistence that the divine nature is unprecedented and Forrest’s acceptance that we do not know the divine nature even if we are warranted in his unrestricted consciousness hypothesis—becomes vanishingly small. In reply, Forrest concedes the point that the ineffability he accepts and his position of anti-supernaturalism are in tension. He moves to treating anti-supernaturalism as methodological rather than as a substantive constraint on our conception of God. He suggests two reasons for this: one, respect for the theistic tradition’s view that we cannot know the divine nature; second, epistemic modesty: it is presumptuous to assume God to be the kind of entity of which there is not much we are ignorant (Forrest 2001, p. 102). On the other hand, one might have reservations about how important the distinction is between a methodological principle and a substantive constraint on a concept of something.

Other Arguments About God

Wynn and the Design Argument

This is a good opportunity to introduce another important contribution to Australasian philosophy of religion in the 1990s: the work of Mark Wynn. Wynn attempts a revival of the design argument along the lines of the abductive epistemology that Forrest develops.

Wynn’s basic endeavour is to revitalise the design argument by means of two moves. The first is to situate it as a tenable hypothesis for some, as a better explanation of certain features of the universe—in particular, the pervasive beauty of the world and the universe’s propensity for richer and more complex forms of material organisation—than naturalism. The second is to formulate his design arguments or hypotheses so as to bring out this emphasis on the goodness of the world. For the central idea of his arguments is that the world exists because it is good that it exist. Hence, God as designer is argued to be a better explanation of the goodness of the world than naturalism.

In connection with the first move, his book God and Goodness (1999a) has four sections. The first spells out the explanatory superiority of the theistic design hypothesis over its naturalist rival. The second considers the force of disvalue in the world as evidence against the design hypothesis and argues that it is not decisively against it. The third part takes it that the argument so far cannot be claimed to show that the hypothesis that the world exists because it is good that it does is more likely true than not. So instead, in this part, Wynn seeks to offer certain moral considerations to show that the balance of probabilities is not clearly against the hypothesis. Central here is his argument that trust in the basic goodness of the world is reasonable, taking moral considerations about trust into account. The final section takes up Wynn’s concern to connect the concept of God to which design-from-goodness arguments give rise to themes in religious life such as worship and salvation. The central thrust here is to articulate a concept of God which extends the conception of him as good from the first three parts and as a religiously adequate object of devotion and ultimate commitment.

The second move—the arguments that design hypotheses are better explanations of goodness in the world—is crucial. Without them, the design hypothesis does not pull its weight. The first consideration concerns the pervasive beauty in the world and builds on an argument of F.R. Tennant’s 1930 Philosophical Theology. The basic thought is that nature’s capacity pervasively to induce in us positive aesthetic responses is better explained if there is a God who makes the world to be this way pervasively beautiful than it is on the assumption that such aesthetic responsiveness is caused by selective mechanisms in evolutionary history (the biophilia hypothesis). Moreover, the universe’s being such that it develops towards richer and more complex material structures is better explained by the design hypothesis than by naturalistic explanations. Wynn seeks also to bolster this inference with a design hypothesis to explain the suitability of the universe for life.

The biophilia hypothesis claims that preferences for biodiversity and natural environments are likely or expected to be coded for in the genes as we evolved in settings with many species and in natural environments. Wynn carefully discusses various empirical works which purport to support and extend this line of thought. I would note that one does not have to be impressed by such attempts to explain our aesthetic preferences (for savannah, no less, if some of this empirical work is to be believed). In relation to the emergence of richer and complex material structures in the universe, Wynn is careful to state his target: he seeks to explain why it is that by the operation of natural laws the universe develops towards richer and more complex structures, when, as it were, for all that the operation of laws of nature requires such developments would have no significance.

Further, Wynn argues that the design hypothesis is a better explanation of the goldilocks facts about the universe that make it suitable for life, since naturalism either has no explanation at all or, with the multiple universes hypothesis, is more extravagant and less clearly explanatory. I find his arguments here less than convincing overall. I will start with the last claim, regarding the multiverse hypothesis. The goldilocks universe is one of the many, many possibilities for the way the Big Bang would develop. I take it that part of the force of trying to think of the physics of the initial singularity from which the Big Bang emerged in quantum-mechanical cosmological terms is to represent this idea. Roughly, there is a state vector for the Big Bang which incorporates a probability distribution over possible ways the universe could emerge to be from the Big Bang. This does not require multiple universes, it seems to me. What we have is simply a statistical explanation. Individual events in statistical explanations do not have explanations. The requirement for an explanation is rejected by the claim that all there is the state vector of the initial singularity. The universe with our goldilocks facts is one of those events in the space of possibilities over which there is a probability distribution. That we emerged is just chance. It is not that chance fails to explain, as if this itself is a failure and the emergence of our universe from the Big Bang should be explained. It is rather that there is no explanation for this kind of event—it is one possible event in the probability distribution. And that’s all there is to say.

Of course, we have reason to be grateful that this world is pretty good (or as good as it is, indeed a gift) and mostly exquisitely beautiful. But this is not weird or in need of special explanation. That what is good and beautiful is good and beautiful is not at all undermined by its origins in chance or in the brute bumping and grinding of subatomic particles up through long-chain proteins according to the laws of nature. We are what we are even if we descended from apes—but we aren’t apes. Just so, the sunset is exquisite even though the reds and so on are effects of light refraction and our visual systems. The point here seems to weaken the force of Wynn’s arguments for he requires that the God of the design hypothesis act through the causes of nature and their laws. Wynn complains a number of times that naturalism lacks explanations for transitions in the world—from chemicals to life, from life to sentience, from sentience to sapience. But naturalism does not pretend to have such explanations—it is a philosophical thesis and it is science which tries to explain these things nowadays inasmuch as they have explanation. That is what naturalism says.

There is a telling passage on p. 54 of Wynn’s (1999) book, where he describes his approach (as following Holmes Rolston’s) and points out that he is not arguing that God intervenes at these junctures to ‘upgrade’ the chemical soup to being a self-replicating, living system. Instead, he aims to try to explain why nature is such as to develop as it does given that, as far as the laws of nature are concerned, there is no necessity that life, sentience, etc., should develop. However, this God does not act so as to intervene and override the laws but acts through them: ‘there must be some naturalistic account of the tendency of the world to act as a value generator, since this tendency is deep-seated, and accordingly, somehow written into the material order itself’. So Wynn’s approach ‘is not committed to the insufficiency of scientific explanations of these matters’. Yet, within a few lines he goes on to say: ‘Rolston’s [and Wynn’s] view invites us to suppose that scientific explanations of natural value generation can take what forms they may, providing we recognise the insufficiency of Darwinian accounts’ (p. 54). But the sufficiency of Darwinian accounts, give or take some details of evolutionary biology, is exactly what Wynn’s approach is committed to if he is committed to scientific explanations and God’s working through them, not against them. One has to fish or cut bait here. For once we have the explanations from these scientific principles, we understand how goodness and beauty came about.

Now, that is not to say that goodness and beauty somehow ‘aren’t there’, or are not objective, or are ‘nothing but’ electromagnetic radiation displaying the effects of scattering through the atmosphere. ‘Objective’ is said in many ways. Goodness and beauty do not have to be ‘reducible’ to be ‘out there’. And they can arise from natural causes and be what they are—just as we arise from the apes but are what we are. So there is more to say about goodness, beauty and even the holy, after we have given naturalistic explanations of them. For goodness is not an illusion, nor is beauty, nor, I think, is holiness and the sacred. We need to describe them and limn their structures from within the perspective they furnish us on the world. But natural explanations are enough explanation of their origins. God does not seem to add further intelligibility to their explanation. It seems to me, her role, if she exists, is different. Again, this is not to say that there is not value in the kind of natural theology Wynn and others do. But that value arises after one believes in God: God casts a light to see the natural order in. But she does not add lumens to the light nature casts on itself so that we see it.

Oppy and Ontological Arguments

The ontological argument received detailed and very thorough attention from Graham Oppy during our period. He presents his 1995 work, Ontological Arguments and Belief in God, as an instalment in a large project of expounding and defending agnosticism on religious questions. In this book he argues that ontological arguments do not provide agnostics with good reason to change their view about God’s existence. His view is therefore negative about the value or persuasiveness of ontological arguments.

Ontological Arguments and Belief in God consists of 12 chapters of 199 pages and a 136-page appendix of ‘Literature Notes’ mainly discussing in some detail recent literature on ontological arguments. In Chap. 1, Oppy provides an historical overview of the argument and its detractors, covering Anselm, Descartes, Leibniz, Hume, Kant and the positivists. Oppy distinguishes six kinds of ontological argument: definitional, conceptual, modal, Meinongian, experiential and ‘Hegelian’. The distinction is based to some extent on the character of the crucial moves in the argument but also, in the main, on Oppy’s diagnosis of their characteristic weaknesses. Each kind of ontological argument is given a chapter. There follow chapters on applying the fruits of these discussions to the arguments of Anselm and Descartes, including a characterisation of a general objection to ontological arguments (Chap. 8); on an examination of miscellaneous alleged general objections to ontological arguments (Chap. 9); on whether existence is a predicate (10); on parodies of the ontological argument (11); and on whether there might be other uses to which ontological arguments might be put. Oppy formulates his conclusion; thus, (i) ‘I conclude that there are no ontological arguments that provide me with a good reason to believe that God…exists’ (Oppy 1995, p. 198); (ii) ‘I also conclude that there are no ontological arguments that provide me with a good reason to believe that God…does not exist’ (1995, p. 198); (iii) ‘I conclude that there are perfectly general grounds on which I can dismiss the possibility of a dialectically effective ontological argument’ (1995, p. 198); and (iv) ‘I conclude that there is no other use to which ontological argument can be adapted by theists or atheists: Given their dialectical impotence, there is no other purpose that they can successfully achieve. Thus I conclude that ontological arguments are completely worthless…’ (1995, p. 199). These different conclusions are based on different aspects of the discussion in the book.

The basic problem, according to Oppy (in §8.4), from which ontological arguments suffer is this. The terms (names: ‘God’; definite descriptions: ‘the maximally perfect being’, ‘that than which no greater being can be conceived’, etc.) which occur in such arguments range over the collection of things that are supposed to include God, and either do or do not occur within the scope of a sentential operator such as ‘according to the definition…’ or ‘according to the theistic story…’. Where the argument occurs within the scope of such an operator, the non-theist can agree that the conclusion (‘according to the definition of “God”, God exists’) follows from the definition (‘according to the definition, God is that than which none greater can be conceived’). However, she is not thereby given reason to move from her non-theism or to think that her non-theism has to this point been irrational. For one can accept that something exists according to its definition or in some story about the universe, without accepting that there is anything that satisfies the definition or that that story about the universe is true. Of course, the theist wants more of this argument: they want to ‘detach’ the proposition that God exists from the operator ‘according to the definition’ and assert it free of such a limitation. But if the operator is intensional, this kind of detachment will not be valid. However, if the operator is extensional, the non-theist can complain that the argument begs the question on the point at issue between her and the theist—that is, whether there is a thing satisfying the definition (…the story is to be believed…). On the other hand, if the terms do not occur within the scope of such a sentential operator, the situation is similar. The non-theist can resist the thought that she has to accept the unqualified premise of the argument.

It is rather a cavil to complain that occasionally Oppy gives too short a shrift to some moves that the theist could make so that the ontological argument could fare better under criticism. The book is a wealth of highly detailed, sophisticated and fair discussion of as wide a variety of ontological arguments as has ever been collected. This comprehensiveness alone assures the study its place as the starting point for future examination of ontological arguments. One may have reservations about its account of begging the question in the dialectic of debates over the argument (1995, p. 53ff.). One may have the view that there is more value to be wrung from ontological arguments for theism than Oppy takes seriously (cf. Gale 1998, p. 717; Taliaferro 1997, p. 553f.). One might have the view that there is a version of the argument which Oppy does not take seriously enough and which escapes the basic kind of critique (Wierenga 1998, p. 164). Or one may not find Oppy’s epistemological sympathies convincing.⁸ But such reservations about Oppy’s argument do not undermine its fundamental achievement.

Pantheism

Lest it be thought that theism was the only doctrine of God which received attention in our period, we note the defence of pantheism propounded by Michael Levine and further discussions of pantheism by Oppy and Forrest later in the decade.

Levine laments the focus on Judeo-Christian theism in English-speaking analytic philosophy of religion. In Pantheism: A Non-theistic Concept of Deity (1994), he seeks to defend a doctrine of God which represents a genuine alternative to this concentration on theism (see Levine 2000, in which he argues that contemporary Christian analytic philosophy of religion ‘lacks vitality, relevance and “seriousness”’). He seeks to develop a doctrine of pantheism which is defined in an open enough way to include thinkers whom he takes to be paradigmatic pantheists, such as Spinoza, Walt Whitman and Lao-tse. However, he also takes Hegel, Plotinus, some pre-Socratics, John Scotus Eriugena, Giordano Bruno and Shankara to be more or less pantheistic in their thinking. Quite some burden of his discussion is to distinguish pantheism from confusions about it and to make clear its difference from atheism and theism (1994: Part I). To this end, he delineates a form of pantheism in which the existence of God is insisted upon, thus distinguishing it from atheism (cf. Spinoza’s treatment as an atheist by many theistic critics of the seventeenth century). Further, it is argued that God is not to be understood as a being with personal attributes, thus distinguishing pantheism from theism as well as panentheism (1994, pp. 11–12, 148; panentheism being the view of Forrest 1996a, pp. 202, 233ff.). Rather, God is the (impersonal) all-encompassing Unity of the world. The universe, in its unity, has a divine character (‘character’ here having nothing to do with virtue). This unity does not involve any monism, as it might if the universe is taken to be a single substance (1994: §2.3). Instead, the unity of the world is evaluative in nature. This is not to say that God, as the unity of the world, is morally good (1994, p. 316), though it does seem that the unity of the world is the moral good (cf., e.g., 1994, pp. 209, 238, 242).

Levine argues, convincingly, that pantheism is not implied by theism and that it is no claim of pantheism that the world is God’s body. He discusses some ‘problems of pantheism’—creation, evil, ethics and ecology and salvation and immortality (1994, Chap. 4). The pantheist, Levine argues, does not obviously need a doctrine of creation (p. 180), but if he is to have one, it is not unexpected that it would be emmanationist (p. 196). Evil is not a problem for pantheism, as it is for theism. Critics who have suggested that it is more of a problem for pantheism than for theism have assumed that pantheism must be monistic, which is false; they have imported assumptions about divinity which are properly theistic (1994: §4.2). Pantheists may not even choose to use the concept of evil widely at all, preferring to speak of moral right and wrong and perhaps reserving ‘evil’ for particularly egregious wrongs (pp. 197f.). Pantheists, like theists, will tend to be ‘moral realists’ in metaethics (p. 219). Also, pantheism can be the basis of a more adequate and much needed ecological ethics which is in no way anthropocentric (1994, §4.3). Pantheists, further, reject the notion of personal immortality (pp. 242, 248ff.). Achievement of the human good involves ‘a certain kind of relation to the Unity in order to live “properly”’ (p. 242).

Of particular importance is Levine’s aim to bring out the religious, moral, and spiritual significance of pantheism as an alternative to theism and not merely leave it as an abstract intellectual hypothesis about God or nature. Since divine Unity is non-personal, worship is not appropriate (1994, p. 362). Prayer is particularly misplaced (p. 315), though a respect, reverence, or honour towards divine Unity is apt, as found in, for instance, devotion to the universe, nature observation, or art (p. 314). Pantheist practice is aimed at establishing ‘right relation’ to the Unity. The means for this are partly individualistic (pp. 346, 352) but generally will be ‘contemplative or meditative’ (p. 342). And while achieving the right relation to the Unity involves other people (p. 348), it is no accident that pantheism has not been practiced as a church-style community (pp. 357ff.).

Levine’s discussions of pantheism are developed in a careful dialectic with numerous thinkers and ideas. His aim is not to define a new kind of pantheism but to achieve clarity about it and to locate it as a viable alternative to theism for modern thinking and informed people. He does not argue for his substantive positions as often perhaps as one would like. His careful articulation of a distinctive doctrine of God that avoids certain of theism’s challenges, while showing that the pantheistic God can take up quite a few of the theistic God’s roles while simply abandoning others not worth having played, is of the first importance. Levine’s work requires serious engagement and debate. Such serious attention has been paid by Forrest (1997) and Oppy (1997), but space precludes discussion of it.

Morality and Method

The 1990s might be thought of as the decade of ethics. For during this period, the field of ‘applied ethics’ became a distinct area of work for philosophers or a specialism with a somewhat complex relation to philosophy but typically involving philosophers. New ‘centres of ethics’ grew up and became involved in the professional moral cogitations of various professions, notably medicine and other healthcare-related professions. But also, teaching, research and consulting work with other professions were to be found. In addition to the (pre-existing) Centre for Human Bioethics at Monash University, we find the Plunkett Centre for Ethics in Healthcare at Australian Catholic University, with support from St. Vincent’s Hospital in Sydney; the St. James Ethics Centre, working with the business (and later the sporting) community; the Caroline Chisholm Centre for Ethics; the Southern Cross Centre for Ethics in Adelaide; and the Centre for Philosophy and Public Issues in Melbourne (which later developed into CAPPE, the Centre for Applied Philosophy and Public Ethics, funded in 2000 as an ARC Special Research Centre with branches at the University of Melbourne and in Canberra). Philosophers were the directors of all these centres. The mission of each centre involved addressing some area of particular ethical pressure in the wider community, and the centres put philosophers into fruitful and challenging dialogue with the professions and other groups. Philosophers such as Phillip Pettit and Bernadette Tobin worked on the Australian Health Ethics Committee’s various subcommittees on developing important statements of policy and protocol for the handling of sensitive ethical topics in the delivery of healthcare. Others, such as Max Charlesworth, worked with the NHMRC on topics such as resource allocation and its challenges under the impetus of downward pressures on healthcare spending by government. Such work became more mainstream in the profession. Some philosophy PhD graduates went on to work as clinical ethicists in hospitals around Australia, and the phenomenon became much more widespread of individuals who already were qualified in various professions—but especially medicine—returning to university to take philosophy, and particularly ethics, courses to become qualified participants in the ethical debates of their professions. Many, such as the neurologist Grant Gillett of Christchurch, became a philosopher in his own right (see Gillett 1992).

This continued a wave of interest among the professions in philosophical ethics that began in the 1980s. It developed to a peak in the 1990s. Many new courses in applied ethics appeared in the listed offerings of departments around Australasia, along with informal courses, intensive courses, and the like. New associations for the debate and study of ethics in the public and professional life of the region were founded. The Australasian Bioethics Association began life in the early decade, though it has merged since into the Australasian Association of Bioethics and Health Law. The Australian Association of Professional and Applied Ethics was founded at a conference held in Newcastle University in 1993, which had in turn developed out of a conference in Sydney the year before on ‘Teaching Applied Ethics’. Over time, however, the involvement of philosophers in these associations has waned somewhat.

There were, no doubt, various influences behind these developments. One was simply the importance and difficulty of the questions which society was facing after the 1980s ‘greed is good’ mania and the challenges of new biomedical technologies. One might suspect there was something of a wave of ‘ethics-chic’ after the perceived amorality of some public players during the 1980s (Bond, Skase et al.). Particularly, the challenges of new biomedical technologies led to the rise of ethics committees to oversee the ethical quality of biomedical and other research. Additionally, the economic reforms of the 1980s under Labor and the recession of the early decade (which we ‘had to have’, if Keating was right) probably enlivened the sense that ever-escalating public spending had to be halted. Given the political consensus then for smaller government and the intrinsic justification and voter popularity of Australia’s and New Zealand’s systems of healthcare funding, the problems of resource allocation became prominent. Traditional ethical voices on questions like these—the churches, legal opinion leaders and political commentators—were having less traction with the public. Philosophers had already been contributing to academic and popular debate about such social questions, though in small numbers. The call grew for greater public engagement by philosophers and others to act as public intellectuals. Philosophers prominent in this connection were Peter Singer and Raimond Gaita.

Other matters arose during the 1990s which attracted philosophers’ attention: with the Mabo Decision, the push towards Aboriginal reconciliation gained momentum (cf. Australasian Journal of Philosophy 2000); there was the Ormond College sexual harassment case, the matter of Helen Demidenko’s The Hand that Signed the Paper (1994), and so on. Thus, philosophy was beginning to have a greater role in engaging public ethical debates and influencing the ethical thinking of the wider community and of opinion-formers outside academia. Perhaps a less noble, but no less real, influence on these developments was the felt need to make philosophy more relevant to the wider public for reasons of protecting the discipline from the depredations of funding cuts imposed by government, especially after 1995. If a case could be made for the discipline’s value to the public as a contributor to public debate about ethics, social issues, and public policy, there might be less blood drawn by the cuts. Bioethics boomed but so did applied ethics quite generally. One way or another, all these influences, and no doubt others, led to a greater role during the 1990s for philosophy in the public sphere, principally in ethical debate, but not only there (cf. e.g. Malpas and Solomon 1998; Schmidtz and Goodin 1998). There is an interesting fuller history of this aspect of our period to be written.

In what follows, however, I am not going to attempt even the beginnings of this history. Rather, I will summarise three important contributions to the discussion of public ethics. I will then explain two lines of development in metaethics during the period which represent very different conceptions of what philosophy is.

Public Ethics

Max Charlesworth has defended a pluralist, multiculturalist liberalism as a suitable public philosophy to inform decisions needing to be made about bioethical issues in society. In Bioethics in a Liberal Society (1994), he deplores the tendency towards paternalism and undue interference of the state, the medical profession, and others into areas of decision-making which are the proper arena of individual freedom and autonomous choice. Utilitarian thought comes in for severe criticism in this connection. Charlesworth argues that autonomy is fundamental to social ethics because without autonomy informing action, one does not perform ethical acts (1994, p. 11). People should be as free as possible to make their own choices without state, legal, or other interference. Charlesworth discusses death and dying, assisted reproduction, and resource allocation in the light of this fundamental conception. He is not unaware, however, of various challenges to this kind of liberalism. But he argues that his variety of liberalism does not lead to relativism and does not require a kind of autonomy which does not demand of us to live altruistically. However, in his view, the state has no role in making us good or in unduly cutting off options for our autonomous choice. He therefore defends the right of homosexual couples to have access to assisted reproductive technology to form their families, at taxpayer expense. He also argues against the use of quality-adjusted life years, the WHO, and the New Public Health approach, as well as the Oregon experiment of ‘community consultation’ for solving problems of resource allocation—and his main concern with these approaches is that they do nothing for individual autonomy.

A second significant contribution to thinking about public ethics was Robert Goodin’s 1995 Utilitarianism as a Public Philosophy. Goodin’s conception of the place and justification of utilitarianism is quite distinctive. Less a comprehensive moral doctrine about the nature of right and wrong and what is morally relevant, Goodin’s book argues for the justification of utilitarianism precisely as a public philosophy: other things besides the consequences for individuals’ welfare matter in the private life and for ethics as a task for each one of us in living our lives well, but for public policy and public officials, utilitarianism provides the most satisfactory framework. In this argument, Goodin is still concerned to fend off criticisms against utilitarianism and the claims made for its rivals—his particular targets are libertarianism and the deontological emphasis on intentions. However, the fundamental question concerns public policy and the role of the state. Here he defends an important role for the state and the public official in shaping individual conduct and, in particular, protecting individuals from each other and themselves. He therefore argues in favour of certain paternalistic laws such as the requirement on motorcycle riders to wear helmets and the prohibition and control of dangerous substances. Arguably, on this point, he underestimates the place of the idea of individual liberty—of the freedom to do things that risk or cause harm to oneself. Goodin argues that paternalistic laws are justified in those instances where we are not in fact the best judges of what is best for ourselves (even though generally we are the best judges of what is good for ourselves)—e.g., due to weakness of will, common psychological mistakes judging probabilities, and a bias towards the present. He doesn’t fully address the liberal (more than libertarian) idea that even so, the individual has the right to be silly, self-negligent, and so on. Sometimes, we value noninterference ahead of thriving more fully and wish the state to butt out when it seeks to interfere on our behalf. Goodin’s principal focus is particular problems rather than finding unifying principles to deal with them. His discussions are often insightful and illuminating.

In How Are We To Live? (1995), Peter Singer addresses an educated general public on the alternative to living in the manner extolled by the 1980s: a life of materialistic self-interest and greed. Part of the book’s discussion is to make the point that the ‘greed is good’ credo is historically specific—to the West of the recent past—and therefore not compulsory, and, further, that living ethically is in no way blocked by our genetic endowment (1995, Chap. 5) or by rational decision-making (1995, Chap. 7). He discusses how not living an ethical life can cause harm to others and oneself (Chaps. 2–3), and he points to the possibility that capitalist societies (e.g., Japan) can be based on values other than self-interest (Chap. 6). He urges that what recommends making the ‘ultimate choice’ (1995, p. 4) to live ethically rather than any other way is that it provides the firmest foundations for genuine self-esteem and for being all we can be (1995, p. 218). He provides a number of examples of people who have made this choice and he brings out the satisfactions it has furnished them with, precisely in how their attention is dominated by the values themselves which they serve rather than the pay-offs such service afforded them. Of course, examples of some people for whom living ethically was satisfying in this way are hardly an argument that it would also be satisfying for all of us.

Singer’s book is not for philosophers. Too many points remain unargued—e.g., why contrast an ethical life with one lived serving a specific community or a noble tradition, where this may involve raising a family, contributing to knowledge, beautifying the world, or living a religious life? It isn’t as if the only options are either living the ideally (negative) utilitarian life (i.e., a life dedicated to the removal of suffering) or being a crassly materialistic, self-absorbed go-getter. Why must my life, if it is to be ethical, involve imagining ‘myself in the situation of all those affected by my action (with the preferences that they have)’? (Singer 1995, p. 174, emphases mine). Maybe some preferences themselves are horrid and unethical. And, though all human beings are my fellows, my responsibilities in life cannot extend very far. Surely I must make the best of the choices afforded me by my special responsibilities and relationships, even if other choices could produce more overall good for the world. Why is it not the vice of moralism to prejudge negatively those whose informed choices lead them to engage focally with things other than solving the world’s problems, important as this is? They may have talents suiting them for these other things (of course, it matters that it is, e.g., art, political service, or some real excellence rather than, say, being the best hit man in the state), and they may have enough to deal with in their lives (e.g., a difficult family, a history of poverty, and low expectations), or they may simply feel that their own community has supported them and they wish to give something back to it. So, why do they do wrong not to think of a group who has done nothing for them, even though they are their fellow human beings? Still, Singer’s book represents an important expression of the idealistic spirit of the 1990s after the self-absorption of the previous decade, and it calls people’s attention to the importance of living morally well for living well simpliciter.

Metaethics

Jackson, the Canberra Plan, and Smith on the Moral Problem

In many areas of philosophy, the principal dividing line over method has been over whether there is anything worth calling ‘conceptual analysis’ that has a serious stake in what philosophers ought to be doing. For instance, Devitt and Sterelny, inter alios, have championed the view (roughly) that since there is no real distinction between analytic and synthetic truths, and since ‘pre-philosophical linguistic intuitions’ are hardly determinate or stable on the kinds of problems philosophers worry about, there is no serious reason to be concerned with ‘conceptual analysis’. Proponents of the Canberra Plan differ.

The term ‘Canberra Plan’ was first used in a pejorative way in drafts of a paper by O’Leary-Hawthorne and Price (1996). The term was reclaimed over time by exponents of the Canberra Plan with the aim of collecting some shared aspects of their approach to philosophical analysis (cf. Nolan 2010). The shared approach consists of the following ingredients.

First, some background. One takes a fundamental ontology of what exists. The typical fundamental ontology among practitioners of the Canberra Plan is physicalism. One then examines the place one can find in the fundamental ontology, in its own terms, for the items from some other vocabulary. Typical examples of ‘problematic’ vocabularies include semantic properties: truth, meaning, reference; psychological vocabulary: intentional states with their content, qualia, free will, and so on; and moral properties: right, wrong, good, bad, and so on. The basic question, then, has the following form: Given the fundamental ontology of physics which includes volume and mass, is density something further to these two properties or does it, more or less, ‘come for free’ once we have those two properties? Is having a natural motion away from the centre of the earth like density, or should we dispense with it entirely? Since density comes for free, we can keep it. And since natural motion away from the centre of the earth does not fit with our physics, we eliminate it. So, does intentional content ‘come for free’, or is it more like natural motion away from the centre of the earth? Density makes the cut because we can express just how it supervenes upon mass and volume, without being identical to either of these alone. How do reference, intentional content, right and wrong, etc., fare by this criterion?

The Canberra Plan is a way of addressing this problem. In general, its aim is conservative, that is, to keep as much of our familiar worldview as possible. We have seen Jackson in section “The Mind: Intentionality and Mental Causation” above developing his theory of intentionality. The background to that theory is this idea that we find in the Canberra Plan. The method being proposed, then, is this. First, we need an understanding of what is being attributed to something when we attribute the problematic property to a subject. That is, we need to articulate the implicit theory underlying commonsense practice in our ordinary use of the relevant terms for the item. The terms involved are taken to be treatable as implicitly definable by a defining theory which can be teased out through thought experiments (Jackson 1998e, f). This process of explication involves the collection of ‘platitudes’ which capture the ordinary use of the term. ‘Platitude’ here does not have its ordinary meaning—the theses or claims about the items being defined need not be trivial or obvious, they need only be well entrenched. This means, roughly, that they are claims about the analysandum and related concepts with which all, or most, competent users of the vocabulary being analysed will agree, even if this agreement may take some reflection and deliberation in the light of possible cases (Jackson 1998a, p. 46ff.).

Once one has assembled enough platitudes, one manipulates the claims in the collection of platitudes to formulate a role for the problematic concept. Part of the power of the method is derived from the fact that because any given term is taken to be implicitly defined by a defining theory underlying the relevant conceptual practice of competent users of the vocabulary in question, there is no need for explicit definition, and the analysis can be perfectly relaxed about interconnections among concepts in the network of concepts to which the analysandum belongs. The concept is located holistically by its interconnections with other concepts. Such a strategy of articulation invites an analysis of a given concept in terms of its role in the network of concepts, including relations to entry-level concepts, relations to other theoretical terms in the network, and relations to exit-level concepts. The identity of the concept is its place in this network. In the case of mental items like belief, the entry-level concepts are to do with perceptual inputs; connections with other theoretical concepts are to do with how the concept of belief is related to other concepts such as desire; and the exit-level concepts are some specification of behaviour. In the case of mental theoretical terms like belief and desire, the role derived from the collection of platitudes is a causal one: beliefs cause or would cause behaviour which, if they are or were true, will or would satisfy desire. In more general terms, the role is quite abstract and is expressed in terms that are neutral relative to the fundamental ontology or any of its rivals.

Once we have the role specified, we seek the things in the world, from our best knowledge of it, which actually fill the role given in the analysis. The identification of what it is that occupies this role goes to the matter of fundamental ontology. For Canberra Planners, this is physicalism, so they will look to fundamental science (however more exactly that is defined) to fill the roles specified by the analysis. However, this is not necessary in order to execute the basic strategy of the Plan (Jackson 1998a, p. 26). Typically, Canberra Planners take the implicit theory articulated by means of a role in a conceptual network as giving the meaning of the analysandum. Further, this role will generally fix the referents of the analysandum concept. At this stage, the possibility arises that there is not something in the resources of the fundamental ontology which exactly fills the role in all its clauses as specified by the articulated commonsense theory embedded in the practices of competent speakers of the vocabulary. However, Canberra Planners do not require purity in this matter. Near enough filling of most important or enough clauses of the implicit theory articulated in the analysis suffices. (How much of the implicit commonsense role is required is another question).

Once the referent in the actual world is fixed, in typical examples the analysandum is taken to be (or becomes) a modally rigid term. Hence, (simplifying) the way things are with the best deserving role filler from the fundamental ontology entails the way things are with the analysandum item and its related concepts in the theoretical network implicit in the competent use of the vocabulary. This secures supervenience⁹—incorporation in the ontology without extravagant postulation (cf. density). Thus a place for the problematic vocabulary within the fundamental ontology of the metaphysics is assured. Insofar as the analysis of concept C in terms of role R furnishes analytic truths, we could say that it represents a kind of commonsense functionalism about C. Further, inasmuch as we can argue that it is analytically true that C is whatever in fact best fills R, the method provides a way of doing a priori metaphysics. The role of conceptual analysis in yielding R is modest, however, and exposes C to the risk of elimination if it turns out that nothing in the fundamental ontology is a best deserving filler for the role conceived in commonsense.¹⁰ Important elements of the commonsense worldview—intentional content, truth, meaning, free will, moral properties, etc.—get to stay in the worldview provided by complete science for, or if, they are entailed by it.

The term ‘Canberra Plan’ is also sometimes used for a set of views commonly associated with philosophers connected to the philosophy program at the Research School of Social Sciences (RSSS) of the Australian National University. The doctrines include physicalism, two-dimensionalism in semantics, four-dimensionalism about material objects and time, and (for some) metaethical descriptivism. Two-dimensionalism is quite important here, but discussion would take us too far afield.¹¹ It clarifies that the entailment of the ways things are in terms of the problematic vocabulary by the way things are in terms of the fundamental ontology is not an a priori deduction (as if by a conceptual connection between primary and secondary intensions). Rather, it is necessary truth preservation where such necessary truths depend on the way things turn out to be in the actual world, as a finding of empirical discovery (Jackson 1998a, p. 25ff.). That said, the Canberra Plan strategy does not require, for all the vocabularies it is applied to, that the identification of items referred to by the terms of the problematic vocabulary with items in the vocabulary of the fundamental vocabulary be parsed modally as rigid designation. For the question of the rigidity of folk terms is not always determinate (see Jackson 1998a, p. 144, in connection with moral terms).

We saw earlier (in section “The Mind: Intentionality and Mental Causation”) Jackson’s way of implementing the Canberra Plan for intentional content. We will now move to the application of the strategy to morality. To do so, I will discuss Michael Smith’s The Moral Problem (1994) and make certain comparisons with Jackson’s moral functionalism.

Smith argues that morality present us with a problem. When we reflect on our ordinary practices with moral terms, two lines of thought emerge very naturally:

• 1. (Objectivity of moral judgement) Moral judgements of the form ‘It is right that I φ’ express a subject’s beliefs about an objective matter of fact, a fact about what it is right for her to do.

• 2. (Practicality of moral judgement) If someone judges that it is right that she φ then, ceteris paribus, she is motivated to φ.

  In addition to this, there is ‘the standard picture of human psychology that we get from Hume’:

• 3. (Humean psychology) An agent is motivated to act in a certain way just in case she has an appropriate desire and a means-end belief, where belief and desire are, in Hume’s terms, distinct existences.

Together, these three give us ‘the moral problem’: if one and two are both true, there are at least some beliefs that are necessarily connected with desires, which, if three is true, is impossible (one standard solution: so reject 3 as philosophical dogma: Platts, McDowell). one and three imply that two is false and moral judgements are not essentially practical (second standard solution: externalism). And two and three imply that one is false and that nothing in morality requires the idea of moral facts (third standard solution: expressivism). Yet, as far as our understanding of morality and its place in the natural order is concerned, it seems desirable to hang on to all three in some way (Smith 1994, Chap. 1, esp. 11ff.). Moreover, each of the standard solutions fails to do justice to all the platitudes about morality or reasons explanations of action (Smith 1994, Chaps. 2–4).

Smith’s solution has a number of steps. The first is to distinguish two aspects in the Humean picture: a picture of motivation and a picture of the nature of our normative reasons for doing things. The former is the standard account of the explanation of action in terms of beliefs and desires. This psychological Humeanism stays. Smith argues that only a Humean theory of human motivation can account for why (motivating) reason explanations for action are a species of teleological (goal-oriented) explanation (1994, §§4.6, 4.7).

The latter is the idea that desires are beyond rational criticism: desires are basic and, being distinct existences from beliefs, are aspects of ourselves we find ourselves with in ways that elude rational evaluation (cf. Smith 1994, p. 9, §4.2). This goes. Instead, Smith develops an anti-Humean conception of normative reason for action. It is based on the thought that normative reasons are the deliverances of deliberation on our values, and valuing is a species of believing something desirable (Smith 1994, pp. 147–151). Proposition C2 is central: if an agent believes she has normative reason to φ, then she should rationally desire to φ (1994, p. 148). A platitude Smith finds in Korsgaard (1986) provides the key: what it is desirable that we do is what we would desire to do if we were fully rational. The analysis, then, goes like this: one has normative reason to φ in circumstances C if and only if, were one completely rational, one would desire to φ in C. (C2, he argues, follows from this).¹² Facts about normative reasons are facts about what we would desire to do if we were fully rational (Smith 1994, pp. 151–180).

Thus, practical rationality, which is intrinsically connected to motivation but can break down, explains action as far as it does—depending on how rational we are. And it is facts about what we would desire in the circumstances that are the objects of the beliefs which are our normative reasons. This is anti-Humean because (i) it means that our normative reasons, our values, beliefs and so on, can play a proper role in the explanation of behaviour; (ii) it furnishes us with a critical perspective on our basic, actual desires, showing us that we can have reason to change them; (iii) it brings out the categorical character of our normative reasons—their rational force is independent of the desires we actually have; and (iv) this fact alone shows that moral judgement is a social matter in the sense that no one person is more likely than others to be able to work out authoritatively what we have best normative reason to do in a given set of circumstances. Next, Smith has to distinguish the specifically moral reasons from other normative reasons. This he does in terms of the kinds of substantive considerations which enter into specifically moral deliberation (avoiding harm, keeping promises, telling the truth, and such) as yielded by the relevant platitudes. Additionally, Smith argues in Chap 6 that there are moral reasons. I shan’t rehearse this. The last bit of work here is to explain how this solves the moral problem. How does this theory account for the consistency of one, two, and three?

The central move is this. Moral rightness in circumstances C is what we would desire in C if we were completely rational, where this is a desire for something of the right substantive kind for the normative reason to be a moral reason. Thus, one is true as what we would desire in C if we were completely rational is an objective fact of the sort that can be expressed in the beliefs about normative questions which we have. Further, such beliefs are connected appropriately to motivations, so that to have such a belief is to be such that one would desire the relevant thing, action, attribute, or whatever, if one were fully rational. So, two is true. Finally, one and two imply that what explains one’s φ-ing, where one believes that it is right to φ, is this belief and the motivation to φ, if one is completely rational. This is consistent with the motivation in question being the desire to φ, which one will have if one is rational, given the belief. Of course, one might not be rational, in which case one will have some counter moral desire which is a distinct existence form the belief that one should φ. So three is true. The moral problem is solved.

Jackson and Pettit’s so-called moral functionalism is a more abstract account of morality than Smith’s (Jackson and Pettit 2004d; Jackson 1998a, pp. 129–38, 140–50).¹³ It is committed to fewer substantial claims. Both accounts are committed to cognitivism about moral judgement—that it is belief and so represents ways things might be, and so is apt for truth (Jackson et al. 2004; cf. Jackson 1998a, pp. 115–17). Smith’s one, above, says this. Both accounts share a general naturalism about reality. Smith is interested in developing a conceptual analysis of moral concepts, as is Jackson (and Pettit). The two theories both collect platitudes in the service of doing so. More substantively, Jackson finds much that is congenial in Smith’s analysis of moral concepts. (a) Convergence in the maturation of folk moral theory is important (Jackson 1998a, p. 137; Smith 1994, §6.3). (b) Jackson favours identifying the moral property, for instance, of rightness with the realiser of the rightness role which moral functionalism characterises and sees this as echoing Smith’s insistence that the target of moral judgement and motivation is not so much what is right as such but the goods that make something right (Jackson 1998a, p. 137; Smith 1994, pp. 75f., 83–91). (c) Jackson prefers a Humean version of the content strategy for explaining how taking up a moral judgement about some matter is to take up an essentially directed attitude in relation to it. He sees affinities between this and Smith’s theory of normative reasons which directly connects a deliberator to what she has motivation to do, if she is rational (cf. Jackson 1998a, pp. 155–57; Smith 1994, §§5.4, 5.8–5.10).

However, despite important similarities and considerable shared ground concerning method, Smith demurs from Jackson’s ‘analytical descriptivism’. Jackson’s moral functionalism leads him to an identification of moral properties, via reductive analysis, with descriptive properties. Indeed, against the Cornell moral realists (such as Boyd and Railton), Jackson argues that the metaphysical descriptivism they share furnishes sufficient materials to construct analytical descriptivism (Jackson 1998a, pp. 146–50). In contrast, Smith argues against Jackson’s position when he rejects definitional naturalism (1994, pp. 44–56, 161–64 [fifth question]) because it is vulnerable to what he calls the ‘permutation problem’. Instead, Smith argues that we can get the following from his ‘summary-style, non-reductive’ analysis:

• Conceptual Claim: Rightness in circumstances C is the feature we would want acts to have in C if we were fully rational, where these wants have the appropriate content.

• Substantive Claim: Fness is the feature we would want acts to have in C if we were fully rational, and Fness is a feature of the appropriate kind (as shown in the platitudes regarding the substance of moral judgements).

• Conclusion: Rightness in C is Fness. (Smith 1994, p. 185.

This, Smith suggests, is a broadly naturalistic analysis of moral properties in two senses. First, it permits the identification of the moral property with a natural property in the circumstances. Second, since fully rational creatures are naturalistically realised, the properties referred to in the analysis do not require any non-natural properties, even though it involves a certain idealisation and full rationality may not itself be naturalistically definable (Smith 1994, p. 186). However, it is non-reductive and eschews definition.

The Canberra Plan is not a monolithic approach to philosophy. Smith is not an instance of it in this last point. And while this is important, so much else of his view ticks important boxes in the Canberra Plan model. Perhaps we should say that the school is more loosely defined than most of us think? Perhaps we should not consider Smith’s metaethics a good enough example. Not much depends on such matters. What is perhaps interesting is the shared resistance to particularism in metaethics among Jackson and Smith (see Jackson et al. 2004, read at a conference held at Australian Catholic University, Ballarat, in April 1998 concerning how thin moral philosophy can afford to be). It is not obvious that one like Smith who resists naturalistic reduction, though not naturalism, about morality, need have any scruples about particularism. Jackson’s reductive take on the supervenience of ethics needs to resist particularism. I do not see why Smith needs to. Indeed, if he shares the view expressed in the joint paper with Pettit and Jackson, his resistance to reductive naturalism in The Moral Problem would seem to be undermined, and he could represent a less qualified place in the implementation of the Canberra Plan. Perhaps the question is whether one wants to do ‘serious metaphysics’ (Jackson 1998a, pp. 4–5) or get the understanding of morality right.

Raimond Gaita

Gaita is clear what he wants to do, and it is not serious metaphysics in the sense of the Canberra Plan or in any sense for that matter. Gaita arrived back in Australia to be the foundational professor of philosophy at Australian Catholic University in 1994. He is best known for his prize-winning memoir occasioned by the death of his father, Romulus, My Father (1996). His major works of philosophy during the decade were Good and Evil: An Absolute Conception (1991) and A Common Humanity: Thinking about Love and Truth and Justice (1999).

I will try to work my way into Gaita’s conception of moral philosophy by beginning with his examination of Plato’s Gorgias (Gaita 1991, Chap. 15). The question he poses is: what is it that distinguishes Socrates from Gorgias? The conventional philosophical answer is that Socrates, being a philosopher, is concerned with truth and reason, whereas Gorgias, being a rhetorician, is concerned with persuasion with an eye to inducing his audience to believe what will help Gorgias achieve his ends. However, this answer, though true, doesn’t explain why Plato works so hard to distinguish them. What explains this goes to what Socrates’ being concerned with (or for) truth and reason amounts to. It is this latter that Gaita wants to reclaim for moral philosophy and more generally for philosophy itself.

Callicles, like Polus, is impressed by Gorgias. Callicles is impressed that Gorgias is always able to answer any question he puts to him. He is impressed by the fact that Gorgias cannot be surprised by any question. It seems he has a ‘mastery’ of all subject matters and is quick on his feet. Still, Socrates is unimpressed and continues to claim that oratorical quickness and the masterful command of knowledge should not be taken as the criterion of worthiness. He chides Polus for giving speeches instead of answering briefly and to the point and with what he really believes, not ‘mere words’.

It is clear from the Alcibiades’ speech in the Symposium that Socrates was capable of ‘casting a spell’ over his associates. It left Alcibiades somewhat disconsolate that he himself was unable to rise to what it was in Socrates that exerted such attraction that made him compelling to Alcibiades. It is tempting to see the difference between the force of Socrates’ presence and that of Gorgias in terms of the former’s ‘concern for truth and reason’ and the ‘latter’s concern for persuasion in the service of achieving his ends, come what may’. Indeed, Socrates presents Polus with an argument that it is better to suffer evil than to do it. Polus remains unconvinced. He agreed to the premises, to the steps of the argument and to its conclusion, but he suspects that he has been tricked. Assume Socrates’ argument is sound (for here that is not so much the issue). Let’s say Polus accepted its soundness. Let’s say that he professed the argument to whoever was prepared to listen and that he affirmed its conclusion. Given the man Polus is, this profession can be expected not to go deep. His acceptance of the argument is of a piece with his being impressed by what it is in Gorgias that impresses him. Polus’ profession of the argument would be ‘mere words’.

Compare a story Arendt tells about Eichmann in the last moments before his execution. Eichmann had always been ‘elated by’ what he called ‘winged words’ of oratory:

He began by stating emphatically that he was a Gottgläubiger, to express in common Nazi fashion that he was no Christian and did not believe in life after death. He then proceeded: ‘After a short while Gentlemen, we shall meet again. Such is the fate of all men. Long live Germany, long live Argentina, long live Austria. I shall not forget them’. In the face of death, he had found the cliché used in funeral oratory. Under the gallows, his memory played him one last trick; he was “elated” and he forgot that this was his own funeral. (Arednt 1964, p. 252, cited in Gaita 1991, p. 278.

This disconnect between the reality facing Eichmann and his words and the reality of what Polus has assented to and Polus’ words is radical. However, ‘the reality facing Eichmann’ and what I’ve expressed as ‘the reality of what Polus has assented to’ are not on all fours. It can be required of one to think a long time about and explore the implications of some claim one has come to believe. But this isn’t like ‘the reality facing Eichmann’. Eichmann’s error is not one of failing to see what the proposition expressing his death sentence implies or what its truth-conditions are. His error is that he fails to rise to the gravity of his situation as a convicted murderer about to be executed for his deeds. His failure is that he can only manage cliché, that he fails to see what he has done and been convicted for, and that he continues to wish for the prospering of Nazi Germany, with all that means for the Jews. Eichmann either is unable or unwilling to face the reality of his situation in all these ways. He has failed the test of being open to the repentance his crimes demand and to being exposed in this to his captors, his victims. He has failed to hear the voices of the dead in his conviction and sentence.

Polus is similar. If we thought of philosophy and so the power of Socrates’ presence, in terms of the concern for truth and reason, and contrasted this with oratory and so the power of Gorgias’ charm and personality as being concerned merely for persuasion in the service of achieving his ends, whatever they are, we will miss the important similarity of Eichmann’s and Polus’ relation to their situation. Polus has understood the argument and assented to it, where the argument is conceived as a series of propositions and reasoning on propositions. But it doesn’t ‘go into him’. His uptake of it is shallow and at the same level as ‘mere words’. He doesn’t ‘get it’. His grasp of the propositions does not imply that he is to be expected to think he should lay down his life for Socrates if that is what not doing evil meant. Yet, one imagines, Alcibiades would, even though in so many other ways he is a ratbag (and even if we have qualms about Alcibiades’ doing this, these will be out of concern for whether he would rise to the kind of sacrifice for Socrates that would be worthy of him, rather than merely offering a self-preening grand gesture). Something more has to happen to Polus for him to get the argument in the way Socrates intends.

This is not to say that Polus cannot change, that the conviction that it is worse to do evil than to suffer it could not become something deep for him, that he couldn’t come to ‘get it’. But it would not be the force of the proposition, its truth-conditions, or the rationality of the argument (all of which can be abstracted from the fact that it was Socrates that led him to accept the proposition or argument in question) that pulled off this change in him. Of course, in a person who cares about philosophy, argumentation would not be irrelevant to any change that person undergoes. Part of the concern for reason and truth is a concern for truth-conditions, validity of reasoning and so on. A philosophical Polus (which he isn’t) would see such concern for the argument as part of what he is open to learning from Socrates in their conversation. But it is not the whole of it. For Polus ‘grasps the proposition’ involved. Or more precisely, he ‘kind of’ grasps it. He clearly does not see it for what it is—he thinks he may have been tricked, he either cannot or refuses to trust it. In a sense, he clearly does not understand it, as is shown up in precisely his thinking he may have been tricked even though he accepted the premises, recognises that the steps in the reasoning are valid, and ‘grasps the proposition’ expressed in the conclusion. To become a ‘real philosopher’ he has to come to see that what he experienced as a trick is not a trick, and that what he understands when he ‘gets it’ is not the same thing as what he had thought had tricked him. But to see this, to change in this way, is to see that he has to learn from Socrates—that man who could make a good speech but in doing so argued for the rightness of his deeds as what he had to do even if it meant receiving the condemnation of others. Understanding what the reasoning and the proposition mean involves Polus in being moved to see the authority in Socrates’ life—he brooks no bullshit, he thinks hard and he means it. No Gorgias, this fellow. He’s not about getting his way by manipulating his audience with fine speeches and taking advantage of their gullibility. Philosophy is more than simply a concern for truth and valid reasoning, as that could be prised off what Socrates stood for. It is standing for the good that as the man or woman you are you stand for. It is standing by the authority of one’s life as an interpretant of one’s moral claims. That’s what Alcibiades gets and it is what he sees in Socrates but cannot live up to. That is what Polus cannot see and so he thinks he may have been tricked even though he grasps all the propositions and valid steps of reasoning.

If Polus comes around to trusting the reasoning, it matters that it was Socrates, that man, whose authoritative example moves him to this understanding. For what he is convinced of and that it was Socrates who convinced him of it are inseparable. It is as much the authority of Socrates’ life (standing up against the Athenians when they violated their own law to convict the generals after the debacle of Salamis; refusing to escape from his death sentence for to do so would be to betray his mother Athens, even though it had unjustly condemned him; never afraid of personages and reputation in discussion and always after the truth; and so on) as what he formulates as his sincere belief (the proposition) that Polus doesn’t understand and would have to come to understand if he were to change and no longer suspect that he has been tricked. And indeed, in Polus’ situation, what reason and truth are is not a separate part of philosophy to trusting Socrates and his call to Polus to be the man, that is, the best and most intellectually honest man, he is (Socrates’ ‘call to seriousness’) (Gaita 1991, p. 282). Socrates does not complain that Polus refuses to reason and pursue truth—though this is clearly what he expects of him. Rather, he complains that he refuses to converse, that is, to be open to be engaged with Socrates himself, rather than hiding behind fine words which ring hollow and are not his own and which ultimately miss the point. Truth and reason are more than propositional in this picture. They involve putting oneself at the mercy of earnestly engaged conversation and putting oneself into one’s thought. This is what Eichmann’s words fail to do, and it is what Polus is finding too challenging to do.

So, the form of understanding that it is better to suffer evil than to do it does not (merely) consist in grasping a disembodied proposition whose meaning can be abstracted from Socrates’ life and is utterable in all semantic seriousness by, say, a political spin doctor on the job. One also has to see what it means in the sense of what it comes to in the life of one whose belief in it is informed by his living up to it, as is the case with Socrates. One has to understand what it means to do evil to another and what it means to suffer such evil at another’s hands. One has to see what it comes to in a life which testifies to its force. The reason why Plato works so hard to distinguish Gorgias from Socrates is that they are so easily confused: both can speak well, both are good reasoners, both are quick on their feet, both understand difficult and technical propositions as they need to, both can get others to see things their way and so on. And if it suited him, no doubt Gorgias would argue for the view that it is better to suffer evil than to do it. Still, at the end of the day, the force of Gorgias’ expertise and the power of his charm and personality are a false semblance of Socrates (Gaita 1991, p. 280ff.), for they differ in precisely what matters to the genuine understanding that Socrates has of what it is to suffer evil and to do it.

Gaita works hard to bring out in several ways what this understanding consists in. Examples are central to bringing it out. Gaita refers, for instance, to the remorse eventually felt by Raskolnikov in Dostoevsky’s Crime and Punishment, which reveals to Raskolnikov the significance of his murders, what they really are, as contrasted to corrupt forms of understanding. Remorse is central for this understanding. The ambition of typical philosophical moral theories to provide disciplined and explanatory accounts of moral thinking is shown to fail by such moral emotions. In his remorse, Raskolnikov’s lucid understanding of the horror of his murders would be parodied if understood in terms such as these: ‘My God what have I done. I have been a traitor to reason. I have violated rational nature in another!’ or ‘My God what have I done. I have violated my freely chosen and universally prescribed principle that one shouldn’t kill people under circumstances such as these’ (Gaita 1991, p. 33). Instead, what remorse (not its counterfeits or corruptions) reveals is the enormity and horror of one’s action in taking another’s life. In religious terms, it is natural to put this in terms of the sacredness of human beings. Whether or not this is the best way to put the thought, the thought is a general one which extends to what is evil in (certain forms of) racism, what is evil in rape and so on. Another human being is a precious centre of irreducible value that engages one and represents a particular kind of limit on one’s will, of a kind that is unconditional (Gaita 1991; cf. esp. 282). It does not matter how morally wicked they are, how little their lives have any prospect of meaning or of contributing to their fellows, how pathetic and compromised their life and abilities have become, or how beneath contempt we think them. They are our fellow mortal human beings. This thought conditions everything. Remorse and this sense of the unconditional preciousness of each human being as one with whom we share the great facts of life (being born of a woman, suffering, sexuality, mortality, etc.) are mutually elucidating.

Of course, just as with a philosophical conversation with Socrates, the meaning of things and the emotional tone of our understanding of things can go awry. There are false semblances of remorse; remorse has its corruptions. Correspondingly, there are false renderings of the ‘moral content’ of what is understood in moral understanding. This is the wisdom in Socrates’ dictum that one who knows the good must do the good. Thus, consider the typical contrasts between the person of virtue, the strong-willed person and the weak-willed person. All three in some sense know, for instance, that ‘it is wrong to lie to another to avoid trouble’. The weak-willed person will fail because she fears the trouble. Her fears ‘override’ her moral knowledge. The strong-willed person will not lie because doing so is wrong even though she fears the trouble and is sorely tempted to lie. The virtuous person will not lie even though she understands the trouble involved and would rather avoid it, but she is not tempted to lie and so does not lie. True, her reason is that it would be wrong to lie. True, she differs from the strong-willed person in not being tempted to lie. True, there is a sense in which all three understand the same thought. However, there is more to it in the case of the good person. To be sure, one difference is the shape of her motivation compared to the strong-willed person—she has no temptation to lie. But, many are inclined to say, that difference in the shape of her motivation does not go to a difference in the content of what the three understand. The fact that the same proposition ‘it is wrong to lie to another to avoid trouble’ is understood by all three and yet the three differ motivationally shows that the motivation and the moral understanding are distinct.

Though Gaita probably would not credit Aristotle with what I’m going to observe (1991, p. 25, 239 ff.), even Aristotle, who is the source of these distinctions, does not think of the relation between motivation and understanding like this. For him, in the virtuous person, it is one state of the soul that is ‘orektikos nous or orexis dianoetike’—that is, appetitive understanding or deliberative desire (Nicomachean Ethics, book 6, ch. ii, 1139b 4–5). That is to say, weakness of will is as puzzling for Aristotle as it is for Socrates precisely because intellectual understanding of what morality demands and the moral motivation to do it are the same thing in the good person and explain each other. There is something about lying to another to avoid trouble that the virtuous person grasps but which eludes even the strong-willed person. Aristotle does not serve making this point well with his talk of ‘not using the knowledge’ in his account of akrasia—as if the virtuous person uses something that the weak-willed person understands but fails to. But this overlooks the fact that that contrast is between the strong-willed person and the weak-willed person, not the virtuous person and both of them. The virtuous person, lacking temptation, is one in whom, as Gaita would put it, (proper) emotion is a form of understanding. What they see in virtuously grasping that it is wrong to lie to avoid trouble is not simply a proposition, it is rather the meaning of telling a lie in these circumstances—what it shows about what one thinks of the person one lies to, what it shows about what one thinks of the meaning of what one has done which causes the trouble and so on. One who understands this is one who is struck by the reality of the other, engaged by that reality and by what having done what one has done means and what lying to them would mean to them. It is to be moved by this and thus to be so unimpressed by the prospect of lying even to avoid the trouble that it has no force for one.¹⁴

To summarise, Gaita argues that if we are to understand how ethical thought works, we cannot deal merely with thin moral concepts such as ‘ought’, ‘should’, ‘x is right’ and ‘x is obligatory’. We need to draw from the rich fund of thick moral notions familiar from virtue-talk and from our language for various kinds of actions and sentiments. Among other things, these notions furnish us with a vocabulary of genuine virtue, and the counterfeits of virtue, and vice. To limit our critical vocabulary to the language of propositional evaluation and thin moral concepts, in abstraction from the ways we live that are embodied in thick moral concepts, is radically to distort moral understanding (e.g., witness the parody of Raskolnikov’s remorse above).

Secondly, these concepts require interpretants from life. This has two dimensions. It means that moral concepts are ineluctably social and interpersonal, though in a particular way that is easy to lose sight of if we resort to the theoretical mode too quickly. For what is required is that sense of the reality of another human being by whom one is engaged and who, as precious, is a limit on one’s will. Further, this understanding requires the authoritative example of those whose understanding of the moral meaning of things reveals this meaning for us. Socrates’ expectation of Polus in philosophical conversation is one kind of example. Gaita’s famous example of the nun non-condescendingly caring for the mentally ill is another (1999, p. 18ff.). What is understood in such examples is conditioned by the authoritative example—of Socrates or of the nun. The social nature of moral understanding, then, is not just that ‘moral concepts are social creations’ (as if other concepts are not) or that ‘morality is a solution of social coordination problems’. It is, more fundamentally, that our exposure to others in engaged dialogue, their call to us to be morally serious, and their revelation to us by their example of a goodness that goes beyond virtue are central to what we understand in ‘grasping moral concepts’.

Third, this understanding is a matter of being moved emotionally. Moral meaning is not merely propositional in the sense of what can be abstracted and formulated in a proposition as to what, for instance, the virtuous, strong-willed and weak-willed person can all be described as understanding. Rather, emotional responses are forms of understanding (or misunderstanding). Raskolnikov’s (clear sighted, non-corrupt) remorse reveals for him what he had done.

Fourthly, this line of thought opens up a notion of understanding in which what is understood and the one who understands it and reveals it for us in her authoritative example (like the nun or Socrates) cannot be prised apart—a realm of meaning as authentic individuality answerable to the thick concepts of appraisal but understood in one’s own voice (cf. Gaita 1991, Chap. 9, 308ff., 1999, pp. 237–58).

So, the question is what difference this makes to doing philosophy or, more particularly, to how to do moral philosophy. I will approach this question by considering how far Smith and Jackson might go in appropriating aspects of Gaita’s view of the nature of moral understanding.

Smith’s anti-reductivism perhaps gives him more scope than Jackson has to accommodate the various claims about moral thought that Gaita makes. For instance, Smith’s summary analysis of moral rightness in circumstances C appeals to the feature we would desire acts to have in C if we were fully rational. There is no aspiration in Smith to achieve more than circumstance-bound identifications of what moral rightness is (1994, p. 185). Presumably, then, in a case-by-case way, Smith could exploit the detailed sensitivities of concepts such as courage, venality, sentimentality and so on, in order to determine what the features are which make acts right. This could seem natural given that Smith puts a certain stress in the analysis of moral rightness on the substantive appropriateness of the features, given the circumstances. And it seems obvious that a similar ‘relevance to circumstances’ condition characterises the virtues and like concepts—e.g., patience is not the relevant virtue to be engaged when a quick decision has to be made about how to protect someone from an imminent danger. Jackson, on the other hand, makes clear that moral functionalism leaves it an open question whether thick or thin moral concepts are ‘central’ to moral understanding (Jackson 1998a, pp. 135–37). Both Smith and Jackson may see the motivational internalism of Gaita’s view congenial, though to be sure they would put more emphasis on desire than on emotion as constituting the essential connection of the ‘content of moral thought’ to motivation (Smith 1994, but cf., e.g., 127ff. and Jackson 1998a, pp. 154–60). Again, Gaita clearly rejects the theory conception of our understanding of moral language. This might put him closer to Smith, who also rejects the network analysis of analytic descriptivism that Jackson defends. Gaita’s account of our understanding of moral concepts puts great emphasis on the revelatory power of authoritative examples such as the nun and Socrates. This resonates to some extent with the role Smith gives to paradigms in the acquisition of concepts, including normative concepts (Smith 1994, p. 55; cf. 163ff.). Finally, Gaita’s insistence that moral goodness is sui generis has certain echoes with Smith’s claim that ‘very little outside the sphere of the normative is required to define the normative’ (Smith 1994, p. 163).

However, as compared to the differences between Gaita and both Smith and Jackson, these similarities are relatively superficial. Fundamentally, Gaita insists that moral judgement does not stand in need of the kind of grounding which either Smithian metaethics or Jacksonian metaphysics seeks to give it. In particular, commonsense moral understanding does not require the kind of convergence on a future mature moral outlook that both Smith and Jackson posit as if the morality we have in our moral traditions is not ‘mature’. Further, the conception of sensibility as a form of understanding opposes the Humean understanding of intellectual knowledge of semantic meaning as a distinct existence from motivation, desire and so on. Again, Gaita’s conception of moral goodness as sui generis distances Gaita from Smith’s and Jackson’s naturalism and descriptivism. Gaita writes:

I would argue that the epistemic grammar of moral descriptions involves what I have been calling ‘authoritative disclosure’, as I tried to instance it in the example of a woman whose love for her unborn child had the authority to reveal to another what it would be to have an abortion. This is at least in harmony with an important empirical truth: our thought is thought in a tradition, and shaped by those we respect and admire: we learn by being moved and we learn, or try to learn, when we may trust what moves us and ourselves in being moved. To be sure, that is (or ought to be) critical respect and admiration but that does not mean that we seek a transcendental vantage point in order to assess all opinion. If there is no such vantage point from which we can ‘see the world as from no place within it’ and all moral thought is not a reductive elaboration on what is indisputable, then our thoughts are inescapably, dialogically, in the midst. There is no other place from which we can have anything to say, or anything to learn. (Gaita 1991, pp. 141–42)

Moral philosophy, then, can only proceed from the middle of things so as to articulate a moral point of view or form of life by reflection on a rich fund of examples of authoritative disclosure of the moral shape of things. This might be thought of as explanation by careful description rather than explanation by metaphysical analysis or addressing particular problems set by the dialectical context in the literature. Art and creative imagination have a constitutive place in doing such philosophy rather than an accidental, external role of furnishing examples (though they can do that). Indeed, art and creative imagination ground and inform the interpretation of critical concepts, bringing out moral and conceptual possibilities as a constraint on whatever theorising might be thought necessary to do, rather than being criticised by or founded on theory (cf. Gaita 1991, Chap. 17). Theory—understood as a corrective, explanatory external set of principles that explains, measures or evaluates our practices—does not figure significantly in this. Rather, it is the truthful, lucid articulation of what is revealed in authoritative examples that matters.

The point could be extended to other branches of philosophy, especially those most closely connected to moral philosophy, such as philosophy of mind and philosophy of religion. It is an important alternative to the mainstream in philosophy of mind, for instance, to think of pity as a form of understanding, a cognitive achievement itself, rather than as an emotion contingently associated with the cognitive understanding of what someone did, or to think of understanding what someone said as a form of engagement with an individual whose virtuous life informs what he said. Similarly, it is an important alternative to the way in which the afterlife is understood in mainstream philosophy of religion to think of pitying the dead and wanting to act for their sake as something whose rationality does not depend on an ‘ontology of the dead’ unless it is superstitious (Gaita 1991, pp. 137–140) and to think of the fear of death as something requiring a clear-eyed personal response to one’s own mortality (1991, pp. 306–309) in which the reality/appearance distinction concerns a lucid and authentic individuality more than the grasping of facts.

Other Work in Australasian Philosophy in the 1990s: Some Highlights

This chapter is already long. I want to acknowledge before ending something more of the range of philosophical work done by many philosophers in the region during our period. I apologise to those I do not mention.

In other contributions to the philosophy of mind, Paul Redding’s work on the nature of emotion should be mentioned (Redding 1999). It is also worth mentioning that the functionalist materialist consensus did not have it all its own way. Both Peter Forrest and Grant Gillett contributed papers to Howard Robinson’s 1996 collection, Objections to Physicalism (cf. Forrest 1996b; Gillett 1996).

In other work in moral philosophy, Catriona Mackenzie was developing a feminist line of thought concerning autonomy as relational autonomy (cf. MacKenzie and Stoljar 2000). Karen Green’s work on feminist themes in history of philosophy and ethics is worth mentioning here. Justin Oakley sought to relocate the emotions in the centre of our understanding of morality in part through his work on virtue ethics (Oakley 1992), and with Dean Cocking, he published widely on friendship. As with Oakley, Christine Swanton was one of a number of Australasian philosophers working on virtue ethics (see, e.g., Swanton 1992, 1997).

In epistemology, Tony Coady’s book, Testimony: A Philosophical Study (1992), was a significant contribution to the revival of debate concerning the epistemic value of testimony. Also, Stephen Hetherington (1992, 1994, 1998) began a series of papers critical of mainstream debates and positions in epistemology, developing a somewhat revisionary conception of the place of sceptical arguments in epistemology.

The history of philosophy was particularly well served in the 1990s. Work by Gaukroger on Descartes continued, culminating in Descartes: An Intellectual Biography (1995). Stephen Buckle emerged as a further important scholar of modern philosophy. He published Natural Law and the Theory of Property (1991) and began work on Hume’s philosophy. Udo Thiel’s work on Locke is also noteworthy. Research in ancient philosophy flourished with Harold Tarrant, Rich Benitez and Dirk Baltzly all producing important work. Also worth mentioning are Roy Perrett’s work on Indian philosophy (1998) and Martin Tweedale’s on medieval philosophy (1999a, b).

In philosophy of language and logic, Graham Priest’s work on true contradictions was extended into an account of the concept of limits of thought in Beyond the Limits of Thought (1995). Also, Greg Restall emerged during the 1990s as a new talent in logic, in particular developing substructural logics.

In metaphysics, David Armstrong continued his work of the 1980s on the actualist account of possibility and extended it to the idea of truth making, based on an articulated ontology of states of affairs (Armstrong 1989a, b, 1997; Campbell et al. 1993). John Bacon published his work on tropes and property instances as an alternative to taking universals as primitive (Bacon 1995). Further, Peter Menzies’ work on causation, laws of nature and related topics was being published during this decade (Menzies 1993, 1996; Menzies and Price 1993).

In philosophy of science, Bigelow and Pargetter’s Science and Necessity (1990) opened up the decade along with Brian Ellis’ Truth and Objectivity (1990). In Newcastle, John Wright and Cliff Hooker made important contributions—Wright on realism and explanations (1997), Hooker on a systems approach to understanding scientific reason (1995). Alan Chalmers extended his work (Chalmers 1990).

Finally, I will mention Phillip Pettit, whose work in the 1990s ranged across the philosophy of mind and metaphysics, the theory of explanation and moral theory and also included the political foundations of republicanism (1997) and social ontology (1993).

There is much more that one should have commented on than I can manage here. Limitations of space and the knowledge and ability of the author preclude discussion of further topics or authors. This should be taken as no indication of anyone’s significance to the Australasian scene in the 1990s. Suffice it to say that philosophy in Australasia during the 1990s thrived, and did so even though conditions in Australian universities, at least, were not always easy (arguably, New Zealand universities underwent similarly difficult conditions in the 1980s). Philosophers took up more public roles during the 1990s than had been the case before. Moreover, as I have tried to stress, a certain maturity has developed in Australasian philosophy, as is evidenced by the enhanced diversity and the continued vigour and rigour of the discipline. I think this is reflected in the very fact that there is a concern for our own history as a discipline (cf. Pybus 1993; Franklin 2003). Far from the history of philosophy being a nicety of ‘merely historical’ interest, knowing how we got to where we are is internal to knowing who we are and to giving it significance (cf. Campbell 1992; Passmore 1993; Smart 1993).