Two kinds of non-scientific naturalism

Two kinds of non-scientific naturalism Jonathan Knowles, Philosophy Department, Norwegian University of Science and Technology, 7491 Trondheim, Norway (Preprint, final version in M. Milkowski & K. Talmont-Kaminski, eds., Beyond Description: Naturalism and Normativity , College Publications 2010) Naturalism is a widespread position in contemporary analytical philosophy. It is not however a very specific one, embracing rather a range of different views each of which stress ideas and commitments that others calling themselves naturalists disavow. I take it that any naturalism worthy of the name must, at a minimum, subscribe to the idea that philosophy and science are not thoroughly disjoint enterprises, and that philosophy must defer to what we learn from science where there is, uncontroversially, overlap (such as the in question of the Euclidian versus nonEuclidian nature of space). (Cf. Price 2004: 71.) However whilst many naturalists would see in this track record grounds for a scientific metaphysics and epistemology, others are more chary. They urge us to ask what a reasonable understanding of the achievements of science actually commits us to by way of a general philosophy. Science has, it seems reasonable to say, placed us human beings firmly in the realm of space and time like all other organisms on earth. At the same time, we evince a range of properties and engage in a wealth of activities that seem very difficult to understand through science. We think, we act, we pass moral judgements, we appreciate music and visual art, we seek enlightenment through the study of other cultures and epochs, we carry out science (!). In relation to all of these things we discuss with each other using semantically structured language. There is nothing obviously unnatural about any of all this, seen as the activities of a particular, albeit rather remarkable kind of biological entity; but nor is there obviously a way or even a 1 - 11/21/13 Two kinds of non-scientific naturalism need for understanding its significance and status in scientific terms – through a kind of experimental or theoretical enquiry. Relatedly, there is no obvious route to or need for understanding the objects of the everyday, common sense world in scientific terms – from hats and tables to unfinished symphonies. Common sense and/or broadly humanistic enquiry will suffice. I dub views that stress this line of thought non-scientific naturalism (henceforth NSN), to be contrasted with the more usual scientific naturalism (henceforth SN). Both are naturalistic in that they disavow the existence of supernaturalistic, transcendent and/or immaterial entities and forces, thereby paying due respect to what science has taught us about the universe and our place within it. Both also accept the relevance of science and scientific understanding to at least some traditional philosophical problems. Where they differ is over acceptance of the following: that scientific knowledge is (or in some non-trivial sense could become) exhaustive of what there is to be known; that all facts are ultimately scientific ones; and/or that scientific knowledge and/or facts have a fundamental priority over other forms of knowledge/facts. Anything worthy of the name ‘scientific naturalism’ would at a minimum have to hold to the last of these disjuncts, whereas NSN would reject all of them. NSN as it exists in the contemporary literature is to a large extent the upshot of recent critical attention towards what can seem to be, and its supporters certainly present as, this highly committal idea of SN. For non-scientific naturalists (NSNists, who oppose SNists), it is implausible to hold that all or even the most fundamental facts are scientific. At the same time, rejecting this ‘dogma’ need in no way threaten the naturalistic credentials of one’s view as far a commitment to the wholly physical and biological constitution of human beings and their surroundings goes. 2 / 21-11-13 Two kinds of non-scientific naturalism The critical attention towards SN such considerations have occasioned is very much to be welcomed. The idea that science can answer all our questions, or what this even means, has certainly been subject to too little scrutiny both in the popular press and in more standard naturalistic (often reductionistic) projects within philosophy and cognitive science. My view nevertheless is that NSN’s resistance to the hegemony of science is ultimately unmotivated. The most fundamental reason for this is NSN’s background allegiance to science, which, as I shall seek to show, renders it dialectically inferior to SN. This does not mean that I see SN as thereby vindicated. It may turn out, in the end, that there is no vindication of any overall metaphysical or philosophical position to be had (something I strongly suspect is the case). However, since this is a view I seem to share with many so-called non-scientific naturalists, it is interesting to see how their stand against science breaks down – from the inside, so to speak. I have presented arguments against NSN previously (Knowles 2006), a paper which in turn presupposes arguments in Knowles (2002). In the sequel I partly build on but also seek to clarify and extend these arguments. Further, I distinguish between two kinds of non-scientific naturalism, the second of which requires a substantially different dialectical tack to defeat (there are of course subdivisions within these two broad categories but in the space available here I will have to paint with a broad brush). The first kind of non-scientific naturalism, which I dub expansionist NSN, is exemplified paradigmatically by John McDowell with his notions of second nature and liberal naturalism (cf. especially McDowell 1994, 1997, 2004). Similar ideas are defended by a number of McDowell’s (originally) Oxford-based colleagues, such as Jennifer Hornsby with her notion of naïve naturalism (1997, forthcoming), David Wiggins (cf. e.g. 1987) and Greg McCulloch (cf. e.g. 2002). These views herald from 3 / 21-11-13 Two kinds of non-scientific naturalism the philosophy of mind and metaphysics, but the overarching idea is also evinced in recent epistemological work, such as that of the Susan Haack (1993) and Richard Feldman (2001 – a paper entitled ‘We are all naturalists now’…), as well as informing the general tenor of many other contemporary critiques of SN (cf. e.g. the papers in de Caro & Macarthur 2004). Expansionist non-scientific naturalism seeks, as the name suggests, to expand or liberalise the notion of the natural to encompass as such realms of being – knowledge, rationality, morality and so on – that are often taken as problematic in unreduced form on a more standard construal of what is natural. The second kind of non-scientific naturalism I dub pragmatic non-scientific naturalism, and can be seen as represented by thinkers such Richard Rorty (e.g. 1998), Huw Price (op. cit., forthcoming) and (perhaps) Hilary Putnam (1981, 2004). Pragmatic non-scientific naturalism seeks to deflate the significance of what is usually regarded as the bench-mark of the real, i.e. science, regarding this as just one amongst many competing vocabularies for coping in and with everyday life. At the same time, many of these thinkers also stress the naturalistic credentials of their position – Rorty, for example, in that he explicitly avows that we are biological beings, which explains why our needs in relation to the world are fundamentally pragmatic. In saying this Rorty echoes Quine’s talk of the ‘reciprocal containment’ of science and epistemology (Quine 1969: 83). However, for Rorty, precisely this aspect of our nature – along with the many negative upshots of reductionist projects within the standard naturalistic camp – also renders the search for an absolute bench-mark of the true and the real forlorn. Rejecting this involves a fortiori rejecting the idea that science in particular ‘is the measure of all things’ (Sellars 1963: 1), that it ‘limn[s] the true and ultimate structure of reality’ (Quine 1960: 221). 4 / 21-11-13 Two kinds of non-scientific naturalism In the following I seek to counter these two movements. Part I lays out my case against expansionist NSN (henceforth NSN simpliciter, since it is arguably the more standard and straightforward such position), part II against pragmatist NSN (henceforth NSNP). In the latter I focus on Price’s work, which to my knowledge presents the most honed version of NSNP in the literature, as well as being pitched at a level of sufficient generality to be representative of a broad range of pragmatist positions.1 Before getting under way, I should note that I do not exclude that there might be significantly different forms of NSN than those I consider in this paper. At the same time, the two I critique do it seems clear together constitute a large and significant faction within contemporary philosophy, evincing at root the same kind of general scepticism towards the hegemony of science. Part I: Expansionist Non-Scientific Naturalism At the most general level, there are basically two kinds of argument that are deployed against SN by NSNists (i.e. expansionist NSNists). The first is most characteristically associated with McDowell (opp. citt.) and his followers. McDowell argues that it is mere prejudice to associate the natural with the ‘realm of natural law’, as he puts it, and that we should instead widen our conception of the natural to encompass phenomena associated with our peculiar kind of mindedness as such. McDowell builds on the well-known ideas of Sellars and Davidson that mental phenomena belong to ‘the logical space of reasons’ (Sellars 1963), and hence are in principle irreducible to physical events (Davidson 1980). But for McDowell this incommensurability is not just a matter of modes of understanding; we are in 1 Or so it strikes me. Price relates his own brand of pragmatism to Rorty’s in Price (2008). Rorty endorses Price’s naturalism in Rorty (2007). 5 / 21-11-13 Two kinds of non-scientific naturalism possession of a substantive, albeit ‘second’ nature which involves sensitivity to reasons or, to use the Kantian term, ‘spontaneity’ (rather than mere ‘receptivity’, as is the case in animals). To think saying this involves a lapse into some kind of tanscendent dualism betrays a tendentious and non-obligatory conception of the natural. The second argument is less metaphysically committing, though it can be seen as aiming at much the same conclusion.2 In effect what it contends is that SN is, even from its own naturalistic perspective, either implausible (or at least unjustified) or vacuous. For example, a common variety of SN is physicalism: the view that all properties and objects are fundamentally physical in nature and that everything else we want to countenance as real has to in some way be discerned within the physicalistic universe (cf. Jackson 1998). However, it seems unlikely that science itself furnishes us with anything like sufficient reason to believe in anything beyond a very weak and unthreatening kind of supervenience physicalism, and certainly not the reductionist variety just outlined (see e.g. Dupré 1993; Kornblith 1994.) Physicalism (in this sense) does not seem to be acceptable by naturalists’ own lights and hence should be discarded, whilst weaker varieties hardly enunciate a constraint on any explanatory undertaking or set of concepts. A similar argument can be run with respect to method (or explanation, which I treat under the same heading for simplicity here). A methodological SNist is one who imposes the method of science on all genuinely cognitive enquiry.3 The problem is, 2 The following argument seems to be essentially that used by Haack (op. cit.) and Feldman (op. cit). It is also at work in several papers from a recent volume largely sympathetic to NSN (of both varieties) edited by de Caro and Macarthur (2004); cf. e.g. Putnam (2004), Dupré (2004), Stroud (2004), 3 Such a naturalist is not necessarily opposed to peculiarly philosophical or metaphysical enquiry, but will nevertheless insist that such enquiry employs basically the same methods as those of science (e.g. inference to the best explanation). 6 / 21-11-13 Two kinds of non-scientific naturalism again, that no usefully precise such method, peculiar to all and only science, seems to exist (cf. Knowles 2003: ch. 4.5, Chalmers 1999: ch. 11 for overviews of the relevant arguments and literature). However, without either a metaphysical or methodological basis of some relatively committing kind, it seems SN has very little bite and cannot require us to understand e.g. moral or mental talk in ‘naturalistically kosher’ terms. The conclusion NSNists draw from these considerations is very close to McDowell’s: that what generally goes under the name of (natural) science – physics, chemistry, biology, branches of psychology and social science – does not in fact have any necessarily privileged role to play in delimiting what exists and what can be known. History, literary criticism and other humanities and non-natural social sciences, as well as everyday common sense explanatory practices, have just as much right to serious consideration as natural science – without their being anything ‘supernatural’ about their subject matter. What are we to make of these two (admittedly abstract but nevertheless widespread) arguments? Note to begin with that they are in tension with one another. McDowell’s argument presupposes a certain general conception of what science is, whereas the second presupposes that such a conception essentially does not exist. If we accept, as I think we should, the latter disjunct, then McDowell’s argument is decisively weakened insofar as he cannot appeal to a neatly circumscribed ‘realm of natural law’ with which to contrast the ‘logical space of reasons’ (or ‘the realm of spontaneity’). Of course, he can (and does) also appeal to what he thinks is distinctive Philosophers who espouse conceptual analysis as the method of philosophy (e.g. Jackson op. cit.) are of course not methodological SNists. Often these people are physicalists, but those who are not might it seems still consider themselves scientific naturalists (e.g. David Chalmers). My concern here is with naturalists who essentially balk at the idea of science as the mark of the natural, rather than those might who espouse the idea of a special philosophical method in addition to that of science. 7 / 21-11-13 Two kinds of non-scientific naturalism about the latter realm – that it involves explanations of a special, normative sort. We shall return to this idea below. What of the second argument? According to this, there is simply nothing within extant natural science that ‘puts pressure’, as it were, on traditionally nonscientific endeavours, such as (say) history and common sense psychology. However, though there may be no principled, a priori pressure, it doesn’t of course follow that there is no a posteriori pressure. Below, I will seek to show that what we have learned within contemporary cognitive science suggests a way of understanding the above endeavours in a way that does render them in a significant way subordinate to natural scientific knowledge. Before doing that however it is vitally important to be clear about the dialectical situation here. We must not forget that NSNists defer to science when it comes to physics, chemistry, biology and at least many parts of scientific psychology. At the same time, they want to draw a line somewhere between these disciplines/subdisciplines and full blown scientism – a line which demarcates what, for simplicity’s sake, can be broadly characterised as questions about ‘the mind’, properly so-called. However, on the face of it such a demarcation can seem very arbitrary. Here is Hornsby, considering how we should rule out non-physical, supernatural entities from our ontology: ‘[T]he easy answer is that, a hundred and fifty years on from Darwin, it is a reasonable view that nothing more than natural processes in a world of regular material stuff has ensured the presence here of beings with everyday psychological properties. The alternative can nowadays seem simply unmotivated. [...] I suggest that in the twenty-first century, rather little needs to be said in favour of a rather unexciting sort of physicalism.’ (Hornsby forthcoming: 13-14) 8 / 21-11-13 Two kinds of non-scientific naturalism There is something disingenuous about this statement viewed as a defence of a weak kind of physicalism amenable to NSN (how Hornsby intends it). Nothing very much needs to be said in favour of the general scientific world-view for us to be entitled to it, we are told. But one shouldn’t need reminding that this is precisely an issue of hot controversy across the globe even today, and amongst inhabitants of our own ‘secularized’ societies. Now it may be that there is no ultimately satisfying defence of science vis à vis alternative ‘world-views’ – that little here needs or indeed can in the final analysis ‘be said’. Nevertheless, given we do pin our hopes to science when it comes to physics, chemistry and biology, why the chariness when it comes to psychology and cognitive science? Why are the latter, but not the former seen as up for grabs: as intellectual pursuits one might reasonably demur at (at least to some significant extent)? The rhetoric of NSN is replete with examples of this kind of strategy, whereby science and its basic ontological picture is presented as uncontroversial but also – and, it is often suggested, in virtue of that very fact – impotent in relation to answering questions of a more philosophical nature. But that misrepresents the dialectical situation: science in general is an achievement that we have to stand up for. Of course, more specific reasons for scepticism towards a fullblown cognitive science do exist. The present point is that the natural scientific worldview doesn’t ‘come for free’, and that therefore all naturalists must start with a predisposition towards its general reliability and explanatory potential. Another preliminary point needs to be registered before we continue. This is that being a SNist need not involve holding that everything can be explained scientifically or that everything there is to be known can be known scientifically. There may for reasons famously averred by Chomsky (e.g. 1988) and Fodor (1983) be 9 / 21-11-13 Two kinds of non-scientific naturalism limits on what we can understand scientifically (candidates include consciousness, language-use and scientific reasoning). However, these limits would not enunciate an alternative non-scientific kind of understanding that we are in possession of. NSNists need to establish more than that there are limits to what science can explain. Moreover, they need, as we have seen, to do this consonant with accepting and indeed promoting the achievements of science generally. Let us now return to the relationship between scientific psychology, on the one hand, and humanistic understanding and common sense psychology on the other. The latter ‘disciplines’ are often seen as non-scientific insofar as they employ a certain mode of understanding or method allegedly peculiar to the human sciences: the method of Verstehen (‘understanding’), as it is often termed (due to its prevalence in hermeneutically-inspired German philosophy). Exactly what Verstehen is has perhaps never been laid out wholly explicitly, but that something like it exists, in some form or other, is not something I want to question. It seems clear that we humans do deploy an intuitive form of understanding for predicting and explaining our fellows, revolving around the concepts of action, belief, meaning and (something like) desire that constrain one another holistically. In employing this understanding, we do not invoke, at least explicitly, laws or principles of explanation. It also seems reasonable to think that the study of history (etc.) extend this kind of everyday understanding to more recherché topics of study, such as temporally or spatially remote cultures, texts etc. However, the question for us is whether this intuitive ‘hermeneutical’ framework can not be accommodated and accounted for within whatever uncontroversially natural scientific approaches to the mind are on offer today, or at least which it is reasonable to think will be on offer in the near future. If it can, though Verstehen could perhaps still be regarded as a special method for the human sciences, its 10 / 21-11-13 Two kinds of non-scientific naturalism viability as such would be dependent on facts revealed by natural science, and in that sense the latter will presumably have to be viewed as providing our most fundamental conception of human nature – from within which the activity of Verstehen is itself an object for scientific explanation. To be a little bit more concrete, the idea within contemporary cognitive science is that we possess various intuitive, perhaps even in a sense innate ‘modules’ of understanding – naive (or folk) psychology, physics, biology and so on – whose use in unrefined form is a cognitively useful and significant activity for our organism.4 We have a need to and therefore have evolved or developed a capacity for understanding other agents (inter alia) without engaging in science (a capacity that history and the rest simply extend to more recherché matters). Exactly what form such a module might take, what the capacity consists in (knowledge or off-line processing) and so on are issues of hot controversy within contemporary cognitive science. Nevertheless, the basic idea of folk psychological explanation and prediction being something we have a naturally explicable capacity for is not controversial. The availability of this kind of explanation, coupled with NSN’s general commitment to science, surely puts a deep dent in the idea of NSN insofar as this denies all fundamental knowledge is scientific. Even so, it does not challenge the autonomy of Verstehen per se. For some NSNists, Verstehen and science answer to fundamentally disparate explanatory projects. Here is a famous passage from McDowell: [P]ropositional attitudes have their proper home in explanations of a special sort: explanations in which things are made intelligible by being revealed to be, or to 4 In relation to folk psychology, see e.g. Wellman (1990). For a more general overview of the area, see e.g. Carruthers (2006): ch. 3. 11 / 21-11-13 Two kinds of non-scientific naturalism approximate to being, as they rationally ought to be. This is to be contrasted with a style of explanation in which one makes things intelligible by representing their coming into being as a particular instance of how things generally tend to happen. (McDowell 1985: 389) According to this contrast, Verstehen is not (merely) an intuitive explanatory framework or a kind of proto-science; the insight it provides is of a fundamentally different kind from that of science. However, this line faces many and I think decisive problems, as several authors have remarked (e.g. Henderson 1993, Knowles 2002). Unfortunately the issue is too large to go into in detail here, but two main points that can be registered. The first is that the position again involves a specific conception of scientific explanation that we saw above reason to be suspicious of. Science is not a priori restricted to causal-cum-law like explanations, a fortiori nor is cognitive science. Indeed some recent work aims precisely to take account of Verstehen’s explanatory features – often described as characterizing a so-called ‘personal’ level of explanation – within a nonreductive but nevertheless fully scientific framework (cf. Hurley 1998). The second problem is that I think, in the final analysis, we can make nothing of the idea of primitively and irreducibly normative explanation, that is, of explanations that adumbrate what ought to be the case as explanatory – in and of itself – of things that are the case, such as someone’s having done or thought something. This is plausibly the only idea that might really distinguish what people have called ‘Verstehen’ from a more general notion of explanation employed in science or in relation to everyday events (like windows smashing); but it is not, unfortunately for this way of defending NSN, one that it seems we can give a coherent sense to (cf. Knowles 2002). 12 / 21-11-13 Two kinds of non-scientific naturalism A possible response to this, discernable in the work of Hornsby (1997), is that what distinguishes folk psychology from science is something that distinguishes explanation at the level of ordinary, common sense objects and events generally – persons, actions, windows smashing, cricket balls etc. – from that of science. Put bluntly, science is concerned to show how very small events can generate ordinary sized ones, but there is no reason to think common sense, macro-level explanations, of either physical or psychological events, will be reducible to such micro-processes. But this line too faces insuperable problems. Firstly, there is a much greater acceptance that science acts as a corrective to common sense in the case of physics than in psychology. This undermines the parallel Hornsby builds on: whilst it is reasonable to think (as things stand) that at least one perfectly good explanation of why John threw the cricket ball was that he wanted to hit the stumps, it is not nearly so clear that the breaking of the window is as well explained by saying merely that the cricket ball hit it. Secondly, the argument, even if accepted, does not in any case undermine SN, at least as we have been developing it, since this eschews the kind of reductive programme for science the argument presupposes. Finally, much actual science (geology, zoology etc.) seems precisely concerned with macro-level objects and states of affairs as such. I conclude that the idea of a thoroughly autonomous level of common sense explanation for a thoroughly autonomous common sense world is without plausibility (cf. Knowles 2004).5 5 A slightly different response, also discernible in Hornsby’s work, would stress the idea of folk psychology as essentially involving talk of persons (though without any commitment to specifically normative explanation). This I think is not wholly implausible, but in light of a) the fact that animals would also seem to qualify as persons in the relevant sense b) the availability of non-reductive empirical work incorporating precisely the idea of personal-level explanations (Hurley op. cit.), this also strikes me as inadequate in holding SN at bay. 13 / 21-11-13 Two kinds of non-scientific naturalism Many NSNists will of course still demur at the idea of a scientific intentional psychology – or for that matter at a scientific explanation of our folk psychological practices. John Dupré (2001) sees such projects as a manifestation of atavism, bolstered by an ungrounded prejudice against the humanities. But, rhetoric aside, my plaint here has been to sketch an underlying dialectical situation that makes it hard to see how NSN can avoid the slide into SN. From the perspective of science generally, which NSNists endorse, I take it as uncontroversial that contemporary cognitive science both a) comprises a genuine research programme for explaining the cognitive basis of our folk psychological practices b) encourages a not wholly pessimistic attitude concerning the possibilities for scientifically understanding intentional psychology itself. Given a), NSN is already compromised insofar as scientific knowledge must then be viewed as fundamental relative to whatever disciplines our folk psychological capacities underlie (history, literary theory etc.). As for point b), this is formulated, deliberately, to allow pessimism to be something one might reasonably mount a case for. However, given that cognitive science itself opens for such pessimism (cf. the discussion of Chomsky and Fodor, above), NSN must be confessed to be on dialectically very thin ground. From within the general scientific world-view that NSNists themselves endorse, the derision often heaped on ambitious cognitive science can be explained without endorsing any alternative, non-scientific modes of explanation. NSNists cannot by contrast explain away what is surely a reasonable scepticism towards the idea of non-scientific modes of understanding, once McDowell’s idea of a fundamentally different kind of insight and Hornsby’s notion of an autonomous common sense world are rejected. Dividing through, what remains is the reasonable scepticism. 14 / 21-11-13 Two kinds of non-scientific naturalism Needless to say, this doesn’t end the debate between opponents and proponents of (ambitious) cognitive science, but then that would hardly be something a single paper could aim to do. What I do think I have shown is that, conceived at a certain level of abstraction, the ball is in the NSNists’ court. The very general arguments that its supporters standardly employ against SN are inadequate to dislodging its place as our default form of naturalism. Part II: Pragmatist non-scientific naturalism I turn now from expansionist to pragmatic non-scientific naturalism (NSNP), a position I see as most clearly articulated and defended in a couple of recent papers by Huw Price (opp. citt.). In Price (2004) it is argued that a proper deference to science on the part of philosophers does not entail the position most commonly known as ‘scientific naturalism’ in contemporary analytical philosophy. According to this position, which Price himself terms object naturalism (henceforth ON), all there is is the world studied by science and all knowledge is scientific knowledge. The challenge for ON is to show how talk that on the face of it does not latch on to this world – talk about causation, modality, mathematics, intentionality, consciousness, morality and so on – nevertheless does so, or else can, with justification, be regarded as essentially meaningless, false, or at least ‘second rate’. Price is highly suspicious of ON. However, he does embrace a position he calls subject naturalism (henceforth SuN), according to which ‘philosophy needs to begin with what science tells us about ourselves’ (ibid.: 73) – including our linguistic practices. According to Price, SuN is not, as one might initially surmise, a corollary of ON, but rather a view potentially, and perhaps even actually, in a position to undermine ON. In a word, science itself 15 / 21-11-13 Two kinds of non-scientific naturalism may or even will show us that there are limits to what can be understood scientifically. Price argues as follows. On the standard ON-strategy, we begin with talk of something, call it ‘X’, and the question arises as to what place such talk can have in relation to the world described by science. What, in particular, are the occupiers of the causal roles identified by this talk? However, this strategy assumes that our talk of ‘X’ links us to something out there in the world, X itself. Of course, it is a platitude that ‘X’ refers to X (or Xes). However, as the possibility of minimalist or deflationist theories about reference should remind us, this is not necessarily a substantive truth, as opposed to a mere bi-product of the existence in our language of a device that allows us to say things about what people say, about words in general, and so on. To say “‘snow is white’ is true”, for deflationists, is not to say anything more about the world than to say ‘snow is white’. Something similar, it seems, might well apply to ‘reference’, and if so there is no automatic route from semantic platitudes to theories about how a certain kind of role is filled in reality. Price thinks this kind of semantic deflationism is a live option when one takes a SuN-approach to language. Moreover, semantic deflationism avoids a problem for a more substantive naturalistic semantic theory which Paul Boghossian has pointed out, namely that, in being empirical, it leaves open the possibility that there may be no determinate semantic relations; in thus opening for the empirical possibility of semantic irrealism substantive semantics is incoherent, for it implies the same holds for the semantic terms themselves, which would disallow a formulation of the irrealistic thesis (cf. Boghossian 1990).6 Moreover, even if this objection fails and one 6 Boghossian’s paper also argues that deflationism is incoherent because it must take an irrealist stance with respect to non-deflationary truth and reference, which presupposes a robust notion of truth. Price (following a number of others) argues that 16 / 21-11-13 Two kinds of non-scientific naturalism could motivate a non-deflationary semantics, there is still what Price calls ‘Stich’s problem’, to the effect that questions about what there is in the world aren’t determinate given what we say about them (cf. Stich 1996). Stich used to worry whether or not there were beliefs, and gave a negative answer based on the falsity of folk psychology relative to a mature cognitive science. But does that really show there are no beliefs – or just that beliefs are very different from what we took them to be? We seem to be left with an indeterminacy in reference of a kind that again undermines the ON-approach to naturalization. Now merely adopting a SuN approach to language is not to embrace semantic deflationism – though Price seems to suggest that it might well lead to this (2004:. 82). Moreover, as Price goes on to argue, if semantic deflationism is embraced, then the standard arguments for ON fail to go through, for then we cannot frame the kind of question object naturalists (such as Lewis and Jackson) standardly ask: What are the truth-makers for such and such semantic roles?7 Ontological questions become instead internal to discourses, or perhaps lapse entirely. Science itself, then, as applied through the project of SuN, will show us that there are certain limits on what we can expect to explain scientifically. This suggests that, insofar as giving answers to questions about what exists and what is true are central to the project of ON, then ON deflationism can get around this objection by not asserting that there either exist or do not exist such substantive relations. 7 Price also considers a non-semantically based version of ON that starts not with language but the world, so to speak (ibid.: § 7). The problem with this is that insofar as what we are seeking to motivate is a metaphysical picture (‘the world is as described by science’), it fails to provide any overall motivation for this of the kind we had with the semantic picture, where truth-makers are seen as physical occupiers of causal roles. A possible response (made by the author in discussion with Price) would simply be to stress the role of causation as a criterion for the real, but as Price pointed out this merely begs the question of the ontological primacy of causal talk (cf. also Price forthcoming on ‘eleatic naturalism’). 17 / 21-11-13 Two kinds of non-scientific naturalism is misconceived. Moreover, insofar as ON just is SN – our scientific naturalism – then so is the latter. In my view, however, it would be a mistake to identify SN with ON. Moreover, I think SuN is unconvincing as a non-scientific naturalistic alternative to SN. To begin to see this, consider first what seems to be an important ambiguity in the notion of SuN, as Price deploys it. As noted, SuN receives its central treatment in Price (2004), but he also discusses it in Price (forthcoming), where the emphasis is rather different. In Price (2004), the main idea seems to be that SuN has a substantive philosophical role to play – that of undermining ON. The problem with this is that it doesn’t seem to involve any kind naturalistic investigation at all. Price clearly wants SuN to involve recommendation of such investigation, such that it will be an open question whether an investigation of our linguistic practices will yield a substantive or non-substantive semantics: ‘The subject naturalist’s task is to account for the use of various terms – among them the semantic terms themselves – in the lives of natural creatures in a natural environment.’ (ibid.: 82). She may in this task ‘simply find no need for an explanatory category of semantic properties and relations’ (ibid.). However, it turns out that this need not and had better not be construed as ‘an investigation into whether there are semantic properties’ (ibid., my emphasis), on pain of falling foul of Boghossian’s argument (see above). But where does that leave us? It seems it leaves us in the position of having to say that SuN amounts to a more or less a priori insight that naturalistic representational semantics is bankrupt. Maybe that is right (as several other thinkers have argued). What I don’t see is how it is something recognizably scientific, i.e. an empirical study, that is meant to be showing us this. Now SuN is clearly intended by Price to be a position that involves naturalistic investigation into the different functions of different discourses play in our lives – 18 / 21-11-13 Two kinds of non-scientific naturalism moral, aesthetic, causal, mathematical and so on. He might thus say that it does indeed have an a priori element, as outlined above, but that rejecting substantive semantics still leaves a lot of work to be done in saying how our different discourses function. As he expounds upon it in Price (forthcoming), SuN is the project ‘of explaining our linguistic practices – for that way, if all goes well, lies a scientific foundation for the suggestion that different parts of language serve different functional ends’ (ibid.: 24). Here Price also makes clear that he sees our linguistic practices as essentially ontologically committing, and hence plurality at the level of discourse amounts to a kind of ontological plurality, something which can sound very anti-scientistic (i.e. contrary to the hegemony or priority of scientific knowledge). However, if we bracket the ontology issue for the moment, what does the project of SuN as just described really amount to? I would seem to amount to a kind of naturalized epistemology of the kind cognitive science seeks to undertake, some of whose elements we outlined Part I. In other words, it involves seeing language’s conceptual functions in relation to our overall functioning as a biological organism. Moreover, if this is right, it surely thereby involves giving the discourse of scientific enquiry a priority over other discourses, something that any genuinely non-scientific naturalism would abjure. This may sound too quick in neglecting the ontological commitments of the different discourses. Doesn’t plurality here really sound the death-knell for a thoroughgoing scientific world-view? Well, I am prepared to admit that from, say, a moral or mathematical perspective our ontological commitments will look rather different than from a natural scientific one. I can also allow, for the sake of argument, that our talk in each case is equally cognitivist in its structure, allowing the construction of logical argument and reference to worldly properties that underscore 19 / 21-11-13 Two kinds of non-scientific naturalism the point of such argumentation. Nevertheless, if the point of all this in the different cases is something the scientific discourse is going to tell us and inform us more deeply about, then surely we are tacitly assuming the priority of that body of knowledge, i.e. what we call ‘science’. Not just what we want to say exists within each discourse or language game, but also the fact that we make such commitments at all is in thrall to the scientific – in this case, psychological or anthropological – story about why we do all this. It is misleading, then, to sum up, as Price does, by saying that ‘science is just one thing among many that we do with the linguistic tools of ontological commitment’ (ibid.) – to say that science presents us with just another ontological view from just another perspective. For the very notion of what such views and interests are is, by Price’s own reckoning, one that is beholden to scientific understanding. In a word: Given questions about ontology are posterior to, since responsible to questions about, explanation, then plurality with regard to the former does not impugn a scientific naturalism which is (in a certain important sense) monistic with respect to the latter. On this second reading of SuN, then, it does emerge as a kind of naturalism, but it does not enunciate any deviation from SN – from which it clearly follows, insofar as SuN is not ON, that SN does not equal ON either. Now of course, whether there really can be a scientific representational semantics is big issue, and not one I can hope to say anything useful about here. I am in fact sympathetic to deflationism and something like Price’s pragmatist line on ontological commitment, but that is not to the point in the present context. What is to the point is that SuN either is not distinct from SN or else isn’t a way of being a naturalist at all, but merely a way of registering a certain skepticism to a certain kind of naturalistic project (perhaps in and of itself consistent with SN). 20 / 21-11-13 Two kinds of non-scientific naturalism Since I take Price’s views to provide an explication of the essentials of the view I have called NSNP, I also conclude that the latter fails to represent any significant alternative to SN. Taken together with the conclusion of Part I, nonscientific naturalism generally fails in a plausible way to distance itself from SN and must be considered undone. Stepping back from the details, I think we can see a link between the two parts of the paper. Underlying Price’s papers is a scepticism to the project of substantive naturalistic semantics and the associated idea of metaphysical realism. This is perhaps most clearly evinced in Price (forthcoming), where he attempts to show that Quine’s criterion for ontological commitment is either super-scientific and non-naturalistic or else lacks any kind of philosophical significance and is merely an internal affair: a matter of a science finding its own postulates reasonable. Ultimately, what one can countenance as existing must assume a certain practice, but practices are many and varied and not all of them scientific. Hence it can look as if SN is going to be in trouble, in lacking the wherewithal for vindicating the idea that all significant ontological commitment is to scientific objects and thereby all significant knowledge scientific. As we have just seen, however, this commitment is a not mandatory component of SN, as witnessed by the availability of Price’s SuN which, though clearly not the same as ON, seems only nominally distinct from SN. But even without the idea of SuN (only fully developed in Price 2004), a reasonable question to ask Price would seem to be: What positive programme is one left with on your conception of ontological commitment? Should we stick at being idealists or ontological relativists of some stripe? Most philosophers, in any case, would want to be allowed to say that we humans are – really, really are – fully contained in the world of spacetime and evolved organisms, even though they will also say we can make sense of it 21 / 21-11-13 Two kinds of non-scientific naturalism in various ways, perhaps some non-scientific. Thus are we returned to the dialectic played through in Part I above, the upshot of which was that non-scientific naturalism is not a dialectically stable position. Conclusion None of what I have said in this paper comprises an argument for SN that a sceptic of science or of naturalism generally would have to accept. However, the difficulty of maintaining a naturalism that does not fundamentally defer to science does strike me as a significant result. As a final note, one should not get the impression that I place absolutely no weight on broadly epistemological arguments for naturalism, of the kind for example which says that natural science offers the hope of the most all-encompassing, coherent and consistent ‘picture of the world’ that we might ever hope to find. Having said that, there are almost certainly limits to this kind of argument, revolving centrally around the extent to which science can show itself to be precisely such a picture (cf. Knowles 2008). If scientific practice can show itself resourceful in the face of challenge, albeit only in a local, piecemeal fashion, that is perhaps some reason to stick with it. Ultimately, however, a thoroughgoing scientific naturalism will I think need to ally itself to a kind of quietism, according to which one is to a certain extent simply beholden to the practices of the dominating cognitive pictures and practices of one’s epoch and culture (cf. Knowles 2003: ch. 1 for further discussion). This need not, however, be viewed as a concession, for such quietism is arguably a concomitant of any position worthy of the epithet ‘naturalism’. Moreover, if, as seems reasonable to hold, science and scientific understanding do represent such a dominating practice in contemporary Western society, then the best option will arguably be to work within 22 / 21-11-13 Two kinds of non-scientific naturalism this framework – not in the reductive way some naturalists tend to do, but nevertheless such that one doesn’t neglect or renounce sources of insight which one’s own practices in any case implicitly endorse.8 References Boghossian, P. (1990) ‘The status of content’ The Philosophical Review, April: 15784. Carruthers, P. (2006) The Architecture of the Mind: Massive Modularity and the Flexibility of Thought, Oxford UP. Chalmers, A. (1999) What is This Thing Called Science? 3rd edition, Open University Press. Davidson, D.(1980) Essays on Actions and Events, Oxford UP. de Caro, M. & Macarthur, D. (2004), eds., Naturalism in Question, Harvard UP. Dupré, J. (1993) The Disorder of Things: Metaphysical Foundations of the Disunity of Science, Harvard UP. (2001) Human Nature and the Limits of Science, Oxford UP. (2004) ‘The miracle of monism’ in de Caro & Macarthur (2004). Feldman, R. (2001) ‘We are all naturalists now’, ms., delivered at APA meeting 2001 (available at http://www.ling.rochester.edu/~feldman/papers/naturalism.htm). Fodor, J. (1983) The Modularity of Mind, MIT Press. Haack, S. (1993) Evidence and Inquiry, Blackwell. Henderson, D. (1993) Interpretation and Explanation in the Human Sciences, State University of New York Press. Hornsby, J. (1997) Simple Mindedness, Harvard UP. (forthcoming) ‘Physicalism, conceptual analysis and acts of faith’ in I. Ravenscroft, ed., Minds, Worlds & Conditionals: Essays in Honor of Frank Jackson, Oxford UP (page references to ms.). Hurley, S. (1998) Consciousness in Action, Harvard UP. Jackson, F. From Metaphysics to Ethics, Oxford UP 1998. Knowles, J. (2002) ‘Is folk psychology different?’ Erkenntnis 57: 199-230. (2003) Norms, Naturalism and Epistemology: The Case for Science without Norms, Palgrave Macmillan. (2005) ‘Varieties of naturalism’ in W. Østreng, ed., Synergies. Interdisciplinary Communications 2003/2004, Centre for Advanced Studies, Norwegian Academy for Science and Letters, Oslo. (2006) ‘Non-scientific naturalism?’ in S. Pihlström, R. Vilkko & H. Koskinen, eds., Science – A Challenge to Philosophy? Peter Lang. 8 Earlier versions of this paper were presented at KNEW’06 in September 2006 in Kazimierz, Poland, at Apeiron’s ‘Aporetisk aften’ at NTNU in November 2006, and at the Fifth Conference of the Spanish Society of Logic, Methodology and Philosophy of Science in Granada, Spain, December 2006. I would like to thank Bjørn Ramberg and (especially) Huw Price for their comments on the first occasion, Kevin Cahill and Siv Dokmo for their comments on the second, and Jesús Vega Encabo for his comments (both oral and written) on the third. Thanks finally to David Macarthur for written comments on the version presented at KNEW’06. 23 / 21-11-13 Two kinds of non-scientific naturalism (2008) ‘Is naturalism a threat to metaphysics?’ Norsk filosofisk tidsskrift, 43, 23-32. Kornblith, H. (1994) ’Naturalism: both metaphysical and epistemological’ in P. French, T. Uehling, & H. Wettstein, eds., Midwest Studies in Philosophy, Vol. XIX: Philosophical Naturalism, University of Notre Dame Press. McCulloch, G. (2002) The Life of the Mind: An Essay on Phenomenological Externalism, Routledge. McDowell, J. (1985) ‘Functionalism and anomalous monism’ in B. McLaughlin & E. Lepore, eds., Actions and Events. Blackwell. (1994) Mind and World, Harvard UP 1994. (1996) ‘Two kinds of naturalism’ reprinted in his Mind, Value, and Reality, Harvard UP 1998. ‘Naturalism in the philosophy of mind’ in de Caro & Macarthur (2004). Price, H. (2004) ‘Naturalism without representationalism’ in de Caro & Macarthur (2004). (forthcoming) ‘Quining naturalism’ (page references to ms.). (2008) ‘One cheer for representationalism’ in R. Auxier, ed., The Philosophy of Richard Rorty, Open Court Library of Living Philosophers XXXII. Putnam, H. (1979) Reason, Truth and History. Cambridge UP. ‘The content and appeal of naturalism’ in de Caro & Macarthur (2004). Rorty, R. Truth and Progress, Cambridge UP 1998. (2007) ‘Naturalism and quietism’, in his Philosophy as Cultural Politics, Cambridge UP. Sellars, W. (1963) ‘Empiricism and the philosophy of mind’ in Science, Perception and Reality, Routledge. Stich, S (1996) ‘Deconstructing the mind’ in his Deconstructing the Mind, Oxford UP. Stroud, B. ‘The charm of naturalism’ in de Caro & Macarthur (2004). Wellman (1990) The Child’s Theory of Mind, MIT Press. Wiggins, D. Needs, Values, Truth, Blackwell 1987. 24 / 21-11-13 Two kinds of non-scientific naturalism Abstract In this paper I critically evaluate two different kinds of what I call ’non-scientific naturalism’ (NSN). ‘NSN’ denotes views which, whilst not supernaturalistic or opposed to science, do oppose standard versions of naturalism which see scientific knowledge as exhaustive of what can be significantly known or at least as in some way fundamental. I argue that, at least in relation to the two broad ranges of positions I consider, NSN suffers from dialectical weaknesses which makes its resistance to the hegemony of science unmotivated. The first kind of NSN is represented by e.g. John McDowell’s liberal naturalism and Jennifer Hornsby’s naive naturalism. I call this kind of view expansionist NSN, since it seeks to motivate an expansion of what we regard as the realm of the natural beyond what is comprehensible to standard science, most notably to accommodate the specifically human trait of rational mentality. The other type of non-scientific naturalism I term pragmatic NSN, which is exemplified paradigmatically by the vocabularism of Richard Rorty. The idea here is that as a matter of scientific fact language has a far greater range of functions than scientifically describing the world, which moreover has no particular priority over these other functions. I argue that expansionist NSN’s idea of the need for an expansion of the notion of the natural is not well-grounded. McDowell et al are right to oppose physicalist-cum-reductive models of scientific explanation as adequate for understanding all natural phenomena, but there is no reason why scientific naturalism should be committed to such models. In response supporters of NSN will question whether scientific naturalism thus understood can non-arbitrarily demarcate science from what is generally regarded as non-(natural) science, such as history, social science and perhaps even philosophy. I argue that this tactic fails. In relation to pragmatist NSN I focus on some recent work by Huw Price, since he explicitly defends a pragmatist version of naturalism, ‘subject naturalism’, which he distinguishes from what he calls ‘object naturalism’ and equates with standard scientific naturalism. Object naturalism presupposes a substantive semantic starting point for enquiry, something Price argues is suspect from the point of view of a scientific account of our use of language (a subject naturalist perspective). However, Price’s subject naturalism turns out to presuppose a priority for scientific approaches that reveals that it is not really distinct from what is reasonable to regard as a scientific kind of naturalism, a view which mutatis mutandis should not to be equated with Price’s object naturalism. 25 / 21-11-13