Attached is a link to my slides from a talk I presented at the Humanist Convention in April of 2017 (last week as at the time of posting). They summarise some of my recent thinking about metaphysical naturalism, an argument in defence of which constitutes the majority of the talk. I hope they may be of use and interest to some. Eventually I will get around to writing up my thoughts into a proper article, which I will post on my blog. I anticipate, however, that it will be significantly more technical than these slides, so these may make a good ‘first introduction’ to some of these issues for people with less philosophical background.
In this piece I consider the two related concepts of ‘moral facts’ and ‘moral obligations’, contrasting them within theistic and naturalistic worldviews. I first consider what is meant by ‘moral facts’, and argue that, subject to a certain clarification regarding the meaning of ‘mind-independence’, objective moral facts can exist within a naturalistic framework, as facts concerning states of affairs relating to idealised human desires. I then consider the concept of ‘moral obligations’, and argue that such obligations may be consistent with naturalism depending upon how the notion of ‘moral duty’ is interpreted. I also argue, however, that the concept is essentially unnecessary in a naturalistic worldview, as it does little beyond what is done by the concept of ‘moral facts’. I conclude with some analysis of how theists and naturalists may respond to the moral skeptic, arguing that neither can provide moral motivation to the skeptic on the basis of reason alone.
The first question to be considered is whether or not ‘moral facts’ exist. For a moral fact to ‘exist’, what I mean is that the proposition in question is true. Thus, the question I am asking is whether any propositions about moral states of affairs are true, a view called moral realism, as opposed to error theory, the position that all moral propositions are false. (There are also so-called non-cognitivist positions which hold that moral statements are not propositions at all. I will not address such views in this piece.)
To facilitate clarity, let me propose a working definition of objective moral facts:
(1.0) An objective moral fact is an evaluative proposition concerning the ‘rightness’ or ‘wrongness’ of some action or outcome, which is true irrespective of the mental state (opinion, belief, etc) of any person.
In my personal view, I think it unlikely that objective moral truths as defined in (1.0) exist, as I believe that the rightness or wrongness of any action is always ultimately determined by the mental states of human beings (and potentially other sentient creatures too, but I’ll leave that out of the discussion for now). According to the view that I lean towards, moral facts are propositions concerning the maximal fulfilment of idealised preferences, considered from a social point of view (see my earlier piece describing Railton’s Reductive Naturalism for more detail).
In keeping with this view, I would propose a new definition of moral facts:
(1.1) An objective moral fact is an evaluative proposition concerning the ‘rightness’ or ‘wrongness’ of some action or outcome, the truth of which is not dependent upon the mental state of the agent making the moral judgement in question.
The difference between (1.0) and (1.1) is that, while (1.0) requires that the truth values of all moral propositions be independent of the mental states of any person (i.e. they are facts about nature itself independent of humans, or perhaps something beyond nature), (1.1) only requires that the truth values be independent of the mental state of the person evaluating the claim. Thus, if one person or group thought that genocide or female genital mutilation or ignoring the poor were not morally wrong, by this view they would be mistaken. They would be mistaken because they hold a false belief concerning the truth value of a certain moral proposition, which proposition derives its truth value from particular states of the world concerning which states of affairs would be conducive of the maximal fulfilment of idealised desires, from a social point of view. Moral facts are thus facts about the world with objective truth values independent of the mental states of those evaluating the truth of the claims.
Also note that according to my preferred view, it is even possible for everyone to be mistaken about moral facts. This is because the truth value of moral propositions does not depend (primarily, though it may have some relevance in some cases) upon people’s opinions concerning the truth value of the proposition. Rather, moral propositions derive their truth value from states of affairs concerning idealised preferences of agents considered from a social point of view. It is perfectly possible for entire societies to hold systematically mistaken views regarding such idealised preferences – indeed, I think I can cite some plausible historical examples, though I won’t do so here because I’m fairly sure that doing so would distract the discussion. The main point to note is that, although according to my preferred position, moral facts are subjective in the sense that their truth value is dependent upon the mental states of humans, and not merely upon natural states of affairs outside of humans (as are, for instance, many scientific claims), they are also objective in the sense that their truth value is not determined by the attitudes or preferences of the person making the judgement, or even the collective judgements of a society, since it is possible for an entire society to hold mistaken views concerning what would best satisfy idealised preferences from a social point of view.
Having considered objective moral facts, what can we make of the idea of moral duties? It seems that the mere existence of moral facts, absent certain further assumptions, need not necessarily imply any moral duties – after all, there are any number of other propositions which are objectively true, but nonetheless do not entail any duties.
Let me (tentatively) define moral obligations as follows:
(2.0) A moral obligation is a duty to act in a certain way that arises as a consequence of one or more objective moral facts.
While I think this definition goes some way towards capturing our primitive notion of ‘moral obligation’, I am left rather unsatisfied. I still find it very hard to understand what is meant by this notion of a ‘duty to act’ -what does it mean to say that we have a duty to do something? Sometimes duties are acquired on the basis of someone accepting a formal or informal position of some authority and responsibility, and explicitly or implicitly promising to act in a certain way in fulfilment of this role. It seems, however, that this does not really capture the inherent proscriptivity entailed by our concept of ‘moral duties’. That is, we would generally want to say that there is no action that one needs to take in order to acquire moral duties, nor is there any way of eschewing them, as would be possible for other duties by, for example, stepping down from the role in question.
The idea of ‘moral duties’ seems to be that, in some sense, we “must” act in a particular way, regardless of whether or not we want to, or whether or not we agree, or even whether or not we even know about the duty (though some may perhaps dispute this last point, at least my naive notion of ‘moral duty’ would say that even, for instance, feral children would have moral duties, even though they would presumably have no notion of the concept of morality). But what does it mean to say that we “must” act in a certain way? Obviously this doesn’t mean that we are literally unable to act differently, because quite clearly it is possible to act immorally.
One possible answer, traditionally advocated by some theistic philosophers, is to ground the notion of ‘moral obligation’ in the commandments of God. That is, moral obligations are injunctions to act in a particular way which are made by God, and are (ultimately) enforced by God through some sort of final judgement. The notion of ‘moral obligation’ is thus analogous to that of a legal obligation – both derive from some external authoritative source, both are binding regardless of our particular attitudes or opinions, and both are ‘enforceable’ in the sense of there being consequences for disobedience.
This would lead to a definition something like the following:
(2.1) A moral obligation is an enforceable injunction to act in a certain way, deriving from some legitimate authority ‘external’ to human preferences or opinion.
I think there are various problems with approaches such as this to ground moral obligations on God’s commandments. For example, I think it is at least plausible that one may acknowledge an injunction to come from God, but still question whether or not obeying is the right thing to do. It seems to me that God could at least potentially be evil, and that therefore moral duties are not constituted solely in the injunctions of God, but have reference to things outside of God as well. I’m not saying these and other issues are necessarily insoluble, nor do I wish to get distracted into an extensive debate about them here. I just wanted to flag them as being tangentially relevant before moving on.
Let us suppose, however, that we can develop a consistent and plausible theory of theistic moral duties which circumvents some of the difficulties mentioned above. Can the same be done from within a naturalistic worldview? I think doing so is at least conceptually possible – it seems for example that a principle like karma would go some way towards meeting the criteria set out by (2.1), and at least some understandings of karma see it as essentially a completely natural phenomenon. However, I personally do not believe in karma, or any such natural process like it. As such, I would lean towards the view that, if there is no God, then moral obligations as defined in (2.1) do not exist.
In essence, I lean towards the view that the notion of ‘moral obligations’ is essentially unnecessary in a naturalistic universe, and is really only the cause of conceptual confusion. I believe, as I argued above, that objective moral facts as defined by (1.1) are perfectly capable of existing in a naturalistic universe, and that there is simply no place for or need of ‘moral duties’ that go beyond moral facts. So, for example, I do not believe that it is necessary to interpret a statement like ‘you should behave in this way’ as a statement about the existence of moral obligations or duties. Rather, I think it is perfectly consistent and sufficient to interpret this as an assertion of the proposition ‘behaving in this way would promote the maximal fulfilment of idealised preferences from a social point of view’, as well as an injunction to act in accordance with this fact (this notion of moral assertions constituting injunctions relates to another model in meta-ethics that I find persuasive, R. M. Hare’s Universal Prescriptivism, however I do not wish to get too distracted discussing that in detail).
Furthermore, I do not believe that the mere truth of particular moral facts provides any rational obligation to act in accordance with them. That is, those who ignore morality are not necessarily irrational, they are just immoral. Do I believe that the truth of moral facts provides any moral obligation to act in accordance with them? It depends upon what is meant. If by ‘moral obligation’ one means something like (2.1), then no, I do not think moral facts entail moral obligations (since the moral facts are not injunctions from an external authority in the sense required for moral duties). On the other hand, I think a lot of people talk about ‘moral obligations’ more loosely as essentially referring to the same thing as ‘moral facts’, and in this looser sense I do tend to think that moral duties exist, because (as I argue above) I tend to believe that moral facts exist.
Responding to the Moral Skeptic
So where does all this leave us? Certain theists tend to phrase this discussion in terms of having a response to the ‘moral skeptic’, who when confronted with a moral claim, asks question like ‘why should I?’ or ‘what privileges your view over mine?’ I believe that, working within the framework of Railtonian Reductionism that I have outlined here and elsewhere, the naturalist can provide answers that are at least as satisfactory as those the theist can give (I personally think they are much better, but that’s a stronger claim I won’t attempt to defend in full in this piece).
The theist could answer (something like) ‘you should because God commands it, and he is our creator and so has legitimate authority over such things’. It seems to me, however, that the moral skeptic could acknowledge that God exists and mandates particular commandments, but still either dispute that they are morally obliged to follow these commands, or even just fail to care about divine moral obligations, and not feel motivated to live up to them. It seems to me such a person has not committed any mistake of rationality here – they just don’t care what God has to say on the matter, and so far as I can tell this violates no precepts of sound reasoning. It may, of course, make them an immoral person, but there seems little else the theist could say to motivate or convince them.
The naturalist could answer (something like) ‘you should because doing so would better promote the fulfilment of idealised desires from a social point of view’ (this is often described less verbosely using language like ‘promoting human flourishing’ or ‘maximising wellbeing’). Of course, the moral skeptic could acknowledge this to be the case, but still dispute that they have any moral obligation to promote human flourishing, or simply fail to care and see no reason to act in accordance with any moral facts or duties that may exist. As before, such a person may be immoral, but as far as I can see they have not committed an error of rationality, and as such there seems little else that could be said to motivate or convince them.
Thus, at the end of the day, I think that neither the theist nor the naturalist can convince the moral skeptic to follow the precepts of morality using reason alone, which perhaps they may antecedently have wished to do. I think, however, that this inability should not come as a great surprise, as to suppose that rationality and moral motivation are inextricably linked in this way would be to believe that the most rational people are also the most moral, a view which seems highly dubious at best, nor indeed does it even seem consistent with our naive notions about morality. As such, I think that what Adam Smith described as ‘moral sentiments’ are very important – not to ground the existence of moral truths as such, but rather to provide a basis for our caring about them and acting in accordance with them. I think this is necessary regardless of whether one believes in God or not.
In this piece I provide a critique of the Cosmological Argument portion of this video (see here https://www.youtube.com/watch?v=57Vr1Bkmvhk), a talk called ‘A Case for Christianity’ which has recently come to my attention. I argue that the speaker’s defenses of inference to the best explanation, the contingency of the universe, and the principle of sufficient reason, are all inadequate, and fail to properly consider plausible alternatives and counterexamples. I also discuss the comparative abilities of theism and naturalism to offer an ‘explanation’ for the origin of the universe, arguing that the speaker’s case for theism’s superiority is not well supported by the arguments he uses. Finally, I make some brief comments in response to the fine-tuning argument. Note that I do not discuss the historical arguments made in the second part of the talk, as I have addressed these in much more detail here (http://goo.gl/KCrJgL).
Inference to the Best Explanation
The speaker begins by appealing to ‘inference to the best explanation’, claiming that we use this sort of inference in science, history, and everyday life all the time, and that therefore it is valid. I believe that his argument is problematic for a number of reasons. First let me consider the two ‘examples’ that he gives to illustrate inference to the best explanation.
His first example is of electromagnetic theory providing evidence for the existence of electrons. Here, however, he simply presupposes the truth of scientific realism (that is, the notion that theoretical entities like electrons ‘really exist’, and are not merely useful fictions). Not only is scientific realism highly controversial, but it is also precisely the point he is attempting to establish here. That is, in order to accept that electromagnetic theory provides strong evidence for the ‘real’ existence of electrons, one would already have to accept the philosophically controversial premise that abductive arguments (another name for inference to the best explanation) are a valid method of reasoning about what is true. But this is precisely the point that the speaker is attempting to establish by citing this example. Thus the argument is question-begging.
In the case of the open window example, the speaker is confusing abduction and induction. Induction refers to the process of inferring that because something is often the case, or has often happened in a particular way, that therefore it is probably likely to happen similarly in this particular (new) case. That is a different type of argument to inference to the best explanation, but is precisely the type of reasoning being used in the window example. Thus, this second example also fails to support the speaker’s argument about the validity of inference to the best explanation.
Aside from the flaws of his examples, there is a deeper problem with the speaker’s argument – he fails to provide a proper definition of what they mean by ‘explanation’. It does no good to say ‘explanation tracks truth’ when it is not at all clear what ‘explanation’ actually means, or what one looks like. At various points throughout the talk he speaks of explanations as providing ’causes’ of something, as giving ‘a reason why’ something happens, and also of being able to fit with empirical data. These are all different notions of explanation (and there are many more that are debated in philosophy). Before any sensible argument can be made about what inferences can be drawn on the basis of explanations, it is first necessary to provide at least a reasonably clear explication of what exactly is meant by this term. Otherwise, things that one claims as being ‘explanations’ may not actually be explanatory at all (a potential issue with some of his later arguments). In sum, the speaker simply does not address these issues in sufficient depth (or really even allude to them at all), and thus they fail to make their case for the validity of abductive arguments.
A final problem with inference to the best explanation, which the speaker also does not address, is that at best all that such arguments can tell us is that when some explanation is superior to another, then we can infer that the state of the world ‘corresponding to’ that explanation is more likely. We cannot actually say how much more likely it is without knowing more about the comparative explanatory power of the competing explanations. It could be the case that even the best explanation available is so poor, is such as bad explanation, that the corresponding state of the world is still not very likely.
The Contingency of the Universe
The speaker argues that the universe is probably contingent, because the universe is simply the sum total of everything in the universe, and as far as we know everything in the universe is contingent. There are several flaws with this argument.
First, we simply do not know very much about the large-scale structure, origin, and nature of the universe. We do not know what was possible and what wasn’t – the science (and philosophy) of these matters is a long way from being settled. For the speaker therefore to simply assert that ‘as far as we know everything is contingent’ grossly overstates the extent of our knowledge, and dismisses too readily the high levels of uncertainty that remain.
Second, the speaker actually gives no reason as to why the universe should be contingent even if all of its constituent components are contingent. This is simply the fallacy of composition. He does acknowledge that it isn’t logically necessary that this be the case, but then he simply brushes off this objection and asserts that ‘it is a real stretch’ to argue that the universe could be necessary even though all its constituents are contingent. Why? No argument is given. Indeed, there seem to be many obvious counterexamples where properties of the whole are not manifested in any of the parts. For instance, cells are alive, but cells are made up of nothing but atoms, which are not alive. Words have meaning, but words are made up solely of vibrations of air or dots of ink, which do not have any meaning associated with them individually. To give another example, we would have to ‘go and look’ to see if any particular book was in a library – that fact would be contingent. But it would not be a contingent fact that a library contains books of some sort, or else it would not be a library at all.
For these reasons, the speaker fails to establish their conclusion that the universe is contingent.
The Principle of Sufficient Reason
The speaker argues that all contingent facts must have some reason or explanation as to why they are the case, a notion which is called the ‘principle of sufficient reason’. He argues that this principle underpins essentially all of science, and that rejecting it leads to nonsensical conclusions. However, I think the speaker fails to establish their argument about PSR, for the following reasons.
First, he is not clear about exactly what constitutes a ‘reason’. Is it a physical cause? A non-physical cause? An explanation? A purpose? What exactly? It seems difficult to take the argument very seriously when it is not even clear what claim is actually being made. On a related point, even the notion of causation itself is philosophically problematic, as David Hume and others have noted. To this the speaker makes no reference at all, and seems content merely to take the concept of ‘causation’ as an unproblematic given.
Second, the fact that something like the PSR (arguably) ‘underpins all of science’ does not imply that it is everywhere and always true. The author falls into the same trap that he accuses the naturalist speaker of falling into, namely of assuming that because a given concept sometimes works or is successful in a particular sphere (in this case science), it therefore follows that it is universally applicable. That simply does not follow. It could be the case that science works well for questions where PSR (or something like it) is applicable, and does not work well for questions where it does not. One can also raise the deeper question of whether science actually provides ‘reasons’ or ’causes’ at all, rather than merely describing empirical regularities (again, as argued by Hume). These are complex and much-debated questions in philosophy, but the speaker ignores them, and simply adopts as ‘obvious’ particular simplistic answers which, conveniently enough, also support his argument.
Third, to reject the PSR does not imply ‘nonsense’. It merely is to say that we do not properly understand abstract and difficult concepts like ‘causation’ well enough to make confident claims about them.
The Failure of Naturalism
The speaker then proceeds to argue that naturalism is unable even in principle to provide an explanation for the origin of the universe, as naturalistic explanations can only refer to physical laws, which themselves did not exist before the universe and hence cannot be appealed to in an explanation of it. A few responses are in order here.
First, the line of argument being made here is very dubious. It seems that the speaker is saying that we could tell that naturalistic explanations could never explain the origin of the universe, even before we had even tried to construct any, or test them to see if they work. He is saying that even in principle they simply cannot yield such an explanation. Looking back over history, it seems this line of argument that science ‘cannot possibly even in theory’ explain any given phenomenon has fared very poorly, the most obvious example being vitalism and explaining the unique nature of living beings. In general, I think it is wise not to place great confidence in armchair philosophizing arguments about what science can and cannot explain ‘in theory’. Their track record seems to be very poor indeed.
Second, it is not at all clear the a naturalistic explanation would require physical laws. When we begin talking about things that existed “before” the universe began, and how the universe could have come into being, we are so far outside of the realm of what we can understand, of what we can know about with any confidence, and so far beyond the bounds where our intuitions are useful, that it is just not at all clear what a naturalistic explanation of the origin of the universe would look like, or what it would need to appeal to. The fact that the speaker cannot now imagine how such a thing could be developed is simply an example of the fallacious argument from lack of imagination.
Third, the argument here relies on the notion that the universe is contingent, and that contingent things require explanations, both premises which, as I argued above, are questionable at best.
God as an Explanation
In this section, the speaker argues that theism provides a satisfactory explanation for the origin of the universe, on that basis that it was created by an uncaused non-physical mind. There are many problems with this proposed explanation.
First, no non-physical substance, entity, or process, is known uncontroversially to exist. The only non-physical things that we think ‘exist’ are abstractions, like nations or languages or mathematical theorems. But God is not supposed to be an abstraction; he is supposed to be a ‘real’ non-physical entity. It is certainly possible that such entities exist, but outside of the question of God, we do not have any other good reasons to believe that such things are exist at all (indeed, the very notion may be incoherent – this is debated). In contrast, we know that physical processes and entities are real (or, at least, we know this with a fairly high level of confidence, philosophical skepticism notwithstanding). For this reason alone, I think it is reasonable that naturalistic causes be granted higher plausibility when considering questions such as how the universe came to be.
Second, even if we are to accept non-physical causes, there seems to be no reason to accept this particular one that the speaker presents. Instead of a non-physical uncaused mind, could we not instead posit a non-physical uncaused substance called ‘vitalic phlogiston’, which gives rise to the universe as a product of the fluctuations of its internal harmonic vibrations. It seems there is an almost limitless number of potential non-physical ‘explanations’ (again, a problematic term the speaker does not properly define) for the origin of the universe. Why should we prefer Christian theism over any of these others? One may argue that additional criteria or evidences are available with provide such reasons, but in that case it seems that the cosmological argument by itself is not actually doing very much ‘work’, so to speak, of providing support for theism.
Third, the speaker’s claim that the rebuttal that ‘minds are complex’ necessarily assumes materialism, seems to be rather a stretch. In fact, it seems perfectly reasonable to say that, even if materialism is false, minds are still very complicated things. Minds are capable of a wide range of thoughts and behaviours, which are often very unpredictable and interact in complex and unexpected ways. One can go on and on listing various ways in which minds are complicated, none of which depend in any way on the notion that minds must be material. The argument is not that minds are complex because brains are complex. The argument is that minds are complex precisely because, by their nature, by the definition of what we mean by ‘mind’, a mind is an intricate, multifaceted, and hence complicated thing. Merely stating that ‘God is non-material and therefore simple’ does not address this point, and is little more than argument by assertion.
The Fine-Tuning Argument
The speaker ends with an argument that further evidence for divine creation can be gained from the fact that the universe is, despite apparently enormous odds against it, capable of sustaining intelligent life, a fact which is a natural corollary of the theistic explanation, but not of any naturalistic explanation. One can question this argument on a number of grounds.
First, it is by no means established that the universe is in fact actually ‘fine-tuned’ for life. Certainly some scientists and philosophers think that this is the case, but there are also many who do not (e.g. Victor Stenger). As I argued before, we simply do not know enough about the laws of nature, how they interact, why they are as they are, and what else could have been possible, to make any confident claims about ‘fine-tuning’.
Second, even if the universe is fine-tuned, the speaker does not adequately consider potential naturalistic explanations for this. He too readily dismisses multiverses, which, although doubtless sound absurd to a layman, are nonetheless taken very seriously by a large number of physicists and philosophers, and are widely considered to be a powerful, plausible explanation for a wide variety of phenomena (including many apparent paradoxes of quantum mechanics). This of course is not proof that multiverses really exist, but I think the notion cannot be dismissed nearly so readily as the speaker does. A second, totally independent possible naturalistic explanation is the various forms of the anthropic principle. Although this sort of anthropic reasoning is highly controversial, so too is the existence of God, so it seems unreasonable and unfair to dismiss such potentially powerful alternative explanations arguments so readily.
Overall, contrary to the speaker’s argument, it is not clear that theism has the unique advantage of being able to explain the apparent ‘fine-tuning’ of the universe far better than can naturalism.
In my view, the speaker fails to establish his argument. He makes too many quick leaps of logic on the basis of questionable premises, without adequately considering possible objections, alternate explanations, or rebuttals. The speaker is also far too ready to make confident conclusions about difficult questions, such as the nature of causation and the origin of the universe, despite the fact that we simply do not know very much at all about these matters, or even how to think about them properly. Overall, the claims made about the likely existence of a creator God are not justified by the equivocal and incomplete nature of the reasons provided.