Background
This essay is an attempt to help rehabilitate Thomas Aquinas's reputation – and Aquinas's reputation still needs it. Aquinas has, of course, long been followed and defended by legions of Roman Catholic scholastic philosophers. More recently, Aquinas has been rediscovered by analytic philosophers of religion, mining medieval ore for their contemporary analytical mills. Almost without exception, however, these philosophers are themselves Christians or at least sympathetic to Aquinas' Christianity.
The situation is very different in secular philosophical circles. There, Aquinas is largely of historical interest, and limited historical interest at that. At best, he is seen as a comparatively unoriginal ‘synthesizer’; at worst, a vandal who took over Aristotle's gigantic achievement, only to adulterate it with theological dogma. Probably, this secular reputation still owes much to Bertrand Russell's dismissive treatment in chapter XII of his History of Western Philosophy. There, Russell makes some even-handed individual remarks about Aquinas, but his overall attitude is illustrated in this famous passage:
There is little of the true philosophic spirit in Aquinas. He does not, like the Platonic Socrates, set out to follow wherever the argument may lead. He is not engaged in an inquiry, the result of which it is impossible to know in advance. Before he begins to philosophise, he already knows the truth; it is declared in the catholic faith. If he can find apparently rational arguments for some parts of the faith, so much the better; if he cannot, he need only fall back on revelation. The finding of arguments for a conclusion given in advance is not philosophy, but special pleading. I cannot therefore, feel that he deserves to be put on a level with the best philosophers either of Greece or of modern times.Footnote 1
These criticisms have long rankled with me, so recently I undertook to defend Aquinas from them.Footnote 2 In doing so, I made three small points and one large one: First, I suggested that Russell's criticism takes no account of the nuanced things about the relation between faith and reason that Aquinas actually said (e.g., in Book I, chs. 2-12 of SCG). Second, I pointed out that Russell does not adduce a single case where Aquinas actually fails to follow the argument where it leads, and, on the contrary, that Aquinas is rather scrupulous about following arguments where they appear (to him) to lead.Footnote 3 Third, I pointed out, as Anthony Kenny does, that this criticism is a bit rich, coming from someone who spent many years and several hundred pages trying to prove that 2 + 2 = 4, something (presumably) he believed before he began to philosophise. Footnote 4
My main point, however, was that Russell's criticism that Aquinas ‘does not set out to follow the argument wherever it may lead’ enshrines an arbitrary and unreasonable epistemic principle. The idea of ‘following the argument wherever it leads’ may look innocent, even bland, but it isn't. To see that it isn't, one needs to notice three things about it: it is a normative principle, a tracking principle and a maximizing principle. It is a normative principle in that it lays down a rule about how we ought to manage our beliefs. It is a tracking principle, because it requires our philosophical beliefs to track (Russell says ‘follow’) a certain something. It is a maximizing principle because that certain something is a maximal property, namely the property of being the conclusion of the best argument available. This latter point may not be immediately obvious, but it (or something very like it) must be so, or else Russell's objection would not get off the ground. After all, Russell's objection to Aquinas cannot be that his conclusions are based on no reasons at all – he implies they are often based on Christian scripture or Aristotelian principles – but that they are not based on the best reasons. For Russell, the best reasons are of course the best philosophical arguments, where an argument is understood as an intellectual consideration that could in principle be expressed in terms of premises, conclusions and an inferential relation between them. Arguments of this type are evaluated primarily in terms of the truth or reasonableness of their premises and the kind of support these premises provide for the conclusion in question, so the best argument will presumably be the one that exhibits these characteristics to a higher degree than other available arguments.Footnote 5
Combining these normative, tracking and maximizing characteristics in a single standard, Russell's criticism presupposes an epistemic principle which could be formally expressed as: ‘Direct Argument Maximalism’ (‘DAM’):
DAM: Belief B is epistemically permissible for S at t iff B has maximal argument value for S at t, where:
B has Maximal Argument Value for S at t iff no incompatible belief has higher Overall Argument Value for S at t; and
The “Overall Argument Value” of B for S at t= df the balance of the value of the arguments for belief B over the value of the arguments against B, for S at t; where:
The Value of an individual argument for B for S at t is some function of the degree of justification for S at t of that argument's premises and the degree of truth-preservingness of the relation between that argument's premises and conclusion.
A striking feature of Russell's view, when expressed in this way, is how closely it resembles another normative, tracking, maximizing principle, namely utilitarianism:
DU: A is permissible for S at t iff A maximises utility, such that:
A maximises utility iff no other action (open to S at t) produces more utility than A, where:
The “utility of A”=df the balance of pleasure over pain for the aggregate of sentient creatures affected by the consequences of A.
According to the utilitarian, the problem with commonsense moral beliefs is that it they involve commitments for or against certain kinds of actions (e.g., killing the innocent), but since the property of, say, not killing innocents does not necessarily coincide with the property of maximising utility, such commitments keep our actions from tracking utility. Similarly, according to Russell, the problem with religious philosophers such as Aquinas is that they are committed to particular religious beliefs before argument begins, and these religious commitments keep their beliefs from tracking the best arguments.
Given these similarities, I argued that DAM is the epistemic counterpart of Direct Utilitarianism, that it suffers from the same problems as Direct Utilitarianism, and that it is no less controversial in epistemology than Direct Utilitarianism is in ethics.Footnote 6 Russell still has his admirers, however, so my criticism of him (and my defense of Aquinas) struck a nerve in at least one of them. Graham Oppy, for example, replied that I ‘completely misrepresent the position which Russell develops in the final five paragraphs of his chapter on Aquinas’, and that my ‘complaints against Russell are quite without substance’Footnote 7. The present paper is a continuation of that discussion and a further opportunity to consider Russell's criticisms of Aquinas in detail.
Russell's Complaint against Aquinas: Another Interpretation?
In his reply to me, Oppy does not defend DAM; indeed, he identifies additional problems with it. Instead, Oppy argues that I am wrong to suppose that Russell's complaint against Aquinas requires DAM; that, on grounds of charity in interpretation, ‘it is absurd to think that Russell would even implicitly have committed himself to such a stupid epistemological doctrine’; and that, all of my other criticisms of Russell no longer apply since he is not committed to DAM.Footnote 8 I shall show that Oppy's reinterpretation of Russell is incorrect and that all of my original criticisms of Russell are still very much to the point.
Russell probably disliked all sorts of things about Aquinas's method and conclusions, but the particular objection I had in mind is expressed specifically in his claims that Aquinas ‘does not, like the Platonic Socrates, set out to follow wherever the argument may lead’, that in Aquinas ‘the appeal to reason is, in a sense, insincere since the conclusion to be reached is fixed in advanced’, and that ‘no Catholic is likely to abandon belief in God even if he becomes convinced that St. Thomas's arguments are bad; he will invent other arguments, or take refuge in revelation.’Footnote 9 As noted, I take it that part of the objection in those passages is that Aquinas has antecedent theological commitments for or against certain beliefs, and these prevent him from tracking the only thing that matters in philosophy, viz., the conclusions of the best arguments. Let us call this the ‘Failure to Track Objection’. Even if this were not Russell's complaint against Aquinas, it would still be worth responding to, as it is commonly levelled against religious philosophers generally, but it is also the natural way of reading Russell in the passages quoted.Footnote 10
Graham Oppy disagrees: he claims that I have attacked a straw man, as Russell's objection here is ‘not that you believe without argument (or reason); rather the objection is that you should not to pretend have grounds or reasons other than those which you actually have’. According to Oppy, ‘…what Russell is urging is that, when we do have reasons for our beliefs, we should give those reasons when we try to defend those beliefs, rather than cast around for other things which might be used as premises in arguments to the conclusions we favour’, and that to do otherwise is a kind of ‘perversion of reasoning’, and manifests a kind of philosophical bad faith or insincerity.Footnote 11 Let us call this the ‘Insincerity Objection’. (Along the way, Oppy discusses what features make for greatness in a philosopher, the extent to which originality is one of those features, and whether philosophers working in a commentary tradition are likely to achieve such originality to a high degree. There is much of interest in what Oppy suggests on these matters, though I suspect that a case can be made even for Aquinas's originality.Footnote 12 Here, however, I propose only to consider whether Oppy is right about the Insincerity Objection.)
Does Russell mean the Insincerity Objection?
It certainly appears that Russell is criticizing Aquinas for ‘not following the argument wherever it leads’ (i.e., the Failure to Track Objection), and Oppy gives little reason to suppose that, despite the appearances, it isn't Russell's objection, except for the suggestion that there is another interpretation of Russell here. But notice: even if it can plausibly be made out that Russell does mean the Insincerity Objection, it does not follow that Russell doesn't also mean the Failure to Track Objection. It is possible, after all, that Russell is making more than one objection in this passage. More importantly, there is good reason to suppose that Russell does not mean Oppy's Insincerity Objection. To show this, however, we have to get clear on the Insincerity Objection and how it is supposed to work. I suggest that it can be summarised as follows:
-
1. Whenever S believes P for a reason, there will always be some reason, R, that is S's real or main reason for believing P.Footnote 13
-
2. Whenever S believes P and gives a philosophical argument for P, then the premises of that argument should express R. (Anything else would be insincere.)
-
3. At many points, Aquinas believes some thesis, P, for some reason, R, but gives an argument for P, the premises of which do not express R.
-
4. Conclusion: at many points, Aquinas's arguments are insincere.
By way of comment, I will observe that (2) is a ‘norm of sincerity’ for philosophical argumentation; (1) is a presupposition of that norm; and (3) is an assertion of the ‘facts of the case’, according to which Aquinas is allegedly in violation of the norm expressed in (2).
If this is the Insincerity Objection, we should not, according to the principle of charity, attribute it to Russell, since it is a bad one. First, with regard to (1), the presupposition that for Aquinas's beliefs, there will always be some real or main reason for those beliefs: Oppy gives us no good reason for supposing that this is true, even for beliefs which have some reason or other. It may be true of some beliefs, but for a wide range of beliefs, over-determination of reasons for belief seems quite likely. Why, e.g., do I believe the Pythagorean Theorem? For lots of reasons: I first learned about it in what seemed to me an authoritative textbook; then several different teachers told me about it; then one of these teachers led me through the proof in discursive steps; then I saw a proof in ‘diagram form’ with a triangle and nine squares, sixteen squares, and twenty five squares along the sides of that triangle; then I worked through the proof myself; then I became aware of how many of my beliefs would be overturned if the theorem were false, and so on. Which of those is the real or main reason for my belief in the Pythagorean Theorem? Why suppose that any of them is?
We can make some headway, by distinguishing between the originating grounds of belief, the episodically occurrent grounds of belief, the structurally occurrent grounds of belief, and justificationally available grounds. For some individual beliefs, these will be the same, but for others these will be different, and for many, these will be shifting back and forth. Even if we identify, say, the structurally occurrent grounds of belief as ‘main’ or ‘real’ reasons, we have no reason to suppose that always, or most of the time, or even often, there will be some one main or real reason; this will likely be especially true of theoretical beliefs, whose interconnection with other beliefs is strong and extensive.Footnote 14
Premise (3) fares little better: we are owed a convincing example of a case in which Aquinas gives an argument for P, where his real grounds are clearly other than one's expressed in the premises. Even where the arguments seem bizarre to Russell or Oppy or most modern readers – and I have to say, the argument about incest to which Russell alludes seems a bit odd to me – that is not sufficient reason to infer that it isn't one of Aquinas's real reasons, or even his main real reason.Footnote 15
Premise (2) is a norm limiting the arguments that philosophers may offer, but it needs only to be stated to be seen as arbitrary and unreasonable. Why must philosophers offer only arguments whose premises express their ‘real’ or ‘main’ reasons for their belief in the conclusion (assuming such exists)? What sort of ‘must’ is this: moral? Or epistemic? Or some other sort? Doesn't the appropriateness of the premises offered depend partly on the purposes of argumentation in a particular context? Of course, if the purpose of philosophical argumentation is not knowledge or true belief or justified belief, then it is easy to see why one might offer arguments whose premises do not express one's real reasons for belief. But even in truth-oriented conceptions of philosophy, there will be contexts in which we wouldn't do this: for example, when we wish to persuade someone who would not grant those premises, and not to beg the question against them! As long as the premises of that argument express what we take to be a genuinely good reason for belief, there is nothing insincere or otherwise inappropriate in offering such an argument, even if they're not our real or main reasons. Indeed, this is precisely the context of the Summa Contra Gentiles, and – this is the important point here – Russell explicitly acknowledges this fact:
St. Thomas's most important work, the Summa Contra Gentiles, was written during the years 1259-64. It is concerned to establish the truth of the Christian religion by arguments addressed to a reader supposed to be not already a Christian; one gathers that the imaginary reader is usually thought of as a man versed in the philosophy of the Arabs. He wrote another book, Summa Theologiae, of almost equal importance, but of somewhat less interest to us because less designed to use arguments not assuming in advance the truth of Christianity.Footnote 16
Again, Russell characterizes Aquinas's project in the SCG as follows: ‘My purpose (he says) is to declare the truth which the Catholic Faith professes. But here I must have recourse to natural reason, since the gentiles do not accept the authority of Scripture.’Footnote 17 Far from being insincere, this might be thought a mark of respect to his interlocutors, and in any case the only dialectically intelligible strategy in such a context.
I conclude that either Oppy is right and Russell meant the Insincerity Objection but forgets something he takes pains to establish twelve paragraphs earlier, or Oppy is wrong in attributing the Insincerity Objection to Russell in the passage under discussion. For my part, I incline to the second possibility. Why else would Russell note, without apparent irony, that:
All these arguments on sexual ethics … appeal to purely rational considerations, not to divine commands and prohibitions. Here, as throughout the first three books, Aquinas is glad, at the end of a piece of reasoning, to quote texts showing that reason has led him to a conclusion in harmony with the Scriptures, but he does not appeal to authority until his result has been reached.Footnote 18
Oppy has not given any positive reason to suppose that Russell did not mean the Failure to Track Objection, and his alternative interpretation simply cannot be sustained on Russell's own terms.
Oppy's Other Arguments (And My Replies)
If my arguments above are right, then my secondary criticisms of Russell's assessment of Aquinas come back into play, as these bore mainly on Russell's complaint that Aquinas fails to follow the argument wherever it leads. Let us review these other criticisms briefly, beginning with the most straightforward.
1. I noted that, ‘Russell does not adduce a single case where Aquinas actually fails to follow the argument where it leads’.Footnote 19 Oppy replied that this is ‘irrelevant to Russell's complaint’, since his complaint is not (as I had suggested) the Failure to Track Objection but the Insincerity Objection; i.e., ‘Russell's complaint is that Aquinas's entire project aims at finding arguments for conclusions whose real justification lies elsewhere: in that sense, Aquinas almost never follows arguments where they lead.’
Reply: If I am right so far, and Russell really is making the Failure to Track Objection, then of course it is relevant whether Aquinas really does fail to follow the argument where it leads. Russell has not adduced a single case where the argument plainly leads one way, but Aquinas goes another. Even in cases where Aquinas believes something on the basis of revelation and Russell thinks he shouldn't, this reflects not a failure to follow the argument, so much as Aquinas's and Russell's disagreement over the several sources of justified belief (i.e., of ‘epistemic permissibility’).Footnote 20
2. I noted, further, that this particular criticism of Russell's ‘takes no account of the nuanced things about the relation between faith and reason that Aquinas actually said (e.g., in Book I, chs. 2-12 of SCG)’.Footnote 21 Oppy replied that this is, ‘…irrelevant to Russell's complaint. For one thing, Russell's complaint has nothing to do with Aquinas's theory about the relation of faith and reason; rather it concerns the practice which Aquinas adopts. For another thing, it won't do to say that it was an article of faith for Aquinas that very many things are supposed to be knowable in the light of reason.’Footnote 22
Reply: I have argued that there is nothing wrong with Aquinas's offering arguments whose premises do not express Aquinas's ‘real’ or ‘main’ reasons for believing those theses, especially when he is trying not to beg the question against interlocutors who would not share his premises. If I am right, the only other way in which Aquinas could be insincere would be if he himself doesn't think those arguments are good ones. He couldn't think those arguments were good ones, e.g., if they concerned propositions he believed could be known only via revelation. This is why it is relevant that Aquinas takes pains to distinguish between things that can be known via reason and things that can be known only via revelation, and to give philosophical arguments from generally available premises only for the former.
Moreover, if I am right that Russell is making the Failure to Track Objection, part of his motivation for this objection might be his perception that Aquinas's philosophical arguments are offered for conclusions given in advance, when these conclusions are something that cannot be known in advance of philosophical argumentation. Here again, Aquinas's claims about what can be known via revelation and via reason are relevant. Russell may disagree with Aquinas about these classifications – indeed, Aquinas may well be mistaken, but it does not follow that Aquinas is engaged in any philosophical funny business by offering the arguments he does.
3. Quoting Anthony Kenny, I observed that Russell's complaint that Aquinas does not follow the argument wherever it may lead ‘is a bit rich, coming from someone who spent many years and several hundred pages trying to prove that 2 + 2 = 4, something (presumably) he believed before he began to philosophise.’Footnote 23 Oppy replied,
That the complaint which Russell makes against Aquinas can be turned against the author of the Principia Mathematica is, I think, plainly mistaken. The aim of the Principia is to show that mathematics can be reduced to logic. Russell believed this could be done, and set out to show that it could be done. There is nothing in the Principia, which suggests that Russell was arguing for claims which he believed on other grounds. (Of course, Russell had independent grounds for believing 2 + 2 = 4; but the point of the proof in Principia is not to justify that belief. Rather, the point of the proof is to show that that mathematical belief can be derived from purely logical premises.)Footnote 24
Reply: Originally, I quoted Kenny's quip more as a wisecrack than anything else, but I now think it is completely correct and deeply revealing. Oppy is right, of course, that Russell's immediate goal in the Principia Mathematica was to show that mathematics can be reduced to logic, but his paramount reason for wanting to show this was epistemological. From his youth, Russell had been obsessed with the goal of proving that mathematical knowledge was certain, as he acknowledges in numerous autobiographical passages, e.g.:
My original interest in philosophy had two sources. On the one hand, I was anxious to discover whether philosophy would provide any defence for anything that could be called religious belief however vague; on the other hand, I wished to persuade myself that something could be known, in pure mathematics if not elsewhere … . As regards, the foundations of mathematics, I got nowhere. In spite of a strong bias towards empiricism, I could not believe that ‘two plus two equals four' is an inductive generalization from experience, but I remained in doubt as to everything beyond this purely negative conclusion.’Footnote 25
Logicism in general and Russell's analytical techniques in particular were motivated largely by a desire to ground more and more knowledge on fewer and stronger axiomatic foundations, and were inspired by the work of Frege and Peano. Russell notes:
Having reduced all traditional pure mathematics to the theory of the natural numbers, the next step in logical analysis was to reduce this theory itself to the smallest set of premises and undefined terms from which it could be derived. This work was accomplished by Peano. He showed that the entire theory of the natural numbers could be derived from three primitive ideas and five primitive propositions in addition to those of pure logic. These three ideas and five propositions thus became, as it were, hostages for the whole of traditional pure mathematics. Their logical ‘weight,’ if one may use such an expression, is equal to that of the whole series of sciences that have been deduced from the theory of the natural numbers; the truth of this whole series is assured if the truth of the five propositions is guaranteed, provided, of course, that there is nothing erroneous in the purely logical apparatus which is also involved.’Footnote 26
Moreover, Russell's epistemological motivation for logicism was not idiosyncratic: Frege explicitly states that it was part of his motivation as well:
I became aware of the need for a Begriffsschrift when I was looking for the fundamental principles or axioms upon which the whole of arithmetic rests. Only after this question is answered can it be hoped to trace successfully the springs of knowledge upon which this science thrives.Footnote 27
On this view, it may be possible to know mid-level mathematical truths in advance, but the mathematics as a whole is in better epistemological shape once the logicist programme is completed in Principia Mathematica. If this is Russell's view, however, he should not criticize Aquinas for attempting something similar (i.e., showing that something antecedently believed rests on a secure foundation which is ‘clear to the light of reason’).
Of course, Russell acknowledges that, in the Summa Contra Gentiles, Aquinas gives philosophical arguments that do not appeal to the Christian revelation because he is arguing with ‘gentiles’ who do not accept that revelation. In the Summa Theologiae, however, Aquinas is not arguing with gentiles, and so in this case one could be forgiven for wondering why he bothered to argue for such truths, if these could be known independently of philosophical argument. As I have already suggested, various reasons may exist for giving philosophical arguments for propositions which can be known independently of such arguments: they may overcome passing doubts in oneself or others; they may exhibit the interconnectedness of one's beliefs; they may deepen one's understanding of the beliefs in question; and they may raise one's degree of justification for those beliefs to an even higher level; they may even shed light on those premises. Indeed, in the preface to their Principia, Whitehead and Russell claim that:
…the chief reason of any theory on the principles of mathematics must always be inductive, i.e., it must lie in the fact that the theory in question enables us to deduce ordinary mathematics. In mathematics, the greatest degree of self-evidence is usually not be found quite at the beginning, but at some later point; hence the early deductions, until they reach this point, give reasons rather for believing the premises because true consequences follow from them, than for believing the consequences because they follow from the premises.Footnote 28
This may be as true in philosophy as it is in mathematics, at least on Aquinas's view. I suspect that part of the reason Russell does not see this is because he thinks that Aquinas cannot have knowledge of substantive truths of religion on the basis of revelation, and so he must await the outcome of philosophical argumentation. Indeed, this is suggested in Russell's claim that Aquinas ‘is not engaged in an enquiry, the result of which it is impossible to know in advance.’ This, more than the Insincerity Objection, lies at the bottom of Russell's complaint against Aquinas: Russell thinks these results cannot be known in advance; Aquinas thinks they can.Footnote 29
