The purpose of this essay is to take issue with two aspects of Marilyn Adams's monumental work William Ockham. Part I deals with Ockham's ontology, arguing (i) that Adams does not sufficiently appreciate the use Ockham makes of the prinicple of ontological parsimony in his attempt to refute the thesis that there are extramental universals or common natures and (ii) that she sets an implausibly high standard of success for Ockham's project of showing that the only singular entities are substances and qualities. Part II argues that Adams fails to provide a convincing defense of Ockham's 'anti-secularist' answer to the question of how Christian thinkers should react to prima facie conflicts between the deliverances of faith and the deliverances of reason.

Marilyn Adams's massive work William Ockham is the best comprehensive study of Ockham's thought ever written in English or, as far as I know, in any other language.1 Without a doubt, it will be the standard secondary source on Ockham's philosophy and theology for a long time to come. 

Among the numerous virtues of Adams's book is its sustained (and, to my mind, highly successful) attempt to root out many of the now tiresome misrepresentations of Ockham's writings which continue to be passed down from generation to generation by historians of philosophy, natural science, and theology.2 Shorn of these misinterpretations, Ockham's intellectual legacy turns out to be far less titillating than the wholesale subversion of Christian Aristotelianism that he is commonly credited with (or blamed for). Indeed, Adams's work makes it abundantly clear that Ockham's own ostensible agenda is a distinctly conservative one for an early fourteenth-century thinker, viz., to synthesize Aristotle's philosophy with the Catholic faith. Nonetheless, within the context of medieval Aristotelianism Ockham is a brilliant and in many ways controversial thinker, and by setting the record straight on just which positions Ockham actually held, Adams has cleared the way for genuinely fruitful historical and philosophical discussions of his thought. My hope is that the present essay will merit this description. 

My purpose is to take issue with Adams on two large and relatively independent /318/ topics, viz., Ockham's ontological reductionism (Part I) and his conception of the relation between faith and reason (Part II). Section A of Part I is devoted to showing that Adams fails to recognize the central role played by the principle of ontological parsimony in Ockham's attempt to refute the claim that there are extramental universals or common natures, while in section B I argue that she burdens Ockham with an unjustifiably high standard of success in his attempt to establish the thesis that the only singular entities are substances and qualities. In Part II I propose to show that Adams does not provide a convincing vindication of Ockham's answer to the question of how Christian philosophers ought to deal with apparent conflicts between the deliverances of faith and the deliverances of reason. My strategy here is to develop in some detail the striking contrast between Aquinas and Ockham on this important issue and to argue on that basis that any persuasive defense of Ockham against Aquinas must go deeper than Adams suggests.

Ockham is obsessed with ontology, so much so that his singlemindedness at times distracts him from deeper and more pressing matters. Still, it is in doing ontology that he is at his philosophical best. 

As Ockham sees it, ontology is the site of the two worst blunders in philosophy. The first is the postulation of extramental universals: 

Some claim that besides singular entities there are universal entities, and that singular entities as conceptualized are the subjects in singular propositions and universal entities as conceptualized are the parts of universal propositions. But this opinion, inasmuch as it claims that in addition to singular things there are extramental entities that exist in those things, I regard as altogether absurd and as destructive of Aristotle's whole philosophy, of all knowledge, and of all truth and reason, and [I believe] that [this opinion] is the worst error in philosophy--an error rejected by Aristotle in Metaphysics VII--and that those who hold it are incapable of knowledge.3

The source of many errors in philosophy is the claim that a distinct signified thing always corresponds to a distinct word in such a way that there are as many distinct entities being signified as there are distinct names or words doing the signifying.4 

A. Phase one

The first phase, discussed by Adams on pp. 3-141, centers on the category of substance and, more particularly, on the classical question of whether natural kind terms in this category, i.e., species and genus terms such as 'human being' and 'animal', signify common, as opposed to singular, entities. Since none of his opponents subscribes to it, Ockham feels no need to refute Platonic realism, according to which natural kind terms signify eternally and necessarily existing universal entities that are exemplified by singular substances, but do not exist in those substances.5 Instead, he zeroes in on 'moderate realism', a cornerstone of which is the negative semantic thesis that natural kind terms, whatever else they might signify, do not at any rate signify singular substances as such. 

Adams recounts Ockham's criticisms of no fewer than six versions of moderate realism. The most straightforward among them, viz., those championed by Walter Burleigh and the early Duns Scotus, maintain that a singular substance, in addition to having physical constituents (matter and form or, in the case of a spiritual substance, form alone), also has metaphysical constituents that are signified by the natural kind terms true of it. For instance, the matter/form composite which is Socrates has distinct metaphysical constituents answering to each of the common terms 'human being', 'animal', 'living thing', and 'substance'. These constituents, known as common natures, are thought to provide a ground for metaphysical definitions and essential predications, and in this way to underwrite the possibility of scientific knowledge, which within an Aristotelian framework is just the knowledge of essences or natures.6

Other versions of moderate realism, including those espoused by Thomas Aquinas and Henry of Harclay, resist precise description. They seem very close to Ockham's own conceptualist theory in their ontological commitments, though not in what they say about the signification of natural kind terms.7 /320/ 

I will confine my remarks here to the two more straightforward versions. According to Burleigh, a natural kind term signifies a common nature that exists whole and entire in each singular substance of which the term is true. Accordingly, Socrates's humanity is numerically identical with the humanity of every other human being. Corresponding theses hold for each of the common natures signified by the terms 'animal', 'living thing', 'substance', etc.--thus the need for a principle of individuation that makes singular substances of the same species distinct from one another. What's more, each common nature is 'really' (i.e., numerically) distinct from all the others--and this immediately raises the question of how the singular things constituted by these discrete common natures can possess the strong intrinsic unity Aristotle attributes to primary substances. Ockham concocts a wide array of objections, some of them ingenious, against Burleigh's position, but Adams believes this to be the one form of moderate realism that can withstand Ockham's onslaught (pp. 38 and 67-69). 

A bit more subtle is Scotus's early theory, according to which the common natures signified by the kind terms true of a given singular substance are entities that (i) in themselves have a unity less than numerical unity and are hence shareable, (ii) are 'contracted' by an individual difference (a thisness or haecceity) to yield the singular substance, which has full numerical unity, and (iii), when contracted by the individual difference, are (a) really identical with one another and with the individual difference and yet (b) 'formally' distinct from one another and from the individual difference. ((iiia) is meant to preserve the unity of the substance.) On this theory Socrates's humanity is really distinct from Plato's humanity and really identical with, though formally distinct from, both Socrates's animality and Socrates's individual difference. In response, Ockham argues at length against the very possibility of an extramental distinction between real entities which is merely a formal and not a real distinction. After thoroughly examining these arguments, Adams concludes that they undermine both the earlier and the later accounts Scotus gives of the formal distinction (pp. 46-59). 

What surprises me is that Adams not only plays down Ockham's use of the principle of ontological parsimony in the reductionistic program as a whole, but completely ignores its role here in the first phase. This principle, known popularly as Ockham's razor, dictates that entities are not to be multiplied beyond necessity. True, Ockham does not explicitly mention the principle in the long tract on universals and common natures in Ordinatio I.8 Yet he in fact employs it there as his most potent weapon against both Burleigh and Scotus. Or so, at least, I will now argue. 

Ockham's clearest formulation of the principle goes something like this: When a number n of entities (or types of entity) is sufficient to make a proposition p true, then it is gratuitous to posit more than n entities (or types /321/ of entity) in order to account for p's truth.9 So stated, the principle constitutes a methodological constraint on the construction of ontological theories. The ontologist, on Ockham's view, aims to determine which general categories of being must be posited in order to render true the propositions included in the generally agreed upon 'data' of ontology. The entities thus posited are the things signified by--in technical terminology, the significata of--the categorematic terms occurring in those propositions. Given this conception of ontology, the postulation of universals or common natures is justifiable only if one can argue persuasively that such entities alone can adequately serve as the significata of the natural kind terms contained in the data propositions:

[A universal] entity should not be posited except to preserve the essential predication of one thing with respect to another, or to preserve scientific knowledge of things and the definitions of things. These are the arguments suggested by Aristotle on behalf of Plato's view.10

Both Ockham and his interlocutors, then, presuppose that ontology is a theoretical or postulational, rather than a purely descriptive, enterprise. I do not know exactly how to characterize the difference between postulational and descriptivist approaches to ontology, but it is only with the help of some such distinction that we can understand why certain philosophers take Ockham's razor to be an utterly obvious methodological constraint in doing ontology, while others deem it just as obviously irrelevant to the task of delineating the most general categories of being. Numbered among the latter is Nicholas Wolterstorff, who maintains that ontology is "descriptive, not explanatory"11 and says this about Ockham's razor:

[One] reason for the reluctance to admit that there are predicable universals is adherence to the popular but puzzling dictum that we ought not multiply entities beyond necessity. Seldom is it clear what is meant by this. What might sometimes be meant by it is that we ought not to say that there is or exists a certain entity, or that there are or exist entities of a certain sort, unless we find it necessary to do so. Or what might be meant is that we ought not to refer to entities unless we find it necessary to do so.... [But] what we wish to do [as ontologists] is find out whether there are predicable universals; whether for this or that purpose it is necessary to say that there are, or to refer to them, is really quite a different matter.12

Be that as it may, my purpose here is not to assess or even to clarify the debate between postulationalists and decriptivists. I merely want to show how Ockham uses his razor against Scotus and Burleigh. The postulational conception of ontology enables us to distinguish two different kinds of arguments that a conceptualist like Ockham might employ against realists. First, there are direct arguments, which try to establish that the various theories propounded by realists suffer from insuperable deficiencies. One might argue, as Ockham does, that the realist theories have absurd consequences or that they fail to save the data. As noted above, Adams judges that these direct arguments succeed against Scotus but not against Burleigh. However, in the Ordinatio discussion of universals Ockham also employs indirect arguments, i.e., arguments meant to show that even if the realist theories are internally coherent and adequate to preserve the data, those data can nonetheless be accommodated equally well by a coherent theory that does not posit any common natures at all. He argues, for example, that common natures are not required to undergird the truth of propositions about how substances resemble one another in various ways and to various degrees; similar arguments focus on essential predications, metaphysical definitions, and the possibility of scientific knowledge.13 What follows, given the principle of parsimony, is that even if Ockham's direct arguments against Burleigh and Scotus fail--indeed, even in the absence of such arguments--he still triumphs as long as he can show that his own more economical theory posits enough significata to make the data true. Ockham's indirect arguments are in fact very powerful; as I see it, they effectively defeat the straightforward versions of moderate realism. But my main point here is simply this: Ockham clearly presupposes, as Adams should have realized, that the principle of parsimony stands behind these arguments. 

Ockham's own positive account of universals repudiates the notion that singular substances have metaphysical constituents of the sort described above. Every real entity is singular in and of itself. So no problem of individuation is generated by the mere fact that kind terms apply to singular substances, and there is no need to posit individual differences that contract common natures to yield such substances. But how is it that certain mental, spoken, and written terms, though singular in being, are universal in signification? And what do they signify if not common natures? To take the second question first, Ockham replies that what they signify are none other than the singular entities of which they are truly predicable.14 For instance, the kind term 'human being' signifies all and only the singular substances of which 'human being' is truly predicable, and it is a common or universal term because it has the capacity to signify many human beings in such a way that it signifies no one of them more than any other. In response to the first question, Ockham asserts that a mental common term has the capacity to signify many singulars by its very nature, whereas spoken and written /323/ common terms have this capacity only because of the linguistic conventions by which they are subordinated to the corresponding mental terms. (On pp. 121-141 Adams lays out the problems attendant upon this doctrine of the natural signification of mental terms.) 

So the conclusion of the first phase of Ockham's ontological program is that every entity is singular. The only universals are universal terms, and they are universal not in their being but in their signification alone. 

B. Phase two

Now for the second phase of the program, which Adams addresses on pp. 143-313. Medieval Aristotelians generally hold that every singular entity is either (i) a substance, (ii) a part of a substance (whether an integral part or an essential part such as matter or form), or (iii) an accident apt by its nature to inhere in a substance. In the second phase, Ockham turns to accidents and tries to establish that the only accidents are qualities--where, as noted above, the parallel semantic thesis is that all terms in the categories other than substance and quality are connotative rather than absolute.15

According to Ockham, connotative terms signify substances and qualities while connoting various conditions of applicability that do not in themselves implicate the existence of any entities besides substances and qualities. So, for instance, no singular entities distinct from substances or qualities serve as the significata of, say, relative terms such as 'mother', 'similar to', 'equal to', and 'to the left of', or of quantitative terms such as 'seven', 'double', 'point', 'line', and 'surface', or of temporal terms such as 'now', 'yesterday', and 'ten years from now', or of action terms such as 'cause', 'create', and 'generate' or their passive counterparts. To support this contention, Ockham proceeds through the accidental categories one by one, employing both direct and indirect arguments to impugn the postulation of relations, quantities, times, positions, actions, etc., as distinct entities.16 The arguments are by and large as fascinating as they are complicated, but in the end he invariably concludes that propositions in which the relevant connotative terms occur require for their truth no ontological furniture other than substances and qualities. 

Adams supplies a detailed analysis of Ockham's arguments concerning the key categories of quantity and relation. However, in assessing this second phase of Ockham's program (pp. 287-313) she sets a criterion for success which is too high both in itself and from Ockham's point of view. According to Adams, Ockham must show here, as he does not, (i) that the mental language--which Adams claims to be 'logically perspicuous'--contains no connotative terms at all, and hence (ii) that connotative terms are eliminable without loss from spoken and written language. Her claim, in other words, is that Ockham's ontological reductionism entails a linguistic reductionism as well, so that the second phase of his program succeeds only to the extent that he can provide plausible /324/ translations, duly shorn of connotative terms, for all propositions in which such terms occur. 

To be sure, Ockham does explicitly maintain that every connotative term has a nominal definition that expresses its conditions of applicability, and he also asserts that the mental language contains no distinct synonymous terms. Adams makes the further assumption that a connotative term is synonymous with its nominal definition and infers that on Ockham's reckoning every spoken or written proposition involving a connotative term corresponds to a mental proposition that contains only the nominal definition of that term. She then suggests, without corroboration from any text, that nominal definitions themselves contain no connotative terms and concludes that Ockham at least "should have" held that all connotative terms are eliminable: 

Ockham maintains that all connotative terms are susceptible of nominal definitions.... And a development of his logic ... would conclude that all connotative terms are synonymous with complexes of primitive syncategorematics and absolute terms that name substances or qualities. If so, Ockham should have ... regarded all connotative terms and hence all terms in the categories other than substance and quality as in principle eliminable in favor of such complexes. (p. 298)

So Ockham, in contrast to Adams, does not think that the success of the second phase of his program hinges on his being able to show that every categorematic term in the mental language is absolute and not connotative. Nor, it seems to me, can one argue plausibly that he should have thought otherwise. He clearly does believe that the truth of propositions containing connotative terms can be preserved without assigning any entities other than substances and qualities as the significata of those terms, and from this belief it follows that any entity signified by a connotative term is also signified by an absolute term in either the category of substance or the category of quality. But I find it exceedingly difficult to detect a nexus between this thesis and the claim that all connotative terms are eliminable. 

Consider a simple example. Ockham claims, plausibly, that the written proposition

(1) Socrates and Plato are similar to one another in being wise, 

(2) Socrates is wise and Plato is wise, 

Now it seems clear that in order to make this argument, Ockham does not have to assert that (1) and (2) say exactly the same thing in exactly the same way; nor does he have to maintain that the term 'similar' signifies Socrates and Plato in the same way that the proper names 'Socrates' and 'Plato' do or in the same way that the term 'wise' does. In short, he has no reason to deny that (1) and (2) correspond to distinct mental propositions or that the term 'similar' occurs in the mental language. He needs to claim only that, contrary to what his opponents contend, the term 'similar' signifies just substances and qualities, though in its own distinctive way. 

Of course, Ockham must supply stories like the one about the term 'similar' for at least a wide range of connotative terms. Perhaps he does not go far enough in discharging this duty. Still, the point of such stories is simply to clarify the ontological ramifications of the use of connotative terms and not to show that the mental language contains no such terms. This may mean that /326/ the mental language fails to qualify as a 'logically perspicuous language' in Adams's sense. But is Ockham's project thereby rendered any less interesting or promising? I think not.

Unlike Aquinas, Ockham never wrote anything resembling a treatise on the relation between faith and reason. Yet a tolerably clear picture of his thinking on this matter emerges from those texts in which he deals directly with topics in natural and revealed theology. Adams devotes a lengthy chapter (Chap. 22, pp. 961-1010) to faith and reason, concentrating especially on the question of how Christian thinkers should deal with prima facie conflicts between the deliverances of faith and the deliverances of reason. In the end, however, she allows Ockham to beg the question against St. Thomas and overlooks, or at least mutes, fairly obvious criticisms of Ockham himself. 

Chapter 22 concludes with a debate between Ockham and his Thomistic critics, but here I will bypass the critics and go straight to the tract on faith and reason found in the opening nine chapters of St. Thomas's Summa Contra Gentiles. Although I agree with St. Thomas on the issue under dispute, my purpose here is not to defend him against Ockham, but only to grant him a fair hearing and to show that Adams's attempt to vindicate Ockham fails.
 

A. St. Thomas on conflicts between faith and reason

St. Thomas portrays his own project of explicating the Christian faith and refuting objections to it as a continuation of the 'Gentile' philosopher's quest for wisdom, i.e., for a systematic understanding of "the truth which is the origin of all truth, viz., the truth that pertains to the first principle of being for all things" (Chapter 1).21 This pursuit of wisdom, identified on Biblical grounds with the search for God and simultaneously identified with what Aristotle calls First Philosophy, is the "most perfect, noble, useful and joyful of human endeavors" (Chapter 2), mainly because the limited grasp of 'divine truth' possible in this life furnishes us with a foretaste of that evident and face-to-face knowledge of God which is, according to Christian revelation, the principal constituent of ultimate human fulfillment. 

Notice that philosophy, understood expansively as the endeavor to articulate and defend a comprehensive metaphysical vision of the world, is free to, indeed obliged to, draw upon every source of truth available to us as human beings. St. Thomas distinguishes two ways in which divine truth is made manifest to us, viz., through revelation and through natural reason, where the latter ostensibly includes every source of truth distinct from Sacred Scripture and the teachings of the Church. He realizes that many will balk at his unabashed insistence that Christian revelation counts as a legitimate source of truth, but since this issue does not separate him from Ockham, I will simply /327/ ignore it here. St. Thomas is in any case more concerned with another question. Mindful of the metaphysical achievements of Plato, Aristotle, and their philosophical progeny, he asks whether reason can serve as an alternate source of the truths revealed to us by God and, more specifically, whether reason can demonstrate such truths by arguments from evident premises. The answer is both yes and no:

In those things that we profess about God there are two types of truths. For there are some truths about God that exceed every capacity of human reason, such as that God is [both] three and one. But there are other truths that natural reason is also capable of arriving at, such as that God exists, that there is one God, and others of this sort. Indeed, philosophers, led by the light of natural reason, have proved these truths about God demonstratively. (Chapter 3) 

In Chapters 3-6 St. Thomas addresses several questions immediately prompted by this distinction between the mysteries and the preambles: Is it reasonable to believe that there are truths about divine matters which in principle exceed our natural capacities for systematic understanding? Wasn't it pointless for God to reveal truths that natural reason is capable of establishing on its own? Is it proper for God to demand that we accept on faith propositions that reason cannot even in principle attain to? Isn't it frivolous and intellectually irresponsible for us to assent to the mysteries of the faith? I will not deal directly with these questions here, since, once again, they do not divide St. Thomas from Ockham. 

We can begin to approach the genuine differences between Ockham and Aquinas by observing that the distinction between the mysteries and the preambles suggests a second conception of philosophy which is narrower than the one adumbrated above. On this conception, philosophy draws its premises from natural reason alone and is thereby set off from theology, which takes /328/ revealed propositions as its starting points and tries, within the limits of human finitude, to understand them systematically. This distinction between philosophy and theology became pivotal in the thirteenth century when Aristotle's works flooded into European universities, and since then it has served within Catholic universities as the theoretical foundation for the separation of philosophy departments from theology departments. 

St. Thomas singles out this narrower sense of philosophy in part because it helps him clarify what he regards as the proper posture Christians should assume toward secular learning in general and secular philosophy in particular. The history of Christianity has been marked by recurrent and bitter disputes over this issue. From the earliest times some Christians (I will dub them 'anti-secularists') have denounced secular 'wisdom' as an adversary of Christianity. They have sternly warned fellow Christians about the pitfalls of syncretism, and they have acerbically asked why, if not because of an obsequious (and typically futile) desire to curry favor with intellectually prestigious unbelievers, a Christian might want to study, say, the books of Aristotle with the same intensity as the books of Sacred Scripture. They recall that when St. Paul preached in Athens, he was ridiculed by the philosophers, who in their pride preferred the wisdom of the world to the wisdom of God (Acts 17:16-34). What, they ask disdainfully, has Jerusalem to do with Athens? Christianity is itself a philosophy or wisdom that competes with secular philosophies and aims to displace them. (Observe the fallback here to the broader conception of philosophy.) 

I do not mean to insinuate that Ockham is a full-fledged anti-secularist. He does not, for instance, spurn efforts to articulate Christian doctrine with the help of conceptual resources borrowed from secular philosophy.23 Nor does he repudiate in theory the natural theologian's attempt to show that at least some revealed truths can be established on grounds that unbelievers as such should or at least can accept. 

However, as we shall see shortly, he does evince anti-secularist leanings on one important issue. All sides agree that because human reason stands in need of the illumination of faith, it is not surprising that philosophers who operate in ignorance of revelation often come to conclusions that are contrary to the faith. According to anti-secularism, however, a Christian is not obliged to refute such conclusions on their own terms, i.e., by appealing only to the deliverances of reason. Indeed, anti-secularists allege that in many cases a philosophical (in the narrow sense) refutation may well be impossible, given that human sinfulness has rendered reason unreliable. Perhaps this means that Christian doctrine will inevitably appear foolish in the eyes of secular philosophers. So be it. The Christian's task is to emulate St. Paul, who preached the Gospel in its own terms and on its own terms even to the intellectually sophisticated Athenians. /329/ 

It would be a mistake to suppose that St. Thomas does not feel the force of these considerations or that he does not recognize a grain of truth in them. To the contrary, he issues regular warnings of his own about the frailty of human reason in its postlapsarian state and about the intellect's susceptibility to prejudices and distortions that are induced by the affective disorders attendant upon human sinfulness. Nonetheless, he maintains that the effects of sin do not prevent reason from functioning as an independent and inherently trustworthy measure of both truth and intellectual virtue. Just as, appearances sometimes to the contrary, there can be no genuine conflict between the moral law imposed upon us by God and the standards of moral perfection intrinsic to human nature, so too there can be no genuine conflict between our divinely imposed obligation to accept revealed truths and the standards of intellectual perfection intrinsic to human nature. For the deliverances of faith and the deliverances of reason both emanate from the same mentor: 

The teacher's knowledge contains the very same thing that the teacher introduces into the soul of the student--unless [the teacher] teaches deceitfully, which cannot be said of God. But the cognition of naturally known principles is implanted in us by God, since God Himself is the author of our nature. So these principles are also contained in the divine wisdom. Therefore, anything contrary to principles of this sort is contrary to the divine wisdom; so it cannot come from God. Therefore, those things that are held by faith on the basis of divine revelation cannot be contrary to natural cognition. (Chapter 7) 

Our intellect is bound by conflicting considerations in such a way that it cannot proceed to a cognition of what is true. Therefore, if contrary cognitions were instilled in us by God, our intellect would thereby be hindered from cognizing the truth--[an effect] that cannot come from God. (Chapter 7) 

From this it clearly follows that whatever arguments might be propounded against the doctrines of the faith, they do not proceed correctly from first principles, known per se, which are implanted in [our] nature. Thus, [such arguments] do not have the force of demonstration, but instead are either [merely] probable arguments or sophistical arguments. And so room is left for answering them. (Chapter 7) 

No philosophical argument or theory that entails a conclusion contrary to the faith is warranted to so high a degree as to render its philosophical competitors (including its own negation) rationally unacceptable. 

Now while revelation enhances a Christian thinker's ability to identify false philosophical conclusions, it does not by itself supply a philosophical (in the narrow sense) justification for rejecting the arguments that lead to those conclusions. Only natural reason can do this. Further, the project of replying to such arguments on their own terms is, according to St. Thomas, a demand of intellectual virtue for Christians as a community (though not for each individual) and an integral part of the Church's mission to reach out to those intellectually sophisticated unbelievers who accept none of the theological authorities Christians typically have recourse to:

Some of the [Gentiles], such as the Mohammedans and the pagans, do not agree with us on the authority of any Scripture by means of which they could be won over--in the way that we can argue with Jews by appealing to the Old Testament and with heretics by appealing to the New Testament. But [the Mohammedans and pagans] accept neither [the Old nor the New Testament]. Therefore, it is necessary to revert to natural reason, which everyone is compelled to assent to--although in divine matters reason is wanting. (Chapter 2)

B. The Ockhamistic alternative

Let us now return to Ockham. As regards natural theology, Ockham is decidedly less sanguine than his predecessors about what reason unaided by revelation can demonstrate in the strict sense about the existence and nature of God. It is demonstrable, he believes, that there is a being such that no being is prior to or more perfect than it, but it is not demonstrable that there is just one such being.25 Moreover, although there are 'probable', i.e., plausible, philosophical arguments to the effect that one or another of the divine perfections is actually possessed by some being, unaided reason cannot demonstrate in the strict sense that any being has any of these perfections. 

Nonetheless, it is worth noting that by contemporary standards Ockham is positively bullish on natural theology. Today it would be remarkable indeed to find a theistic philosopher who claims to have demonstrated what Ockham explicitly asserts to be demonstrable, viz., the existence of a being than which none is more perfect. Ockham is no sceptic regarding natural theology, and only historical shortsightedness could lead one to think otherwise. 

As I intimated above, the real chasm separating Ockham from Aquinas appears only when we turn to Ockham's views about how tensions between faith and reason are to be resolved. Several examples come immediately to mind, but I will focus on the doctrine of the Trinity, since Adams discusses it in some detail (pp. 996-1007). According to this doctrine, a singular divine nature with just one intellect and will is shared by three distinct divine persons. The tension first arises when Ockham, after arguing on philosophical grounds that there are no real relations and that relative terms are all merely /332/ connotative, concedes that the doctrine of the Trinity entails or at least strongly suggests that the three divine persons are constituted by real relations of knowing and loving which they bear to one another. 

Given this apparent conflict between reason and faith, St. Thomas would maintain that the philosophical theory (call it T) leading to the conclusion that there are no real relations contains a 'philosophical' flaw, i.e., an error or infelicity that can at least in principle be discovered and rectified by natural reason itself. In a case like this, revelation guides reason by prompting the reexamination of a philosophical theory that it has exposed as unsound. 

Ockham, however, does not see things this way. To be sure, T contains a flaw, because it yields the false general conclusion that there are no real relations. But it is only through revelation that we can so much as detect T's falsity, and so we should not expect the flaw in T to be one that reason can rectify on its own. What's more, the fact that T contains this sort of flaw does nothing to alter its status as the only rationally acceptable account of relations. Thus, although we must reject T in all its generality, the proper course is to accept a modified, less general, version of T (call it T*) that applies to all and only those cases about which revelation has nothing contrary to say--even while we admit that there is no philosophical flaw in T and no philosophical justification for preferring T* to T. 

To relate this a bit more perspicuously to our discussion of St. Thomas, let us suppose that a secular philosopher invokes T to pose an objection to the doctrine of the Trinity. This philosopher argues that since, according to T, there are no real relations, the doctrine is false. 

In response, St. Thomas will maintain that the objection must emanate from a theory that can reasonably be impugned and rejected on philosophical (in the narrow sense) grounds alone. In keeping with what was said above, I take this to mean that T is not so highly warranted on philosophical grounds as to rule out its competitors as rationally unacceptable. The Christian philosopher is thus charged with carrying out a careful critique of T and, if possible, constructing a philosophical alternative to it. 

Ockham, by contrast, seems prepared to hold that T, despite entailing conclusions contrary to the faith, is indeed warranted to such a degree that it renders its philosophical competitors unacceptable on unrevealed grounds alone. His response to the secular philosopher goes like this: "I know by faith that T is mistaken, even though we share no common ground upon which I can argue my case against T in a way that has some purchase on you. But because T is the only acceptable philosophical account of relations, I do not propose to jettison it entirely. Instead, I will substitute T* for T, so that we can agree at least on all those cases that divine revelation does not speak to. You might find this response deficient and even a bit annoying, since I have not tried to refute your objection directly. But in this instance such a refutation is impossible." /333/ 

As Adams reports (pp. 999-1003), Ockham responds in like manner to the objection that the doctrine of the Trinity violates general laws governing the concept of identity. Since he believes that no philosophical account of identity which accommodates the doctrine is rationally acceptable, he simply replies that principles such as the transitivity and symmetry of identity, while they apply to all other cases, do not apply to the Trinity. For in this one instance, known to us only by revelation, there are three distinct things (the divine persons constituted by the relations), each of which is nonetheless identical with the one divine nature. So we know by revelation that there is an exception to the general principles in question, even though we have no philosophical warrant for countenancing this exception. In contrast, St. Thomas tries to show that the doctrine does not in fact breach any evident laws concerning identity. 

C. Adams's defense of Ockham

I turn now to Adams's defense of Ockham. She begins by conceding that on Ockham's view there can be ultimate conflicts between faith and reason: 

Authority sometimes implies ... conclusions that are contrary to reason--e.g., that three things are one thing, or that there are relative things really distinct from absolute things. Ockham always allows the claims of reason and experience to be defeated by contrary pronouncements of the Church, which should lead 'every thought captive'.... Church pronouncements usually restrict themselves to the narrowest possible subject matter. They do not stipulate, for example, that whenever things are related, there is a relative thing or things really distinct from the relata that relates them. To admit the latter is, for Ockham, to take a step outside the bounds of reason. To grant the fully general thesis [that whenever things are related, etc.] would be to multiply miracles without any necessity. Ockham's method is thus to subordinate reason and experience to Church authority, while keeping violations of reason and experience to a minimum. (p. 1009) 

Adams, however, seems simply to assume with Ockham that there can be ultimate conflicts between faith and reason, and then asks how a Christian thinker should react to them. For example, suppose that such a thinker tentatively agrees with Ockham that all the philosophical theories about relations which cohere with the doctrine of the Trinity are themselves rationally unacceptable. What then? Adams limits the choices to two, the worthier of which is Ockham's:

The question is how a theologian should respond to a philosophical theory that accommodates doctrinal claims ..., when after careful and honest assessment he finds it to violate the canons of reason and experience. He can reject the theory while treating the doctrinal theses as exceptions--which Ockham does. Or he can conclude that he himself is incapable of evaluating the general theory even though he finds it contradictory and/or unthinkable. In preferring the former strategy to the latter, Ockham is expressing a certain confidence in his own ability to assess philosophical theories. But if this is arrogant, I would ask which of us is in a position to cast the first stone? (p. 1010) 

Ockham's critics charge that his method is tantamount to theologism. For in allowing the philosopher to be informed by theology ... and admitting that Church authorities and our natural faculties sometimes lead us in opposite directions, he compromises the autonomy of philosophy. After reason and experience have reached their conclusion, he must refer to Ecclesiastical pronouncements to see whether and how their results must be qualified or reversed. 

Further, they charge, Ockham's philosophical mistakes led to this deplorable methodology. Had he the insight to see that Aquinas's positions, which harmonize /335/ faith and reason, are correct, or the humility to learn from him, Ockham would have strayed less far, both substantively and methodologically. 

In my opinion, there is a certain perversity in this objection. While ostensibly blaming Ockham for allowing his theological commitments to restrict his philosophical conclusions, it chides him for not regarding disharmony between a philosophical theory and Christian doctrine as a decisive indication of philosophical wrongheadedness. (pp. 1009-1010) 

What I find especially bewildering is that Adams, who throughout the rest of her book takes great care to point out the moot aspects of Ockham's thought, refrains from asking any pointed questions here. For instance, if Ockham concedes that certain evident principles--say, those governing the properties of identity--are, strictly speaking, false, then how can he justify his own ostensible confidence in these principles when they are employed outside of theological contexts? Only in the last paragraph of Chapter 22 does Adams even begin to come to grips with this issue: 

[Ockham] thinks that given who we are, we cannot but employ the laws of thought and various inductive principles as reliable guides. As Augustine taught, reason distinguishes us from the beasts, and God intends us to use it to understand the created world. Nevertheless, given who God is, we shall never be able to grasp Him fully thereby. When all is said and done, the Christian philosopher must join the rest of the Church in thanking God for grace to acknowledge and worship the mystery. (p. 1010) 

Nor does she point out that Ockham's methodology runs the risk of closing off theological inquiry at too early a stage. Ockham seems to assume that the deliverances of faith are easily identifiable. The theologian simply compares them to the theories delivered up by philosophers and makes the requisite modifications in the latter. Revelation thus serves as a check on reason, but reason apparently cannot serve as an independent check on the interpretation of the sources of revelation. (After all, as Ockham sees it, faith sanctions propositions that reason takes to be absurd.) What's more, on Ockham's view reason and faith do not guide one another. Instead, reason reaches its conclusions and then revelation qualifies those conclusions without initiating either a reexamination of the theories that led to the conclusions in the first place or a careful analysis of the deliverances of faith. St. Thomas's alternative seems not only more subtle but much more consonant with the practice of the best Christian theologians, himself included. 

I do not mean to imply that these criticisms are unanswerable within an Ockhamistic framework. I only mean to say that Adams should have addressed them forthrightly. In the final analysis, I found this the least luminous section of an otherwise stellar book.

1. William Ockham (Notre Dame, IN: University of Notre Dame Press, 1987), two volumes, 1402 pp. In what follows unannotated page numbers constitute references to this book. 

References to Ockham will be taken from the critical edition of his philosophical and theological works: Guillelmus de Ockham, Opera Philosophica (hereafter: OP), volumes I-VII (St. Bonaventure, N.Y.: Franciscan Institute Press, 1974-1988), and Opera Theologica (hereafter: OT), volumes I-X (St. Bonaventure, N.Y.: Franciscan Institute Press, 1967-1986). 

2. Here are a few of the common notions about Ockham which Adams exposes as errors: (i) that his nominalist (or, better, conceptualist) theory of universals directly entails conventionalism with respect to natural kind terms and so undermines the possibility of genuinely scientific knowledge (pp. 109-141 and 287-305); (ii) that he posits qualities as entities distinct from substances only because such entities are required by the Catholic doctrine of transubstantiation (pp. 277-279); (iii) that his account of the intuitive cognition of particulars leads to scepticism regarding perceptual beliefs (pp. 588-594 and 625-629); /337/ (iv) that in attributing actuality and extension to primary matter in itself, he anticipates Descartes's identification of matter with extension (pp. 690-695); (v) that his analysis of causation savors of Humean empiricism and engenders scepticism regarding beliefs about particular causal connections (pp. 741-798); (vi) that he is wholly sceptical about philosophical arguments for God's existence (pp. 966-979); (vii) that his Christology succumbs to the Nestorian heresy, according to which there are two persons, as well as two natures, in Christ (pp. 979-996); and (viii) that his views about merit, grace, and predestination are infected with the Pelagian heresy, according to which human beings have a natural capacity to merit eternal salvation (pp. 1295-1297 and 1345-1347). 

3. Expositio in Librum Perihermenias Aristotelis I, Prooemium, § 8 (OP II, pp. 362-363). This translation and those that follow, both from Ockham and from St. Thomas, are my own. 

4. Summula Philosophiae Naturalis III, chap. 7 (OP VI, p. 270). Signification is, strictly speaking, a semantic property that categorematic terms have independently of their occurrence in propositions, while supposition is a semantic property they have as used in propositions. What these terms signify and supposit for (their significata and supposita) are always entities. Syncategorematic terms (propositional connectives and operators, quantifiers, prepositions, conjunctions, etc.) have neither signification (in the strict sense) nor supposition, even though they affect the truth conditions of the propositions in which they occur. For more on signification and supposition, see Ockham, Summa Logicae I, chap. 33 (OP I, pp. 95-96) and chaps. 63-69 (OP I, pp. 193-209). 

5. Medieval scholastics generally hold that Platonic realism runs afoul of the Christian doctrine that everything other than God depends upon God for its existence. For a recent dissenting opinion, see Alvin Plantinga, Does God Have a Nature? (Milwaukee: Marquette University Press, 1980). 

6. Metaphysical definitions are those formulated in terms of genus and difference. See Ockham, Summa Logicae I, chap. 26 (OP I, pp. 84-89) and Quodlibeta Septem V, ques. 15 (OT IX, pp. 538-542). 

Essential or per se predication has two modes. In the first mode, the predicated term is included explicitly or implicitly in the subject's metaphysical definition, e.g., 'Human beings are animals' or 'Human beings are substances'. In the second mode, the predicated term connotes an essential attribute not included in the definition, e.g., 'Salt is soluble in water'. Scientific knowledge of a substance includes knowledge of everything that is true of it per se. In "The Necessity of Nature," Midwest Studies in Philosophy 11 (1986): 215-242, I have defended this conception of scientific knowledge and the corresponding account of the natural modalities. 

7. According to St. Thomas, 'human being' signifies an abstraction, human nature, which in itself is not properly said to exist or to have any sort of unity. Instead, it always has, per accidens, one of two types of existence: multiple existence outside the mind in individual human beings as that which is essential to them as human beings, or unitary existence in the mind as a universal mental term, 'human being', that is predicable of many. See De Ente et Essentia, chaps. 2-3. 

8. See Scriptum in Librum Primum Sententiarum: Ordinatio, dist. 2, ques. 4-8 (OT II, pp. 99-292). /338/ 

9. See, for example, Quodlibeta Septem IV, ques. 24 (OT IX, p. 413): "When a proposition is made true by things, if two things are sufficient for its truth, it is superfluous to posit a third." 

10. Scriptum in Librum Primum Sententiarum: Ordinatio, dist. 2, ques. 4 (OT II, p. 122). 

11. On Universals (Chicago: University of Chicago Press, 1972), p. xii. 

12. Ibid., p. 127. 

13. See especially OT II, pp. 122-127 and 210-212. 

14. An objection: If common terms signify singular things, then their signification will constantly change as singular things come into and pass out of existence. Ockham replies that a common term signifies all the things of which it can be truly predicated. (See Summa Logicae I, chap. 33 (OT I, p. 95).) On pp. 400-416 Adams discusses the question of whether this and other claims Ockham makes about tense and modality entail the existence of merely past, merely future, and merely possible entities. 

15. Ockham does not claim that every term in the category of quality is absolute. A quality term is connotative if it can come to be true or cease to be true of a substance simply because of a spatial rearrangement of the substance's parts. So quality terms like 'curved', 'dense', etc., are connotative, whereas terms that signify colors, habits, dispositions, powers, etc., are absolute. See Summa Logicae I, chap. 55 (OP I, pp. 179-182) and Quodlibeta Septem VII, ques. 2, (OT IX, pp. 706-708). 

16. The accidental categories are all discussed in Summa Logicae I, chaps. 44-62 (OP I, pp. 132-193). Also, a tract on relations is found in Quodlibeta Septem VI, ques. 8 - VII, ques. 8 (OT IX, pp. 611-730), while quantity is treated at length in Quodlibeta Septem IV, chaps. 23-28 (OT IX, pp. 406-445) and in the Tractatus de Quantitate and the Tractatus de Corpore Christi (both in OT X). 

17. Ockham typically uses corresponding pairs of abstract and concrete terms--and not connotative terms and their nominal definitions--as paradigms of synonymous terms. He asserts (implausibly) that according to Aristotle kind terms in the category of substance and their abstract counterparts, e.g., 'animal' and 'animality', are synonymous, since they signify exactly the same things in exactly the same way. See, e.g., Summa Logicae I, chap. 7 (OP I, pp. 23-29) and Quodlibeta Septem V, ques. 11 (OT IX, pp. 523-528). On the other hand, there are places in which Ockham indicates that sameness of nominal definition is at least a necessary condition for synonymy. See, e.g., Summa Logicae III-2, chap. 14 (OP I, p. 529) and Quodlibeta Septem IV, ques. 7 (OT IX, p. 334). 

18. OT IX, p. 676. See also Expositio in Librum Porphyrii de Praedicalibus, chap. 2, §2 (OP II, pp. 31-32). 

19. See Expositio Super Libros Elenchorum II, chap. 16 (OP III, p. 302). 

20. OT IX, pp. 582-584. 

21. In Chapter 9 St. Thomas divides knowledge of the first principle of all being into three parts: (i) knowledge of God in Himself, (ii) knowledge of the procession of creatures from God, and (iii) knowledge of the ordering of creatures to God as an end. These topics define the first three books of the Summa Contra Gentiles. So the knowledge of God in the final analysis includes a knowledge of all creatures as well. /339/ 

22. This helps explain why St. Thomas denies that one and the same person can have both faith (fides) and systematic knowledge (scientia) with respect to the same proposition. The point is sometimes rendered into English as the claim that it is impossible for anyone both to believe and to know the same proposition, but this is rather misleading given the standard use of the terms 'believe' and 'know' in contemporary epistemology. 

23. An animus against the intrusion of secular philosophy into theology characterizes many of the most important and influential reactionary movements in Church history, e.g., the fourth- and fifth-century resistance to the conciliar definitions of the doctrines of the Incarnation and the Trinity, at least some elements of the thirteenth-century opposition to Aristotle, sixteenth-century Lutheranism's call for a return to the Bible, and twentieth-century Barthian neo-orthodoxy. 

24. I am glossing over many complications that a full account of the relation between faith and reason would have to deal with: How much certitude must a philosophical or scientific theory have before it necessitates the reformulation of a doctrinal statement that it appears to conflict with? And how far can such a reformulation go before it is no longer a reformulation but a repudiation of the doctrine in question? These problems are exacerbated by the fact that what reason normally yields are probabilities rather than certainties. St. Thomas would win too easy a victory if he only had to show that philosophical or scientific theories which seem to conflict with doctrine are not demonstrated in the strict sense. But neither should one be forced on pain of irrationality to accept the most probable or most popular current theory. I suspect that, as with scientific rationality, advances in our understanding of theological rationality will depend on close and sophisticated studies of concrete historical cases. 

25. See Quodlibeta Septem I, ques. 1 (OT IX, pp. 1-11). For more discussion of the arguments for God's existence, uniqueness, and infinity, see Scriptum in Librum Primum Sententiarum: Ordinatio, dist. 2, ques. 10 (OT II, pp. 337-357) and Quodlibeta Septem II, ques. 1 (OT IX, pp. 107-111); III, ques. 1 (OT IX, pp. 199-208); and VII, ques. 11-18 (OT IX, pp. 738-779). 

26. See, e.g., Summa Theologiae III, ques. 2, art. 5, ad 1. 

27. David Hume, Dialogues Concerning Natural Religion, ed. Richard Popkin (Indianapolis: Hackett Publishing Co., 1980), p. 7.