Jump to ContentJump to Main Navigation
Knowledge, Belief, and GodNew Insights in Religious Epistemology$

Matthew A. Benton, John Hawthorne, and Dani Rabinowitz

Print publication date: 2018

Print ISBN-13: 9780198798705

Published to Oxford Scholarship Online: March 2018

DOI: 10.1093/oso/9780198798705.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 22 January 2021

Testimony, Error, and Reasonable Belief in Medieval Religious Epistemology

Testimony, Error, and Reasonable Belief in Medieval Religious Epistemology

(p.29) 2 Testimony, Error, and Reasonable Belief in Medieval Religious Epistemology
Knowledge, Belief, and God

Richard Cross

Oxford University Press

Abstract and Keywords

Aquinas generally adopts a fallibilist epistemology, according to which it is often impossible to have good internalist justification for a belief. In line with this, he adopts a fully externalist account of the reasonableness of divine faith. Faith is justified if and only if it is caused in the believer by God. Scotus is more optimistic about the prospects for internalist justification generally. Hence, he believes that it is possible to have justified belief even on the basis of merely human testimony. The views that the two thinkers adopt on the theological question are thus wholly parasitic on prior epistemological commitments.

Keywords:   Aquinas, Scotus, fallibilism, internalism, epistemic justification, religious belief

My purpose here is to consider the question of the reasonableness of Christian belief as it was approached by two medieval theologians, Thomas Aquinas and Duns Scotus. I choose these two thinkers primarily because they provide strikingly contrasting accounts, albeit that Scotus directly targets some of the central theses of Aquinas’s view on Christian faith. But what is notable about the differences is that they directly track different general epistemological positions accepted by Aquinas and Scotus for philosophical reasons quite independent of the theological question. Basically, Aquinas holds that the only ground for Christian belief is divine testimony; Scotus, contrariwise, holds that Christian belief can be made fully credible simply on the basis of human testimony. Both thinkers accept a hierarchy of epistemic attitudes—doubt, suspicion, belief (opiniomere belief), faith (fides), and knowledge (scientia). And both thinkers believe that certain cognitive processes, in certain situations, are liable to produce the wrong belief. But here is the difference: Scotus is much more optimistic than Aquinas about our capacity to diagnose such malfunctioning cases, and thus consciously to correct the initial output belief. He is thus in general far more optimistic about an internalist account of what it is for a belief to be reasonable. As we shall see, he is, in line with this, quite optimistic in general that scepticism can be refuted. Aquinas’s approach is very different. He is far more interested than Scotus is in the question of error, and adopts a globally fallibilist epistemology, believing that our natural cognitive processes are sometimes tainted by malfunction that we can neither be aware of nor prevent. Aquinas in effect moves towards an error theory that would if successful undermine any attempt at providing internalist justification for religious faith. Religious faith—specifically orthodox Christian belief—is supposed to have the highest degree of certainty (i.e. is supposed to be formed on the basis of a process that cannot go wrong), and this is why Aquinas holds that the only ground for Christian belief is divine testimony: only if the relevant faith is caused by (p.30) God can the process that causes it be maximally reliable. Aquinas in effect analyses the different epistemic attitudes externalistically, in terms of the reliability of the various processes that produce the relevant doxastic output. And the divine origin of that process is not transparent to us. Scotus’s optimism about the prospects for internalist justification more generally allows him to adopt a view of reasonable Christian belief grounded merely on human testimony—on evidence that is internally accessible to us.

2.1 Thomas Aquinas

2.1.1 Epistemology, cognitive theory, and error

Thomas Aquinas summarizes much of the earlier epistemological tradition in distinguishing between various epistemic attitudes: doubt, suspicion, belief (opiniomere belief), faith (fides), and knowledge (scientia)—in ascending order of epistemic value.1 For non-theological epistemological purposes, we can ignore faith: this is according to Aquinas a specifically theological category, and I return to it below. Doubt, suspicion, and belief (opinio) are grouped together as attitudes that ‘lack firm assent (firmam assensionem)’. Doubt is simple agnosticism: the mind fails either to assent or dissent; in suspicion the mind inclines more to one side than the other on the basis merely of ‘some slight sign (aliquo levi signo)’; in belief the mind inclines more to one side than the other, but ‘with fear that the opposite may be the case (cum formidine alterius [partis])’.

Knowledge (scientia) involves ‘firm assent’.2 Aquinas often speaks as though it is restricted to what we would label the logically necessary.3 Such propositions ‘compel assent’:4 it is not possible to grasp the terms without assenting to the proposition. But Aquinas also holds that various contingent particulars can be known in this way too: specifically, ones that are accessible introspectively: ‘Knowledge (scientia) of the soul is (p.31) most certain, because each person experiences in himself that he has a soul, and that the activities of the soul are in him.’5 Indeed, Aquinas thinks that these kinds of object compel assent, in the sense that it is not possible for someone with access to the objects not to know them. But the objects of knowledge here are not propositions; so, I suppose, strictly speaking, the only propositions that can be the object of knowledge are logically necessary ones.

The contrasting cases (belief, suspicion, doubt) have as their objects those propositions about extramental reality that fail to be logically necessary—that is to say, merely metaphysically necessary ones and contingent ones. These propositions do not compel in the way that the objects of knowledge do. In accordance with this, Aquinas claims that there is a sense in which these latter attitudes are voluntary. I do not think that he holds us to have a choice in assenting and dissenting. Rather, since the object (the evidence) does not compel assent, some further causal input is required: and the only available locus for this causal input is some agency on the part of the subject, namely the will.6

Should we understand these epistemic attitudes internalistically or externalistically? There has been some debate about this in the literature, and it is hard to say for sure (though I will make a proposal in a moment). One thing to be clear about: even though Aquinas is not always very explicit on this issue, this is not because he is unaware of the distinction. He makes it, for example, in the context of his discussion of the kinds of certainty that attach to Christian faith:

Certitude can be thought about in two ways. In one way, [something is certain] from the cause of the certitude: and in this way that which has a more certain cause is said to be more certain. … In another way, certitude can be thought about on the part of the subject. And in this way that which the human intellect more readily grasps (plenius consequitur intellectus hominis) is said to be more certain.7

(p.32) The first sense of ‘certitude’ is externalist: it is the epistemic quality attaching to outputs that reliably track inputs—here, a cause that is more ‘certain’ reliably produces the appropriate belief-output. The second is internalist and psychological: it is the felt level of credence attaching to a belief. As we shall see later in the chapter, Aquinas is discussing here a case in which these two kinds of certitude can come apart: a belief can be maximally certain in the first way—maximally reasonable from an externalist perspective—without attracting a high credence level.

Why are certain propositions the objects of knowledge, and others the objects merely of belief? The reason (the only one Aquinas offers) is that there is always scope for error in the latter cases. Aquinas explicitly addresses the knowledge/belief distinction in the context of the possibility of error only in the case of testimonial evidence.8 In this context, he distinguishes two kinds of certainty: the kind of certainty that is had ‘without fear of error’, and what he labels ‘probable certainty’: an epistemic quality attaching to beliefs that are true ‘in the greater number of cases’, even though they may turn out to be false ‘in the minority’.9 The context of the discussion is the varying degrees of credibility attaching to the evidence of greater or lesser numbers of witnesses in a legal case. Human acts are ‘contingent and variable’, and for this reason in such cases the kind of credibility that attaches to necessary truths—certainty ‘without fear of error’—is unattainable. But this does not render it unreasonable to adopt a particular belief—e.g. about the guilt or innocence of the accused. Rather, in such cases ‘the certitude of probability suffices’.10 Aquinas holds that the credibility of witnesses is increased by (among other things) their number, consistency, moral standing, intellectual capacities, and lack of evident bias or interest.11

It is the fact that the beliefs that are formed on the basis of testimony are true ‘for the most part’—and thus liable to undiagnosed error in a minority of cases—that gets them their lower epistemic status. And I think that this explains his views on the lower epistemic status of all of those sorts of natural cognitive outputs that fall short of scientia.

As Aquinas sees it, the process of forming beliefs merely on the basis of sensation involves the following stages: sensation; a sequence of sensory ‘judgments’ about the content of the sensation; (propositional) belief about the content of the sensation. Aquinas implies that error can occur at all of these stages—and hence that it is not possible for the deliverances of the senses to have the kind of reliability that attaches to knowledge in his technical sense. Indeed, operations further along in the causal process are more apt to malfunction than those earlier in the process. I examine the various features of the process in turn.

Each of the five external senses has a proper object, the feature of a material object that explains the fact that it can be sensed under a particular sensory modality—colour, for (p.33) example, in the case of sight, and sound in the case of hearing.12 The possibility of error at this bottom-most step is real but minimal:

A sense is always true with respect to its proper sense objects, or else has little falsity. For just as natural powers are deficient with respect to their proper operations only on rare occasions (in minori parte), because of some damage, so too the senses are deficient with respect to a true judgment concerning their proper objects only on rare occasions, because of some damage to the organ. This is evident in the case of the feverish, to whom sweet things seem bitter because of the tongue’s disorder.13

Aquinas talks about ‘judgment’ in this case. But he is evidently not talking about something propositional. He is talking about a sub-group of what we would call ‘seemings’: sensory seemings. Aquinas sometimes talks about these kinds of seemings as ‘certain’: ‘we function with certainty regarding some actual object … when sensing it.’14 But it turns out that even occurrent sensations are subject to error. When the organ does go wrong, the error consists in the object’s form being

received in the sense differently from the way in which it exists in the sense object. (I mean differently with respect to species, not with respect to matter—for example, if the flavour of something sweet were received on the tongue as though it were bitter. With respect to matter the sense always receives differently from the way in which the sense object has it.)15

All cognitive acts, according to Aquinas, consist in the form of the external object being somehow ‘received’ in the cognizer without making the receiver an instance of the relevant kind. A form F-ness, that is to say, can be received in two ways: in matter, such that it constitutes with matter an instance of F-ness; or in some kind of cognitive receptor, such that it fails to constitute the receptor as an instance of F-ness.16 In the passage just quoted, the ‘species’ here is something like the content of the perception: error occurs when the content does not match up to the object.

Now, the five senses are capable of sensing material substances as such, but only under the senses’ proper objects. As Aquinas puts it, following Aristotle, material substances are sensible only ‘accidentally’, in the sense that the properties under which the object is sensed are accidental to it.17 Substances are in this sense the per accidens objects of the senses. Not only are they the per accidens objects of the external senses; Aquinas holds that they are the per accidens objects of some inner non-intellectual (p.34) cognitive power responsible for causing an inner representation of the substance as a particular of a given kind—variously the so-called ‘cogitative’ power18 and the imagination.19 This kind of cognition is merely perceptual awareness, not a propositional judgment or belief.

Aquinas discusses the possibility of error in the case both of the per accidens object of the external senses and of the object of the imagination. On the external senses, Aquinas at one point concedes that an external sense can sense the per accidens objects of sense, but that when they do so the sensation is subject to considerable danger of malfunction.20 Elsewhere, however, Aquinas writes in ways that suggest that the external senses, in the absence of any further power, are necessarily incapable of avoiding deception:

With respect to per accidens … sense objects, the senses are deceived. Sight is in this way deceived if a human being wishes to make a judgment through sight about what that colored thing is or where it is. Someone is likewise deceived who wishes to make a judgment through hearing about what it is that makes the sound.21

Aquinas devotes considerable time to malfunction in the imagination. And this turns out to be important, because the imagination turns out to be centrally relevant to our intellectual, propositional, beliefs about particulars. He supposes that the imagination is subject to the following kinds of processes: picturing the proper sense objects in the presence of such objects; picturing the proper sense objects in the absence of such objects; picturing substances (per accidens sense objects) in the presence of such objects; and picturing substances in their absence. In each case, the picturing is causally dependent on some initial sensation:

The movement of imagination that is brought about by an act of sense different from those … senses [viz. of proper and per accidens objects] in the way that an effect differs from its cause.22

Because of this causal dependence, the imagination is more prone to deception than the external senses:


Falsity (which consists in a sense’s being unlike its sense object) can occur in imagination more easily than in a sense. This is because (i) an effect is weaker than its cause, and (ii) insofar as something is extended farther from its first agent, to that extent it receives less of that agent’s force and likeness. For there is falsity when the sense object’s form is received in the sense differently from how it exists in the sense object.23

But this proneness comes in degrees. In relation to the imagination’s picturing the proper sense objects in the presence of those objects, Aquinas maintains that this picturing ‘is true for the most part’.24 In relation to its picturing the proper sense objects in their absence, ‘it is possible to be deceived … for we sometimes imagine absent things as white, even though they are black’.25 What Aquinas means is just that we can misremember how things really are (or were). In relation to the per accidens objects, the picturing ‘can be false whether the sense object is present or not’—presumably in, among other things, cases in which we simply cannot make out what the object of the senses actually is—for example, ‘when it is far off’.26

The material about the imagination is important for the question of the credibility of propositions ascribing universal predicates to contingent individuals (e.g. ‘Socrates is a man’), since the mechanisms that form these are no more reliable than the activity of the imagination. (This is not, of course, to say, that the level of credibility might not be high; it is just to say that it will fall short of any of the levels of credibility of elements in the process prior to imagination.) As Aquinas sees it, the sensory operations just described somehow ground the further, intellectual, judgment that such-and-such is the case. According to Aquinas, no cognition of a universal can be had without our forming unified particular (sensory) representations of particulars: and so in the process of cognizing universals we automatically cognize particulars. The process is not particularly clear, but here is what he says:

Our intellect knows directly the universal only. But indirectly, and as it were by a kind of reflection, it can know the singular, because … the intellect, in order to understand, needs to turn to the phantasms in which it understands the species [viz. the universal]. … Therefore it understands the universal directly through the intelligible species, and indirectly the singular represented by the phantasm. And thus it forms the proposition ‘Socrates is a man.’27

The idea is that any kind of universal-concept formation involves the activity of the imagination, picturing a particular of the relevant kind. And this particular representation is evidently able, somehow, to figure in the content of propositions about particulars. I assume that error in an individual’s empirical beliefs is to be grounded either in sensory malfunction (in the way just described), or in some mistake in inductive reasoning. Aquinas never discusses the issue to the best of my knowledge.

(p.36) But Aquinas does have things to say about error in propositional beliefs about universals. We form universal concepts by abstraction from sensations of particulars.28 Simple cognition of a universal consists in a universal form’s inhering in the mind in such a way as not to constitute the mind as an instance of the cognized kind; and since there is no restriction on the content of possible thoughts, it makes no sense to talk about error or malfunction in such cases.29 But abstraction does not immediately get us knowledge of substance-kinds. We begin (according to Aquinas) from very general concepts, and proceed to form more specific concepts by further analysis.30 The end result is the formation of a proposition, a real definition of a substance kind—something that we would think of as metaphysically necessary, and Aquinas would call ‘self-evident in itself but not to us’.31 These propositions do not fall under scientia strictly speaking (which has as its domain merely the logically necessary). But since they are metaphysically necessary, such propositions—e.g. ‘Water is H2O’—can form premises in ordered sets of syllogisms that are what Scott Macdonald has labelled ‘non-paradigmatic scientia’.32 Someone who knew a priori the nature of water would know without investigation that water is H2O, and would thus treat ‘Water is H2O’ as a premise in strictly scientific reasoning. But we are not such privileged cognizers, and there is from our cognitive standpoint no scientia strictly speaking of ‘Water is H2O’. All such non-paradigmatic cases of scientia have an epistemic status lower than scientia strictly speaking. In such cases, we have merely belief, albeit of a superior kind:

Just as in those things which, in natural matters, act for the most part, a particular degree is found (because the stronger a natural power is, the less frequently does it fail in its effect), so in the progression of human reasoning, which lacks utter certainty, a particular degree is found, in so far as it approaches more or less to perfect certainty. For through this progression there is sometimes brought about a belief (fides) or opinion (opinio), on account of the probability of the propositions from which it proceeds—even if knowledge (scientia) is not brought about: because the mind totally inclines to one side of a contradiction, albeit with fear that the other side [may be true].33

The idea is that a high epistemic status attaches to the abstractive and inductive generalizations that lead to knowledge of real definitions. (Note that ‘fides’ here is simply synonymous with ‘opinio’: it is not being used in a technical sense.) But error can of course occur in this process:

The intellect … may be … deceived in the quiddity of composite things, not by the defect of its organ, for the intellect is a faculty that is independent of an organ; but on the part of the composition affecting the definition, when, for instance, the definition of a thing is false in relation to something else, as the definition of a circle applied to a triangle; or when a definition is false (p.37) in itself as involving the composition of things incompatible; as, for instance, to describe anything as ‘a rational winged animal.’34

We can be simply mistaken about what water is. And not only this; we can mistakenly believe that there is some kind that turns out to be metaphysically impossible.

2.1.2 Christian faith

These epistemological principles have a direct impact on Aquinas’s views on the credibility of Christian faith. To understand why, we need to keep in mind an assumption about Christian faith that was, as far as I know, common in the Middle Ages and beyond: that Christian faith must be as epistemically robust as scientific knowledge:

The act which consists in having faith (credere) has firm adherence to one side. Someone with faith has this [firm adherence] in common with someone who knows a first principle or a conclusion from such a principle (sciente et intelligente), even though his cognition is not made perfect through manifest vision.35

Given that, according to Aquinas, beliefs held on the basis of human testimony cannot attain this level of epistemic security, it follows that Christian faith cannot be grounded in any kind of human testimony. This is indeed precisely what Aquinas argues: the kinds of external reasons that can be given for the Christian faith do no more than show that the contents of the Christian faith are ‘not impossible’:

Arguments which induce one to accept the authority of faith are not demonstrations that can lead the human intellect to scientific vision (visionem intelligibilem). For this reason these things do not cease to lack apparentness. But they remove impediments to the faith, by showing that the things posited in the faith are not impossible.36

The contrast is between scientia and entailments from things thus known, on the one hand, and faith on the other: equally certain, even though only the former is had on the basis of compelling evidence.

But in what sense certain? In one important passage, part of which I quoted above, Aquinas distinguishes between internalist and externalist senses of certainty in this context:

Certitude can be thought about in two ways. In one way, [something is certain] from the cause of the certitude: and in this way that which has a more certain cause is said to be more certain. And in this way, faith is more certain than the three things just mentioned [viz. the intellectual virtues of wisdom, understanding, and scientia]. In another way, certitude can be thought about on the part of the subject. And in this way that which the human intellect more readily grasps (plenius consequitur intellectus hominis) is said to be more certain. And in this way, since (p.38) those things that pertain to faith are above the human intellect whereas those things which are subject to the three things just mentioned are not, faith is less certain.37

According to this passage, then, the certainty of faith is determined on wholly externalist grounds—that it was brought about by a cause that cannot be defective. This is clearly one way of overcoming the kind of fallibilism that attaches to non-scientific cognition outside of theological contexts. From the internalist point of view, Christian faith is not maximally certain. But it is, nevertheless, such that there is no possibility of it being mistaken: it is certain in an externalist sense. And it is so only because it has a supernatural cause.

Aquinas argues that faith is a habit: a disposition to assent to a given set of claims.38 The idea is that someone with such a disposition would, when presented with one of that set of claims, automatically assent to it. So infused faith is what John Hawthorne has, in a piece of rare historical research, labelled a ‘hyper-reliable’ habit (a hyper-reliable intellectual virtue, we might say, or something like an intellectual virtue), preventing the person who has it from firmly assenting to any ‘proposition that contradicts true sacred doctrine’.39

Where does this leave the will? Aquinas holds, as we have seen, that all epistemic states that fall short of scientia, the involvement of the will is necessary. And so it is here:

A created intellect is determined to one side not by reason, but by the will. And for this reason assent is here taken to be an act of the intellect as it is determined to one side by the will.40

Now, on the one hand, faith is wholly the gift of God. Aquinas thinks it necessary to say this in order to avoid the Pelagian heresy.41 The Pelagian view is that human beings have an active role in their own salvation, either causing themselves to be saved by their actions, or actively cooperating in their reception of divine grace. But there are marginal cases: what about the case, for example, in which the independent human role in accepting the gift of grace consists merely in being free to resist grace but nevertheless not resisting it? Aquinas thinks that even this minimal view is heretical: faith is wholly a gift of God, something infused by him (in the jargon).42 God’s action is supposed to be non-coercive. But it gains this feature in the case under discussion precisely by God’s causing someone to will in a certain way. Faith is infused by God, by his moving the will to accept it—and since the gift is accepted in accordance with will, the acceptance is not coerced. But this is fully in accord with Aquinas’s general discussion of grace: grace is accepted in virtue of God’s moving the will to accept it. In the reception of grace:

there is an interior act of the will, and with regard to this act the will is moved, and God is the mover; and especially when the will, which hitherto willed evil, begins to will good. And hence, to the extent that God moves the human mind to this act, we speak of operating grace.43

(p.39) God gives the relevant habit to the believer by moving the believer’s will to accept the habit. Once the habit is possessed, the habit causes the will to accept—and the intellect to assent to—the contents of Christian faith.

Part of the content of Christian faith is that it is revealed by God, and adopting an externalist reading of Aquinas can help us respond to an objection that has been raised—and that, as we shall see, Scotus raises—against Aquinas’s view. At one point, Aquinas makes the following observation:

If, in faith, we consider the formal aspect of the object that aspect in virtue of which faith is held, this is nothing other than the first truth: for faith assents to something only on the grounds that (quia) it is revealed by God.44

On an internalist reading, this would make it look as though Aquinas means to affirm that our knowing that such-and-such was revealed by God serves as grounds for accepting it. And this raises two problems. The first is the one that Scotus later raises: it is surely circular. (I return to this below.) The second has been raised by John Lamont. If it were the case that we believe something on the basis of testimony that we have good reason to suppose is infallible, there should be no place for the will in Christian faith:

It is quite possible for us to see that there is a contradiction implied in God’s speaking falsely, and hence that it is impossible that this should happen. Since this is so, why should faith differ from knowledge? And why should it be voluntary?45

Lamont calls this an ‘unresolved tension in his [viz. Aquinas’s] conception of faith’.46 But the worry relies on our being antecedently aware that the content of the faith is revealed by God. This would indeed seem to make the view susceptible to Lamont’s objection. Aquinas does not posit such awareness, however. According to Aquinas we can come to accept that the contents of the faith are revealed by God only once we have faith: it is, as it were, part of the content of the faith:

The light of faith brings it about that we see those things that are believed. For, just as through other habits of the virtues a human being sees that which is appropriate to him through that habit, so through the habit of faith the mind of a human being is inclined to assent to those things that pertain to right faith, and not to other things.47

I assume that, among the things to be believed is the claim that God reveals them. Now, it is true enough that Aquinas is not as clear as we might like on this matter. But charity perhaps suggests that we should adopt an externalist reading of the last but one passage quoted from Aquinas: claiming that ‘faith assents to something only on the grounds that it is revealed by God’ means no more than that the habit of faith (including a disposition to believe that the faith was revealed by God) is caused by God alone. The will (p.40) is involved in accepting the habit, antecedently to assenting, on the basis of the habit, to the various propositions that comprise the content of the faith.

In any case, Aquinas thinks that empirical evidence is on his side in the matter. He notices that two people might be presented with the same evidence—miracles, say, and human testimony—but yet only one of them accepts the faith for which these things are evidence. So this kind of evidence cannot be sufficient for belief, and there must be some divine or supernatural explanation for the different responses—i.e. infused faith.48

Given all this, it follows that someone accepting the tenets of the Christian faith, but in such a way that this acceptance was not brought about by God, cannot have faith at all:

Whoever does not inhere, as in an infallible and divine rule, in the teaching of the Church, which proceeds from the first truth manifest in the Holy Scriptures, does not have the habit of faith, but holds those things which are of the faith in some manner other than through faith. … So it is clear that a heretic on one of the articles of faith does not have faith about the other articles, but a particular belief (opinionem) according to his own will.49

The heretic is in a triple bind: he lacks (i.e. ignores) all the relevant evidence, he lacks sufficient reason to believe any of the Christian faith, and he lacks God’s grace moving the will to assent to the evidence.

Aquinas pursues to its absolute logical conclusion his insight that human testimony cannot be sufficient for faith. It is not the Prophets, Apostles, and Evangelists who are relevantly trustworthy: it is God alone. As we shall see, Scotus takes a very different line, apparently holding that the heretic’s beliefs are just as credible as the Catholic’s. This is because Scotus takes a much more optimistic line about testimonial credibility in general than Aquinas does.

2.2 Duns Scotus

2.2.1 Against scepticism

I think it is fair to say that Aquinas has very little concern for these epistemological questions as such; what captures his interest are the theological, moral, and legal consequences of the epistemological questions. For example, he says nothing systematically about the reliability of sensation—I have had to rely on scraps from the De anima commentary. The situation is very different in Scotus. Sceptical concerns came to the fore in the thinking of an immediate predecessor of Scotus’s, Henry of Ghent, who believed that the only way to counter scepticism was by an appeal to divine illumination. Scotus argues, contrariwise, that an appeal to divine illumination does not solve sceptical worries, and instead attempts to show how such concerns (p.41) can be best countered by a thoroughgoing epistemological naturalism. At any rate, he gives the matter sustained attention in a reasonably well-known question, one which Robert Pasnau has described as ‘by far the most sophisticated of its kind in the Middle Ages’.50 I return to this discussion in a moment.

Scotus uses the same terms for the various epistemic attitudes that we find in Aquinas—my concern here will be with knowledge (scientia), faith (fides), and belief (opinio). There is no doubt that these are often construed by Scotus externalistically—epistemic attitudes with or without the possibility of error. Billy Dunaway makes the point very effectively in his chapter in this volume, and we shall see a number of examples below. But he understands some of them internalistically—in particular, in cases of beliefs based on human testimony. Given that—as Scotus supposes—human testimony is the relevant source for Christian faith, he has an internalist account of the rationality of belief. Certitude in relation to faith is something that we ‘experience in ourselves’:51 it is a felt credence level. The thrust of Scotus’s anti-sceptical arguments is to try to provide reasons for believing that our basic cognitive mechanisms are either reliable or correctable; so we should think of the general thrust of Scotus’s epistemology as leading in an internalist direction—and thus in sharp contrast to Aquinas’s externalism.

The basic gist of the anti-sceptical arguments—which have been discussed extensively in the literature, and which I will not treat in any detail here—is that our cognitive mechanisms can be shown to be by and large reliable, and that in cases of deviant outputs it is in principle always possible to put in place correctives.52 Just as for Aquinas, Scotus holds that it is impossible to grasp the terms of a self-evident proposition without assenting to it.53 Any proposition derived from such a composite by syllogistic reasoning will also be equally certain, on the basis of the dictum de omni.54 Equally, it is impossible to be mistaken about one’s own subjective experience, or about the identity of one’s own soul: not even God could substitute my soul for another and it still be the case that I would have the same self-awareness.55 Scotus takes Aristotle to teach that there is a further mental condition about which we cannot be mistaken: namely, being awake. And Scotus also seems to accept this, though as far as I can see only on the basis of Aristotelian authority.56

Sensation can be shown to be reliable, too. Scotus considers two cases. In the first, all of the senses apparently yield the same ‘judgment’ about external reality: which is to say that the deliverances of the senses are all in conformity with each other. In the (p.42) second, this condition does not obtain. Scotus discusses the first case by appealing to a principle of inductive reasoning (I label it ‘SPIR’ (‘Scotus’s Principle of Inductive Reasoning’)) that he takes to undergird empirical knowledge in general:

[SPIR] Whatever comes about (evenit), for the most part, from some non-free cause, is the natural effect of that cause.57

In the case at hand, the basic idea is that if such-and-such a visual experience (for example) regularly occurs in the apparent presence of such-and-such an object, then the experience is caused by the object (and thus veridical). I assume that the point about multiple sensory modalities is that these provide a way of confirming, other than by sight (for example), that such-and-such an object is present to sight. (Or perhaps that the uniformity provides some inductive reason to accept the veracity of the senses in the relevant case.)

In the second case, Scotus holds that the intellect will have (or can get) a principle enabling it to adjudicate between the different senses.58 He gives two stock examples. First, we see a semi-submerged stick apparently broken.59 But we know through various sensory modalities—‘both sight and touch’—that a stick is harder than water, and we have a self-evident principle that harder things are not broken by softer things. (Self-evident because it is, I suppose, part of the meaning of ‘hardness’ to be thus unbreakable.) So we have a way of adjudicating between the different deliverances of the different senses (sight and touch).60 Secondly, the sun (and things far away in general) look to us smaller than they really are.61 But ‘a measure applied to an extended object is always equal to itself’, and, ‘both sight and touch attest to the fact that the same measure can be applied to something that is seen, whether the thing is near or far’.62 So the idea is that we always have the means to discern when we are subject to some kind of illusion—there is always in principle a mechanism to enable us to discern and correct deviant outputs. Scotus does not mean to suggest that it is easy, or that failure to attend sufficiently might not result in error. But sufficient care can enable us to avoid error. (Scotus even allows exceptional cases in which the nature of the experience intrinsically involves a lack of certainty—though, I assume, such an experience can still allow for certainty about the credence level that attaches to it.)63

Scotus devised SPIR to underwrite inductive empirical generalizations overall—the fourth case he considers in his sequence of anti-sceptical arguments. He argues that SPIR is more-or-less self-evident, in the sense that it is a matter of the definition of ‘natural cause’ that it cannot regularly produce effects of a different kind from those (p.43) that it is apt to produce. Experience gives us applications of the principle. Consider a substance that can be subject to many different kinds of accidental modification. Suppose that this substance regularly produces an effect irrespective of the configuration of accidents it possesses. We can infer that the effect is the natural effect of the substance, the one it is apt to produce.64 Scotus goes on to outline a procedure of inductive reasoning. Take a fact to be explained, and think through kinds of possible hypothetical explanation. Then identify a thing in the world that instantiates one such explanation. Scotus’s example (taken from Aristotle) is a lunar eclipse:

When an experience of the conclusion is accepted—for example, that the moon is sometimes eclipsed—and we suppose that the conclusion is true, we look for the cause of the conclusion by the way of division. And when we arrive, from the experienced conclusion, to principles known from their terms, then the conclusion can be known more certainly (namely, in the first genus of knowledge, because deduced from a principle that is known per se) when derived from the principle known from its terms. For example, this is known per se: an opaque body, interposed between a visible object and a light source, prevents the multiplication of light to the visible object. And if it were found by division that the earth is such a body, interposed between the sun and the moon, then [the conclusion] will be known by a propter quid demonstration (because through the cause), and not merely by experience (as the conclusion was known before the discovery of the principle).65

Scotus uses this account to explain how inductive knowledge allows us to establish middle terms in Aristotelian scientific syllogisms—indeed, the scientific nature of syllogistic reasoning requires major premises that express necessary a posteriori truths (or, if you prefer, metaphysical but not logical necessities).

In line with all this, Scotus of course accepts some kind of Aristotelian abstraction—forming general concepts from particular experiences66—though in contrast to Aquinas he holds that the process of conceptual analysis typically starts not with a generic universal concept but with a universal concept of a species under a nominal definition; further empirical input and conceptual analysis allows progress towards a real definition—as in the passage just quoted.67 Again, I think Scotus would note that the fact that such work is hard does not mean that it is not possible to get certainty, in much the way outlined.

In all of these cases, the question at issue seems to be the reliability of a given cognitive process—albeit that this reliability is something that is accessible to us (this, after all, being the point of Scotus’s arguments). Scotus’s treatment of testimonial beliefs is somewhat different, developed along more directly internalist lines—though no less epistemically optimistic. As we shall see, this combination of philosophical claims (p.44) (optimistic internalism) leads Scotus, at least in his earlier discussion of the issue, to a quite different account of Christian faith from that espoused by Aquinas.

Both points are made in the following passage:

I believe (credo) that the world did not begin at the same time as I did, not because I know (scio) that it preceded me (because there is no knowledge (scientia) of the past, according to Augustine), or because I opine (opinor) that the world preceded me; but I assent to the claim that the world preceded me firmly, by acquired faith (fidem acquisitam), on the basis of the testimony of others (ex auditu aliorum), whose veracity I believe firmly; neither do I doubt that the world preceded me, or that there are parts of the world that I have not seen, because I do not entertain doubts about the veracity of those who tell me these things and who assert that these things are true. For this reason, just as I do not hesitate about their veracity (which is as it were the premise (principium)), so too neither do I hesitate about what they say (which is as it were the conclusion that follows).68

What Scotus is talking about here is subjective certainty—a credence level. And his optimism (ceteris paribus) about testimonial beliefs leads him to treat such beliefs as having a credence level equal to that attaching to knowledge (scientia). As he puts it elsewhere, ‘we can believe in the testimony of others, and even so firmly that that belief is called … knowing (scire)’.69 Scotus talks about such a belief as acquired faith (fides)—a category that does not exist in Aquinas’s account of natural cognitive processes. At various places in his work, Scotus gives some examples of fully credible beliefs based on testimony: here, his belief that the world existed before he did, and that there exist parts of the world that he has not seen; elsewhere, more specifically, that Rome exists (given that Scotus has never been to Rome),70 and, more generally, beliefs about ‘histories and things written about wars and other events, which are written in chronicles’;71 and in a passage I quote below he adds ‘that there is anywhere in the world that you have not been, [and] that this person is your father and this your mother’. The idea, I take it, is the rather commonsensical one that it would be simply unreasonable for Scotus (or indeed anyone similarly placed) to entertain doubt about these matters. The rest of the quoted passage explains how acquired faith—certain belief on the basis of merely human testimony—is possible. Basically, in cases where we unhesitatingly accept the veracity of our informants, we unhesitatingly accept their testimony. Scotus holds that this applies to very commonplace beliefs; I assume that there is a presumption in favour of the veracity of our informants: we automatically trust them unless there is a defeater. Note that the belief in testimonial evidence is inferential: what is basic is our belief that informers are ceteris paribus reliable.

(p.45) Scotus is clear that all of these beliefs—beliefs that fall short of scientia strictly speaking—require the role of the will. But he states explicitly the view that I ascribed above to Aquinas (see n. 6, this chapter): that the role of the will cannot be to make someone believe something. My will cannot, to use Scotus’s example, make me believe that the number of stars is even. The will is merely a ‘general mover moving [the intellect] (generalis motor movens)’. To the extent that the evidence does not compel assent, conscious causal input on the part of the agent is required: and all such causal input is ascribed to the will.72

2.2.2 Christian faith

As we have seen, Aquinas holds that infused faith secures belief ‘without the fear of the opposite’—i.e. without the possibility of error.73 And I have just shown that Scotus construes the certainty attaching to faith internalistically, not externalistically. Given his optimism about the possibility of subjective certainty attaching to human testimony in general, it is no surprise that he thinks of acquired faith as providing the relevant kind of certainty, even in the absence of infused faith. As he puts it:

On account of the credibility (credulitatem) of the revealed articles, it is not necessary to posit infused faith so that a human being might firmly believe all the revealed articles, and be determined to one side without fear of the opposite; neither can its necessity be deduced on the grounds that acquired faith (fides acquisita) is above opinion (opinio) (which assents to one side of a contradiction with fear of the opposite) but below knowledge (scientiam) (which is from the evidence of a scientific object).74

Acquired faith is something available on the basis simply of human testimony, and Scotus goes on to give examples of other things that it is rational to believe without doubt simply on the basis of human testimony (I have discussed some of these above). Just like these cases, then, acquired faith here is held to have a greater degree of credence than Aquinas’s belief (opinio), and to have this credence on the basis of the credibility of its contents. Indeed, it follows from Aquinas’s view that there could be no such thing as acquired faith: on the basis of human testimony we could never have more than opinio.

Scotus holds that it is possible to acquire faith on the basis of the Church’s testimony,75 or simply on the basis of someone’s preaching,76 without any appeal to grace or infused faith. At one point, he attempts to show that it is possible, without infused faith, to have beliefs that have both the same content as those had by infused faith, and an equal (p.46) degree of subjective certainty as had with infused faith. He gives two rather vivid examples. The first concerns someone believing simply on the basis of human testimony:

Suppose a Jewish child were brought up among us, and taught in our divine way, but without ever being baptized. He would, by faith acquired through listening, believe and assent to all the things which we assent to, just as I, by faith acquired through listening to others (namely, my parent), whose veracity I believe, believe that many times have passed, and that the world did not begin with me; and, from the report of those worthy of belief, [believe] that Rome (which I have never seen) exists. So too I firmly assent, by faith acquired through listening, to the things revealed in Scripture, by believing in the Church that approves the truth of these authors.77

Here, Scotus compares the credence level of theological beliefs acquired in the way described with the credence level of standard human testimonial evidence. Note that Scotus’s argument here requires a different view of the epistemic status of Church authority from that accepted by Aquinas: for Aquinas, the Church’s teaching amounts to divine testimony; for Scotus, it is merely human.78 (We will see later in this chapter something analogous in the case of scriptural teachings.)

The second is more difficult:

Suppose that there is someone baptized, who believes all the articles [of faith], but later falls into error about one of them: for example, about the Trinity of persons and unity of essence. He nevertheless feels (sentit) himself to assent firmly to all the other articles of belief, just as before (because it is not necessary, if he errs in one article, to err in all of them, because heresies are not necessarily connected). Therefore he believes the other articles without infused faith (since infused faith is destroyed when heresy appears, otherwise faith and heresy would be compatible). Therefore he believes the other articles by acquired faith prior to the corruption of infused faith, because he believes the other articles even with the error in relation to one of them. Neither is acquired faith produced in the third instant, in which infused faith is destroyed by heresy, for that error was not a disposition for the production of faith at that time; rather it was a disposition against the production of faith; therefore it existed with infused faith prior to the error. And thus it is certain that acquired faith is in us.79

The discussion here is a little obscure. Basically, Scotus has in mind a three-stage process: orthodox belief, heretical belief, and the destruction of infused faith. The second and third stages are temporally simultaneous, but there is a causal relation between them (heretical belief causes the destruction of infused faith). Scotus has two arguments in favour of faith acquired on the basis of human testimony. First, there is no phenomenal difference in the person’s credence before and after the switch from orthodoxy to heresy; secondly, the advent of heresy does not seem to be a plausible cause of any kind of true faith (as it would have to be a cause of acquired faith if that did not previously co-exist with infused faith).

(p.47) But why in any case should acquired Christian faith attract the same credence level as beliefs to the effect that I was born, or that Rome exists? Scotus’s strategy is to attempt to show that the Church and the Bible are highly credible witnesses. The arguments strike me as a rather motley crew. The point that I wish to draw attention to is their functional role in Scotus’s overall argument about the credence level of acquired faith. As Scotus sees it, ‘there are eight ways of rationally convincing’ people who do not accept the totality of the Catholic faith;80 he later adds two more.81 First, the fulfilment of scriptural prophecies.82 Secondly, the internal consonance of the biblical texts, which requires the guidance of ‘a higher intellect’.83 Thirdly, the authority of the sacred writers: they claim to have access to divine revelation in a way that allows us to infer that they were not deceived;84 they underwent great persecution, and so would not be lying for some kind of gain.85

Fourthly, the attentiveness and care which the Jewish and Christian communities put into establishing the authentic contents of the Bible. I discuss this in more detail since, among other things, it lays out an epistemological principle undergirding political life. The argument has two parts, both of which anticipate in rough form some recent developments in social epistemology: first, that we should in principle trust the communities that we belong to (else no political life would be possible); and secondly that we should particularly trust the judgments of such communities if there is evidence that they have taken great care in coming to those judgments. Of course, the relevant community is the Church, and the judgment the Church’s approbation of the Bible:

[If] you will believe nothing about a contingent thing that you have not seen—and thus will not believe that the world was made before you, or that there is anywhere in the world that you have not been, or that this person is your father and this your mother—this incredulity destroys all political life. So if you want to believe something about a contingent thing that is not and was not evident to you, then what is maximally credible is the community, or those of whom the whole community approves, and especially when a reputable (famosa) and honest community receives with maximal diligence those things that are to be approved. Such is the Canon of Scripture. But there was such solicitude among the Jews about the books to be included in the Canon, and such diligence among the Christians about the books to be received as authentic, that no such solicitude has been found about any writing to be deemed as authentic—especially since such sacred communities cared about these scriptures as containing things necessary for salvation.86

Here, the basic idea is that we should believe what our community believes unless we can find a defeater: there is, in other words, a presumption of rationality in such cases.

(p.48) Fifthly, the ‘reasonableness of [the Bible’s] contents’.87 Sixthly, ‘unreasonableness’ of other belief systems (Scotus mentions: pagan idol worship; a Muslim eschatological expectation of a life devoted to food and sex; the empty nature of Jewish rituals after Christ; the Manichaean (i.e. Cathar) belief in a supreme (or at least very great) evil principle. Elsewhere of course Scotus has a proof for the existence of God, making atheism unreasonable too, as he would think).88 Seventhly, the ‘stability of the Church’: Scotus contrasts to Christianity Jewish communities, which ‘do not retain their vigour’, and Islam, which, as he thought, had been decisively defeated (by combined Mongol and Monophysite Christian forces) at the Third Battle of Homs in 1299.89 Eighthly, the evidence of miracles.90 Ninthly, the evidence of unbelievers (Josephus and the Sibylline oracles). Finally, evidence of God’s activity among those who seek for him:

God is not absent from those who seek salvation with their whole heart. For many who very diligently seek salvation are converted to this sect; and the more earnest they are in seeking, the more they are confirmed in this sect, and quick in penitence, and are changed from malice to a good life; and thirdly, many endure pains for this sect with great exultation of spirit. None of these things would be probable (probabilia) unless God singularly approved this sect, dependent on Scripture, and ordained it to salvation.91

What, then, of infused faith? Scotus ends up changing his mind on the nature and necessity of infused faith. His earlier account, in the Lectura, which has been the focus of my discussion thus far, he does not see how there could be such a thing as a hyper-reliable habit, and thus has difficulty fitting infused faith into his overall scheme at all. Later, in the Quodlibet, he modifies his account of what an intellectual habit could be, and comes to accept something like Aquinas’s view—though explicitly without abandoning his optimism about acquired faith.

In the Lectura, Scotus describes two views on infused faith—one that construes the relevant certainty internalistically and subjectively, and one that construes it externalistically and objectively. The first one takes Aquinas’s claim that the certainty of infused faith is grounded on the fact that God reveals the contents of that faith. As Scotus presents the theory, infused faith has two components: it has particular contents (the articles of faith), and it causes the believer to assent to those contents for the reason that God asserts them (and, in turn, grounds this assent on belief in God’s veracity).92 Scotus objects: the contents of the belief are not a priori, and so the belief, to be reasonable, requires some further ground. But what grounds it cannot be divine (p.49) testimony, since accepting that the belief derives from divine testimony is itself a consequence of the belief. Aquinas’s view, on this internalist reading, is viciously circular.93 (Of course, if I am right, Aquinas would not accept this reading. As John Lamont has rightly argued, Scotus’s responses ‘do not adequately get to grips with the Thomist position that they oppose’.)94

Neither can the grounds for faith be some acquired belief, since then the credence level attaching to infused faith could be no greater than the credence level attached to acquired faith—contrary to the hypothesis.95 (Again, the idea is that the grounds for certainty need to be accessible to the believer.) Secondly, we apparently rely on human testimony for the belief that such-and-such was revealed by God—for example, ‘because John [the divine], or some other apostle, I do not care which, says that this [viz. that God is a Trinity and unity] was revealed to him’. From which Scotus infers that the credence level of infused faith depends on the credence level of acquired faith.96 (There is a deep disagreement here about the nature of the Scriptures (and, derivatively, Church teachings): are they God’s words, or merely witnesses to God’s words? Aquinas affirms the former, Scotus apparently the latter.) Scotus’s overall conclusion is that this first view is untenable: ‘Therefore it seems that infused faith should not be posited in the way in which [this theory] posits it.’97

The second view is that infused faith consists simply in a divinely bestowed propensity to believe particular propositions—the articles of faith—much like Aquinas’s hyper-reliable habit. (I number the sentences for ease of reference.)

[1] God immediately infuses into us a habit of faith, inclining our intellect to assent perfectly to the articles of faith.98 … [2] This habit immediately inclines [the intellect] to the articles of faith, just as it would do if they were acquired from articles that are known (cognitis), in the same way (in this respect) as knowledge (scientia) inclines the intellect to things that can be known (cognoscibilia). [3] Yet the things that can be known are present in the habit not in such a way that the habit represents them; rather, they are present through a species. [4] Nevertheless, if the habit remained but the object was not present through a species, and if the species could be removed but the knowledge remain, then the knowledge would incline [the intellect] more to those objects that were at some time present in the species than to any others, such that a human being would more easily understand [the objects] and infer conclusions from them, than he would from other objects of which he had never had a [cognitive] act. [5] Similarly in this case, although the articles to be believed (credibilia) are present not through infused faith (such that it would represent these objects—since no habit does this) but from some other source (through hearing, or through the reading of Scripture), nevertheless infused faith primarily inclines [the intellect] to the articles to be believed, as to its objects, such that faith relates to the first Truth, about which it is (as its first object).99

(p.50) Habits are propensities for certain kinds of behaviour—here for certain kinds of belief. The idea in [1] is that infused faith is one such propensity, given directly by God, inclining the believer to accept the articles of faith. [2] draws an analogy: there is a habit, called scientia, which is a propensity to believe necessary truths and a facility in drawing inferences from them. [3] Cognitive habits such as these are propensities to believe, but they are not themselves representational. There are representational cognitive items, known as intelligible species—the stored cognitive content that is somehow actualized in memory and occurrent cognition—but cognitive habits are not such things. [4] continues the analogy from [2] (though in a rather obscure and pointless way). Suppose we possessed the habit of scientia and had some access to the objects of scientia without our having the species that bears the relevant cognitive content stored in the intellect. We would be in a more optimal cognitive position in relation to these objects than we would be to those which we had never encountered—we would be able ‘more easily to understand the objects and infer conclusions from them’ (since this, after all, is what a habit of scientia makes us fit for). [5] The habit of infused faith is like this: it is not content-bearing (since no habit is content-bearing). But suppose we had access to the objects of faith in some other way (e.g. from preaching or reading the Bible). The habit of faith would then incline us to believe the contents of the preaching or reading, and to grasp that they are revealed by God.

Unlike the first view, Scotus does not unequivocally reject this one in the Lectura—indeed, in an off-hand comment he speaks as though it is true.100 But he offers a sequence of objections to which he does not reply. So the discussion can hardly be called decisive. The basic objection is that there is no case in which someone who is simply presented with particular theological doctrines automatically assents to them—but such assent ought to be the result if the habit of infused faith genuinely does the work that is required of it on this theory. What is required too is the additional teaching that a certain epistemic obligation attaches to believing the articles—that they are ‘things to be believed’. But—supposing that the source of this additional information cannot be divine testimony (on the grounds set out in the refutation of the first view on infused faith)—the source of this information can only be human testimony: thus making infused faith dependent on acquired faith.101 In addition, Scotus worries that infused faith on this view would somehow by-pass the will. If someone presented with Church teachings automatically believed them, belief would indeed by-pass the will.102

In this earlier discussion, Scotus makes it clear that in any case he does not see how such a propensity to believe could reliably result in the relevant output. There could not, in other words, be such a thing as a hyper-reliable habit of the kind proposed by Aquinas. Scotus suggests that a habit of a given kind acts ‘naturally’: that is to say, a (p.51) given input invariably results in a given output. But the output varies with the input: nothing about the habit itself can guarantee that it yields the right output. Scotus’s example is a disposition to believe ‘first principles’: given a false input (e.g. that such-and-such a proposition express a first principle, supposing that the proposition express nothing of the kind), the habit might well produce a false output. Likewise in the case of Christian faith: a habit of faith is such that a true input (e.g. Catholic teaching) will produce a true output, while a false input (e.g. heretical teaching) will produce a false output. And, Scotus reasons, it makes no difference to the reliability of the habit whether or not the habit is produced naturally or supernaturally. Infused faith, as much as acquired faith, could thus yield a false output.103

Aquinas’s hyper-reliable habit is something that is responsive only to one type on input—namely, a true one that has as its content of the Christian faith. In the Lectura, Scotus does not seem to conceive of habits in this way. They are general dispositions to believe certain more broadly construed types of content, and seem to be distinguished in terms of the epistemic attitudes attaching to the relevant outputs (e.g. a habit to assent to things that are the objects of scientia, a habit to assent to things that are the objects of fides, and so on: hence, I think, the example of the first principles, the objects of intelligentia (on intelligentia, see n. 3, this chapter)). But it seems odd to think that not even God could make a habit of the kind construed by Aquinas. And a couple of years later, in the Quodlibet, Scotus concedes that infused faith could be such a hyper-reliable habit:

Infused faith cannot incline to anything false; moreover it inclines by virtue of the divine light, in which it is a participation, and thus [it inclines] to nothing other than what is conformed to that divine light. Therefore the act of believing, to the extent that it depends on this faith, cannot tend to anything false.104

Still, in this text, Scotus remains fully convinced of the subjective certainty of acquired faith, and that such certainty is reasonable.105 He also maintains that the presence of infused faith is not something internally accessible—and hence adds no subjective certainty to acquired faith.106 Acquired faith is objectively certain to the extent that it yields the same output as infused faith; and it is necessarily so to the extent that infused faith and acquired faith are joint causes of the same output.107

How might the earlier Scotus—the one who has no place for infused faith—account for revelation? Quite easily, for Scotus does not believe that divine revelation is anything like infused faith. Rejecting infused faith, or rejecting some kind of infallibilism, does not require rejecting divine revelation. The reception of divine revelation, according to Scotus, involves direct vision of the truths being revealed: it (p.52) is not simply the reception of certain propositions, but an encounter with the realities that the relevant propositions—the articles of faith, for example—are about. People with these direct encounters simply ‘cannot not assent’ to the truths revealed.108 The crucial thing for us is to ensure that we accept the (human) testimony of the recipients of revelation: and Scotus’s arguments are designed to show that we have ways of doing so rationally.109


Bibliography references:

Aquinas, Thomas. Summa Theologiae. http://www.corpusthomisticum.org/iopera.html.

Augustine. 1889–90. De civitate dei XXII, c. 5. Ed. Emanuel Hoffmann, 2 vols, Corpus Scriptorum Ecclesiasticorum Latinorum, 40/1–2. Prague and Vienna: F. Tempsky; Leipzig: G. Freytag, 40/2, 590, ll. 14–18.

Barat, Judith. 1992. Consent: The Means to an Active Faith According to Thomas Aquinas. American University Studies, V/126. New York: Peter Lang.

Brown, Jerome V. 1984. “Duns Scotus on the Possibility of Knowing Genuine Truth: The Reply to Henry of Ghent in the ‘Lectura prima’ and the ‘Ordinatio’.” Recherches de théologie ancienne et médiévale, 51: 136–82.

Cross, Richard. 2014. Duns Scotus’s Theory of Cognition. Oxford: Oxford University Press.

Dougherty, M. V. 2005. “Aquinas on the Self-Evidence of the Articles of Faith.” Heythrop Journal, 46: 167–80.

Duns Scotus. 1639. Opera omnia, ed. L. Wadding. 12 vols. Lyon.

Duns Scotus. 1950–2013. Opera omnia, ed. C. Balić and others. 21 vols. Vatican City: Vatican Polyglot Press.

Duns Scotus. 1997–2006. Opera philosophica, ed. G. Etzkorn et al. 5 vols. St Bonaventure, NY: Franciscan Institute.

Faucher, Nicolas. 2015. “Prêter foi avec parcimonie. Le traitement scotiste de la foi acquise et de la foi infuse.” In C. Grellard, P. Hoffmann, and L. Lavaud, eds, Conviction, croyance, foi. Pistis et Fides de l’Antiquité tardive au Moyen Age. Turnhout: Brepols.

Hawthorne, John. 2013. “Aquinas on Faith and Knowledge.” In John Marenbon, ed., Continuity and Innovation in Medieval and Modern Philosophy: Knowledge, Mind, and Language (British Academy Dawes Hicks Symposium on Philosophy, 2011), 119–33. Proceedings of the British Academy, 189. Oxford: Oxford University Press.

Lamont, John R. T. 2004. Divine Faith. Ashgate New Critical Thinking in Religion, Theology and Biblical Studies. Aldershot and Burlington, VT: Ashgate.

Macdonald, Scott. 1993. “Theory of Knowledge.” In Norman Kretzmann and Eleonore Stump, eds, The Cambridge Companion to Aquinas, 160–95. Cambridge: Cambridge University Press.

Pasnau, Robert C. 1999. A Commentary on Aristotle’s De anima. Yale Library of Medieval Philosophy. New Haven, CT: Yale University Press.

(p.53) Pasnau, Robert C. 2003. “Cognition.” In Thomas Williams, ed., The Cambridge Companion to Duns Scotus, 285–311. Cambridge: Cambridge University Press.

Perler, Dominik. 2006. Zweifel und Gewissheit: Skeptische Debatten im Mittelalter. Frankfurt am Main: Klostermann.

Pich, Roberto Hofmeister. 2012. “Duns Scotus on the Credibility of Christian Doctrines.” In Alessandro Musco, ed., Universalità della ragione: Pluralità delle filosofie nel medioevo, 12. Congresso Internazionale di Filosofia Medievale, Palermo, 17–22 settembre 2007, 3 vols. Palermo: Officina di Studi Medievali, II/2.

Pickavé, Martin. 2010. “Henry of Ghent and John Duns Scotus on Skepticism and the Possibility of Naturally Acquired Knowledge.” In Henrik Lagerlund, ed., Rethinking the History of Skepticism: the Missing Medieval Background, 61–96. Studien und Texte zur Geistesgeschichte des Mittelalters 103. Leiden: Brill.

Pickavé, Martin. 2012. “Human Knowledge.” In Brian Davies and Eleonore Stump, eds, The Oxford Handbook of Aquinas, 311–26. Oxford: Oxford University Press.

Poppi, Antonino. 2014. “Problemi della fede nelle lezioni di Giovanni Duns Scoto (Lectura III, Reportationes Parisienses III).” In Marco Forlivesi, Riccardo Quinto, and Silvana Vecchio, eds, Fides Virtus: The Virtue of Faith from the Twelfth to the Early Sixteenth Century, 261–314. Edited by Archa Verbi—Subsidia 12. Münster: Aschendorff.

Serene, Eileen. 1982. “Demonstrative Science.” In Norman Kretzmann, Anthony Kenny, and Jan Pinborg, eds, The Cambridge History of Later Medieval Philosophy, 496–517. Cambridge: Cambridge University Press.

Siebert, Matthew Kent. 2010. “Aquinas on Testimonial Justification: Faith and Opinion.” The Review of Metaphysics 69: 555–82.

Staudinger, Siegfried. 2006. Das Problem der Analyse des Glaubensaktes bei Johannes Duns Scotus. Veröffentlichungen der Johannes-Duns-Skotus-Akademie für franziskanische Geistesgeschichte und Spiritualität Mönchengladbach 22. Kevelaer: Butzon und Bercker.

Stump, Eleonore. 2003. Aquinas. Arguments of the Philosophers series. London and New York: Routledge.

Vos, Antonie. 1998. “Knowledge, Certainty and Contingency.” In E. P. Bos, ed., John Duns Scotus, Renewal of Philosophy. Acts of the Third Symposium Organized by the Dutch Society for Medieval Philosophy Medium Aevum (May 23 and 24, 1996), 75–88. Elementa: Schriften zur Philosophie und ihrer Problemgeschichte 72. Amsterdam and Atlanta: Rodopi.

Walter, Ludwig. 1968. Das Glaubensverständnis bei Johannes Duns Scotus, Veröffentlichungen des Grabmann Instituts (Neue Folge), 5. Munich: Schöningh.


(1) For this paragraph and the next, see Aquinas, Summa theologiae [= ST] II-II, q. 2, a. 1 c unless otherwise noted. In translating ‘opinio’ as ‘belief’ (rather than the more normal but misleading ‘opinion’), I follow an excellent practice that I think was explicitly initiated by Pickavé (2012, 317). For an account of Aquinas on faith that is not hugely distant from what I offer here, see Stump (2003, 361–88). See also M. V. Dougherty (2005), Barat (1992), and Lamont (2004, 53–73 [on Aquinas], 83–9 [on Scotus]). For a good general overview of the history of some of the philosophical issues involved, see Serene (1982).

(2) The terminology is a little fluid: in non-theological contexts, Aquinas treats faith as synonymous with belief (see Aquinas, Expositio libri posteriorum [= In post. an.] lib. 1, lect. 1, n. 6, a passage I quote below); in theological contexts he makes a distinction, and treats ‘faith’ as a term of art for the beliefs of Catholic Christians. Unless I state otherwise, I restrict the term ‘faith’ in this way in what follows.

(3) Strictly speaking, Aquinas would say that first principles—the objects of intellectus or intelligentia—compel assent immediately; syllogistic conclusions from these principles—the objects of scientia strictly speaking—compel assent only on the basis of the first principles and syllogistic validity: see e.g. In post. an., lib. 1, lect. 7, n. 6. My impression is that Aquinas is using ‘scientia’ rather loosely in the passages under discussion here, to cover all of these cases. I assume that metaphysical necessity is that the opposite of which includes a contradiction, and that logical necessity is that subset of metaphysical necessity restricted to that the opposite of which can be known a priori to include a contradiction.

(4) Aquinas, ST II-II, q. 2, a. 9 ad 2.

(5) Aquinas, De veritate, q. 10, a. 8 ad 8 in opp. See too Aquinas, Sententia libri ethicorum, lib. 3, lect. 3, n. 12: ‘It is not possible to be in ignorance as to who it is who acts, because in this way one would be in ignorance about oneself, which is impossible.’

(6) See Aquinas, ST II-II, q. 1, a. 4 c. In general, Aquinas holds that any conscious human agency is to be ascribed to the will as the executive power. In the case of assent to knowledge, the intellect turns out to be passive because coerced. But in other cases—in which some input from the agent is required—it follows straightforwardly from Aquinas’s action theory that the input must come from the will. Contingency, if it arises at all, is located not in the will but in the intellect: ‘Reason has a power to opposites in relation to contingent things, as is clear in dialectical syllogisms and rhetorical persuasions. But particular operations are contingent, and therefore the judgment of reason is related to diverse things and is not determined to one’: ST I, q. 83, a. 1 c. I suspect that Aquinas does not mean to suggest here that there is a contingent relation between this given informational input and this given epistemic output. Rather, what he means is that an intellect in different circumstances could reasonably respond in different ways to the same informational input. For example, it might be reasonable to respond positively to the testimony of someone reputable, but negatively to the same testimony offered by someone dishonest. But I am not sure about this. At any rate, I do not think Aquinas need be committed to this as a universal rule governing all cases of belief—since, after all, the only general claim he makes has to do with the causal role of the will, not with the contingency of reasons. Note that this interpretation of the role of the will in belief is rather different from the one proposed by Stump (2003, 363–4), which seems insufficiently general.

(7) Aquinas, ST II-II, q. 4, a. 8 c.

(8) For Aquinas on testimony, see Siebert (2010).

(9) Aquinas, ST II-II, q. 70, a. 2 c.

(10) Aquinas, ST II-II, q. 70, a. 2 c.

(11) Aquinas, ST II-II, q. 70, a. 3, c. As a point of terminology: Aquinas does not usually talk about reasonable (rationabilis) belief, but rather about probable (probabilis) belief. (Later, Scotus is happy to use the term ‘reasonable’, as well as ‘probable’, about beliefs.)

(12) See e.g. Aquinas, De anima, q. 13 c; Aquinas, Sententia libri de anima [= In de an.], lib. 2, lect. 13 nn. 384–5.

(13) Aquinas, In de an., lib. 3, lect. 6, n. 661 (Pasnau 1999, 338).

(14) Aquinas, In de an., lib. 3, lect. 6, n. 646 (Pasnau 1999, 331); see too In de an., lib. 2, lect. 13, n. 384.

(15) Aquinas, In de an., lib. 3, lect. 6, n. 664 (Pasnau 1999, 339).

(16) For discussion of this, see Cross (2014, 33). (For convenience, I ignore here the question of the so-called species in medio.)

(17) Aquinas, In de an., lib. 2, lect. 13, n. 387. In what follows, I ignore the so-called ‘common sense objects’ (see Aquinas, In de an., lib. 2, lect. 13, n. 386), partly for reasons of space, and partly because what Aquinas says does not add anything much of philosophical interest to what he says about the proper and per accidens objects of sense.

(18) Aquinas, In de an., lib. 2, lect. 13, nn. 396–8 (Pasnau 1999, 208). The contrast with non-human animals is instructive: they sense an individual ‘not in terms of its being under a common nature, but only in terms of its being the end point or starting point of some action or affection’: Aquinas, In de an., lib. 2, lect. 13, n. 398 (Pasnau 1999, 208–9).

(19) The reasons for this terminological slippage have to do with the context of the discussion, as part of a commentary on Aristotle’s De anima. Aristotle does not have the cogitative power—this is something derived by Aquinas from Avicenna—and the imagination (phantasia) in Aquinas’s reading of Aristotle here does something like the work that the cogitative power does in Aquinas and Avicenna.

(20) Aquinas, In de an., lib. 3, lect. 6, n. 662 (Pasnau 1999, 338).

(21) Aquinas, In de an., lib. 2, lect. 13, n. 385 (Pasnau 1999, 204). Pasnau helpfully comments: ‘The point here may be that the senses are liable to deception when they make judgments about per accidens … sense objects. Or Aquinas may be saying something rather different: that the senses will be deceived if they attempt, without the aid of the higher sensory and intellective powers, to make judgments that go beyond color and sound—judgments, for instance, about what and where the object is’ (Pasnau 1999, 204–5, n. 1).

(22) Aquinas, In de an., lib. 3, lect. 6, n. 664 (Pasnau 1999, 338–9, slightly altered).

(23) Aquinas, In de an., lib. 3, lect. 6, n. 664 (Pasnau 1999, 339, slightly altered).

(24) Aquinas, In de an., lib. 3, lect. 6, n. 664 (Pasnau 1999, 339).

(25) Aquinas, In de an., lib. 3, lect. 6, n. 665 (Pasnau 1999, 339).

(26) Aquinas, In de an., lib. 3, lect. 6, n. 664 (Pasnau 1999, 339).

(27) Aquinas, ST I, q. 86, a. 1 c.

(28) See Aquin as, ST I, q. 85, a. 1.

(29) See Aquinas, ST I, q. 85, a. 6 c.

(30) See Aquinas, ST I, q. 85, a. 3.

(31) Aquinas, ST I, q. 2, a. 2.

(32) Macdonald (1993, 175).

(33) Aquinas, In post. an. lib. 1, lect. 1, n. 6.

(34) Aquinas, ST I, q. 85, a. 6 c.

(35) Aquinas, ST II-II, q. 2, a. 1 c.

(36) Aquinas, ST II-II, q. 2, a. 10 ad 2.

(37) Aquinas, ST II-II, q. 4, a. 8 c.

(38) Aquinas, ST II-II, q. 4, a. 1 c.

(39) Hawthorne (2013, 124), commenting on ST II-II, q. 5, a 3 c.

(40) Aquinas, ST II-II, q. 2, a. 1 ad 3.

(41) See Aquinas, ST II-II, q. 6, a. 1 c.

(42) Aquinas, ST II-II, q. 4, a. 4 c.

(43) Aquinas, ST I-II, q. 111, a. 2 c.

(44) Aquinas, ST II-II, q. 1, a. 1 c.

(45) Lamont (2004, 67).

(46) Lamont (2004, 68).

(47) Aquinas, ST II-II, q. 1, a. 4 ad 3.

(48) Aquinas, ST II-II, q. 6, a. 1 c.

(49) Aquinas, ST II-II, q. 5, a. 3 c.

(50) Pasnau (2003, 301).

(51) Scotus, Lectura III, d. 23, q. un., n. 2 (Vatican, XXI, 97).

(52) In addition to the discussion in Pasnau (2003), see e.g. Brown (1984); Vos (1998); Perler (2006, 33–115); and Pickavé (2010).

(53) Scotus, Ordinatio [= Ord.] I, d. 3, p. 1, q. 4, n. 230 (Duns Scotus [1950–2013], Opera omnia, Vatican, III, 139).

(54) Scotus, Ord. I, d. 3, p. 1, q. 4, n. 233 (Vatican, III, 140).

(55) Scotus, Ord. I, d. 3, p. 1, q. 4, n. 239 (Vatican, III, 145). See too Scotus, Quaestiones super libros metaphysicorum VII, q. 13, n. 158 (Duns Scotus [1997–2006], Opera philosophica, IV, 271).

(56) Scotus, Ord. I, d. 3, p. 1, q. 4, nn. 238–9 (Vatican, III, 144–5).

(57) Scotus, Ord. I, d. 3, p. 1, q. 4, n. 235 (Vatican, III, 142), roughly repeated at n. 241 (Vatican, III, 146).

(58) Scotus, Ord. I, d. 3, p. 1, q. 4, n. 242 (Vatican, III, 147).

(59) Scotus, Ord. I, d. 3, p. 1, q. 4, n. 242 (Vatican, III, 147).

(60) Scotus, Ord. I, d. 3, p. 1, q, 4, n. 243 (Vatican, III, 147–8).

(61) Scotus, Ord. I, d. 3, p. 1, q. 4, n. 242 (Vatican, III, 147).

(62) Scotus, Ord. I, d. 3, p. 1, q. 4, n. 244 (Vatican, III, 148).

(63) See Scotus, Ord. I, d. 3, p. 1, q. 4, n. 239 (Vatican, III, 145).

(64) Scotus, Ord. I, d. 3, p. 1, q. 4, n. 235 (Vatican, III, 142).

(65) Scotus, Ord. I, d. 3, p. 1, q. 4, n. 236 (Vatican, III, 143).

(66) See in particular Scotus, Ord. I, d. 3, p. 3, q. 1, nn. 348–66 (Vatican, III, 209–22).

(67) For the (very dense) discussion, see Scotus, Ord. I, d. 3, p. 1, qq. 1–2, nn. 71–91 (Vatican, III, 49–60). For Aquinas’s account of the genesis of specific universal concepts, see e.g. Aquinas, ST I, q. 85, a. 3.

(68) Scotus, Lectura [= Lect.] III, d. 23, q. un., n. 19 Vatican, XXI, 103–4).

(69) Scotus, Quodlibetum [= Quod.], q. 14, n. 5 (Opera omnia, Lyon: 1639. Vol. XI, 356).

(70) Scotus, Lect. III, d. 23, q. un., n. 15 (Vatican, XXI, 102).

(71) Scotus, Lect. III, d. 23, q. un., n. 4 (Vatican, XXI, 97). This list occurs in the context of an objection; but Scotus never rejects the view that we ‘firmly adhere’ to these things ‘by acquired faith’.

(72) For the material in this paragraph, see Scotus, Lect. III, d. 25, q. un., n. 45 (Vatican, XXI, 173–4). I am grateful to Nicolas Faucher for drawing this passage to my attention.

(73) For a general discussion of Scotus on faith, see Faucher (2015); Walter (1968); Staudinger (2006); and Poppi (2014, 261–314).

(74) Scotus, Lect. III, d. 23, q. un., n. 19 (Vatican, XXI, 103).

(75) Scotus, Lect. III, d. 23, q. un., n. 14 (Vatican, XXI, 100–1).

(76) Scotus, Lect. III, d. 23, q. un., n. 17 (Vatican, XXI, 102–3).

(77) Scotus, Lect. III, d. 23, q. un., n. 15 (Vatican, XXI, 101–2).

(78) On Aquinas’s claim that the Bible (and Church teachings, which simply follow from biblical teaching according to Aquinas) is literally God’s speaking, see Lamont (2004, 13–27).

(79) Scotus, Lect. III, d. 23, q. un., n. 16 (Vatican, XXI, 102).

(80) Scotus, Ord. prol., p. 2, q. un., n. 100 (Vatican, I, 61).

(81) For discussion of all of these arguments, see Pich (2012), II/2, 997–1011.

(82) Scotus, Ord. prol., p. 2, q. un., n. 101 (Vatican, I, 61–2).

(83) Scotus, Ord. prol., p. 2, q. un., n. 102 (Vatican, I, 62–3).

(84) Scotus later claim that it would be impossible to be mistaken about having seen the divine essence (as supposedly happened in the conversion of St Paul): see Scotus, Ord. prol., p. 2, q. un., n. 116 (Vatican, I, 81).

(85) Scotus, Ord. prol., p. 2, q. un., n. 105 (Vatican, I, 65–6).

(86) Scotus, Ord. prol., p. 2, q. un., n. 107 (Vatican, I, 68–9).

(87) Scotus, Ord. prol., p. 2, q. un., n. 108 (Vatican, I, 70–1).

(88) For the non-Christian belief systems, see Scotus, Ord. prol, p. 2, q. un., n. 109 (Vatican, I, 71–3); for the vast proof for God’s existence, see Ord. I, d. 2, p. 1, qq. 1–2, nn. 39–147 (Vatican, II, 148–215).

(89) Scotus, Ord. prol., p. 2, q. un., n. 112 (Vatican, I, 76–7).

(90) Scotus, Ord. prol., p. 2, q. un., n. 113 (Vatican, I, 77–8). Scotus devotes some space to a rather anti-Humean argument—that he has got from Augustine—to the effect that the spread of Christianity without miracles would have been more incredible than the fact of miracles themselves: see Scotus, Ord. prol., p. 2, q. un., n. 114 (Vatican, I, 78–9); see Augustine, De civ. dei XXII, c. 5 (CSEL, 40/2, 590, ll. 14–18).

(91) Scotus, Ord. prol., n. 119 (Vatican, I, 84–5).

(92) Scotus, Lect. III, d. 23, q. un., n. 22 (Vatican, XXI, 104–5).

(93) Scotus, Lect. III, d. 23, q. un., nn. 24–5 (Vatican, XXI, 105–6).

(94) Lamont (2004, 84).

(95) Scotus, Lect. III, d. 23, q. un., n. 26 (Vatican, XXI, 106).

(96) Scotus, Lect. III, d. 23, q. un., n. 33 (Vatican, XXI, 108–9).

(97) Scotus, Lect. III, d. 23, q. un., n. 34 (Vatican, XXI, 109).

(98) Scotus, Lect. III, d. 23, q. un., n. 35 (Vatican, XXI, 110).

(99) Scotus, Lect. III, d. 23, q. un., n. 38 (Vatican, XII, 111–12).

(100) See Scotus, Lect. III, d. 23, q. un., n. 58 (Vatican, XXI, 119). But note that other off-hand comments seem to treat the decisively rejected first view as true: see Scotus, Lect. III, d. 24, q. un., n. 53 (Vatican, XXI, 146); Ord. IV, d. 14, n. 179 (Vatican, XIII, 45–6).

(101) Scotus, Lect. III, d. 23, q. un., n. 45 (Vatican, XXI, 114).

(102) Scotus, Lect. III, d. 23, q. un., n. 46 (Vatican, XXI, 114).

(103) For the material in this paragraph, see Scotus, Lect. III. d, 23, q. un., n. 54 (Vatican, XXI, 118).

(104) Scotus, Quod., q. 14, n. 7 (Wadding, XII, 354).

(105) See Scotus, Quod., q. 14, n. 5 (Wadding, XII, 352).

(106) See Scotus, Quod., q. 14, n. 8 (Wadding, XII, 355).

(107) See Scotus, Quod., q. 14, n. 7 (Wadding, XII, 354).

(108) See Scotus, Lect. III, d. 24, q. un., n. 63 (Vatican, XXI, 148–9); see too Ord. prol., p. 1, q. un., nn. 62–5 (Vatican, I, 38–40).

(109) Thanks to Charity Anderson, Caleb Cahoe, Therese Scarpelli Cory, Keith DeRose, Billy Dunaway, Nicolas Faucher, Alex Hall, John Hawthorne, Tim O’Connor, Blake Roeber, Jeff Russell, Lydia Schumacher, Richard Swinburne, and Christina van Dyke for discussion of some of the issues that arise in this chapter.