Thomas Aquinas thinks that a person can believe the articles of the Christian Creed without violating any epistemic duties, without suffering from cognitive malfunctioning of any kind, and without behaving in a way that reduces the objective likelihood of arriving at true beliefs.¹ Contemporary epistemologists would express this by saying that Christian beliefs, ² according to Aquinas, can be epistemically justified or warranted.³ Using the notion of justification in a broad sense that covers warrant, I will in this article present what I take to be Aquinas’s explanation of why Christian beliefs are justified. I will also argue that his explanation is cogent and defensible in the contemporary epistemological arena. Since the picture of faith that Aquinas presents is somewhat complex, however, a number of different interpretations of the epistemological principles behind it have emerged. After having discussed three common readings, I will suggest a fourth interpretation that I take to be preferable. Given what I call the “Testimonial Interpretation,” Aquinas’s account of faith provides a viable explanation of how Christian beliefs are justified.⁴

However, before I can discuss this explanation and compare the Testimonial Interpretation to alternative construals, it is necessary to give a brief overview of what Aquinas explicitly says about divine faith. This overview will highlight four features of faith that any plausible interpretation must account for.

Faith, for Aquinas, is one of the three theological virtues, and as such a stable disposition (habitus), given to humans by grace, that perfects us and helps us attain our supernatural goal.⁵ All virtues are by their nature ordered towards some act, and the primary act of faith, according to Aquinas, is to believe, or to “think with assent” (ST II-II, q. 2, a. 1; De verit., q. 14, a. 1).⁶ However, faith also requires an external act, which is public confession. The acts of faith, like all acts, receive their identity from their object – what they are directed towards or are about.

Aquinas makes a distinction between the material and the formal object of faith. The material object is God and all things in so far as they are related to God. However, we humans can only think about God and other things by grasping propositions about them, so when considered from the perspective of the believer, the material object of faith are propositions about God and other things in relation to God, primarily those propositions found in the Creed – the articles of faith – and secondarily other biblical propositions. The formal object of faith, on the other hand, is the aspect under which the relevant propositions are to be believed, namely under the aspect of being divinely revealed (ST II-II, q. 2, a. 2; see also ST I, q. 1, a. 3).⁷ Aquinas describes the formal object of faith as “nothing else than the First Truth [Prima Veritas]. For the faith of which we speak does not assent to anything except because it is revealed by God” (ST II-II q. 1, a. 1, co). God reveals the propositions of faith by speaking to us through human intermediaries, who have an instrumental role in the process of revelation.⁸ In Christ, however, God teaches us directly as the incarnate Word.⁹

Now, there is one thing that all propositions held by faith have in common, namely that their truth is not “seen” (visum) by the believer’s intellect.¹⁰ What Aquinas means by this is that a believer neither has a logical demonstration of the truth of the articles of faith, nor do their truth impress itself on his mind in an immediate, nondiscursive way. In other words, what is known by faith is by definition not apprehended through the intellectual virtue of scientia, which is the habit of demonstrative knowledge, or by the virtue of intellectus, which is the habit whereby first principles are immediately known.¹¹

It is important to note, however, that Aquinas’s denial that we have scientia or intellectus of the articles of faith does not entail that we lack knowledge of these articles (Ross, 1985). Although scientia is often translated as “knowledge,” the former is clearly a less inclusive notion than knowledge as we typically understand it today. For example, Aquinas would deny that I have scientia of the fact that Australia exists. Almost all contemporary philosophers, however, would agree that I can know that Australia exists, even though I have never been there. Since Aquinas – as we will see – ascribes a maximal degree of objective certainty to faith, it is clear that faith on his construal would count as a kind of knowledge given most modern conceptions of the latter category.¹²

Now, when we know a proposition by scientia or intellectus, our intellect cannot but assent to it. The intellect’s assent is forced by the evidence, so to speak. In the act of faith, however, it is the will that moves or causes the intellect to assent to what God has revealed. The will’s movement of the intellect, in turn, is motivated by a desire for eternal life and is caused by divine grace. As Aquinas writes: “The act of believing is an act of the intellect assenting to the Divine truth at the command of the will moved by the grace of God” (ST II-II, q. 2, a. 9; see also De verit., q 14, a. 1, co). This means that faith is dependent on grace in two respects: first, because faith as a virtue is infused, and second, because every act of faith is due to the will being moved by grace. Because the movement of grace elevates the will’s act to a supernatural level without undermining its voluntary nature, the act of faith is meritorious.

The peculiar properties of faith led Aquinas to describe it as “a mean between science and opinion,” a mean or midpoint that shares properties with both extremes (ST II-II, q. 1, a. 2 s.c; De verit., q. 14, a. 8). Opinio is ordinary human belief based on inconclusive evidence. Faith is like opinion in that both are of something “unseen” – something non-evident to the intellect, which means that both faith and opinion can only come about as a result of an act of the will. Faith is like scientia, on the other hand, in that it involves firm and unwavering assent to what is believed. Such firmness does not characterize opinion, which always involves some degree of uncertainty or doubt (ST II-II, q 2, a. 1, co; De verit., q. 14, a. 1, co).¹³

Although Aquinas, as we have seen, emphasizes the role of the will and of grace for faith’s assent, he also emphasizes that faith is reasonable, and he does so in two ways. First, he points out that faith is based on God’s own testimony, which makes it epistemically more certain than scientia. Aquinas writes: “Much more is a man certain about what he hears from God, Who cannot be deceived, than about what he sees with his own reason, which can be mistaken” (ST II-II, q. 4, a. 8, ad 2).¹⁴ Secondly, Aquinas emphasizes that there are good apologetic arguments for the claim that what we take to be divinely revealed is in fact divinely revealed. He writes:

For divine wisdom himself … by suitable arguments proves his presence, and the truth of his doctrine and inspiration, by performing works surpassing the capability of the whole of nature, namely, the wondrous healing of the sick, the raising of the dead to life, a marvelous control of the heavenly bodies, and, what excites yet more wonder, the inspiration of human minds (SCG I, c. 6).

Bringing together his appeal to divine authority and to outer and inner signs of credibility, Aquinas writes: “The believer has sufficient motive for believing, for he is moved by the authority of Divine teaching confirmed by miracles, and, what is more, by the inward instinct of the Divine invitation: hence he does not believe lightly” (ST II-II, q. 2, a. 9, ad 3). An interesting question that I will soon address is what role these different elements play for faith’s rational status.

One last thing that is important to mention in the present context is the role played by the gifts of the Holy Spirit in faith’s assent. There are seven gifts of the Spirit, and different gifts are appended to and assist different theological virtues, according to Aquinas. The two gifts that belong to the virtue of faith have the same names as the intellectual virtues intellectus (which is usually translated as Understanding) and scientia (translated as Knowledge). While virtues – both natural and theological – are “principles for self-moved actions,” the gifts are habits that make a person amenable to be moved directly by the Holy Spirit – a kind of movement that we need if we are to reach our eternal destination (Jenkins, 1997, p. 156). The gifts of intellectus and scientia mediate the promptings of the Holy Spirit with respect to faith, so that our grasp of and assent to the articles of the Creed become more perfect than our own reason – even when informed by the theological virtue of faith – is capable of ensuring. As Aquinas writes: “In matters directed to the supernatural end, to which man’s reason moves him, according as it is in a manner, and imperfectly, informed by the theological virtues, the motion of reason does not suffice, unless it receive in addition the prompting or motion of the Holy Spirit” (ST I-II, q. 68, a. 2).

After this birds-eye overview of Aquinas’s teaching on faith, it might be helpful to highlight four features of faith that are most relevant for our present purposes. First, Aquinas holds that we accept the articles of the Creed on account of the authority of God himself, that is, we believe them on the basis of divine testimony.¹⁵ Second, faith’s assent is caused by the will, which acts in view of the divine goodness. Third, the will’s assent is caused by divine grace, which moves the will without compromising its freedom.¹⁶ Fourth, there are good apologetic arguments in favor of the claim that the articles of faith have been revealed by God. Exactly how strong Aquinas takes these arguments to be is somewhat unclear.

Before tackling the question of how Aquinas thinks about the epistemic justification of faith, I must clarify what I mean by epistemic justification. An important, first distinction to make is between justification as an activity and as an epistemic status that a belief – or more precisely, a person’s act of believing a certain proposition – may have.¹⁷ Most contemporary philosophers agree that a person can be justified in believing a proposition without necessarily being able to justify this belief by, for example, giving reasons or producing evidence for it. In this article, I will exclusively be concerned with justification in the former sense, as a status that a person’s beliefs may or may not have. When it comes to the question of what exactly this status entails – what it means for a person to justifiedly believe something – a wide range of theories have been proposed by contemporary philosophers.¹⁸ Since the interpretations of Aquinas’s view that I will discuss appeal to principles taken from such theories, I will not preempt the discussion by declaring my sympathies for any particular theory already at this stage. Instead, I will initially work with a very broad notion of justification that covers what some philosophers refer to as “warrant.” Philosophers who distinguish warrant from justification typically understand justification as a function of whether the subject meets certain internalist requirements,¹⁹ such as respecting certain epistemic norms or fulfilling certain epistemic duties.²⁰ Warrant, on the other hand, has mainly to do with the satisfaction of externalist criteria, such as whether the cognitive processes or faculties that generate a person’s beliefs are objectively reliable, or operate in the way they are designed to operate. ²¹ If a belief is generated by a reliable or properly functioning epistemic process or faculty, it has the status of being warranted, according to this view. In this article, I will not distinguish between justification and warrant. Instead, I will take the notion of justification to be applicable in the context of externalist as well as internalist theories of how beliefs acquire positive epistemic status.²² I will also assume that only beliefs that are justified can constitute knowledge, leaving open – for now – the question of whether the relevant justification needs to be of an externalist or internalist kind, or involve both externalist and internalist elements.

Since the distinction between epistemological internalism and externalism is important in the present context, let me briefly explain how I view it. Very roughly speaking, proponents of internalist theories hold that only factors that are within the subject’s direct cognitive reach can justify beliefs. ²³ Internalists therefore typically claim that being justified in believing a proposition P requires evidence, reasons or grounds for P, where evidence, reasons or grounds are understood to include only items that are accessible to the subject by reflection alone. Externalists, on the other hand, deny that what justifies a belief must be within the subject’s direct cognitive reach. They allow that factors that are external to the subject’s perspective on the world can justify beliefs. Externalists can therefore hold that, for example, the fact that a certain process or method of belief formation is objectively reliable can confer justification on the beliefs that it produces. It does not matter whether the subject knows, or is in a position to know, that the relevant process is reliable. It should be noted that there are also theories of epistemic justification that combine internalist and externalist elements in various ways. A “mixed” theory of this kind can for example say that some essential justifiers need to be internal, while others can be external (Pappas, 2023, Section 4).

Although the contemporary terminology and categories I have just rehearsed have few explicit counterparts in Aquinas’s thought, they are arguably useful tools for unearthing the epistemological principles behind his teaching on faith. This view is shared by many contemporary interpreters, who have suggested ways of explicating Aquinas’s understanding of faith in modern epistemological terms. I will now discuss three such attempted explications, which I – following John Jenkins – will call, respectively, the Naturalist Interpretation, the Voluntarist Interpretation, and the Supernatural Externalist Interpretation. After having critiqued these interpretations, I will proceed to suggest a fourth way of construing Aquinas’s epistemology of faith that better accounts for the various elements in his understanding and also makes his view more defensible in the contemporary epistemological arena.

According to the Naturalist Interpretation (which could also be labeled the “Evidentialist Interpretation”) at least some persons assent to the articles of the Creed because, first, they accept a cluster of arguments from natural theology, second, they believe on the basis of human testimony and other evidence that miracles and other signs have occurred in biblical history and in the history of the Church. From these considerations, they infer that God has made a revelation in history, which is contained, in its essence, in the Creed. Since what God reveals cannot be false, belief in the Creed is justified (Jenkins, 1997, Chapter 6.2).

This interpretation, suggested by philosophers of religion such as John Hick and Terrence Penelhum, focuses on the last of the four features of Aquinas’s view of faith that I enumerated above, namely his claim that there are good arguments in favor of the faith (Hick, 1988, pp. 20–21; Penelhum, 1977; see also Pojman, 1986, pp. 32–40).²⁴ There are many problems with the Naturalist Interpretation, however. It is very unclear, for example, whether it is capable of accounting for the necessity of grace for the act of faith, given that the publicly available evidence is supposed to be sufficient for convincing an unprejudiced person beyond reasonable doubt (Jenkins, 1997, p. 165).²⁵ It is also unclear how the interpretation can account for the faith of those who are ignorant of the apologetic case for God’s existence and divine revelation. Perhaps the most serious problem, however, is that the Naturalist Interpretation cannot explain the epistemic certainty that Aquinas ascribes to faith (especially in De veritate). If the epistemic basis for faith is arguments from natural theology and the like, then it seems that faith should have no more certainty than is warranted by the human reasoning behind it. Aquinas, however, says that “faith is more certain than all understanding and scientia,” and he also argues that “human reason is very deficient in things concerning God” (De verit., q. 14, a. 1, ad 7; ST II-II, q. 2, a. 4, co). Hence, as an interpretation of Aquinas, it seems that the naturalist model fails. Although this failure seems to have escaped a number of non-specialist scholars in the wider field of philosophy of religion, the defects of the Naturalist Interpretation are for the most part clearly recognized by Aquinas-scholars. I will therefore not spend any more time on the problems that this interpretation faces.

The second model is the Voluntarist Interpretation. It is defended in different versions by James Ross and Eleanor Stump, who both emphasize the second feature of Aquinas’s teaching on faith – the will’s role in the assent of faith (Ross, 1985; 1986, pp. 228-229; Stump, 2003, Chapter 12).²⁶ Proponents of the Voluntarist Interpretation claim that Aquinas was an epistemological reliabilist. Reliabilism is an externalist theory of justification according to which a belief is justified if it is produced by an objectively reliable cognitive faculty or process. What it means for a faculty or process to be reliable can be characterized in various ways, but for present purposes it will suffice to say that a cognitive faculty or process is reliable if it produces beliefs with a sufficiently high truth-ratio (see Goldman and Beddor, 2021, Section 1.1).

Now, Aquinas’s view, as we have seen, is that the will moves the intellect to assent to the articles of faith because the will desires something good, namely eternal life with God. Thomas writes: “We are moved to believe what God says because we are promised eternal life as a reward if we believe. And this reward moves the will to assent to what is said, although the intellect is not moved by anything which it understands” (De verit., q. 14, a. 1, co). What proponents of the Voluntarist Interpretation claim is that this way of arriving at Christian beliefs might be (or in fact is) a reliable belief-forming process, designed by God to be conducive to true beliefs. If this is the case, and if reliabilism of the relevant sort is true, it follows that beliefs produced by the will-driven process that Aquinas seems to envision are justified and therefore count as knowledge.

One general problem with this interpretation is that the authority of God seems to play no role for the justification of faith’s assent, which goes against what Thomas holds. According to Aquinas, faith “does not assent to anything, except because it is revealed by God” (ST II-II, q. 1, a. 1, co). This suggests that belief in the articles of faith is justified by the fact that it is God who has revealed them, rather than by the fact (if it is a fact) that a certain will-driven process of belief-formation is reliable. A second problem is that proponents of the Voluntarist Interpretation have difficulties explaining how a will-driven process of the relevant kind can be reliable. James Ross argues that the “biases” of our rational appetite (the will) may be rooted in reality in such a way that will-induced assent in certain cases produces knowledge. The idea here is that certain goods that the will is attracted to “stand in appropriate relations to reality” (Ross, 1986). Unfortunately, Ross does not expand on what those relations are, which makes it difficult to judge whether the model he suggests is plausible. Since desire-driven processes of belief-formation are normally regarded as paradigmatically unreliable, Ross would need to explain what distinguishes the desire-driven process of faith from unsound ways of forming beliefs, such as wishful thinking in general.

Eleonore Stump attempts to improve on Ross’ account in this respect. She argues that Aquinas’s metaphysics of goodness provides a basis for distinguishing faith from mere wishful thinking. Aquinas holds that there is a necessary relationship between perfect goodness and perfect being, and he also holds that perfect being necessarily exists. If this is true, argues Stump, it follows that “an intellect that assents to the proposition that God exists on the basis of the will’s hungering for God’s perfect goodness will be reliably right, because of the connection between goodness and being” (Stump, 2003, p. 368). On the other hand, beliefs formed on the basis of a desire for less-than-perfect goodness will have no reliable tendency to be true, which explains why ordinary wishful thinking does not generate justified beliefs. Things that are less-than-perfectly good need not necessarily exist.

Even if we grant that Stump’s model can explain how a will-driven assent to God’s existence can be justified, it is hard to see how the metaphysical connection between goodness and being can account for the justification of other Christian beliefs – a question that Stump unfortunately does not expand on. For example, while there is a necessary connection between God’s perfect goodness and his existence, there is no necessary connection (or even a probabilistic one) between God’s perfect goodness and his decision to become incarnate. The perfectly good God could have decided never to become man. Stump’s model therefore does not tell us how belief in the incarnation and other contingently true doctrines could be reliably formed on the basis of a desire for God’s perfect goodness (see Jenkins, 1997, pp. 184–185).

Perhaps it could be argued that God’s goodness entails his truthfulness, and that a person who is inclined to believe what God reveals (such as the doctrine of the incarnation) on the basis of a desire for perfect goodness will be reliably right in their beliefs because God always speaks the truth. The crucial issue here, however, is not whether God always speaks the truth but whether what purports to be a divinely revealed message really is such. With respect to this issue, a desire for perfect goodness seems to be of little or no relevance. Jews desire God’s perfect goodness no less than Christians, but the former do not accept the doctrine of the incarnation as divinely revealed. It is hard to argue, moreover, that they merely in virtue of this fact are irrational or display some other cognitive defect.

At this point in the dialectic, the only option that remains for proponents of the Voluntarist model is to argue that the will-driven process of belief-formation that underlies Christian faith is reliable simply in virtue of God’s supernatural influence on our wills through grace. Even if there is no natural (necessary or probabilistic) relationship between peoples’ desire for perfect goodness and the truth of many Christian doctrines, it could still be the case that God is guiding the desires of certain people in such a way that their wills elicit assent to Christian (and only Christian) doctrines. The distinguishing mark between faith and ordinary wishful thinking, on this view, is the mere fact that God supernaturally guarantees the reliability of the will-driven cognitive process we call faith. Before I discuss this response, it will be useful first to look at a third interpretation of Aquinas that focuses precisely on the epistemic role of God’s supernatural guidance “behind the scenes.”

John Jenkins coined the term “the Supernatural Externalist Interpretation” as a name for his own proposal, but there are other models sufficiently similar to Jenkins’s to be included under the same label (Jenkins, 1997, Chapter 6.4). ²⁷ Although my discussion will focus on Jenkins’ specific proposal as an example, my critique of it will also be relevant for other versions of the same “ideal-type.”²⁸ Jenkins’s model builds on the idea that the infused, theological virtue of faith – together with the related gifts of the Holy Spirit – play a similar role in our assent to the articles of the Creed as the natural, intellectual virtue of intellectus plays for our knowledge of the first principles of human scientiae. Aquinas writes: “The infused light, which is the habit of faith, makes manifest the articles [of faith], just as the intellectual light makes manifest principles naturally known” (Super Sent. III, d. 23, q. 2, a. 1, ad 4).²⁹

Now, the supernatural light of faith does not make the truth of the articles of faith evident to the believer. However, through this light and the gift of intellectus, the believer immediately and non-discursively comes to understand that the articles are divinely revealed and are to be believed (Jenkins, 1997, p. 190). In a second operation of the light of faith and the gift of scientia, the believer spontaneously assents to the articles as true. Since their truth is not evident to the believer’s intellect, however, her assent to them is voluntary and the act of faith meritorious (Jenkins, 1997, p. 196).

What is it, then, that justifies a person’s belief in the articles of faith, on this model? Jenkins claims that Aquinas was an epistemological externalist who takes a belief to be justified if it is produced by a properly functioning cognitive faculty. More precisely, Jenkins holds that a person S is justified³⁰ in believing that p to a degree sufficient for knowledge of p if the following two conditions are satisfied: 1)

S’s cognitive faculties were well designed by God to arrive at the truth with regard to the matter in question in the environment in which S finds himself; and

Now, a person’s belief in the articles of faith is the result of the supernatural light of faith rather than the natural operation of our cognitive faculties. However, the role that the light of faith plays in this respect is part of a divine design-plan, and if a person’s belief in the articles is produced in accordance with that design-plan, she will be justified in believing the articles. This theory is an externalist one, which means that it does not impose any requirement that the believer must know that her cognitive faculties work in accordance with a divine design plan. What is required by the theory is merely that this is actually the case.

Jenkin’s model manages to do justice to many aspects of Aquinas’s teaching on faith, and in its general outlines it represents in my view an improvement over the preceding two interpretations.³² One problem that it shares with the Voluntarist Interpretation, however, is that it underplays the role that God’s authority seems to have for the justification of Christian beliefs, according to Aquinas. For Jenkins, it is the proper functioning of a certain supernaturally charged cognitive process that gives Christian beliefs their status as justified (or warranted), not the fact that those beliefs are held on the basis of God’s own testimony. For Aquinas, on the other hand, it is the latter that seems to be the main justifying factor. Believers are moved to believe by the authority of Divine teaching (ST II-II, q. 2, a. 9, ad 3).

I will shortly suggest an interpretation of Aquinas that accounts for the centrality of precisely this dimension of his thinking. First, however, I will point out another weakness that is shared by the Voluntarist and the Supernatural Externalist interpretations alike. This weakness consists in the fact that both models explain the justified status of Christian beliefs by reference to the alleged fact that God secretly guides the belief-forming processes of Christians in a unique way, seemingly tailor-made for ensuring that Christian beliefs are justified. The Voluntarist Interpretation, for example, needs to assume (as I have argued) that a special divine guidance of the volitional responses of Christians makes their belief-forming process reliable, and this is why their beliefs are justified. The Supernatural Externalist Interpretation, likewise, appeals to the proper functioning of a special epistemic mechanism that only Christians possess. While it of course could be true that Christian beliefs owe their epistemic status to the workings of a unique and undetectable epistemic mechanism of one kind or another, it would arguably be more fitting if God were to convey his message by making use of an epistemic mechanism that is publicly available and has natural, non-Christian counterparts. This consideration acquires additional weight against the background of a principle that Aquinas frequently refers to, namely that “divine providence provides for each thing according to the mode of its nature” (SCG III, cap. 150).³³

Here is a suggestion about how we could understand the justification of Christian beliefs in a way that perfectly accords with this principle. Suppose that Christian beliefs are justified simply because they are based on supremely reliable testimony (God’s own). In that case, those beliefs would owe their epistemic status to an epistemic mechanism or source of justification that all human beings are heavily dependent on in various contexts (testimony). The difference between Christians and non-Christians in this scenario is that the latter fail to avail themselves of the specific source of testimonial knowledge that Christians draw on. This difference could be explained in terms of the causality of grace: Christians are caused to assent to God’s testimony by grace, and by assenting to it, they get access to the knowledge that it makes available. The role of grace, however, is on this construal mainly causal rather than epistemic. What justifies the assent that Christians give to God’s testimony is not (primarily³⁴) the fact that this assent is caused by grace in some special way (although it is caused by grace). What justifies the assent, instead, is the fact that the testimony assented to is supremely reliable and makes knowledge available.

In what follows, I will suggest an interpretation of Aquinas along these lines. Since this interpretation is more “economical” when it comes to positing uniquely Christian epistemic mechanisms, and since it is also capable of accounting for everything that Aquinas explicitly says about the nature and workings of divine faith, it is in my view preferable to previous interpretations.

This model presupposes a philosophical view of testimonial knowledge called anti-reductionism. ³⁵ Testimonial knowledge is knowledge acquired from the words of other people or, as the case may be, from the words of God mediated through human spokespersons such as the authors of Scripture. Almost all philosophers agree that testimonial beliefs can be justified and constitute knowledge, but there are two main schools of thought concerning how this justification works – reductionism and anti-reductionism.

Reductionists hold that testimonial beliefs are justified by reference to non-testimonial sources of justification. In other words, the justification that hearing a piece of testimony confers on a person is wholly reducible to other, more basic sources of justification, such as perception, memory and inference. In the words of Goldberg, “a hearer is not epistemically justified in accepting … another’s testimony unless she has (inductive or a priori) reasons, ultimately not themselves based on still further testimony, for regarding the testimony she confronts as credible” (Goldberg, 2007, p. 143).

Anti-reductionists have a less individualistic focus. Instead of saying that a person can rationally believe a piece of testimony only if she has evidence for the reliability of the witness, thinkers like Thomas Reid (1997, Ch. 6.24), Michael Dummett (1994), C.A.J. Coady (1994) and John McDowell (1998b) argue that testimony itself is a basic or sui generis source of epistemic justification. Some anti-reductionists understand this to mean that trust is always the default position. Instead of saying that a person can rationally believe a piece of testimony only if she has evidence for the reliability of the witness, anti-reductionists of the “deafaultist” stripe argue that an assertion is belief-worthy until shown otherwise.³⁶ Unless a listener has reason to suspect that a testifier might be deceptive or incompetent with respect to the subject matter he testifies about, the listener is rationally justified to believe what he says. If the speaker is in fact knowledgeable and speaks the truth, what the listener acquires from his words counts as knowledge. Testimonial beliefs on this account have basically the same epistemic status as memory-beliefs. Most people would presumably admit that I am justified in believing what my memory tells me, unless I have reason to suspect that I might remember wrong, and the same applies to testimonial beliefs, according to “defaultists.” There are also other sorts of anti-reductionists besides defaultists, for example philosophers who hold that a certain kind of interpersonal relationship between speaker and hearer “provides the hearer with a non-inferential reason to believe the speaker” (Siebert, 2016, p. 565).

The main argument in favor of anti-reductionism in general is the fact that the alternative view, reductionism, seems to have skeptical implications. If we can only acquire knowledge from testimony in cases where we have good evidence for the trustworthiness of the testifier, then it is doubtful that we have very much testimonial knowledge at all. For example, high school students often have no idea of who the authors of their physics or history textbooks are, but most people still assume that the students acquire knowledge from these sources in high school. ³⁷ However, if rationality requires that I have positive evidence for an author’s trustworthiness if I am to acquire knowledge from what he has written, then it seems to follow that I acquired very little, if any, knowledge in high school, despite being a rather diligent student. As many philosophers now acknowledge, it seems impossible to account for the justification of many of our seemingly impeccable testimonial beliefs by reference only to non-testimonial evidence, such as our own perceptions, memories, and the inferences we can draw from them. Against the background of this, anti-reductionism seems to be an attractive position (Matilal and Chakrabarti, 1994), and it is this general paradigm that I will assume in the present context.³⁸

In my view, the best anti-reductionist account of testimonial knowledge has been suggested by the philosopher John McDowell (1998b). McDowell holds that testimony, like perception and memory, can forge a cognitive link between a person and an objective fact, so that knowledge of this fact becomes available to the person through the words of another. Of course, sometimes people testify falsely, and in these “bad” cases knowledge only appears to be available. We cannot always know, by our own autonomous powers, when we are in a “bad” case, since it might be subjectively indistinguishable from a “good” case in which veridical testimony really provides access to the world beyond our immediate perceptual reach. However, when we in fact are in a good case, the beliefs we acquire from testimony are justified and constitute knowledge, according to McDowell. The justified status of such beliefs, moreover, does not depend on our possession of evidence for the trustworthiness of the speaker.

This theory of testimonial knowledge entails that it is not wholly up to us whether our testimonial beliefs are justified or not. Since we cannot always tell when we are in a bad case, we are to some extent at the mercy of epistemic luck – what McDowell calls “favors from the world.” This dependence is not something we should regret, however. The idea that we can be in full control of our own epistemic rationality – that we can always ensure that the beliefs we hold are justified – is an illusion, according to McDowell. Rather than kindling this illusion, we should acknowledge that we cannot function epistemically in a fully autonomous way, and that our own epistemic standing can be dependent on the way the world happens to be, including on whether the testimonies we encounter are reliable or not (McDowell, 1998a).

However, since acquiring knowledge from testimony is an exercise of reason, we need to evaluate the testimony of other people – as well as what purports to be divine testimony – in light of our background knowledge and be on the lookout for signs of deception or incompetence. We must, in other words, be aware of the risks we expose ourselves to when we trust others and be sensitive to information that indicates that their testimony may be unreliable. McDowell refers to this kind of sensitivity as “doxastic responsibility,” and he holds that it as a necessary condition for acquiring knowledge from testimony. A person who blindly believes everything he hears fails to acquire knowledge, because he does not respect the norms of rationality that are intrinsic to the pursuit of truth. However, if a person respects those norms and believes a piece of testimony that is actually reliable, what he acquires is knowledge. He does not necessarily need to possess, in addition, positive evidence for the reliability of the witness.³⁹

Now, it is not entirely clear what view of testimony Aquinas held. According to Matthew Siebert, Aquinas overall view is “pluralist,” which means that he combines reductionist and anti-reductionist stances to testimonial justification in different cases (Siebert, 2016).⁴⁰ When it comes to divine testimony, however, Aquinas clearly seems to take an anti-reductionist stance. He seems to hold that the mere fact that God has said something constitutes a sufficient justification for believing it, irrespective of whether the hearer has sufficient evidence that the speaker indeed is God. ⁴¹ This means that it is not the case that the hearer’s justification for believing divine testimony is constituted by an inferential argument from prior knowledge of the testifier’s divine identity (although the hearer must believe that the testifier is God, according to Aquinas). Rather, it is the objective fact that the testimony is God’s and therefore reliable that makes the beliefs that the hearer forms on its basis justified (Lamont, 2004, p. 63). In this respect, Aquinas’s view of the justification of faith can be described as “externalist,” or at least as involving externalist elements.⁴² As McDowell might put it, believers cannot autonomously guarantee their own epistemic standing with respect to purportedly divine testimony but are dependent on “favors from the world” (or in this case from God) in order to be justified in their beliefs.⁴³

This, I contend, is a plausible framework of interpretation for Aquinas’s view. According to him, Christian faith exists “only when one believes for this reason … that it is said by God, which is designated by calling it credere Deo. And this specifies faith, the way any cognoscitive habit has its species from the reason [ratione] by which it assents to anything” (Super Ep. S. Pauli, lect. 4, n. 1, quoted in Siebert, 2016, p. 570).⁴⁴ Note that Aquinas here does not say that our reason for believing P is that we believe (perhaps on very solid grounds) that P is said by God. What he says is that our reason [ratio] for believing P is that P is actually said by God, which is a fact about the world external to our minds.⁴⁵

Suppose, now, that God supernaturally, by means of operative grace and the light of faith, makes a person inclined to believe that it is God who has communicated the Christian message. God could make a person inclined to believe this by making it seem very plausible to her that this is the case, as well as by making her desire that the Christian message be true. In order to make the fact of divine revelation seem plausible and attractive to a person, God may have to make certain other beliefs that the person has seem less plausible to her than before, and God could do this in a gradual way by drawing the person’s attention to certain facts or arguments, by enlightening her about certain misconceptions or reasoning mistakes, and by affecting the emotional coloring of various conceptions and beliefs. As we know, the process of coming to faith is usually not instantaneous but often takes the form of a gradual change of perspective on the world, a change that also involves a person’s will and motivations.

Now, when this gradual process has reached a certain point, it might seem very plausible to the person that Christian beliefs are revealed by God, and she can then, moved by grace and assisted by the relevant gifts of the Holy Spirit (intellectus and scientia) grasp and assent to these beliefs as revealed by God. This assent will be voluntary, since it is not forced by the evidence. The assent will also be doxastically responsible if the believer indeed finds it very plausible that Christian beliefs are revealed by God, and if the fact that she finds this plausible is not due to any neglect of epistemic duties on her part, which need not be the case if Christianity is in fact true. What justifies the believer’s assent to the Christian message, however, is not the fact that it seems plausible to her, or that her assent is caused by habitual and actual grace (auxilium) and perfected by the gifts of the Holy Spirit.⁴⁶ Her belief in that message is justified, instead, because it is based on divine testimony, which as such is supremely reliable.⁴⁷

This brief sketch raises of course many questions, and it would need to be elaborated in various respects. In the context of the present article, however, there is only room for a couple of clarifying remarks. First, it is important to take note of the crucial difference between justification and doxastic responsibility. These notions might easily be confused, but they perform essentially different functions. On the model I have proposed, justification is not a deontological affair (a matter of doing one’s epistemic duties) but a matter of being related in the right way to the objective facts – in the present case, to facts made epistemically available through reliable testimony. Only if a person exercises doxastic responsibility, however, is she in a position to appropriate the rational justification that a piece of reliable testimony may confer on a belief. This is because justification (and the knowledge it enables) is a “standing in the space of reasons,” as McDowell puts it, and having such a standing requires the subject to take reasonable measures in order to avoid being misled by potentially false testimony. Exercising doxastic responsibility, therefore, is a matter of fulfilling one’s epistemic duties as a precondition for acquiring knowledge from testimony. However, since a “good case” – a case in which one responsibly believes reliable testimony – may be subjectively indistinguishable from a “bad case” – a case in which one responsibly believes unreliable testimony – there are no measures available that can guarantee that one never ends up in a bad case. Nevertheless, on the suggested model, the “good case” is essentially different from the “bad case” with respect to the subject’s epistemic standing or status (whether she is justified or not). This means, as we have seen, that people are to some extent dependent on what McDowell calls “favors from the world” (epistemic luck) at the level of epistemic rationality. We cannot autonomously guarantee that we are in a “good case” and that our testimonial beliefs are justified. What we can guarantee is merely that we obey the dictates of doxastic responsibility.

McDowell thinks that we have no choice but to admit this dependence on factors beyond our control, because the only alternative is to restrict the “space of reasons” to inner, subjective items, which purportedly are such that we cannot be wrong about their true nature. The problem with this move is that it opens up an unbridgeable epistemic chasm between the subjective sphere – where we take ourselves to be in autonomous control – and the objective world. This way of conceiving things is the cause of many of the skeptical worries that have plagued the Western epistemological tradition. Aquinas would undoubtedly agree with McDowell that the picture that generates these skeptical worries is false, and that we consequently must acknowledge that the “space of reasons” extends beyond the subjective realm. A necessary consequence of acknowledging this, however, is that we cannot reasonably claim to be infallible about our own epistemic rationality. If the “space of reasons” includes our relations to the objective world, and if we are fallible about how we relate to the objective world – which we clearly are⁴⁸ – then it follows that we are also fallible with respect to whether our beliefs are justified or not. This framework is congruous with Aquinas’s view that “what is in itself the more certain” – namely, the knowledge of faith – “may seem to us the less certain on account of the weakness of our intelligence” (ST I, q. 1, a. 5, ad 1). Divine testimony provides a supreme justification for our beliefs, but this does not mean that it is impossible for us to be fooled by a subjectively indistinguishable “bad case” where divine testimony seems to be available but in fact is not.

A second important clarification concerns the relationship between the justification that God’s public testimony confers on the believer and God’s interior activity in the soul. Although grace and the gifts of the Holy Spirit play a causal rather than epistemic (justificatory) role in the story I have told above, nothing prevents a proponent of the Testimonial Interpretation from admitting that the interior workings of grace also contribute to the justification of Christian beliefs to some extent.⁴⁹ The core claim of the Testimonial Interpretation is that the divine public testimony by itself constitutes a sufficient source of justification for Christian faith (a source which is only accessible, however, to persons who exercise doxastic responsibility). The truth of this claim does not require us to deny that some other source of justification, such as the reliability of graced faculties, could also make a significant epistemic contribution. In other words, the justification of Christian faith could be overdetermined. In principle, it would be consistent with the Testimonial Interpretation to hold that the interior workings of grace and the public divine testimony each provides (by themselves) a fully sufficient justification.

Although many of its details would need to be fleshed out, I submit that the theory I have proposed can be defended as a reasonable epistemological model. I also think that this theory does justice to the various elements of Aquinas’s teaching on faith, and especially the four features listed in the section “Aquinas on Faith” above. The theory portrays Christian beliefs as justified directly by divine testimony, which satisfies the first feature. The model furthermore portrays the assent of faith as voluntary and due to the influence of grace – these were the second and third features. The fourth feature – the existence of good apologetic arguments for the fact of revelation – is congenial to the suggested model in virtue of its emphasis on doxastic responsibility. If there is a strong apologetic case to be made for Christianity in light of public evidence, then it will be easy for God to make the fact of divine revelation seem plausible to any person without deceiving her or clouding her judgment. It should be noted, furthermore, that the notion of doxastic responsibility is relative to a particular person’s epistemic position. What doxastic responsibility requires of me might be different from what it requires of you, depending on our respective levels of knowledge and our situatedness in the world. It is not the case, therefore, that every person needs to have knowledge of apologetic arguments in order to be able to believe, in a doxastically responsible way, what God has revealed.⁵⁰