Virtue epistemology

To start with, notice that you don't necessarily know that p if you truly believe that p on account of a reliable belief-forming process.Footnote ⁴ At least this is what virtue reliabilists say against more austere forms of process reliabilism in epistemology.

For example, the thermostat on the wall might reliably indicate the temperature of your room, such that not easily do you consult the thermostat and thereby form a false belief about the ambient temperature. But while consulting the thermostat to determine the temperature might be a reliable belief-forming process in this way, it might not be reliable for reasons you'd expect. For suppose that the thermostat is broken, so that it randomly reports fluctuating temperatures throughout the day. As it happens, though, you have a hidden friend who is controlling the temperature of your room, always adjusting it so as to fit the reading on the thermostat. If you were to form a true belief by consulting the thermostat in this case, it seems that you wouldn't know what you thus believe, despite your believing what you do on the basis of a reliable belief-forming process. Plausibly, that's because your getting it right on this occasion isn't owing to an exercise of your cognitive abilities, so as to manifest your epistemic agency. Rather your getting it right is entirely to do with the help of your mysterious friend.Footnote ⁵

It's in light of cases like these that virtue reliabilists supplement reliabilist theories of knowledge with an ability requirement. It's not enough for knowledge that you truly believe something as a result of just any reliable belief-forming process. Rather that process must be linked to a display of some ability on your part to get things right on the matter in question, so that when you believe correctly as a result of this process you're creditable for the cognitive success.

It might be tempting to think that the reason why you don't know in the thermostat case is that while you're reliable about the temperature this has little to do with any reliable belief-forming process located in the head – there's no reliable neural belief forming process that's involved in the right way in the causal explanation as to why you get it right about the temperature. You might think that so long as the reliability of a belief-producing process is located in the head in this way then any true belief it produces would satisfy the ability constraint.

This is wrong, however. For not all reliable belief-forming processes whose reliability is ‘internal’ in this way are such that were you to truly believe something on their basis, we'd credit your success to your epistemic agency. That is to say that not all executions of a reliable neural belief-forming process count as a display of some cognitive ability of the sort at issue here (i.e. an ability whose exercise can underwrite knowledge).

For example, you might find yourself (truly) believing a whole host of things for reasons you don't understand. That might be because you've suffered a brain lesion that, as it happens, reliably produces true beliefs on these subjects. But although this looks to be an ‘internally’ reliable belief-forming process it doesn't amount to a cognitive ability on your part, at least not of the sort of interest here. That's because we wouldn't credit you for getting right the things you believe on the basis of this unusual neural belief-producing process, consisting as it does in clear cognitive malfunction.Footnote ⁶

The moral seems to be that you can't know something unless you believe it on the basis of a reliable belief-forming process, such that your getting it right on the basis of that process is properly creditable to your cognitive agency, where it isn't sufficient for these purposes that the reliability of the belief-producing process at issue be ‘internal’ or in the head of the subject. Indeed, neither does it seem necessary, as we'll come to find below.

Plantingianism about faith

Unfortunately for the Plantingian virtue religious epistemologist, the ability constraint on knowledge seems to conflict with the idea that you can know things on the basis of faith. I'll now proceed to show more specifically why that's the case.

First, note that on Plantinga's view believing something on the basis of faith is a means of knowing religious subject matter. He writes that ‘faith is not to be contrasted with knowledge’ (Plantinga (Reference Plantinga2000), 256), but rather faith is ‘seeing the truth of the central Christian affirmations’ (ibid., 206). It's only by an exercise of faith that one knows, for example, that God is in Christ Jesus reconciling the world to Himself. Otherwise, you might know, for example, that it says this in Christian scripture, or perhaps even that this is something you should believe (if you identify with the Christian tradition). But the claim here is that absent faith you can't believe that it's true in the manner required for knowing that it's true.

Note second, however, that, as Plantinga also writes, faith-based beliefs ‘don't come just by way of the normal operation of our natural faculties – they are a supernatural gift’ (ibid., 245). After all, as he says, faith is a ‘belief producing process’ that ‘consists in direct divine activity’ (ibid., 246; my emphasis). More exactly, Plantinga thinks that faith-based belief (and knowledge) is a product of a ‘three-tired’ (ibid., 244) belief-producing process, consisting of the Bible, the activity of the Holy Spirit, and a resulting doxastic state.Footnote ⁷ This is meant to be the belief-producing process that God organized in order to secure among human beings true beliefs about Himself. In the paradigmatic case, as a subject habitually engages in reading the Scriptures, the text and the Holy Spirit conspire to produce true beliefs in the subject.

Whatever the ultimate merits of these theological claims about faith-based belief, for our purposes we can boil them down to two, and call their conjunction Plantingianism about faith-based knowledge:

1. (I) Faith: Faith is the entire three-tiered process of (1) one's coming to believe some claim regarding God's great scheme for human salvation (2) as the result of the work of the Holy Spirit making one to believe (3) what she finds in Scripture.

2. (II) Faith Alone: Only by believing as a result of the process above can one properly know some claim regarding God's great scheme for human salvation.Footnote ⁸

But now the worry for the Plantingian virtue religious epistemologist is that it hardly seems to be owing to your epistemic agency that you believe something correctly when you believe it on the basis of faith. Rather it seems to have everything to do with the agency, or the activity, of the Holy Spirit – it looks to be an entirely divinely orchestrated event.

And so if the ability constraint on knowledge is correct, then it's hardly clear how you know what you believe on the basis of faith as that's being conceived here. Or in other words: how is one to make sense of the idea that when you believe something truly on the basis of faith, this can be put down to an exercise of your cognitive abilities such that your getting it right is largely due to your epistemic agency? This seems to force a choice between virtue reliabilism and Plantingianism about faith-based belief. Premise II of our argument looks initially compelling: cases of faith-based belief are not cases in which you form true beliefs on account of your cognitive abilities.

Stronger and weaker versions of the ability constraint

But perhaps things are not as desperate as they first appear for the Plantingian virtue religious epistemologist. For it might be that if we get clearer about what the ability constraint on knowledge requires, we'll find that it doesn't demand anything that faith-based belief can't be seen to accommodate, even on the Plantingian conception.

Perhaps it seems that cases of faith-based belief are not cases in which you form true beliefs on account of your cognitive abilities only because we're assuming a rather strong conception of what that amounts to. Namely, we're assuming that in such cases you're meant to believe what you do such that your getting it right on the matter in question is primarily creditable to your cognitive agency.Footnote ⁹ Now if this is the correct interpretation of the ability constraint, then faith-based knowledge surely looks impossible. For it seems indisputable that when you get something right by believing it on the basis of faith this isn't primarily creditable to you, since it's most conspicuously owing to the supernatural intervention of divine agency. However, if there are independent reasons to think that this is too strong a construal of the ability constraint anyway, then perhaps there's a weaker version that's weak enough for a virtue-theoretic explanation of faith-based knowledge.

Indeed there do seem to be reasons to suggest that this is too strong an interpretation of the ability constraint. Consider for example a standard case of testimonial-based knowledge. You come off the train in a foreign city and ask a seemingly reliable informant for directions to the museum.Footnote ¹⁰ If they tell you that it's on Third Street, surely you can come to know this by trusting them. But surely that's not in part because your getting it right about the location of the museum is primarily creditable to your cognitive agency. Rather your hitting upon the truth in this regard seems primarily creditable to the epistemic agency of your informant. Nevertheless you come to know what you do, and this suggests that the current interpretation of the ability constraint on knowledge is too strong.

Perhaps we should say instead, then, that when you know that p you must believe that p on the basis of a reliable belief-producing process such that your getting it right is at least significantly creditable to your cognitive agency. This seems weak enough to permit knowledge on the basis of trusting the testimony of others. For surely when you believe what you do in these cases your getting it right is at least significantly creditable to your cognitive agency, if not primarily creditable. You did well enough not to ask directions of a rubbish bin, for instance, and were your informant to betray signs of untrustworthiness you'd do well enough to take notice, and suspend judgment about what they tell you.

But now the question is whether this weakened ability constraint is weak enough for explaining how, in cases of putative faith-based knowledge, you get it right on the matter at issue on account of your cognitive abilities. Can some sense be made of the idea that when you truly believe that p on the basis of faith, your believing truly is at least significantly creditable to your epistemic agency?

I'm not optimistic. In fact there seems to be an important dissimilarity between faith-based belief and testimonial-based belief that suggests that you might not even be significantly creditable for getting something right on the basis of faith. For at least when you know something on the basis of testimony you form your belief of your own accord in a way that's partly explained by, among other things, your actively monitoring your situation, including your being on the lookout for reasons to doubt either the sincerity or the competency of your informant. And so your believing something truly on the basis of testimony enjoins at least this much cognitive ability on your part. It seems entirely natural to say that you have an ability to believe things truly by trusting the say-so of others.

By contrast, however, on the Plantingian conception your believing something truly on the basis of faith is a thoroughly divinely orchestrated event, down to the very cause of your believing what you do. It's not as though human beings are merely unable to discover the truth about religious subject matter on their own (for which reason the Scriptures were organized as a testimony). But neither are they so much as able to believe these truths on their own when confronted with them. Rather, on the Plantingian model it's meant to be the job of the Holy Spirit not merely to facilitate, but actually to create belief on the part of the believing subject. In fact, it's no exaggeration to say that your believing something on the basis of faith is something the Holy Spirit does.Footnote ¹¹

For example, Plantinga writes that the Holy Spirit ‘gets us to accept’ or ‘causes us to believe the great things of the gospel’ (ibid., 254). He even goes so far as to say that the Holy Spirit ‘produces’, even ‘induces’, belief (ibid., 256). I daresay that I was never induced to believe anything by an informant, and that's in part because I clearly have an ability to form beliefs by trusting others.Footnote ¹² And I think the thing to say here is that if Plantinga is right then it's far from obvious that I have a similar ability to form beliefs by trusting the words of Scripture.

I therefore conclude that even if paradigm cases of believing something truly on the basis of faith are cases of believing something truly on the basis of a reliable belief-forming process, they don't look to be cases in which you believe what you do on account of your cognitive abilities, even on a weak conception of what that amounts to. That's because these don't seem to be cases in which your believing truly what you do is even significantly creditable to your cognitive agency, much less primarily creditable. And so I'm not optimistic about retreating to a weakened version of the ability constraint for disputing the idea that when you know something on the basis of faith you don't get it right on account of your cognitive abilities.

Extended cognitive abilities

We found that even on a weakened conception of the ‘ability constraint’ on knowledge believing something truly on the basis of faith just doesn't look to be sufficiently creditable to the exercise of one's cognitive abilities. But it's time to come clean about the fact that up to now we've only had one's organismic cognitive abilities in view – those realized by neural states beneath the skin and skull of the subject. But why think that one's cognitive abilities are all and only one's organismic cognitive abilities? Why not instead allow for cognitive abilities that are extended, or that are realized by cognitive processes driven along by extra-organismic elements?

I submit that in order to get around the argument for agnosticism from the ability constraint on knowledge the Plantingian virtue religious epistemologist should conceive of faith as a form of extended knowledge by virtue of its being produced by an extended cognitive ability. The idea is that people who know things on the basis of faith believe truly as they do on account of a cognitive ability of theirs that actually consists in the activity of the Holy Spirit in conjunction with Scripture. More exactly, the cognitive ability in question is at least a disposition on the part of the subject to execute, in the proper circumstances, a cognitive process whose proper parts include the Holy Spirit and Scripture.Footnote ¹³

On its face this can seem a rather staggering claim.Footnote ¹⁴ For you might think that cognitive processing is an entirely ‘internal’ matter in that it goes on entirely inside one's head. But to insist that the belief-producing process responsible for faith-based belief actually constitutes a cognitive process seems to directly violate this kind of ‘internalism’ about cognitive processing. For it suggests that the factors that actively drive cognition along can include such blatantly ‘external’ things as a book and the agency of another person!Footnote ¹⁵

But actually this is by no means unprecedented. In fact not only is cognitive science already familiar with the idea that something like a book might constitute a cognitive process, but also with the idea that another person might do so as well. To borrow a phrase from Clark and Chalmers (Reference Clark and Chalmers1998, 14), we're learning that when it comes to cognition there is nothing sacred about skull and skin.

Proponents of extended cognition in cognitive science, for example, hold that some cognitive processing involves the use of external artefacts and/or instruments in ways that suggest that these artefacts actually play a constitutive role in driving these processes along. And proponents of distributed cognition go further still, arguing that cognitive processing might depend constitutively in this way not only on external artefacts, but other persons as well. I'll present these two views in turn.

Extended cognition and distributed cognition

To get a grip on the idea behind the thesis of extended cognition, consider what's become a stock case in this regard.Footnote ¹⁶ Otto's biological or organismic memory is failing him. He notices this and so begins to carry with him a notebook in which he jots down the sort of information he would need throughout the course of an ordinary day – the sort of information you or I would simply store in our biological memories. And so when it comes time for you to meet Otto for dinner, while you might remember where you've planned to meet him by consulting your biological memory, Otto will ‘remember’ this by virtue of consulting what he's written down in his notebook. The claim is that since there seems to be no functional difference with respect to how the two of you store and retrieve information we should think of Otto's notebook as actively driving the cognitive processes he employs in figuring out where to meet you for dinner. Andy Clark and David Chalmers write that:

And so the suggestion is that when Otto arrives for dinner according to plan that is in part because he knew where and when to arrive by virtue of exercising a cognitive ability that's extended by virtue of containing the notebook as a proper part. Otto's knowledge is thus extended, since it's dependent upon an extended cognitive ability in this way.

But this isn't the only way ‘extended knowledge’ might be generated. For such knowledge might also be produced on the basis of a cognitive process that isn't merely extended but (also) distributed: a cognitive process consisting not (or not only) of external material artefacts, but also other subjects or persons.

The idea that cognitive processes might be distributed in this way can be motivated in connection with our wanting to understand knowledge on the part of scientific research teams, for example, or else other forms of group knowledge that are distinctively social or collaborative in nature. The idea here is that if we can come to see the individuals of these research teams as collectively realizing a kind of sui generis cognitive process, the way is opened to conceive of the members of these groups as forming a kind of ‘group agent’. And if we can conceive of research teams as group agents, then there's scope for understanding distinctively social knowledge of the sort at issue in terms of the conceptual resources furnished by more mainstream or agent-focused epistemology.Footnote ¹⁷

Alternatively, so-called transactive memory systems are thought to provide support for the idea that cognitive processes might actually be distributed among two or more agents. Consider a case that Wegner et al. (Reference Wegner, Giuliano, Hertel and Ickes1985, 257) describe:

The idea here is supposed to be that neither Rudy nor Lulu individually knows that it was in Peachland where they bought the stuffed Canadian goose. But they do know it collectively, or together, plausibly by virtue of constituting a distributed cognitive process or ability, whereby in cueing one another in the manner illustrated above they're together able to recall some bit of information.

An objection

Now it can seem that these two cases of putatively distributed cognition don't quite seem to illustrate what I'm recommending in connection with faith-based belief: viz., that one of a subject's cognitive processes can be a distributed cognitive process in the sense that it involves one or more subjects as proper parts. For while the research team case and the Rudy–Lulu case illustrate cognitive processes that might include other subjects as proper parts, these cognitive processes don't seem to belong to any one subject as one of their cognitive processes.

This is probably most clear in the research team case, since there the cognitive process at issue belongs not to any individual researcher, but to the collective as a kind of ‘group agent’. But the claim at issue with respect to the cognitive process responsible for producing faith is not that it belongs to some group agent comprising the believing subject and the Holy Spirit – but to the believing subject herself.Footnote ¹⁸

However, matters aren't similarly so clear-cut concerning the Rudy–Lulu case. For example, consider someone who didn't like the idea of ‘group agents’ (perhaps they thought group agents were just too spooky a specimen). What's to prevent her from thinking all the same that the Rudy–Lulu case motivates a kind of ‘externalism’ about cognitive processing that goes beyond the idea that cognition is merely extended to external artefacts? For instance, what would preclude this person from thinking that Rudy enjoys a distributed cognitive process by virtue of its consisting of Lulu as a proper part, and that Lulu enjoys a distributed cognitive process by virtue of its consisting of Rudy as a proper part, and that no cognitive process involving both Rudy and Lulu belongs to some group agent comprised of both Rudy and Lulu? But if there is such a weaker interpretation of such cases available (one that needn't postulate ‘group agents’), then we do already have kinds of cases that illustrate what I want to say about the cognitive process consisting of the activity of the Holy Spirit.

But the more important point is that in order to serve their purpose the ‘research team’ and ‘Rudy–Lulu’ cases needn't strictly illustrate the form of distributed cognition I'm trying to motivate on behalf of a religious believer. It's sufficient that they establish a precedent for the general idea that cognitive processes might consist of other agents, never mind to ‘whom’ those cognitive processes belong. It seems that until the Plantingian virtue religious epistemologist is given some principled reason to think that a cognitive process cannot both belong to some one subject and be distributed, she should feel free to appeal to a process of this sort for making sense of faith-based knowledge.Footnote ¹⁹

In light of the forgoing, then, it's hardly ad hoc for the Plantingian virtue religious epistemologist to claim that the cognitive ability at issue in producing faith-based knowledge actually consists of Scripture and the agency of the Holy Spirit. For the idea that the parts of a cognitive process might include an external artefact (like a book) and/or another person is right in line with recent developments in cognitive science.

But of course not just any ‘external’ belief-producing process qualifies as an extended or distributed cognitive process. There are standards here as anywhere else. What this boils down to is that the process must be appropriately cognitively integrated with the rest of one's cognitive character. And so it remains to determine whether the belief-producing process at issue in the production of faith-based belief plays well the role of an extended cognitive process: or, what amounts to the same thing, whether it can be seen to be appropriately cognitively integrated.

If the mechanism of faith can be seen to be cognitively integrated with the rest of a subject's stable capacities in the way required, then our virtue religious epistemologist is unencumbered to conceive of this process as realizing a cognitive ability on the part of a religious subject. I turn to consider this next.

Glue and trust

Beginning with the ‘glue and trust’ criterion, we can follow Clark and Chalmers in thinking that when determining whether a reliable-belief producing (/sustaining) process should count as an extended cognitive process ‘what is central is a high degree of trust, reliance, and accessibility’ (Clark & Chalmers (Reference Clark and Chalmers1998), 17)Footnote ²⁰ First, the elements comprising a belief-producing process won't count as constituting an extended cognitive process unless they're trusted. For Clark and Chalmers this means that these features are phenomenologically transparent in a way: that a subject ordinarily automatically endorses the deliverances of the belief-producing process at issue, without any especial reflective scrutiny. Moreover, second, the subject should rely on the belief-producing process, where this means that when circumstances call for the sort of information supplied by the belief-producing process, the subject is liable to draw on that source ahead of any others. Lastly, the belief-producing process should be accessible, or readily enough available to the subject. That is to say that the components of the process should be relevantly fixed or ‘glued’ to one's organism.

To illustrate these criteria, recall Otto and his notebook. Plausibly the belief-producing process consisting of Otto's notebook is a good candidate for an extended cognitive process on these criteria. When Otto consults his notebook for information he's previously endorsed and written down he isn't especially critical about what he reads. He doesn't consciously form beliefs about his notebook and its reliability, for instance, when consulting it in this way, much as we don't form similar sorts of beliefs when consulting our biological memories. And so the belief-producing process consisting of Otto's notebook is transparent in this way. Moreover, Otto properly relies on the notebook since whenever he's in a situation in which he'd like to be able to retrieve information he's come across before, he first consults his notebook (he doesn't go about trying to find again the original source of information, he doesn't go about asking others for the information, etc.). Furthermore, it's stipulated that Otto takes his notebook with him everywhere he goes, and so it seems to be readily enough available – available as often as is required – as often as we'd like to have readily available our biological memories.

Does the belief-producing process consisting of Scripture and the activity of the Holy Spirit violate any of the ‘glue and trust’ requirements? Not obviously.

First, this belief-producing process seems to be trusted by the subject in the relevant fashion. In the paradigmatic case, when a subject finds herself believing from having read in her Bible that, say, through Jesus’ death there is redemption and the forgiveness of all wrongdoing (Col. 1:14), she's not especially critical of the reliability of the book, or of the activity of the Holy Spirit in producing in her the impression that it's true. These are transparent players in the relevant sense. It's simply part of the background that she's come to accept that the Holy Spirit is adequate to make her convinced of the things she reads in Scripture, and that the Scriptures are a reliable witness of all they speak to. What she takes herself to be receiving from these sources she swallows down uncritically, as does Otto the information in his notebook.

Second, it isn't obvious that she doesn't rely on this process in the manner required by the ‘glue and trust’ criterion of extended cognition. It's not as though she's no more ready to consult the Bible as she is the Qur'an, the Hadith, or horoscopes on matters pertaining to her faith and religious practice. Furthermore, neither is she prepared to act on the basis of a religious claim unless she takes herself to have learned it on the basis of the belief-producing process at issue – at least not plausibly. To the extent that she is, it seems that the belief-producing process comprising Scripture and the activity of the Holy Spirit isn't being relied on in the relevant way.

Third, it seems that the components of the belief-producing process at issue in faith are as readily available/accessible as required. Little needs to be said to motivate this claim in connection with the Holy Spirit. On standard Christian teaching the Holy Spirit actually resides within the regenerate individual, so that He's as readily accessible as any of one's more ordinary cognitive equipment.Footnote ²¹ However, obviously the Scriptures are not likewise accessible, at least not as accessible as Otto's notebook. The typical religious believer doesn't carry around with her and consult her Bible to the extent that Otto does with his notebook.Footnote ²² And so you might worry that on the ‘glue and trust’ criterion the belief-producing process undergirding faith-based belief is not a good candidate for an extended cognitive process after all.

But I don't think there's any deep problem here. Recall that on the ‘glue and trust’ criterion, in order to count as an extended cognitive process the elements of that process need only be readily enough available to the subject.Footnote ²³

In Otto's case this plausibly requires that his notebook be physically with him at all times, since the putative extended cognitive process comprising the notebook is meant to be substituting for a cognitive process one typically draws on nearly at all times – viz. biological memory.

But the belief-forming process whose elements consist of the Spirit and the Bible is not similarly substituting for any such cognitive process, one that a subject need nearly always be drawing on in the course of an ordinary day. So there's no obvious reason why the requirement that the elements of an extended cognitive process be readily enough available to one should require that one's Bible be as readily available as Otto's notebook.Footnote ²⁴

If defenders of the glue and trust criterion insist to the contrary that ‘readily enough’ requires that the elements of a putative extended cognitive process be as accessible as Otto's notebook is to him, then our response should be that this looks to exclude too much of what appear to be genuine cases of extended cognition (assuming as we are here that there are any at all). Consider, for example, the ways in which one depends on pen and paper in working out a maths problem, or the way in which one interacts with a microscope in conducting one's scientific research.Footnote ²⁵ These are plausibly prime examples of an extended cognitive process in action, and yet one is not always with pen and paper, nor with a microscope.

And so I conclude that there isn't any good reason in light of the ‘glue and trust’ criterion to think that the belief-producing process responsible for producing faith can't constitute an extended cognitive process, and therefore an extended cognitive ability.

Reciprocal interaction

What about the second of the two criteria for extended cognition I listed above: the criterion of ‘reciprocal interaction’? Orestis Palermos and Duncan Pritchard (Reference Palermos, Pritchard and Reider2016, 112) write that, on this criterion, ‘what is required in order to claim that two (or more) systems give rise to some extended or distributed process . . . is the existence of mutual (i.e. non-linear) relations – that arise out of feedback loops – between the contributing parts’.

It's typical to refer to the watt governor to illustrate how this is supposed to work. In terms of dynamical systems theory the watt governor represents a coupled system, whose sub-components consist of the governor and the engine, engaging in certain reciprocal interactions with one another. Palermos and Pritchard explain that the reason why these sub-components form to create a new coupled (or extended) system is that, first, unless we postulate the new coupled (or extended) system we haven't the means to explain certain emerging properties of the coupled system which can't be attributed to any of its component systems; and then, second, because the governor and the engine interact with each other so intimately that these interactions cannot be feasibly described in terms of distinct inputs and outputs from one system into the other. That is to say that at no time is one component system in some state or behaving some way independently of the counterpart system's being in some state or behaving some way at the same time, and vice versa. As Palermos and Pritchard (ibid., 113) write:

In producing religious faith in a subject, does the belief-forming process at issue reciprocally interact with other operations belonging to the subject to count as coupled with the subject in the relevant way? If it cannot be seen that it does, then at least by the lights of the reciprocal interaction criterion the mechanism of faith fails to count as a properly extended cognitive process, and so a fortiori an extended cognitive ability.

Interestingly, however, I think the processes that produce/sustain religious belief in an individual can be seen to cooperatively interact with other features/states of the subject herself in ways that suggest that together they play well the role of a coupled system, as that's been adumbrated above. In order to see this, however, we first need to get more specific about how the Holy Spirit makes one to believe the elements of God's scheme for human salvation when one encounters them in the Bible.

When you believe some religious proposition on the basis of faith, that's because the Holy Spirit ‘gets you to accept’ or ‘causes you to believe’ this. It's something the Holy Spirit does in you. That's what we said above, following Plantinga's lead. However, we left it ambiguous whether this was a basic or non-basic action on the part of Holy Spirit: whether your believing as you do is something He does directly or indirectly. On a more detailed version of how this works – a version Plantinga (Reference Plantinga2000, 305) himself seems open to – your believing an item of faith is something the Holy Spirit does in you indirectly. More specifically, it's by making you perceive the loveliness, attractiveness, or divine beauty of some religious proposition p that the Spirit gets you to believe p.Footnote ²⁶ And, crucially, He does this in response to, or by drawing on, your affective dispositions: whatever capacity you have to love or take delight in the things of God.

For example, take a subject who's convinced on the basis of faith from their exposure to Scripture that through Jesus’ death there is redemption and the forgiveness of all wrongdoing (Col. 1:14). On the more detailed model we're now considering, your believing this is something the Holy Spirit does in you, in response to your affective condition, by making you to perceive the divine beauty or loveliness of the fact that through Jesus’ death there is redemption and the supreme forgiveness of all wrongdoing.Footnote ²⁷ The sense in which the Spirit draws on or responds to your affective capacities is that it's only insofar as you're disposed to delight in the things of God that the Holy Spirit is moved to make you to perceive this proposition to be lovely or beautiful. In this way, then, the production and sustenance of religious faith in an individual cannot be understood apart from the shape of a subject's wider affective profile – consisting of the subject's loves and hates vis-à-vis the things of God.Footnote ²⁸

So far, though, this is only enough to secure a linear relation between states/processes of the subject and the belief-forming process responsible for producing in her religious faith. If these systems are to be non-linearly or reciprocally related, then ideally one's religious faith would function in some capacity as feedback, informing the shape of one's religious affections (which in turn, according to the above account, inform the activity of the Holy Spirit in indirectly causing one to believe religious propositions by making one perceive their beauty).

Interestingly, this seems to be just what happens when we consider the role of one's faith-based beliefs in one's religious practice. For plausibly it's in light of the things one believes by faith that one acts out one's religious life. And surely part of what acting out one's religious life does in one is texture one's religious affections: it enlarges ons's love, adoration, or desire for the things of God. C. S. Lewis reflects something of the same sentiment in Mere Christianity when he writes: ‘Do not waste time bothering whether you ‘love’ your neighbour; act as if you did. As soon as we do this we find one of the great secrets. When you are behaving as if you loved someone, you will presently come to love them’ (Lewis (Reference Lewis2001), 110).

What I mean is that it's one's religious faith that naturally spurs one on in one's religious activity – in one's habitual reading of the Scriptures, communicating with God through prayer, sharing about Jesus’ story with others, and in general participating in the activities of the Church. And surely by engaging in these religious activities one increases in one's love or desire for the things of God – desires and loves which, on the present account, the Holy Spirit then draws on in turn for sustaining or increasing one's religious faith.

And so it seems that religious faith isn't produced as a result of some ‘external’ belief-forming process operating in isolation from the rest of the subject's capacities and behaviours. Rather, the production and maintenance of faith-based belief is the result of a complex two-way interaction between a subject and the mechanism comprising Scripture and the action of the Holy Spirit that is almost helical or dialectical in character.Footnote ²⁹ In the dynamic religious life the two ‘systems’ are ‘mutually interconnected’ in the relevant way: the Holy Spirit is always both simultaneously determining and being determined by states of the subject who has a hand in shaping her affections by exercising her faith in religious practice. Likewise, the activity of the subject is always both simultaneously determining and being determined by the activity of the Holy Spirit, Who's always drawing on the subject's religious affections in producing/sustaining faith in her.

Therefore there seems to be no reason for denying that a subject interacts with the belief-producing process consisting in the activity of the Holy Spirit such that the latter is coupled with the subject in the relevant way, allowing for cognitive integration. Religious faith isn't produced/sustained in a subject apart from the mutual interaction of the subject and the Spirit in this way. The set up is such that you cannot understand the ‘behaviours’ of the Spirit in producing faith in a subject without simultaneously taking into account the behaviours of the subject who's all the time exercising that faith in religious practice, and vice versa. And this is just what we wanted to say of the two systems if one is to be properly conceived of as coupled to the other, according to dynamical systems theory.

Therefore at least with respect to the criterion of ‘cooperative interaction’ there's nothing barring the belief-producing process at issue in the production of faith-based belief from counting as a bona fide extended cognitive process, and so an extended cognitive ability on the part of a subject.

I conclude that there are no obvious reasons for thinking that the belief-producing process at issue in faith-based belief cannot constitute an extended and distributed cognitive process. For we've seen that there's no good reason to think that this process can't be cognitively integrated into one's cognitive character, on any viable conception of what that amounts to.Footnote ³⁰