PHILOSOPHIA CHRISTI
VOL. 8, NO. 2 © 2006
Could the Extended Aquinas/Calvin
Model Defeat Basic Christian Belief ?
ERIK BALDWIN
Department of Philosophy
Purdue University
West Lafayette, Indiana
According to Alvin Plantinga, a belief is warranted if and only if it is
internally and externally rational. Warrant is simply that property of a belief
enough of which, together with truth, makes a belief an instance of knowledge. Internal rationality is a matter of making inferences, deductions, and
connections between the beliefs one holds, and seeking evidence of truth
when appropriate or necessary. One who is internally rational will have correct doxastic responses to experience, the sort of doxastic responses required
by proper function. A person is externally rational if he forms and holds the
beliefs he ought (normatively) to form and hold in virtue of his cognitive
faculties functioning properly in an epistemic environment sufficiently similar to the one for which they were designed (by God or evolution) to operate. In order to develop my argument in this essay, I assume that Plantinga’s
account of warrant as proper function and his views about internal and external rationality are correct.1
Consistent with Plantinga’s epistemology are models he introduces in
order to show how it could be that belief in God, as well as specifically
Christian beliefs about God, could be internally and externally rational and
warranted for Christians in a properly basic way. Plantinga writes that
To give a model of a proposition or state of affairs S is to show how
it could be that S is true or actual. The model itself will be another
proposition or state of affairs, one such that it is clear (1) that it is not
1. I assume that the reader is familiar with Alvin Plantinga’s proper function account of warrant. See Alvin Plantinga, Warrant and Proper Function (Oxford: Oxford University Press,
1993), 151. For his account of internal and external rationality see Alvin Plantinga, Warranted
Christian Belief (Oxford: Oxford University Press, 2000), 110–13. Hereafter, Warranted
Christian Belief is abbreviated WCB.
383
384
PHILOSOPHIA CHRISTI
merely logically but epistemically possible and (2) that if it is true,
then so is the target proposition.2
Plantinga develops two models, one for theistic belief (TB) and another
for Christian belief (CB). The model for TB is called the Standard
Aquinas/Calvin model (or simply the Standard A/C model) and the model for
CB is called the Extended A/C model.
According to the Standard A/C model, belief in God is produced by a
cognitive process or faculty called the sensus divinitatus (SD).3 SD-beliefs
are epistemically basic—warranted in a noninferential manner—and are
formed under certain appropriate conditions. SD-beliefs have a particular
religious content, namely TB. The content of TB is that God is an intellectual, affective, and intentional agent, a person who is all-loving, perfectly
good, all-knowing, and all-powerful. If God exists and has created humans
such that they, by and large and for the most part, form warranted beliefs
about what is going on around them, then humans have warranted beliefs
about a great many things. If God also designed humans such that if they are
functioning properly they tend to form epistemically basic beliefs about God
in appropriate circumstances, such as when gazing at the stars or when looking at a sunset, then humans would form warranted beliefs about God’s
goodness, his omnipotence, and the like in such circumstances. So we see
how it could be that beliefs about God’s attributes can be warranted for
humans in an epistemically basic manner.4
The Extended A/C model is a uniquely Christian extension of the
Standard A/C model. According to the Extended A/C model, beliefs about
the Christian God are produced by means of a three-tiered cognitive process:
the internal instigation of the Holy Spirit (IIHS), Scripture, and faith. IIHS is
a special case of divine testimony through which a human directly comes to
believe the main lines of the Christian gospel, doctrines such as the Trinity,
the Incarnation, Christ’s resurrection, the forgiveness of sins, salvation,
regeneration, and eternal life.5 According to the model, the Holy Spirit instigates the process by which Christians come to form these sorts of beliefs
about God. This instigation can happen when one reads Scripture, but simply hearing the word of God preached at Church or listening to a sermon on
the radio may do just as well. Of crucial importance to the Extended A/C
model is the notion that the Holy Spirit illuminates Scripture for the
Christian, causing the Christian to see its importance and to see that it is true.
The Holy Spirit thus causes people to have faith in God. Following John
Calvin, Plantinga defines faith as “a firm and certain knowledge of God’s
2. Plantinga, WCB, 168.
3. Plantinga describes the SD as “a disposition or set of dispositions to form theistic beliefs
in various circumstances, in response to the sorts of conditions or stimuli that trigger the working of this sense of divinity” (WCB, 173).
4. See WCB, chapter six, “Warranted Belief in God.”
5. Plantinga, WCB, 251–2.
ERIK BALDWIN
385
benevolence toward us, founded on the truth of the freely given promise in
Christ, both revealed to our minds and sealed upon our hearts through the
Holy Spirit.”6 Faith is produced by cognitive capacities or activities equally
involving the intellect, the will, and the affections. Central to the Extended
A/C model is the notion that since humans are sinners God reveals himself
to humans and graciously produces in them the belief that accepting Jesus
Christ’s atoning work on the cross is sufficient for salvation. As such, the
content of faith is simply the central teachings of the gospel contained in the
intersection of the great Christian creeds.7
Plantinga argues that if Christianity is true, given that his proposed conditions for warranted belief are satisfied and both the Standard and the
Extended A/C model are true, then if a subject S has considered how well CB
coheres with her other beliefs, engages in defeating defeaters for CB as they
arise and has been generally epistemically responsible with respect to the
formation of CB, then CB is internally and externally rational and warranted
for S in a properly basic manner.8 That is, according to Plantinga, (1) and (2)
are true:
(1) TB can be internally and externally rational and warranted for the
theist in a properly basic way.
(2) CB can be internally and externally rational and warranted for the
Christian in a properly basic way.
Plantinga’s argument for (1) and (2) is ingenious and has great force.
However, if we grant that (1) and (2) are true there are unexpected consequences, as we shall see in the next section.
The Challenge of Alternative Theistic Extensions
of the Standard A/C Model
In this essay I assume Plantinga’s (1) and that the Standard A/C model
is true. But there are many possible ways to extend the Standard A/C model
so that it covers creedal-specific beliefs about God other than CB. NonChristians may agree on the form of Plantinga’s epistemological program, his
proper functionalism and his Standard A/C model, but disagree with the
Christian about its substance, which of its possible extensions is true. NonChristian theists who accept Plantinga’s epistemology and his Standard A/C
model will deny (2) and in its place affirm some analogue of it; they will sub6. Ibid., 244. Plantinga here quotes John Calvin’s Institutes of the Christian Religion, 3.2.7.
7. Plantinga, WCB, 244.
8. One must clearly keep in mind what Plantinga claims to have shown. Plantinga does not
argue that S knows that CB is true or warranted. Plantinga does claim to have shown how it is
epistemically possible for S to accept CB in an epistemically basic and warranted manner. To
show this he argues for the conditional that if S accepts CB and it turns out to be true, then CB
is warranted for S.
386
PHILOSOPHIA CHRISTI
stitute the CB element of (2) with another set of creedal-specific beliefs
about God. I maintain that the fact that the Standard A/C model can have
multiple extensions undermines the alleged truth of (2) and renders it
implausible to suppose that CB, or any other set of creedal-specific beliefs
extended from the Standard A/C model, can be held by a theist in a properly
basic manner. Accordingly, Plantinga’s contention that CB can be warranted
for Christians in a basic way is implausible, at least for those who understand
the defeater based on multiple extensions of the Standard A/C model I introduce in this essay. Before I develop my argument any further, it is necessary
first to show how it is that non-Christian extensions of the Standard A/C
model can be formulated by non-Christian theists and why this causes problems for Plantinga’s views.
Consider the Jewish, Islamic, and Hindu faiths. TB (or something very
much like it) is held in common by Jewish, Islamic, and Hindu theists, and
their creeds adumbrate, imply, and/or provide raw materials for constructing
Jewish, Islamic, and Hindu versions of the Standard A/C model. And so
Jews, Muslims, and Hindus may rightly think that TB is internally and externally rational and warranted for them in an epistemically basic manner. After
all, Jews, Muslims, and Hindus posit processes by which God makes himself
known to humans, and posit belief forming mechanisms by which humans
come to accept particular creedal specific beliefs about God that are consistent with the deliverances of the SD.9 There is, therefore, good reason to
think that philosophically inclined Hindus, Jews, and Muslims will accept
(1). While individuating models may be somewhat tricky, I maintain that
Christians, Hindus, Jews, and Muslims, in as much as all agree that TB is
true, accept the same model that accounts for how it is that TB could be
externally and internally rational and warranted for them in a basic way.
Christians, Hindus, Jews, and Muslims will have disagreements, yes, but at
the level of abstraction appropriate to the Standard A/C model, there are
none, and so it is reasonable to think that they all affirm essentially the same
model.
Once they accept the Standard A/C model, there is no reason for nonChristian theistic philosophers not to develop their own “Reformed” epistemologies in a way that is consistent with and their particular scriptures and
9. Perhaps it is less clear how the wider Hindu tradition could be made consistent with TB.
However, a careful study of the Dvaita-Vedānta philosophical theologian, Madva, shows this is
indeed possible. Madva affirms that Atman (Self) is distinct from Brahman (Ultimate Reality),
and argues that individual souls (jivas or selves) are never fully absorbed into Brahman, and
hence keep their ontological distinctiveness. He also posits sources of knowledge of God similar to Plantinga’s. Madva’s views, while not universally accepted among Hindus, are not uncommon. Elements of his thought are evident in other threads of Hinduism that are also “friendly”
to their own A/C model extensions, such as the Bhakti and Vaisnava traditions. See Srendranath
Dasgupta, A History of Indian Philosophy, vol. 4, Indian Pluralism (Cambridge: Cambridge
University Press, 1955).
ERIK BALDWIN
387
traditions.10 Accordingly, Jews, Muslims, and Hindus who accept Plantinga’s
epistemology can affirm that their own extensions of the Standard A/C model
could have the same epistemic virtues that CB is alleged to have for
Christians. And so we see that the Standard A/C model can have at least three
extensions in addition to CB, extensions that cover Jewish Belief (JB),
Islamic Belief (IB), and Hindu Belief (HB).11 Thus (2) has at least four
instances:
(2CB) CB can be internally and externally rational and warranted for the
Christian in a properly basic way.
(2JB) JB can be internally and externally rational and warranted for the
Jew in a properly basic way.
(2IB) IB can be internally and externally rational and warranted for the
Muslim in a properly basic way.
(2HB) HB can be internally and externally rational and warranted for the
Hindu in a properly basic way.
I have argued that Christians, Jews, Muslims and Hindus can agree on
similar stories that account for how it is that TB is internally and externally
rational and warranted for them in a properly basic way. But if the Standard
A/C model is true it follows that at most only one of its possible extensions
could be true, since Christian, Jewish, Islamic, and Hindu statements about
God cannot be true concurrently.12 The reason for this is that Christians, Jews,
Muslims, and Hindus accept different and incompatible statements about the
person and work of Jesus, Moses, Muhammad, and Krishna. These religious
traditions have different views about God and his dealings with humans and
accept the authority of different and incompatible holy books and scriptures.
Since God cannot have contradictory or incommensurable properties or
reveal contradictory statements of truth, it follows that God cannot have all
of the properties that Christians, Jews, Muslims, and Hindus claim him to
have where their statements about God really conflict. And yet it is not difficult to see how CB, JB, IB, and HB could be internally rational for those
who affirm them. Surely, if the Standard A/C model is true, God creates
10. Plantinga is aware that his epistemology and his A/C model may be appropriated by nonChristian faiths. He writes, “For religions sufficiently analogous to Christianity, it is possible
that religious believers have access to internally available markers, evidence, phenomenology
and the like for their creedal beliefs, just as Christians have internally available markers, evidence, phenomenology and the like for their creedal beliefs” (WCB, 350). Plantinga gives us no
indication that he finds this problematic in the way that I will show it to be.
11. There very well may be many more possible extensions of the Standard A/C model. On
page 350 of WCB, Plantinga concedes that some forms of Buddhism and some forms of
American Indian religions may have their own extensions of the Standard model. Adding to that
list, it seems to me that one could add some forms of Taoism and Confucianism to the list as
well. I must limit myself to considering just a few possible extensions in this paper.
12. I assume a realist conception of truth, according to which ascribing properties to God is
not radically different than it is for other beings and that religious language and statements in
scripture are coherent and meaningful.
388
PHILOSOPHIA CHRISTI
humans such that their proper doxastic response to experience is to believe
the target propositions of the correct extension. And yet if TB is true, concerning theists who accept TB in the way the Standard A/C model describes,
then only one extension from among CB, JB, IB, and HB could possibly be
externally rational and warranted for theists belonging to these traditions.
Consequently, only theists who accept the true extension of the Standard A/C
model will have external rationality and warrant conferred on their creedalspecific beliefs about God. For these reasons, I conclude that the following
proposition is true:
(3) If the Standard A/C model is true, at most one creedal-specific
extension from among HB, CB, IB, and JB is true; therefore at most
only one extension from among HB, CB, IB, and JB is such that it
is internally and externally rational and warranted for believers.
Plantingans, those who accept Plantinga’s account of warrant and his
Standard and Extended A/C models, surely exist. I have shown how Jewish,
Islamic, and Hindu philosophers can accept Plantinga’s account of warrant
as proper function and Plantinga’s Standard A/C model but accept their own
unique extensions of the Standard model. So, in a sense, I have shown that
al-Plantingas, Al Plantingachandrans, and Al ben Plantingans, proponents of
the Islamic, Hindu, and Jewish extensions of the Extended A/C models,
respectively, could exist. For heuristic purposes, it is helpful to personify
Plantinga’s “Reformed” comrades and dub them al-Plantinga, Al
Plantingachandra, and Al ben Plantinga.13
Plantinga and his comrades agree on many points of theology.
Specifically, they agree that God exists, that God has certain attributes, that
humans acquire knowledge of God in similar sorts of ways, and the like.
They all accept TB and believe that God created humans such that they have
a cognitive faculty (or belief forming mechanism or process) the function of
which is to cause people to accept TB in a properly basic way. They also
agree that the truth of the Standard A/C model (or something very much like
it) accounts for how it is that TB has internal and external rationality and
warrant for theists.
Muslims who accept the Standard A/C model are not likely to disagree
with Aquinas and Calvin concerning TB but they will look to philosophers
in the Muslim philosophical tradition, such as al-Farabi, al-Kindi, al-Ghazali,
Ibn Taimiyyah, and Ibn Hazm, in order to show how it could be that humans
come to hold those creedal-specific beliefs about God unique to Islam in a
way that is consistent with Plantinga’s epistemology.14 It is not difficult to see
13. Thanks to members of the Reformed Epistemology list serve (http://groups.yahoo.com
/group/reformed-epistemology) for suggesting these colorful names for Plantinga’s counterparts.
14. See M. M. Sharif, A History of Muslim Philosophy (Kempton, Germany: Allgauer
Heimatverlag Gmh, 1963), 1:136–155 (Book 2, chapter 7, “Philosophical Teachings of the
Koran”).
ERIK BALDWIN
389
how resources from these philosophers, as well as the Qur’an, the Sunna
(Muhammad’s words and conduct), the Hadith (traditions about
Muhammed’s words and conduct), and the Sharia (Islamic law) can be utilized by Muslims to formulate a uniquely Islamic extension of the Standard
A/C model.15 Likewise, Jewish philosophers can look to Judah Halevi and
Saadya Gaon, or to contemporary Jewish philosophers such as Abraham
Heschel, in order to formulate their own uniquely Jewish extension of the
Standard A/C model.16 Hindu philosophers can draw on many sources that
would give them reason to affirm the essential elements of the Standard A/C
Model as well as their own uniquely Hindu extension of it. They can look to
philosophers belonging to the Nyāya-Vaiśşesika school, such as Gańgeśa,
Guatama, or Raghunātha.17 Perhaps the most promising source for the Hindu
is Madva, a philosopher belonging to the Dvaita-Vedānta tradition.18
While I have not actually developed Jewish, Muslim, and Hindu extensions of the Standard A/C model, I have indicated sources that provide one
with a sufficient idea of how such alternative extensions might be formulated. But it is sufficient for my purposes to show that Jewish, Islamic, and
Hindu philosophers could agree with Plantinga’s epistemological methodology but not on its substance. Thus, I have shown that non-Christian theists
could accept Plantinga’s contention that Aquinas and Calvin are right about
how, in general, it is that humans come to form creedal-specific beliefs about
God in a properly basic manner, but that he is wrong about which of these
creedal-specific beliefs about God are in fact true.
Ironically, the success of the Standard A/C model makes it possible for
it to have multiple extensions. Since we know that if the Standard A/C model
is true then it has at most only one true extension, we also know that all but
one of its possible extensions will be false. Even if we restrict ourselves to
four possible extensions, the antecedent probability that a given extension of
the Standard A/C model is false is relatively high. By “antecedent probability” I mean to pick out the probability that a given extension of the Standard
A/C model is true prior to “looking at the world” in order to either procure
reasons or evidence that supports the truth of one particular extension or in
order to procure reasons or evidence that counts against the truth of other of
15. Some contemporary Islamic philosophers are engaged in a project similar to this. See
Maulana Muhammad Ali, The Religion of Islam: A Comprehensive Discussion of the Sources,
Principles, and Practices of Islam (Columbus, OH: Amadiyya Anjuman Isha’at Islam, 1990);
Bakhtiar Husain Siddiqui, Knowledge: An Islamic Perspective (Washington, DC: Publications
of the Council for Research in Values and Philosophy, 1998); and Absar Ahmad, “Exploring
Islamic Theory of Knowledge,” http://tanzeem.org.pk/resources/articles/articles/absar2.htm.
16. For example, see Saadya Goan’s The Book of Doctrines and Beliefs (Indianapolis, IN:
Hackett, 2002) and Abraham Joshua Heschel’s book God in Search of Man, (New York: Farrar,
Straus, and Giroux, 1955).
17. See the Encyclopedia of Indian philosophies, vol. 1, Bibliography, 3rd rev. ed., ed. Karl
H. Potter (Delhi: Motilal Banarsidass, 1995).
18. See Srendranath Dasgupta, A History of Indian Philosophy, vol. 4, Indian Pluralism.
390
PHILOSOPHIA CHRISTI
its extensions. If a person S is aware of these antecedent probabilities, then,
whichever extension of the Standard A/C model S accepts, S has a reason for
thinking that the model she accepts is probably false, which provides S with
an epistemic defeater for thinking that the target propositions of whatever
extension she accepts are true. This defeater is defeated only if S has some
reason for thinking that the model she accepts is true.19 At this stage, which
extension of the Standard A/C model from among (2CB), (2JB), (2IB), or
(2HB) actually is true becomes relevant. However, Plantinga does not provide
(in WCB) an account or defense of why one should think that Christianity is
the correct extension of the Standard A/C Model, neither does he offer reasons why one should think that its other possible extensions are false. And so
it would seem that Plantingans, at least those that are aware of the argument
developed in this paper, have a defeater for thinking that CB is the correct
extension of the Standard A/C model. In the next section I develop this
defeater in further detail.
A Probabilistic Defeater Based on Multiple
Extensions of the Aquinas/Calvin Model
I think the above argumentation supports the contention that the following propositions are true:
(4) If x is the number of creedal-specific religious traditions that can
appropriate Plantinga’s Standard A/C model and formulate their
own extensions of it, the antecedent probability that the target
propositions of a given creedal-specific extension of the Standard
A/C model, m (CB, JB, HB, IB, and so on), is true is 1/x. [P(m) =
(1/x)].20
(5) For any filling of x greater than 2, the antecedent probabilities are
such that it is more probable than not that the target propositions of
m are false.
That (5) is true is evident. I have already provided preliminary, general reasons for thinking that (4) is true. But the truth of (4) is controversial, and so
further support is in order. Before I offer additional arguments in favor of (4)
it is necessary first to clarify its meaning.
19. S might have some reason for thinking that the other extensions are probably false which
in turn provides S with a reason for thinking that the extension she accepts is probably true. But
this is just a special case of having a reason for thinking that one’s own extension is probably
true.
20. In this essay I assume that the antecedent probability of each possible extension of the
Standard A/C model is equal. Whether the antecedent probabilities of each extension are in fact
equal and what sorts of considerations might raise or lower the antecedent probability of a given
extension are important matters that I plan to consider in more detail in a future paper.
ERIK BALDWIN
391
The expression “P(m) = (1/x)” indicates the antecedent probability that
m, a given extension of the Standard A/C model, is true is equal to 1 divided
by however many extensions of the Standard A/C model comparable to the
Christian extension there are. Accordingly, the expression indicates the likelihood that m is the correct extension of the model prior to considering reasons, in the form of arguments or justifications based on one’s background
beliefs k and evidence base e, that render m more likely to be true relative to
other, comparable extensions of the Standard A/C model.21 (4) takes account
of the fact that since non-Christian theists can extend the Standard A/C
model to cover their own creedal-specific beliefs about God, since only one
such extension can be true, the more extensions of the Standard A/C model
there are, the greater the antecedent probability that a given extension of the
A/C model is false.
Now that the meaning of (4) has been made clear, it is time to explain in
greater detail how accepting its truth provides one with a defeater for thinking that a given extension of the Standard A/C model is true. Let S stand for
a person who accepts Plantinga’s Standard A/C model and some extension of
it who is also aware of the fact that other religious traditions can extend
Plantinga’s Standard A/C model to cover other, mutually exclusive creedalspecific beliefs about God. S knows that since Jewish, Muslim, and Hindu
philosophers may appropriate Plantinga’s epistemology, the Standard A/C
model, and their own unique extensions of the Standard A/C model, S knows
that the extension she accepts might very well be false. S will therefore realize that the antecedent probability of the extension of the Standard A/C
model that she affirms is low. Because of this, S has a reason to doubt
whether her creedal-specific beliefs about God are true. Accordingly, S has a
reason for thinking that her creedal-specific beliefs about God fail to be internally rational for her, and thus has reason to think that her creedal-specific
beliefs about God lack warrant. If S knows all of this and is such that she is
internally rational, S will wonder which extension of the Standard A/C model
is true; it will matter to S a great deal whether S’s claim that she has warranted creedal-specific beliefs about God is accurate. For S to continue to accept
her particular extension of the Standard A/C model knowing that it is more
likely than not to be false is clearly not internally rational. In order for S’s
creedal-specific beliefs to once again be internally rational for her, S must
have some reason for thinking that the extension of the model she accepts as
true is in fact true. If S does not have such a reason then the extension of the
Standard A/C model that she affirms remains unwarranted for her. And so S’s
awareness of the fact that there can be multiple extensions of the Standard
A/C model provides her with an internal rationality defeater for believing
21. Factoring in evidence base e and background knowledge k, we can express the probability of m thus: P(m|ek) < P(~m|ek).
392
PHILOSOPHIA CHRISTI
that the extension of the Standard A/C model she accepts is true, which renders S’s creedal-specific beliefs about God unwarranted.
The reader might not yet be convinced that my argument is sound.
Hence it is worthwhile to motivate its conclusion in another way. Consider
the following story. Suppose that you and five of your friends, all of you honest and trustworthy people who accept Plantinga’s proper function account
of warrant, play a game involving dice. You roll one of the die. Each of you
looks and sees a different number on its face. Now, no one disagrees that
there is a die, that it is white and has numbers on its surfaces, that the die was
cast, or anything like that. You are all quite perplexed, but since you all know
each other to be honest and trustworthy, you rule out the possibility that one
of you is somehow tricking or deceiving the others. Something is obviously
wrong here; everyone knows that the number perceived to be on the surface
of the die is probably not the actual number. In these circumstances none of
you has a good reason to think that your number beliefs are true, and so none
of you would have a warranted belief about what number is showing on the
surface of the die. But since all of you are proper functionalists about warrant, one from among you might be inclined to think something along these
lines: “If my cognitive faculties are functioning as they ought and nothing
else is amiss, then I know the number that is showing on the die, in which
case my friends are surely mistaken and have unwarranted beliefs. So, it’s
not unreasonable for me to believe that my number beliefs might be warranted after all.” But this sort of response is unacceptable. The conditional is true,
but the truth of the conditional is not at issue. What is at issue is the truth of
its consequent. However, for one to fall back on the truth of the above conditional amounts to little more than wishing that its antecedent is true with
the hope that the truth of its antecedent secures the truth of its consequent.
However, even if the antecedent is true, even if you are the lucky one whose
faculties are functioning properly and is unaffected by whatever is going
awry in your epistemic environment, your belief about which number is
showing on the die will nevertheless be unwarranted. The reason for this is
that you have a good reason to think your number belief is false and to continue to believe that it is true in this case is not internally rational. And since
your number belief is not internally rational it follows that it is unwarranted.
A relevantly similar story can be told about a conversation between Al
Plantinga, al-Plantinga, Al Plantingachandra, and Al ben Plantinga concerning whose creedal-specific beliefs are warranted. Suppose this group of
esteemed colleagues meet for dinner and during conversation realize they
affirm the Plantinga’s proper functionalism but accept different creedal-specific beliefs about God. Each thinks that he accepts the correct extension of
the Standard A/C model. During the course of their conversation they realize
that the antecedent probabilities of their respective extensions of the
Standard A/C model are such that it is quite probable that their own exten-
ERIK BALDWIN
393
sion is false. Thus, each sees that at most only one of them has creedal-specific beliefs about God that are true. These Plantingan friends can not assume
the others are lying or untrustworthy, and so something is surely amiss.
Everyone knows that the antecedent probability that one’s own extension of
the Standard A/C model is true is relatively low, and so each of them knows
that (setting aside reasons for thinking that one’s own model is true) that
one’s creedal-specific beliefs about God are probably false. Thus, Al
Plantinga, al-Plantinga, Al Plantingachandra, and Al ben Plantinga each have
an internal rationality defeater for thinking that the target propositions of
their respective models are true. It follows that the target propositions of their
respective models are not warranted for any of them, not even the lucky
Plantingan who is internally and externally rational, whose faculties are
functioning as they ought in an environment for which they were designed.
Clarifications, Objections, and Replies
I have argued that S has a reason for thinking that the extension of the
Standard A/C model she accepts is probably false and that S must adequately address this defeater if her creedal specific beliefs about God are to be
warranted for her again. Serious objections to my argument attack premise
(4). They either attempt to show that my argument fails to generate a defeater
or maintain that the defeater can be adequately dealt with in an epistemically basic way. In this section I focus my attention on objections of these two
types.
Perhaps, some time after becoming aware of my defeater S goes to
church (or mosque, synagogue, or temple) and once again finds herself forming basic belief about God. Suppose that while participating in the services,
S becomes psychologically certain that God is speaking to her in that
“creedal specific” way and is thereby convinced that she has a sufficient reason or ground for thinking that the defeater in question has been defeated. S
might think this on account of the fact that she has experiences that reassure
her that the creedal-specific beliefs about God she holds are true. According
to this objection, the warrant that S has for her creedal-specific beliefs in
these circumstances overpowers or outweighs the defeater that is generated
from the fact that there can be multiple extensions of the Standard A/C
model. Perhaps S can become so assured that her creedal-specific beliefs are
true in such circumstances that the argument in this paper fails to motivate
her to accept the defeater developed in this essay. Thus, although S has not
offered evidence or reasons for thinking that m is true, so the objection goes,
S has no reason to suppose that her creedal-specific beliefs about God are
internally irrational and so there is no reason for S to think they are unwarranted.
394
PHILOSOPHIA CHRISTI
However, if S is reassured of the truth of the target propositions of her
accepted model in this way S does not adequately address the defeater developed in this paper. This is because S’s psychological certitude that God is
speaking to her in that “creedal specific” way is not the same as being epistemically certain of that proposition. The mere fact that she is psychologically certain that her model is true gives her no epistemic reason for thinking
that she accepts the correct extension of the Standard A/C model. According
to Plantinga’s epistemology, our design plan is not aimed at psychological
certainty but truth. And so unless S has a sufficient reason to think that the
model she accepts is true S will continue to have a defeater for thinking that
her creedal-specific beliefs about God are warranted. To be clear, Al
Plantingans, al-Plantingans, Al ben Plantingans, and Al Plantingachandrans
can all go about reassuring themselves that their creedal-specific beliefs
about God are true by going to church, or what have you. But since all of
them can go about this procedure equally or inscrutably well, no progress
against the defeater is made, for S will nevertheless lack a sufficient reason
for thinking that her creedal-specific beliefs about God are (probably) true,
and so the defeater based on multiple extensions of the Standard A/C model
remains undefeated.
Another strategy is suggested by the last paragraph of Warranted
Christian Belief where, regarding CB, Plantinga writes: “But is it true? This
is the really important question. And here we pass beyond the competence of
philosophy, whose main competence, in this area, is to clear away certain
objections, impedances, and obstacles to Christian belief.”22 Following
Plantinga’s lead S might object that even though there are multiple extensions of the Standard A/C model, it is “beyond the competence of philosophy” to provide reasons for thinking that the target propositions of one’s own
preferred extension of the Standard A/C model are true. S could go on to suggest that the appropriate way to handle the defeater in this essay is to fall
back on the epistemic possibility that one’s own model is true and maintain
that the fact that there are multiple extensions of the Standard model does not
provide one with a sufficient reason to think that a given extension of it is
probably false. But S cannot just say this, for that would be simple gainsay.
S must reasonably motivate or ground this claim. Perhaps S can do that by
arguing that the defeater does not arise in the first place, perhaps by maintaining that the arguments in this paper are not strong enough, or that they
are not sufficiently convincing. S might maintain that even though she has
read and understood the argument in this essay her creedal-specific beliefs
about God continue to seem true to her and that if they are indeed true they
are warranted. Consequently, S will continue to think that her creedal-specific beliefs about God are warranted even though she is not able to provide an
argument that they are true or provide reasons for thinking that the
22. Plantinga, WCB, 499.
ERIK BALDWIN
395
antecedent of this conditional is true. The objection here, then, is that my
defeater does not really give a committed Plantingan any reason at all to
think that their creedal-specific beliefs about God are unwarranted and so it
is not contrary to internal rationality to think that one’s creedal-specific
beliefs about God are true. And, as the objection goes, if these beliefs are
true, then they are warranted.
But this response is unconvincing for several reasons. The above
response relies on the conditional truth of one’s creedal-specific beliefs about
God. It claims that the defeater in this paper is not successful and that if one’s
creedal-specific beliefs about God are true then they are warranted. But the
truth of this conditional is not at issue. What is at issue is whether its
antecedent is true, whether or not S’s beliefs about God are in fact true. If S
has no reason for thinking that the antecedent of the conditional is true, then
S has no reason for thinking its creedal-specific beliefs are true. At any rate,
this strategy does not successfully address the defeater articulated in this
essay and S still has a reason for thinking that her creedal-specific beliefs
about God are false. For S to continue to claim that if her beliefs about God
are externally rational and warranted for her without offering a reason for
thinking they are true is an inadequate response to the defeater in question.
Again, S might not be psychologically motivated to accept the epistemic
defeater developed in this essay, but that does not thereby show there to be
some defect in my argument.
A Plantingan might object that a person of faith would not accept (4)
because religious persons are not (or ought not to be) neutral with regard to
whether their beliefs about God are true. Only persons with no beliefs at all
(or very weak beliefs) about which extension of the Standard A/C model is
true would (appropriately) accept (4). After all, when exposed to (4), Hindus,
Jews, Christians, and Muslims (if they are true Plantingans) are likely to
think that God is producing in them their creedal specific beliefs about God
all the same, and they are likely to think that it is a lapse of faith to give up
their beliefs about God so easily. Hence, so the objection goes, it is appropriate for committed believers not to accept (4) on account of the fact that it is
more appropriate for them to accept their creedal specific-beliefs instead.23
I concede that persons of faith may judge it to be appropriate to accept
their beliefs about God in a basic way when they consider (4), and that they
would think that their doing so is a mark of faithfulness. Nevertheless, the
considerations I have offered in favor of (4) call into question whether one’s
basically formed beliefs about God are properly (that is, epistemically) basic.
I have shown that given the truth of the Standard A/C model it is more likely than not that a particular extension of it is false. This probability judgment
is not subjective, but objective; it is not based on what S is willing to or psychologically able to accept, nor is it based on what S feels like or is accus23. William Hasker raised this objection via personal correspondence.
396
PHILOSOPHIA CHRISTI
tomed to accepting or anything of the sort. Surely, faith in God is not merely a matter of being faithfully committed to God; faith is not just emotive or
psychological. Faith in God also contains a belief, or cognitive, component.
And S knows that, given the truth of the Standard A/C model, the creedal specific beliefs she accepts as properly basic are more likely than not to be false.
Hence, S knows that the antecedent probability that CB (or HB, IB, or JB) is
true is 1/4. Accordingly, it is irrational (contrary to internal rationality) for S
to have faith that CB (or HB, IB, or JB) is true in the cognitive sense. S might
psychologically resist this fact. It may be rational and proper for S to go to
church (or temple, mosque, or synagogue) and go on as if CB (or HB, IB, or
JB) were true. Even so, S’s beliefs will not thereby be properly basic or warranted. S cannot insulate or inoculate herself from the epistemic defeater in
(4) simply by supposing that or acting as if CB (or HB, IB, or JB) is true; she
cannot rationally or properly discharge this epistemic defeater by making an
“epistemic leap of faith.” No, (4) gives S a reason for thinking that her faith
in God in the cognitive sense is misplaced, even if (4) does not eliminate S’s
practical or emotive reason to continue to maintain rational faith in God.24
To be clear, I freely admit that persons of faith can have certain experiences that provide grounds for properly basic beliefs about God. Such experiences may have associated with them “internally available markers” that
provide the basis for epistemically basic beliefs about God. But, because it is
more likely than not that these markers do not track truth given that there are
multiple extensions of the Standard A/C model, I have given a reason for
thinking such markers do not epistemically support S’s beliefs about God in
the way Plantinga suggests they do. Hence, S has a good reason to think that
beliefs based on these markers are false. Accordingly, it would seem that in
order for S’s beliefs about God to once again be warranted, S must have
some reason for thinking that these internally accessible markers track truth,
which requires evidence or argumentation.
24. One might object that my argument assumes that religious pluralism is a problem for
Plantinga’s model. But, as Plantinga has argued in WCB, 437–57, the mere fact of religious pluralism does not give the Christian a reason to think her beliefs about God are not warranted.
Thus, continues the objector, the fact that various religions can claim extensions of the Standard
A/C model (that differ in some respects and agree in others) isn’t a problem for Plantinga’s epistemology. I agree that the mere fact of religious pluralism does not generate a defeater for TB.
However, Muslim, Jewish, and Hindu philosophers who accept Plantinga’s proper functionalism and the Standard A/C model can all rightly claim that the fact of religious pluralism does
not generate a defeater for them, too. None of this is problematic. But what generates the
defeater at issue in my essay is that multiple theistic traditions can make use of Plantinga’s
Standard and Extended A/C models apparently equally well. Surely, that there are multiple
extensions of the Standard A/C model is not the same thing as the mere fact of religious pluralism, and so my objection cannot be dealt with in the same manner as can defeaters based on the
mere fact of religious pluralism.
ERIK BALDWIN
397
Conclusion
I think I have sufficiently supported premise (4) of my argument. Hence,
it is reasonable to suppose that if S affirms some extension of the Standard
A/C model m, understands (1)–(5), and is internally rational, then S will
admit that she has an (undermining) internal rationality defeater for the belief
that the target propositions of m are externally rational and warranted for her.
If m is to be warranted for S, then S must defeat the defeater in question. In
order to defeat this defeater S must find some reason for thinking that m is
the true extension of the A/C model (or at least have reasons for thinking that
the other extensions of it are false). But, as I have shown, this implies that S
must have some reason, by way of evidence and argumentation, for thinking
that m is true. That, however, is contrary to the spirit of Plantinga’s A/C models. According to the models S is not supposed to need evidence or arguments
for her beliefs about God to be warranted. Accordingly, I think the following
proposition is (probably) true:
(6) If S is subject to the defeater based on multiple extensions of the
Standard A/C model, for the target propositions of m to be warranted for S, S must have propositional evidence and argumentation that
renders the target propositions of m more likely than not to be true
given S’s evidence base e and background knowledge k.
I do not claim that there are no strategies that could provide S with a
properly basic reason to favor one extension of the A/C model rather than
another, but I think that the prospects are quite slim. I cannot argue for that
point in this paper exhaustively, but a bit of reflection will reveal to the reader that any strategy implemented by the Plantingan that allegedly provides an
epistemically basic reason for thinking that her own extension of the
Standard A/C model is true can be appropriated by al-Plantingans, Al
Plantingachandrans, and Al ben Plantingans equally well. It is easy enough
to see how this would go. Think of any argument that shows how a
Plantingan might have an epistemically basic reason for thinking that her
accepted model is in fact true. Al-Plantinga, Al Plantingachandra, and Al ben
Plantinga could utilize the same formal structure of that argument but in
doing so make different substantive claims. But which substantive claims
about God are true? That question cannot be ignored or set aside. And it is
not plausible to maintain that determining which substantive religious claims
are in fact true is a properly basic matter. Accordingly, there is reason enough
to maintain that if nonbasic reasons for thinking that one’s substantive religious claims are true are disallowed, then Plantinga and his comrades will
not have a sufficient reason for thinking that their own extensions of the
Standard A/C model are true. The defeater in this paper cannot be defeated
given the resources internal to the A/C models.
398
PHILOSOPHIA CHRISTI
Of course, my defeater is not indefeasible. I fully admit that evidence or
arguments that provide S with reasons for thinking that m is true may adequately deal with the defeater in this paper. But Plantinga’s Extended A/C
model claims that the target propositions of m are epistemically basic for S.
And so if S affirms the target propositions of m on the basis of e and k then
the target propositions of m are not properly basic for S. Thus, S is faced with
a dilemma: either S appeals to propositional evidences in order to deal with
the defeater developed in this paper or S does not. The following proposition
describes one horn of the dilemma:
(7) If S does not appeal to propositional evidences and argumentation
in order to determine whether the target propositions of m are internally and externally rational and warranted for her, then the target
propositions of m lack internal rationality for S and are unwarranted for S even if m is true.
Alternatively, S can look for evidence and argumentation that give her a
reason to think that m is the correct extension of the A/C model. This second
horn of the dilemma is captured by the following proposition:
(8) If S does appeal to propositional evidences and argumentation in
order to determine whether the target propositions of m are internally and externally rational and warranted for her, then if the target
propositions of m are warranted for S, they will no longer be epistemically basic for S, which conflicts with the A/C models.
Should S prefer the horn of the dilemma expressed by (7), then her
creedal-specific beliefs about God will be unwarranted and if S prefers the
horn of the dilemma expressed by (8), her creedal-specific beliefs about God
will be warranted but they will not be basic. Presumably, S would prefer that
her creedal-specific beliefs about God are warranted rather than unwarranted, and so S should take the way out of the dilemma described in (8). This
option is expressed in the following proposition:
(9) If S appeals to propositional evidences and argumentation in order
to defeat this defeater, then the target propositions of m are warranted for S but they are not warranted for S in the way that the A/C
models propose.
If (9) is true, then (10) follows:
(10) Any given extension of the Standard A/C model cannot account for
how it is that the target propositions of m are internally rational and
warranted in an epistemically basic way for S—whether or not m is
true.
To be clear, (10) is true on account of the fact that if S’s creedal-specific
beliefs about God are warranted, the way in which they are warranted is not
as Plantinga’s models suppose and the target propositions of m will not be
basic for S. And this will be so no matter which extension of the Standard
A/C model is in fact true. Thus, any particular instantiation of m from among
ERIK BALDWIN
399
HB, JB, CB, or IB that S accepts will fail to be internally rational and warranted in a basic way. The most interesting case, of course, regards CB. Thus,
from (10) we conclude (11):
(11) The specifically Christian extension of the Standard A/C model cannot account for how it is that the target propositions of CB are internally and externally rational and warranted in a basic way for the
Christian—whether or not Christianity is true.
Quite unexpectedly, granting that Plantinga’s Standard A/C model is true
the fact that there are multiple extensions of the A/C model, severely undermines Plantinga’s claim that CB is epistemically basic for many Christians
in the way Plantinga proposes. And so it seems that I have revealed a severe
tension in Plantinga’s epistemology. Internally rational agents look for evidence when appropriate. It is appropriate for S to look for evidence that m is
the correct extension of the A/C model given my defeater. But the A/C models rule out using evidence in order to determine the correct extension of the
A/C model. But it seems that the only option internally rational theists have
at this juncture is to go ahead and rely on evidence. Hence, it seems that an
internally rational agent who accepts Plantinga’s epistemology and has been
exposed to the defeater developed in this paper will conclude that if CB (or
IB, JB, or HB) is warranted for her then it is not epistemically basic. But then
that means Christians (or Jews or Muslims or Hindus) in this position must
either reject or modify the core aim of both the Standard and Extended A/C
Models—that one’s creedal-specific beliefs about God (CB, IB, JB, or HB)
are (or can be) warranted for theists in an epistemically basic way.
In closing, if S is aware of the defeater in this paper and lacks a reason
for thinking that her accepted extension of the A/C model is correct then S
has an undefeated defeater for thinking that her creedal-specific beliefs about
God are true. This defeater can only be defeated if S has a reason for thinking that she affirms the correct extension of the Standard A/C model.
Presumably, in some sense it is ‘up to S’ what to do about this matter, but I
think that it is clear that the rational thing for S to do is to go ahead and find
reasons for thinking that the extension of the Standard A/C model she accepts
is the true one.25
25. I would like to thank Charles Hughes for our many helpful discussions about the material in this paper. Without his help, his mentoring, and his encouragement this paper would not
have been possible. I would like to thank Robert Audi, Michael Bergmann, Nathan Carter,
William Hasker, Daniel Frank, Myron Penner, William Rowe, James Sennett, James Stump,
Michael Sudduth, Michael Thune, and Chris Tucker for making helpful comments and suggestions on various versions and/or sections of this paper and/or the arguments therein. I would like
to offer special thanks to Alvin Plantinga for his writings and for discussing with me issues relevant to my argument and to Richard Swinburne, from whose writings I have learned so much.
I would also like to thank an anonymous reviewer at Philosophia Christi for making some helpful suggestions.