OVERVIEW
The Rationality of Theistic Belief:
Classical Evidentialism and Reformed
Epistemology
The revival of philosophy of religion in Anglo-American philosophy during the last
15 to 20 years has brought with it a renewed interest in the epistemology of religious
belief, broadly understood as a consideration of the justificatory and epistemic
conditions for beliefs the content of which entails the existence of God. What is required
to have a rational or justified belief in God? And what constitutes
religious knowledge? This renewed and vital
interest in re-examining the rationality of belief in God has developed within the larger
context of general epistemology, the nature of rationality and knowledge itself. Seen in
this way the topic of the justification of theistic belief particularizes a broader
interest among many philosophers to analyze and clarify such concepts as justification,
rationality, and knowledge, as well as the relations between them.
I. Two Views on the Rationality of
Religious Belief
A. Evidentialism and Reformed
Epistemology
For much of the history of Western philosophy since the 17th century, the
“rational” status of theistic belief has been closely tied to the possession (or
availability) of rational evidence in the form of (theologically neutral) reasons or
arguments which provide some degree of support for religious and theological propositions.
This tradition of classical evidentialism has,
up through the first half of the 20th century, presented a basic challenge to theists, or
more precisely, a challenge to those theists who believe or claim that their belief in God
is a rational or justified belief. This challenge, roughly put, states: [1] a person is rational or justified in believing
in God only if the person has adequate evidence for his belief in God, where this evidence
consists of other rational beliefs or knowledge of the person and is religiously neutral
evidence (or at least is traceable to evidence which is). This challenge, the so-called evidentialist challenge
to theistic belief, has been the cognitive staple of most modern Western intellectuals
since the Enlightenment. They have taken it, as nearly axiomatic, that the satisfaction of
the evidentialist requirement embodied in the evidentialist challenge is essential to the
rational justification of belief in God. Furthermore,
some have argued the additional thesis: [2]
there is no adequate evidence for belief in God. The conjunction of [1] and [2]
constitutes the evidentialist objection to theistic belief.
Traditionally, philosophical theologians and Christian apologists have responded to
the evidentialist objection to theistic belief by trying to argue that [2] is false on the grounds that there is adequate evidence for the existence of God.
Theistic philosophers have presented a host of theistic arguments: from the existence of a
contingent, complex physical system (the Universe), temporal and spatial regularities
within such a system, consciousness, morality, and religious experience. These arguments
have ranged from strictly deductive arguments (generally proceeding from self-evident or
otherwise evident-to-the-senses principles or ordinary observation) to inductive or
probabilistic arguments (often times with more liberal premises) to show that the
evidentialist challenge can be met head-on by presenting an evidentialist apologetic. For
such philosophers natural theology has been the primary source for answering the
evidentialist challenge (and objection) to theistic belief.
One important approach in contemporary
religious epistemology, and indeed one that has often dominated discussions in the
philosophy of religion for nearly two decades, is Reformed
epistemology. The movement receives its
name - even if an infelicitous one - because of its affinity with claims about the nature
and status of theistic belief and religious knowledge in the Reformed theological
tradition originating with John Calvin. The key philosophers in this movement - William
Alston, Nicholas Wolterstorff, and Alvin Plantinga - have in their own way each developed
the claim, usually taken as the central thesis of Reformed epistemology, that belief in
God can be rational without being based on or otherwise supported by propositional
evidence or argument. More precisely put,
there are some people, at some times and under certain conditions, for whom belief in God
is rational even though their belief in God is not based on evidence in the form of
reasons (i.e., other beliefs or knowledge) which provide adequate support for the belief.
For some people theistic belief is a properly
basic or immediately justified belief.
Reformed epistemologists have challenged the epistemological assumptions of Enlightenment
evidentialism and have presented an alternative way of thinking about the rationality of
religious belief.
As for the conditions under which
theistic belief is rational without propositional evidence, different answers have been
given depending on the particular epistemological framework. Some accounts hold that
individuals are prima facie justified in
holding theistic belief just if they do not have or should not have adequate reasons for
not holding the belief (Wolterstorff 1983a). This is sometimes based on a principle of
rationality that if it seems (epistemically) to a person that X is the case, then (barring special
considerations) it is probably the case that X
(Swinburne 1991, pp. 254-76). This principle
of credulity embodies the idea that our beliefs are innocent until proven guilty. The
contention that belief in God may be immediately justified, though, has received special
impetus from and developed in close connection with the rise of externalist theories of
justification and knowledge. According to the externalist it is possible for something to
confer justification or positive epistemic status without the subject being aware upon
reflection (or even being capable of becoming aware upon reflection) that it possesses
this justificatory or epistemic efficacy. (Internalism
by contrast requires cognitive accessibility to the justifier or justifier’s
efficacy). If a belief’s being the
output of a reliable process of belief formation is sufficient for its being justified or
known, a person may justifiably believe or know theistic propositions if that reliability
condition is satisfied. This will be the case even if the person has no cognitive access
to the faculty responsible for the belief or the reliability of such a faculty (Alston
1991b,e, 1993c; C. Stephen Evans 1995). Sometimes, though, theistic belief has been viewed
in a way analogous to sensory perceptual experience, as having a cognitively accessible
ground in the form of an experiential awareness of God (i.e., God’s being directly
presented to a person’s consciousness), even
though the epistemic adequacy of such a ground is not cognitively accessible (Alston
1991c). Other accounts, while not denying the grounding of belief in God in religious
experience, suggest that belief in God may be immediately rational in the sense that
memory or a priori beliefs are, perhaps as the
result of the proper functioning of one’s cognitive faculties. (Plantinga 1987,
1991).
B. The Religious Epistemology of Alvin
Plantinga
In “Reason and Belief in God” (1983a) and subsequent articles Alvin
Plantinga has presented a systematic exposition and critique of the epistemological
foundations of classical evidentialism, and he has articulated an alternative conception
of the rationality of belief in God - the so-called thesis of proper basicality. As he
sees it, evidentialism is rooted in classical foundationalism, a normative view
about the structure of belief according to which one’s rational beliefs divide into
those which are based on other beliefs and those which are properly basic. The latter are
beliefs not based on other beliefs but nonetheless rational by virtue of being
self-evident, about one’s immediate introspective experience, or evident to the
senses. Since belief that God exists satisfies none of these criteria for proper
basicality, its rationality requires that it be based on beliefs which (i) provide
adequate (deductive or probabilistic) evidential support for it and (ii) where these
supporting beliefs are either themselves properly basic or ultimately based on beliefs
which are properly basic. Moreover, according to Plantinga evidentialism is also rooted in
deontologism.
Historically (from John Locke onwards) the normativity involved in the
evidentialist’s concept of rationality has typically been the normativity of duty and
obligation. A person who believes in God without propositional evidence is somehow
violating an epistemic duty or intellectual obligation. Alternatively, the normativity
involved in the evidentialist position could be (and has been on occasions) thought of as
the normativity associated with the proper functioning of one’s cognitive faculties,
rationality in the sense of freedom from epistemic defect or cognitive malfunction. On
this construal, basic theistic belief turns out to be epistemically defective and so lacks
a kind of epistemic excellence.
Plantinga argues that propositional evidence for theistic belief is not needed for
either epistemic dutifulness or epistemic nondefectiveness. Classical foundationalism
fails to account for the rationality (in either normative sense) of a broad range of our
ordinary, everyday beliefs (e.g., memory beliefs, beliefs in other minds); for such
beliefs are neither properly basic by the classical foundationalist's criteria, nor as
they entailed by beliefs that are properly basic. Similarly the very principle of
classical foundationalism itself does not satisfy the conditions of proper basicality, nor
is it entailed or rendered probable by beliefs that are properly basic. On the first
count, classical foundationalism is false; on the second, self-referentially incoherent.
This philosophical critique of the epistemological framework of classical evidentialism
constitutes Plantinga’s challenge to the evidentialist challenge to theistic belief.
In this way, Plantinga’s anti-evidentialism is parasitic on the recent collapse of
classical foundationalism within general epistemology and the rise of modest versions of
foundationalism. These modest versions of foundationalism do not restrict the class of
properly basic beliefs to those possessing various kinds of epistemic immunities (from
revision, error, or the possibility of doubt), thereby allowing a broader range of
foundational beliefs (often along the externalist lines mentioned above). Plantinga’s
positive position is a version of foundationalism in which belief in God can properly
belong to the foundations of one’s structure of beliefs. To be more precise,
Plantinga argues that [3] some people, at
certain times and under certain conditions, have a rational belief in God (and by
extension other religious beliefs) even though they do not have (or base their belief on) propositional evidence in the form of
(theologically neutral) rational beliefs or knowledge which provide deductive or
probabilistic evidential support for their belief in God, and even if such evidence is not
available in their community.
When Plantinga speaks of theistic belief or belief in God being properly basic, he
is actually thinking of beliefs like, God is forgiving me, God created this, God is
speaking to me, etc, all of which self-evidently entail God’s existence. I will speak
of these theistic beliefs targeted by Plantinga as various beliefs that Pt. Secondly, as for the conditions in
which basic theistic belief is rational, Plantinga emphasizes that these beliefs that Pt,
though not based on reasons, are nevertheless not groundless. He implements the Reidian
distinction between propositional and nonpropositional grounds (or evidence) to urge that
beliefs that Pt are grounded in appropriate widely realized experiential conditions which
trigger the formation of the belief. They are in that way epistemically analogous to
paradigmatic cases of properly basic belief, such as sensory perceptual or memory beliefs,
which have rationality or justification conferring conditions. Thirdly, the criteria for proper basicality must
be arrived at broadly speaking inductively, working from particular cases of beliefs which
we believe are both basic and rational. Such paradigmatic cases of properly basic beliefs
will include sensory perceptual beliefs (e.g., I see the tree), memory beliefs (e.g., I
had breakfast this morning), and beliefs about other persons (e.g., that man is angry).
Plantinga claims that, like these beliefs, belief in God can be (at least for some people
under certain conditions) a properly basic belief. In other terms, his argument may be
stated in terms of showing that there exists a certain parity between theistic belief and other
paradigmatic cases of properly basic beliefs.
Although Plantinga’s central epistemological thesis began as a statement about
the conditions required for being justified in
believing in God, and where “justification” is understood in a deontological
sense to refer to certain epistemic duties, his religious epistemology has (since around
1987) taken a different direction with his development of the highly externalist theory of
warrant, set forth with considerable detail in his 1993 works Warrant: The Current Debate and Warrant and Proper Function, and to be applied to
theistic and Christian belief in Warranted Christian
Belief (forthcoming). Unlike
“justification” Plantinga takes “warrant” to be that quality enough of
which is sufficient (or nearly so) to transform true belief into knowledge. Warrant involves the proper functioning of human cognitive equipment,
the functioning of our belief-forming and belief-sustaining powers, faculties, or
mechanisms in the way they were designed to function, by God or evolution - or both. It
also involves the proper functioning of our cognitive faculties in a specific context: the
appropriate cognitive environment (one sufficiently similar to that for which a
person’s faculties have been designed). Moreover, bearing in mind the distinctly
“epistemic” aim in warrant, given that different parts of our cognitive make-up
have different purposes, warrant must have regard for the proper functioning of a
particular segment of the design plan governing the production of the belief, that part
aimed at the production of true beliefs. Not only so, but the epistemic aim involved in
warrant entails a reliability constraint on warranted beliefs. The cognitive module
parceled out for the production of the belief in question must have a high objective
probability of producing true beliefs when one’s cognitive faculties are functioning
properly in the appropriate conditions. Lastly, the more
firmly a person believes a proposition, the more warrant it will have for him.
The notion of a human design plan has taken Plantinga’s epistemology well into
metaphysics and theology. According to Plantinga the de
jure question about Christian belief is a question about whether Christian belief is
sensible, reasonable, justified or rational. But the answer to this question depends on
what sort of person one thinks human beings are, what sorts of beliefs their noetic
faculties will produce when they are functioning properly. “The dispute as to whether
theistic belief is rational,” writes Plantinga, “can’t be settled just by
attending to epistemological considerations; it is at bottom not merely an epistemological
dispute, but a metaphysical and theological dispute” (1993d, p. 31). Ultimately, what
sorts of beliefs are taken as properly basic will depend on what sort of creature a person
takes human beings to be, and so depends on one’s metaphysical assumptions. Whereas a
naturalist or atheist will tend to require reasons for theistic belief to have warrant, a
theist may indeed have a quite different account of warranted theistic belief. Drawing on
John Calvin and several thinkers within the Reformed theological tradition, Plantinga
argues that, on a theistic view, we may easily think of humans as created in such a way
that they have a natural tendency to form belief in God on the grounds of widely realized
experiential conditions. Plantinga interprets Calvin’s talk about a sensus divinitatis (as well as the inner testimony
of the Holy Spirit) in terms of immediate belief-forming mechanisms - modes of forming
belief in God (and other Christian beliefs) in a basic way. So basic theistic belief can
be warranted and constitute knowledge. Moreover, the account also suggests (as I shall
argue in chapters 3 and 4) that [4] basic
theistic belief is (typically at any rate) epistemically superior to nonbasic theistic
belief by virtue of the theistic design and the degree of warrant conferred on basic
belief in God.
C. The Compatibilist Approach
Although the responses to the claims of Reformed epistemology have been diverse,
there has been a two-fold tendency in the literature critical of the movement's claims,
and especially the claims of Plantinga. First, there has been widespread debate over
whether theistic belief can ever be properly basic (Quinn 1985, 1993, Goetz 1983, Grigg
1990, Meynell 1993). Secondly, many have accepted the Plantingian proper basicality claim
but have sought to modify either the notion of properly basic theistic belief or
evidentialism (or both) in order to bring the two positions into closer relation (Mavrodes
1983, Wolterstorff 1986, Greco 1993, Garcia 1993, Lee 1993, Zeis 1993, McLeod 1994). The
first approach I will call an incompatibilist
account of proper basicality and evidentialism; the second approach, compatibilist Reformed epistemology. Although I
think the arguments against proper basicality are interesting (and some of them quite
challenging to Reformed epistemology), I intend to leave that question out of the picture
and focus rather on the second approach that I think is (at least potentially) far more
philosophically interesting. Moreover, if adequately unpacked, the second approach will
render the critiques of proper basicality fundamentally unnecessary and their motivation
adequately answered. What I set out to do in
this thesis is establish just such a compatibilist theory. I will argue that there is a
form of evidentialism that is compatible with the claims of Alvin Plantinga’s
religious epistemology. This is, of
course, only a first approximation. The plausibility of compatibilism rests on two points:
(1) the explicit and implicit propositions which constitute Plantinga’s religious
epistemology and (2) the nature of evidentialism. Neither of these is easily spelled out,
and therefore considerable effort will be made to analyze both Plantinga’s claims
(and their implications) and the nature of the evidentialist requirement. And the latter
point is just as crucial to the argument as the prior, for not just any form of
evidentialism will do. Some versions to be examined will be shown to fall short of being
epistemically adequate. What must ultimately be argued is that there is an epistemically adequate form of evidentialism that
is compatible with the religious epistemology of Alvin Plantinga.
Moreover, the thesis claim is to be developed within the framework of a modest
version of epistemic foundationalism. Inasmuch as the discussion of religious epistemology
presupposes an array of issues in general epistemology, chapter 1 sets out the basic
epistemological concepts and distinctions that will be employed in the thesis. My goal in
chapter 1 is not to develop a rigorous epistemological theory but to articulate some of
the general features of the epistemological terrain to be traversed in the course of the
work, together with some critical commentary. One
caveat though. Since my aim is to argue for the compatibility of evidentialism and
Plantingian epistemology on the basis of foundationalism, I have chosen to omit discussion
of coherentism, except as it relates directly to epistemic foundationalism. First, space
constraints require delimiting the scope of epistemological discussion to what is most
relevant to the core arguments of the thesis. Secondly, evidentialism has usually been
associated with foundationalism, as has Reformed epistemology. Also, I do not find
epistemic coherentism very plausible. Regardless
of how one spells out the coherentist position, it faces the trilemma of infinite regress,
inappropriate detachment from reality, or vicious circularity. Also, it seems to me that
theistic belief does just as well as any other doxastic candidate on epistemic
coherentism. And although I have not been convinced by those epistemological theories
which attempt a union of foundationalism and coherentism, I will have recourse at points
to pointing out the ways in which coherence may reasonably enter into a foundationalist
account.
The question is, minimally, to what extent the conjunction of something like [3] and [4]
is compatible with some set of evidentialist requirements {R1, . . . ,Rn},
and where each member Ri of the set is severally necessary and jointly
sufficient for an epistemically adequate form of foundationalist-evidentialism.
II. Bi-Level Evidentialism: A Sketch
Although I am specifically interested in Plantinga’s religious epistemology,
the central thesis to be argued will be based on evidentialist requirements for belief
which admit of broader epistemological relevance and application. In one respect the
project seeks to establish evidentialist conditions for theistic belief in a way
compatible with externalism, though many of my observations will have relevance for purely
internalist accounts of justification and/or knowledge. The broader applicability of my
arguments is developed in chapter 8. Among the items I take to be essential to my
evidentialist case are: defeating conditions, overdetermining and partial sustaining
reasons, belief formation and sustenance, conditions for higher-level justification, and
the evidentialist implications of the noetic effects of sin for the design plan and proper
function. Here I briefly sketch the more
significant facets of the arguments to be developed involving these topics.
A. Defeaters, Partial Basicality, and Overdetermination
A staple of contemporary epistemology is the distinction between prima facie and ultima facie justification. The justification S has
for his belief that p may be overridden by reasons to the contrary, and where
“reasons” are understood broadly as inclusive of either experiences or beliefs
(and on some accounts conditions or states of affairs to which the subject has no
introspective access). More precisely, it is recognized (even on many externalist accounts
of justification and knowledge) that justification or positive epistemic status may be
defeated by (i) reasons for regarding a belief as false (rebutting defeaters) or (ii)
reasons for regarding the grounds of a belief as inadequate (undercutting defeaters).
According to some accounts a belief is ultima facie
justified just if it based on an adequate ground and nothing in the totality of a
person’s noetic structure serves as a rebutting or undercutting defeater (or perhaps
that no such defeaters are the sort of thing a person could come to have fairly readily
upon reflection). Although it is common to think of “justification” as what gets
defeated, the idea of proper function yields another sort of defeater that looms large in
Plantinga’s more recent work. These are so-called rationality defeaters, where rationality is
understood in the sense of proper function. Here defeaters (specifically, other beliefs or
experiences of the form (i) or (ii)) defeat a belief
in the sense that they constitute a reason for S to modify his existing noetic structure
by the deletion of a previous belief or by simply holding a belief less firmly. S may
rationally believe that p at time t1, but then acquire a defeater D at time t2.
What this means is that if S is rational (i.e., functioning properly) at time t2
S will either no longer believe that p or believe that p with a less degree of firmness at
t2 than S did at t1.
Two significant consequences follow from this for the rational and epistemic status
of theistic belief. As with beliefs in
general, so with theistic belief in particular - there are putative defeaters. Among these
are: (a) the atheological argument from evil, (b) the projective theory of theistic
belief, and (c) the case for the incoherence of theistic propositions. Given that
Plantinga holds that the more firmly S believes B the more warrant B has for S (assuming
that the other conditions of warrant are satisfied), partial defeaters may reduce
one’s firmness of belief. As a consequence, the degree of warrant had by B is
insufficient for knowledge. In such a case I suggest that for some people reasons may play
a partial causal role in a person’s continuing to be epistemically warranted in their
theistic belief. These reasons may take the form of either undercutting or rebutting
defeater-defeaters (i.e., either reasons to think that the defeater against theism is
inadequately supported or false). The latter obviously translates into reasons for
believing that there is such a person as God, and so natural theology becomes important to
a person’s remaining epistemically warranted in his theistic belief. Whether a
rebutting or undercutting defeater-defeater is required on some particular occasion is a
thorny issue that I will tackle in terms of considerations from subjective and epistemic
evidential probability (e.g., how a person weighs the evidence, his view of relevant prior
probabilities, as well as how strong the argument is given correct inductive standards).
Secondly, since the proper functioning of one’s relevant cognitive faculties
is necessary for warrant and knowledge, this includes the necessity of the proper
functioning of one’s defeater system. This gives rise to what can be called the no-defeater condition: roughly stated, S’s
belief that p is warranted only if S does not have a defeater for the belief that p. Given
the acquisition of a defeater for theistic belief, it may be that a person is no longer
rational in believing some theistic proposition (or at least not believing it with the
same degree of firmness as before the acquisition of the defeater). What is required in
such an instance for a person’s being rational (and so knowing the theistic
proposition) is a defeater-defeater. Although theistic belief may, by virtue of its degree
of warrant as basic, act as an intrinsic defeater-defeater for various putative defeaters,
I present conditions under which this will not be the case. Typically, a defeated theistic
belief will need an extrinsic defeater-defeater. Although Plantinga has argued that only
undercutting defeater-defeaters would be required in such instances, I defend some
scenarios in which rebutting defeater-defeaters will be required. Moreover, Plantinga has
claimed that in the case of undercutting defeater-defeaters, though they may be required
as part of the warrant conferring circumstances, theistic belief need not be based on such considerations (and perhaps should
not). Here I argue that the notion of partial causal sustenance brings intelligibility to
the idea of a person’s basic belief being partly based on reasons in the form of
undercutting defeater-defeaters.
The function of reasons in supplying a person with grounds for remaining
epistemically warranted in their theistic belief suggests that overdetermination plays an important epistemic
role, where this overdetermination is either psychological (refers to the causal source of
a belief) or epistemic (refers to what confers positive epistemic status). In cases where
S’s theistic belief is causally sustained by multiple sufficient grounds, or where S
has other beliefs which are potential grounds for theistic belief, S has epistemically
relevant cognitive resources for continuing in a
state of epistemically warranted theistic belief given the existence of putative defeaters
for theistic belief. To highlight how the importance and even necessity of propositional
evidence may emerge for a person at different points in their cognitive history as a
result of relevant changes in their cognitive circumstance, I present a case for a
two-fold evaluation of noetic structures. The first considers the epistemic status
theistic belief has at some particular time, either the time of its acquisition or
maintenance. The second considers a belief's epistemic status charted through time (a kind
of history of a noetic structure). I will refer to the former as synchronic epistemic status; the latter as diachronic epistemic status. I conclude that there
is a case for regarding noetic structures that are propositionally overdetermined with
respect to basic theistic belief as diachronically epistemically superior to their
non-overdetermined counterparts (where the history of a noetic structure includes
defeating conditions against which basic theistic belief is not an intrinsic
defeater-defeater). Moreover, I will argue that in certain defeating conditions (minus
intrinsic defeater-defeaters) a noetic structure in which theistic belief is partly basic
and partly nonbasic is synchronically epistemically superior at tn (with
respect to theistic belief) to noetic structures in which theistic belief is based solely
on an immediate source at tn. Carefully
spelling out belief and rationality defeating conditions C*, I conclude that for some
people at certain times and under certain circumstances that include C*, their belief in
God is rational (and constitutes knowledge) only if their belief in God is at least
partially causally sustained by reasons in the form of either rebutting or undercutting
defeater-defeaters.
B.
The Evidentialist Implications of Hamartic Cognitive Malfunction
The second major plank in my evidentialist case draws on the notion of cognitive
malfunction, which figures prominently in Plantinga’s epistemological theory. One of the ways in which a belief fails to have
warrant (and so fails to constitute knowledge) is by arising from cognitive faculties
which are not in proper working order, which are subject to malfunction or dysfunction. On
a Christian theistic metaphysics cognitive malfunction is the result (indirectly at any
rate) of sin in the human personality. Within the Reformed theological tradition there has
been much emphasis on what the older Princeton theologians of the 19th and early 20th
century called the noetic effects of sin, a doctrine which has frequently been used to
question either the propriety or usefulness of natural theology. What I will call the
hamartic grounds of cognitive malfunction is simply shorthand for the conjunction of the
principles of the noetic effects of sin and the epistemological theory of proper function.
The central cognitive module of immediate or basic theistic belief formation in Plantinga
is the sensus divinitatis. One of my
evidentialist lines of enquiry (in chapter 5) maps out some plausible ways the sensus divinitatis may be thought to malfunction as
a result of sin. Such an account will show that for some people, whose circumstances
include the exemplification of any one of the several cognitive malfunctions to which the sensus divinitatis is subject, propositional
evidence will be necessary (to varying degrees) if such people are to have epistemically
warranted belief in God. A case for the
existence of other theistic relevant cognitive modules that generate or sustain theistic
beliefs is presented. Even though these mediate sources are subject to similar cognitive
malfunction (as a result of sin) the justification of mediate warrant for some people
rests on a principle of the variable instantiation of the noetic effects of sin. According
to this principle, different cognitive systems will be adversely affected by sin in
different ways and to varying degrees. But unless we postulate mediate sources as well as
immediate sources for theistic belief, it will be hard to account for the fact that
despite the noetic effects of sin, there is genuine natural knowledge of God possessed by
many people. The conclusion is somewhat
unorthodox and may appear somewhat startling: if
man is totally depraved propositional evidence for belief in God is necessary, for some
people will not be able to form epistemically warranted belief in God (or even mere belief
in God) unless they have reasons, such as those supplied by natural theology. Ultimately,
Plantinga’s design plan theory must account for the damage done to the human
cognitive system, and as I shall argue this requires a distinction between a pre-lapsarian
and post-lapsarian design plan. Such an account will impose a restricted strong
evidentialist requirement according to which for some people propositional evidence is
necessary, or necessary and sufficient, for epistemically warranted theistic belief.
C.
Higher-Level Evidentialism and the Pursuit of Reflective Rationality
The evidentialist requirements that develop out of the preceding discussion
contribute to what I will call modest
evidentialism. In chapter 6 I continue my focus on epistemic foundationalism in order
to develop another evidentialist requirement to complete the case for an epistemically
adequate version of evidentialism. Drawing on the multi-level foundationalism of William
Alston, I present an important modification to the theistic foundationalism hitherto
developed, a modification which will allow, among other things, an important distinction
between putative belief in God and belief in the rationality or justification of belief in
God. Even if a person's belief that p is immediately justified (and let us suppose only susceptible to an immediate justification),
this does not rule out finding reasons for the higher-level belief that one is so
justified. Moreover, according to Alston, the only mode of justification for higher-level
beliefs is mediate, they must be based on reasons if they are to be justified. This gives
rise to the strong higher-level evidentialist
requirement. The satisfaction of such a requirement leads to what I will be calling reflective rationality. After articulating the
multi-level scheme, I consider both internalist and externalist versions of reflective
rationality.
In Part III of chapter 6 I develop the implications of higher-level evidentialism
for assessing the relevance of a project such as natural theology. Although the arguments
of chapters 4 and 5 established the kinds of conditions under which natural theology will
be necessary and/or sufficient for epistemic warrant, considerations drawn from reflective
rationality will show that arguments from natural theology are, when appropriately
developed, essential to achieving a form of propositional reflective rationality. Although
one may be reflectively rational about one’s own theistic belief by rationally
considering the adequacy of one’s actual ground for believing it, one may also
exhibit reflectively rationality about a particular theistic proposition (typically the proposition that God
exists) by rationally considering the adequacy of the evidence for that proposition (and
so what would constitute an adequate ground for theistic belief is one were to hold it on
that basis). The basic distinction here falls on differentiating between the kind of
evidence e which provides evidential support for p and the kind of (logical) evidence e*
which provides support for judgements about the force, weight, or adequacy of e for p, as
well as the kind and degree of investigation and critical assessment which goes into
arriving at such judgements. The distinction is developed with discussion of the relevant
categories of epistemic and subjective probability. My conclusion is that since
propositional evidence is necessary for reflective rationality, the resources of natural
theology become crucial for satisfying such an epistemic desideratum, thereby contributing
toward a Reformed conception of fides quaerens
intellectum.
In chapter 7 I focus on the relation between justification at the higher and lower
levels, specifically how being justified in higher-level beliefs contributes to
justification at the lower level. A fairly sophisticated picture of the kind of epistemic
merit achieved by reflective rationality arises from this. I also consider the epistemic
merit of higher-level justification. Even if such a requirement exists, it is important to uncover its epistemic
significance. I suggest that the internalist intuitions which loom large in the
evidentialist tradition are fairly well satisfied at the higher-level in being justified
in believing that one is justified in believing that p. And here one must always have
reasons for belief. Presupposing that individuals are typically critically reflective (or
at least ought to be), and frequently assuming the Cartesian thesis about the mind’s
transparency to itself, the evidentialist tradition has failed to distinguish between
unreflective rationality (having a rational belief that p) and reflective rationality
(having a rational belief that one's belief is rational). Consequently, requirements for
the latter have been made requirements for the former. Although there is no strong
lower-level evidentialist requirement for theistic belief, I shall argue for the
restricted necessity of reflective rationality for one’s being justified in holding
theistic belief(s) given some of the defeating conditions canvassed in the earlier
chapters. This analysis further suggests that the evidentialist tradition has failed to
distinguish between the conditions required for being appropriately related to the
epistemic goal (of acquiring true beliefs and avoiding false beliefs) and those for being
a responsible seeker after such an epistemic state, especially in the presence of certain
defeating conditions.
The conjunction of modest evidentialism with respect to theistic beliefs at the lower level and strong higher-level evidentialism constitutes an epistemically adequate version of evidentialism which is compatible with Plantinga’s religious epistemology - what I will call Bi-Level Evidentialism. Although the evidentialist requirements developed in chapters 4 through 7 are closely tied to Plantinga’s religious epistemology, chapter 8 offers, by way of summary, an account of bi-level evidentialism in which my evidentialist requirements are given broader application in the form of a version of theistic foundationalism of more general epistemological appeal. Bi-level evidentialism establishes that there can be no real Reformed “objection” to natural theology or evidentialism, if the claims of proper basicality are judiciously balanced with a proper conception of the function of reasons with respect to religious belief. Such a conclusion does violence neither to proper evidentialist desiderata nor the sensus divinitatis or the testimonium spiritus sancti so much a part of Reformed epistemology. Moreover, if the arguments developed here are sound, then it follows that the highly externalist features of Plantinga’s epistemology, so frequently taken to marginalize “reasons for belief,” are compatible with the sort of internalist intuitions which gave rise to classical evidentialism in the first place.