Mikael Stenmark on rationality in science, religion and everyday life

Rationality in Science, Religion,
and Everyday Life
Part three of MA Course Knowledge,
Rationality, and Society
Emanuel Rutten
Literature and Schedule
• Literature
– Mikael Stenmark, Rationality in Science, Religion, and Everyday Life
(University of Notre Dame Press: Indiana 1995), Chs. 1-10
• Schedule
Tue Nov 25: Chs. 1-3
Thu Nov 27: Chs. 4-5
Tue Dec 2: Chs. 6-8
Thu Dec 4: Chs. 9-10
• Links to Slides & Questions
– Twitter: @emanuelrutten
Chapter 1: Introduction
Belief formation, revision and rejection
• Part of being human is to have the ability to form, revise and reject various
believes. Our cognitive capabilities include belief regulation processes
• When it is that people form, regulate and reject their believes in a proper,
responsible, or reasonable way? What are the conditions for rational belief?
• Are these conditions the same for all areas of human life (science, religion,
everyday life, etc.)? Are they the same for all cultures? And for all times?
• In order to identify these conditions we have to clarify what rational belief
actually is, and whether it is the same for all areas of life, cultures and times
• The notion of rationality does not apply only to beliefs. Human decisions
and actions can be rational or irrational as well. Yet, our focus is on belief
Science, religion, and everyday life
• Science is taken by many as a paradigm example of rationality. Any model
of rationality must deal with theory acceptance and refutation in science
• Religion originates in existential experiences of joy and suffering, meaning
and alienation, guilt and liberation, etc. We may ask what the conditions
are for reasonable beliefs about the ultimate, the absolute, the sacred
• Everyday life is not an optional area and in it we have clearly the most of
our beliefs. It is therefore also a paradigm case for models of rationality
Theoretical, practical and axiological rationality
• What kinds of ‘things’ can be rational or irrational?
– Beliefs, decisions, actions, behaviors, evaluations, plans, strategies, people …
• What kinds of ‘things’ are typically a-rational?
– Trees, planets, cars, phones, tables, chairs, taste (e.g. pizza, ice-cream), art …
• There are three areas where we can decide what to do. But then there
are three contexts of rationality (Rescher)
Theoretical rationality is about what we should believe or accept
Practical rationality is about what we should do or perform
Axiological rationality is about what we should value or prefer
• Axiological rationality is required since decisions of people can be rational
only in case the ends or aims they try to achieve is in their real interest
Axiom of reasonable demand
• Rationality has to be realistic in the sense that it cannot require more than
what the supposed agent can possibly be expected to do (given the agents
actual ‘resources and circumstances’ or ‘constitution and predicament’).
• In short: ‘ought’ implies ‘can’ (Kant)
• Yet, many models of rationality are about perfect agents. They introduce
agents as purely theoretical abstract constructions (ideal epistemic beings)
• Therefore, many models of rationality are too idealized or utopian to apply
in an interesting and relevant way to actual limited human beings like us
– They imply that we are usually irrational in most things we do, which is an
absurd consequence. Models of rationality must take into account what we
as real human beings can reasonable do and what is not in our power to do
Three ways to construct models of rationality
• But are empirical aspects to be taken into account when proposing models
of rationality? Is it not enough to focus on strictly conceptual and logical
matters? (This question can in fact be asked of any philosophical issue)
• There are three philosophical methods (for creating models of rationality)
The formal approach
The practice oriented approach
The contextual approach
• The first and third are situated at each side of a spectrum. The second
is an inter-mediating position between both extremes
The formal approach (purely a priori)
• Formulation of a model of rationality can be done wholly independently
from the actual practices by applying strikt conceptual and logical tools
• Pure conceptual and logical enquiry has complete authority over the
practices themselves. The actual practices are strictly speaking irrelevant
• The constructed model of rationality has to be appropriate for any
reasonable being and no matter what practice this being is involved in
• In fact, the constructed model of rationality has to apply for all possible
worlds, that is to say, no matter what the world is like
• No examination of (the history of) a practice is required to recommend an
adequate model of rationality for that practice. One model applies to all
• Practices that do not satisfy the pre-established universal and ahistorical
model (standards, criteria) for rationality are refuted as being irrational
The contextual approach (purely empirical)
• Formulation of a model of rationality is totally dependent on the actual
practices. Only an examination of these practices is needed. Nothing else
• The actual practices constitute themselves the only justification for what
is and what isn’t rational within them. External criteria are wholly rejected
• Contextualists merely explicate the standards of rationality that are already
given within the actual practices. They are their own ultimate authority
• The standards for rationality are thus wholly internal to the actual
practices. They are in other words practice-determined
• The standards for rationality can and do in fact significantly differ between
the various actual practices (such as between science and religion)
• Even the nature of rationality (the very meaning of rationality) likely differs
between practices. Rationality just is not the same for all actual practices
The practice-oriented approach (both)
• There is a basis independent of each practice from which we can assess its
rationality and propose rationality criteria for it (contra contextualism)
• It is possible to critically establish from a purely philosophical point of view,
from the “outside” of a practice, standards of rationality for that practice
• Yet, it is required to have substantial contact with each of the practices. We
must take into account its aims, activities and situation (contra formalism)
• The proposed model of rationality for a practice can only be relevant for
a practice if we understand its means and ends. Its function and nature
• We must be sensitive to what is going on in a practice, what one tries to
achieve and what is achievable. This places constraints on proposed models
• The proposed standards for rationality can and do in fact differ between
various actual practices. Yet, its very nature or meaning is likely universal
Rationality and religion
• A key issue is the applicability of rationality to religion. In order to address
this topic properly we need to extend the scope, namely to the applicability
of standards of rationality to both religious and secular views of life
• The discussion of the rationality of religious belief is not only about whether
religious belief meets the criteria or standards of rationality, but also about
what standards of rationality are the approprate ones to use in this context
• The goal here is not to establish whether some view of life is rational, but to
identify the proper standards for assessing the rationality of views of life
• Various models of rationality are investigated in order to assess which of
them is the most appropriate to use for religious belief. We thus examine
the conditions for a discussion on the rationality of religious belief
• One option is the scientific challenge to religious belief. It has it that all
religious beliefs must conform to the rationality standards of science
• A related challenge is evidentialism. It has it that religious beliefs are
rational only to the extend that there is sufficient evidence for them
Models of rationality
• A model of rationality consists of [i] an account of the nature of rationality
together with [ii] an account under what conditions something is rational
(the standards or criteria of rationality)
• The following models can be distinguished
Formal evidentialism – There is one nature of rationality and one set of
standards for rationality. This model is related to the formal approach
Social evidentialism – There is one nature of rationality but different standards
for different practices. This model is related to the practice-oriented approach
Presumptionism – There is one nature of rationality but different standards for
different practices. This model is related to the practice-oriented approach
Contextualism – There are different natures of rationality and different standards
for different practices. This model is related to the contextualist approach
• For social evidentialism and presumptionism one sometimes says that there
is one set of standards (because standards here are still structurally similar)
Chapter 2: The Nature of Rationality
The Nature of Rationality
• Wat is rationality? What does it mean to say that something is rational?
Is it possible to provide a general characterization of rationality?
• Without an answer to this question it will be difficult or even impossible to
identify proper standards (criteria, conditions) for when a belief or action
within a certain practice is rational
• But is there a single meaning of rationality? Perhaps rationality has
different meanings in different domains? In that case it isn’t univocal
Descriptive incommensurability thesis – Rationality has different meanings
in different domains
Normative incommensurability thesis – Rationality should have different
meanings in different domains
Four different aspects of rationality
• We mentioned rationality’s nature and standards. But there are four aspects
The nature of rationality – What is rationality? What is its meaning?
The standards of rationality – What are the conditions for being rational?
The reasons of rationality – What may count as rational evidence?
The aims of rationality – What are the ends or goals of rationality?
• The difference between standards and reasons can also be understood as a
difference between generic and detailed standards. If a practice-oriented
philosopher holds that standards may change between practices, he or
she acutally means the detailed standards and not the generic ones
• Examples of (generic) standards
One must always have sufficient evidence for one’s beliefs (evidentialism)
One may hold-on to one’s beliefs as long as they aren’t refuted (presumptionism)
Four different aspects of rationality
• What counts as rational reasons in one practice (e.g., religion) may not count
as rational reasons in another practice (e.g., science) because these practices
may have different aims. (Thus the fourth aspect is actually about practices)
• Moreover, differences in rational reasons between practices do not always
entail differences in standards between practices. And differences between
standards between practices do not always entail differences in nature
• When we ask for the nature, standards, reasons and aims of rationality, we
refer solely to normative rationality and not to generic rationality
Generic rationality – refers to the capacity for reasoning. An agent has generic
rationality if it has the resources of reason. It thus can be rational or irrational. An
entity without generic rationality is a-rational. It cannot be rational or irrational
Normative rationality – An agent having generic rationality is normatively rational
(hereafter: rational) just in case it excercises its reason properly or responsibly. A
belief or action is normatively rational if and only if it is or can be justified
Deontological rationality
• Any characterization of the nature of rationality needs to take into account
that rationality is a normative concept (‘ought to’) and a term of appraisal
(‘positive attitude, approval’). One way of characterizing it is deontological
• According to the deontological account rationality consists in the fulfillment
of certain duties (responsibilities, obligations) in respect to what one is doing
(e.g., believing, acting).
• In the context of belief (‘theoretical rationality’) these responsibilities are
called epistemic responsibilities. It leads to an ethics of belief
• The epistemic duties are prima facie and not ultima facie. They can be
overridden by other non-epistemic duties in special circumstances
• But what do these duties consist in? This brings us back to the level of
(generic) standards of rationality. Examples include evidentialism (never
belief things without good evidence for their truth) or presumptionism (one
may hold-on to beliefs provided there is not sufficient counter-evidence)
Means-End rationality
• Another characterization of the nature of rationality is means-end rationality.
According to it rationality consists in the efficient pursuit of given ends. There
is thus nothing that is rational per se. Rationality is always goal-relative
• Rationality is about how to achieve our aims as efficient as possible. It does
not tell us what these aims should be. It is about means and not about ends
• To assess whether a certain decision is rational, we not only have to know
people’s goals, but also their resources and situation (r- and s-relativity)
• According to the objective account of rationality what we do is rational
if it is probable that it will satisfy our goals in an efficient manner
• According to the subjective account of rationality what we do is rational
if it seems or appears to us probable that it will satisfy our goals efficiently
– In that case rationality becomes also relative to the information had by the agent
A puzzle concerning rationality
John has been in a state of severe anxiety for years now. Recently he obtained
strong evidence that he is going to die if he does not get rid of his anxiety.
He also has strong evidence that the only way for him to acquire peace of mind
is to believe that the universe will exist forever.
Suppose that John has sufficient evidence for the claim that the universe will not
exist forever. Still, in an ultimate attempt to save his life he starts to try to force
himself psychologically to believe - contrary to the evidence - that the universe
will exist forever.
After two weeks he finds himself in a state of believing that the universe will
exist forever. As a result he obtains the desired peace of mind and thus loses
his anxiety. In this way he manages to save his life.
Now, is John's belief that the universe will exist forever rational?
Can both characterizations be combined?
• The characterization of theoretical rationality as an epistemic duty does in
fact assume a goal: to believe many truths and eliminate many falsehoods
• But then the deontological characterization for theoretical rationality can
be reprashed as follows. People have the responsibility to pursue efficient
means for satisfying their epistemic goals
• In addition to epistemic goals there are also non-epistemic goals (such as
well-being and survival). Epistemic rationality involves only epistemic
ends and non-epistemic rationality involves only non-epsitemic ends
– Are usefulness, simplicity and predictability epistemic goals or not?
• Note that epistemic rationality is not the same as theoretical rationality and
that non-epistemic rationality is not the same as practical rationality (why?)
• The deontological and means-end notions of rationality can be combined:
We have a duty to pursue our epistemic and non-epistemic goals efficiently
– Science and religion have complex (both epistemic and non-epistemic) ends
Some Examples
Mary believes in quantum mechanics because she thinks that there is
good evidence for its truths and her aim is to get in touch with reality
This is an instance of theoretical and epistemic rationality
John believes that the universe will exist forever because it gives him
peace of mind and his aim is to get peace of mind
This is an instance of theoretical and non-epistemic rationality
Brigitte performs a daily walk because it helps her to get new research
ideas, which is very important for her since she wants to study the world
This is an instance of practical and epistemic rationality
Dave engages in sports at his university because it keeps him healthy,
which satisfies his goal of well-being
This is an instance of practical and non-epistemic rationality
Epistemic and non-epistemic reasons
• What these examples show is that we can have epistemic and non-epistemic
reasons for what we believe or for our acts because we can have epistemic
and non-epistemic ends for believing and acting
A good epistemic reason for a belief or act connects the belief in an appropriate
way to the epistemic goals that are presupposed (e.g. believing only truths)
A good non-epistemic reason for a belief or act connects the belief in an
appropriate way to the presupposed non-epistemic goals (e.g. peace)
• If the goal is complex (i.e., consists of both epistemic and non-epistemic
ends), agents are to take into account epistemic and non-epistemic reasons
Towards a holistic rationality
• The combined notion of rationality (“a duty to establish efficient and
sufficient means for our ends”) is still too narrow. For the assessment
of appropriate ends is still out of scope. It is still only about means
• Our only duty is to find efficient means for whatever we want. Reason
does not tell us where to go, it only tells us how to get there. It is the
slave of the passions (Hume)
• Such a limited conception of rationality goes strongly against our intuitions.
How could someone who efficiently chooses his or her total ruin be rational?
We thus need to be able to rationally assess and choose our goals as well
• Plausibly, meaningless, worthless or destructive ends are not rational. Ends
that are not in people’s real or best interests are not rational either. And
the same holds for ends that are contrary to our real needs
• Reasoning about ends is thus closely related to reasoning about values
and preferences. We therefore need to bring in axiological rationality
Towards a holistic rationality
• On a more inclusive conception of rationality a rational person has a duty
to choose appropriate or valuable ends (ends that are in his or her best
interest) and to find sufficient and efficient means to achieve them
• But phrased in this way, the conception is still too narrow. For efficient
means that are destructive, contrary to our real needs or contrary to our
best interests are not rational either. Our means must also be appropriate
– The axiological dimension is thus relevant for our ends and our means
• On the holistic conception of rationality we have a duty to choose valuable
or appropriate ends (ends that satisfy our real needs) and to find sufficient,
efficient and appropriate means for achieving these ends
• The holistic conception of rationality has both a theoretical-practical side or
dimension (conditional: establishing proper means-end connections) and an
axiological dimension (categorical: establishing appropriate or valuable ends)
Chapter 3: Science and Formal
The formal approach to scientific rationality
• The best place to start for introducing formal evidentialism is with science
since most if not all adherents of formal evidentialism take science to be
the paradigm of rationality. So let’s look at how formalists view science
• The formalist holds that philosophy of science is the study of the logical
properties of and the logical relations between scientific propositions
or collections of propositions (theories). It is primarily logical analysis
• The idea is that there is a ‘logic’ underlying the methods of science and
the task of the philosopher of science is to explicate this logic a priori
– Logicality thesis: The rationality of science amounts to a logical system
• There is no need to study actual scientific theories. Simple generalizations
like ‘All swans are white’ properly represent the logical structure of theories
• According to this view philosophy of science is the study on how a rationally
ideal scientist should proceed and what he or she should rationally accept
Scientific evidentialism
• The first requirement is that science demands sufficient evidence (e.g.,
correspondence to empirical data) for all theories it accepts
– Accepted theories should be accepted with a firmness proportional to
the probability assigned to it on the basis of the available evidence
“It is not what the man of science believes that distinguishes him, but
how and why he believes it. His beliefs are tentative, not dogmatic; they
are based on evidence, not on authority or intuition” (Russell)
• The first part is thus an acceptance of evidentialism: It is rational to
accept a theory only if, and to the extend that, there is are good
reasons to believe that it is true. It consists of two principles
The evidential principle: It is rational to accept a proposition only if there
are good reasons (or evidence) to believe that it is true
The proportionality principle: The firmness with which a proposition is
accepted must be in proportion to the strength of the evidence for it
Scientific evidentialism
• But were to start? Typically, it is claimed to start from singular observational
claims such as “At time t and place p event e occured” or “There is a tree in
front of me”. But how to jump from these to the universal claims of theories?
• Besides, according to the evidential principle scientists must also provide
evidence for their observational beliefs. In fact, they must provide evidence
for that evidence as well, and so on. So an infinite regress seems unavoidable
• To prevent such regress the general character of the evidential principle must
be given up. We need to distinguish between non-basic or derived beliefs (in
need of evidential support of other beliefs) and basic beliefs (that do not
need evidential support of other beliefs). This is called foundationalism
• Observational beliefs are then taken as basic (‘foundational’). In fact, it is
not even possible to provide non-circulair evidence for it (why?)
• Revised evidential principle: It is rational to accept a non-basic proposition
only if there are good reasons (or evidence) to believe that it is true
• The relation between our beliefs is asymmetrical. The inferential justification
goes only from basic beliefs to non-basic beliefs. Further, belief A constitutes
evidence (or a reason) for belief B only if it is more basic than B
– Foundationalism differs from coherentism (that does not distinguish basic from
non-basic beliefs)
• Although basic beliefs do not need evidence, they are not groundless. They
can be justified. But this justification is non-inferential or immediate (e.g.,
rests on sense experience)
• What types of beliefs count as properly basic?
Strong foundationalism holds that basic beliefs should be immune to doubt. They
are thus either self-evident (1+1=2) or incorrigible (I am visually aware of ‘green’)
Weak foundationalism holds that basic beliefs must be highly likely true. They are
self-evident, incorrigible or evident to the senses (There is a tree in front of me)
Some philosophers (e.g., Plantinga) go further and also allow for example
memory beliefs as basic. In this and previous case basic beliefs are defeasible
• Fallibilism is the view that all our beliefs might be wrong. Foundationalism
is not to be contrasted with fallibilism (since weak foundationalists are in
fact fallibilists) but with coherentism (no distinction basic and non-basic)
• According to foundationalism a belief is rational only if that belief is either
(a) a properly basic belief, or
(b) can be inferentially supported by properly basic beliefs or other
already derived non-basic beliefs; and
(c) the strenght of the non-basic beliefs are proportional to the
support from the foundation
• Evidence refers to inferential justification of a non-basic belief. Evidence for
such a belief refers to inferential justifiers for it, i.e. the belief is inferentially
supported by other beliefs. Sense experience and memory are examples of
non-inferential justifiers for basic beliefs (these justifiers are called grounds)
– The experience of snowing non-inferentially justifies the belief that it is snowing
– The belief that John is abroad can be inferentially justified from other beliefs
The rules of science
• How can we justify the step from singular observational statements to
the general claims of scientific theories? According to the formalists
there are scientific rules (‘the logic of science’) for rational theory choice
– Scientific rationality is rule-governed or -determined. It’s about rule-following
• Similar to Frege’s deductive logic formalists aimed to establish an inductive
logic for rationally assessing, comparing, accepting and refuting theories
given the available observational data
• Formalists knew that due to the induction problem scientific theories
cannot conclusively be verified by the available empirical evidence. So,
inductive logic will not be able to infer theories with absolute certainty
from observations
• Therefore, acceptable inductive arguments need to show only that the
conclusion (e.g., the theory under consideration) is probable given the
evidence. And the inductive logic must enable the scientist to calculate
a priori precisely the probability of the theory being true given the data
• According to the formalists inductive logic has full authority, not scientific
practice. Today’s dominant model of inductive logic is developed by Bayes
• Scientists are ideal agents (Bayesian agents) that compute the probabilities
of their hypotheses being true (given the evidence) by using Bayes’ theorem
(which can be derived from the axioms of mathematical probability theory)
P(e|h & k)
P(h|e & k) =
x P(h|k)
P( ... | … ) stands for ‘The probability that … given …’
h is the hypothesis (theory) under consideration
e is the available empirical evidence
k is the background knowledge
P(h|k) is called the prior probability (‘the prior’)
P(e|h &k) is called the likelihood
The ratio is called the explanatory power
Objective (Subjective)
Bayesianism takes the
priors to be objective
Popper’s logic of conjectures and refutations
• Popper famously rejected any attempt to develop an inductive logic. We
should not try to support theories (general claims) with observations.
But he agrees that science has a logic and is about following rules
• Popper is thus a formalist because he accepts the logicality thesis
• He proposes the following logic. Evidence should only be used to refute
or falsify theories. This requires only deductive logic since e.g. ‘All swans
are white’ is deductively refuted or falsified by one observed black swan
• Scientists need to conjecture refutable or falsifiable theories and actively
seek evidence that refute or falsify them. Once a theory has survived several
falsification attempts (tests) it is corroborated and can be rationally accepted
• Like all formalists, his account of scientific method is intended to be universal
and ahistorical. He thus accepts the stability thesis of formalism, namely that
the rationality of the scientific method is invariant over time and domain
The rule principle
• Formalists accept the rule principle according to which all rational
beliefs (and actions) must be arrived at by means of appropriate rules
Rules help to establish logical or necessary connections between propositions
Rules help to infer and refute theories from available evidence (observations)
Rules help to reduce or even eliminate arbitrariness (‘predictable outcomes’)
Rules limit room for diverging conclusions (‘universal or objective outcomes’)
Rules can be applied in any possible context (‘practice independence’)
Rules ensure that conclusions folllow necessarily from the available evidence
Rules ensure an outcome in a finite number of steps (‘algorithm’)
Rules application requires no detailed knowledge of the content (‘formal’)
The formalists’ conception of rationality’s nature
• Both a deontological and a means-end conception of rationality can be
found among formal philosophers of science
• With respect to the deontological side scientists have intellectual duties (e.g.,
only accept theories supported by good evidence, try to falsify theories, etc.)
• With respect to the means-end side the rationality of science depends on
whether scientific methods are efficient in achieving the aims of science
Different accounts of the goals of science: truth, predictability, usability, etc.
• It must be possible to compare theories relative to the agreed goal(s). If
there is sufficient evidence that T2 is goal-superior to T1 - and the scientific
community perceives this - than they must abandon T1 in favour of T2
• If the deontological rationality can be thought of as a duty to satisfy certain
epistemic and non-epistemic ends, both conceptions can be combined:
scientists have a duty to establish proper connections between their
scientific theories and their scientific ends
Demarcating science and non-science
• Some formalists have claimed that scientific rationality is the only form of
rationality that exists: “To fail to be scientifc is to fail to be rational”
• Other formalists go less far but are still quite skeptical towards beliefs that
do not meet the(ir) standards of scientific rationality
• But how should science be demarcated from non-science?
• Positivists hold that the demarcation criterion is the same as the one that
they believe separates meaningful from meaningless claims: the verification
principle: an empirical statement has meaning iff it is empirically verifiable
Next to meaningful empirical claims, positivists accept analytic claims. These are
conceptual truths. They are true by virtue of their own meaning (e.g., 1+1=2)
The verification principle is self-refuting though (why?)
• Popper’s demarcation criterion is not meant to draw a line between claims
that are meaningful and meaningless. He holds that non-scientific claims
can still be meaningful. For Popper scientific claims are falsifiable
Expanding to Formal Evidentialism
• Formal Evidentialism is the model of rationality that results if we expand
the formal approach to scientific rationality to all areas of human life
Evidential principle (‘beliefs need good inferential or non-inferential evidence’)
Proportionality principle (‘believe with a firmness proportional to the evidence’)
Logicality principle (‘rationality matter of practice-independent a priori analysis’)
Rule principle (‘belief formation is a matter of following algorithmic rules’)
Foundationalism (‘distinction between basic and non-basic beliefs’)
• Formal evidentialists do not pay attention to the actual agents and the
actual practices. There is no analysis of how people in different practices
actually form and regulate their beliefs. Actual events are mere illustrations
The locus of rationality
• Formal evidentialists embrace the logicality thesis (rationality boils down
to a logical system). So, the locus of rationality is ultimately in systems of
propositions. Only (relations between) propositions can be (ir)rational
– Relations between inferential evidence and theories and relations
between basic and non-basic propositions
• A rational belief is thus nothing more than a belief in a proposition that is
rational(ly justified). And a rational person is nothing more than a person
who believes only rational(ly justified) propositions
• It follows that for formal evidentialism rationality is not a relation between a
particular kind of agent, in a particular kind of situation, and her beliefs.
That’s why concrete practices do not matter for formal evidentialists
– Formal evidentialists embrace person-independent idealized universal models
– Rationality is ultimately connected to propositions, not persons
External and internal rationality
• A distinction can be made between external and internal standards
(criteria, conditions) for rationality
• Internal standards deal with how agents ought to govern their accepted
beliefs. The focus is only on the beliefs present in an agent’s belief system
– Principle of logical consistency
• External standards deal with the relation between an agent’s belief system
and the outside world. It is about what standards a belief must satisfy before
it is allowed to enter or stay in an agent’s belief system
– Evidential principle, Principle of proportionality or Principle of proper basicality
Internal standards for rationality
The consistency principle
Since any absurdity can be derived from a logical contradiction (‘P and not-P’)
it is often suggested as an internal standard that the collection of beliefs of
the agent (‘the agent’s belief system’) must be logically consistent
An agent should not believe both ‘All swans are white’ and ‘There is a black swan’
The principle of deductive closure
It is often suggested as an internal standard that the logical consequences
of what is believed should be accepted by the agent as well
If an agent believes ‘All swans are white’ and the agent also believes ‘John owns a
swan’, then the agent should also accept ‘The swan that John owns is white’
Revised version of both internal standards
• Many formalists acknowlegde that both standards are not feasible because
it is impossible for people (having limited mental powers) to meet them
– People have many beliefs and aren’t able to verify whether their whole belief
system is consistent. Neither can they infer all consequences of their beliefs
– Moreover, it can be in fact argued for that our belief systems cannot be
both reasonable and consistent (how?)
• Yet, both standards can be rephrased as ideals. For rationality it is not
required to actually meet both ideals. It is sufficient to do the best we
can to satisfy these ideals
Revised principle of consistency – People are rational only if they always try their
best to bring it about that they eliminate as many inconsistencies as possible in
what they believe
Revised principle of deductive closure – People are rational only if they always
try their best to bring it about that they accept as many consequences of
what they believe
A problem for ‘deductive closure’
Consider Brigitte. She beliefs only propositions with a probability of being
true of 80% or higher. Suppose that she believes ‘All swans are white’
with probability 80% and ‘John owns a swan’ with probability 80%.
According to the principle of deductive closure she should accept as well
‘The swan that John owns is white’. But this cannot be right, since the
likelihood of this proposition is (0.8)2 and therefore less than 80%
The problem becomes even more pressing if we consider cases with many
different propositions. Holding on to deductive closure would entail that
we have to accept propositons that are in fact highly unlikely, which
does not seem to be a rational thing to do
To follow-up on this. Suppose that Brigitte has beliefs A, B, C and D with likelihoods 80%.
Suppose that A, B and C together entail not-D. Is Brigitte’s belief system contradictory? Is
Brigitte forced to belief a contradiction if we take into account the principle of deductive
closure? And if we take into account Brigitte’s ‘80% rule’ as well?
Chapter 4: The Scientific and the
Evidentialist Challenge to Religious Belief
The 20th century challenge to religious belief
• In the 20th century, much of the discussion about whether or not religious
belief is rational or justifiable took formal evidentialism for granted
• It was not until the end of the 20th century that philosophers of religion
started to realize that the issue of the rationality of religion is as much a
question about the notion of rationality itself as its application to religion
• An early 20th century view of the rationality of religious belief was provided
by positivism. Positivism can be characterized in the following way
A statement is meaningful or has cognitive content if and only if it is analytic
or empirically verifyable (the verification principle of meaning)
A meaningful statement is true if and only if it is logically proven (for analytical
statements) or empirically verified (for empirical statements)
The above two principles constitute the scientific standard of rationality. Thus all
meaningful (and thus all rational) beliefs must be scientific beliefs. Rationality is
scientific rationality. To fail to be scientific is the same as to fail being rational
All non-scientific statements (religion, art, etc.) are nonsense, like “Qg aW Dzz”. So
there are not even false. They are just meaningless – empty of congitive content
The verificationist challenge to religious belief
A statement has cognitive meaning if and only if it is analytic or
empirically verifiable (the verification principle of meaning)
Religious statements are neither analytic nor empirically verifiable
Therefore, religious statements are cognitively meaningless
• So religious statements, such as “God exists”, are not even false. They say
nothing false, because they say nothing at all. They are pseudo-statements
• We must distinguish between the semantical question of religious belief
(are these beliefs meaningful?) and the epistemological question of
religious belief (are these beliefs rationally acceptable?)
• Since according to the verification principle of meaning religious beliefs do
not pass the semantic test, the second never arises. Religion is on this view
thus strictly speaking not even irrational. It is a-rational
The falsificationist challenge to religious belief
• Philosophers of religion like Flew replaced the positivistic verification
principle of meaning with a falsification principle of meaning
A statement is cognitively meaningful if and only if it is analytic or empirically
falsifiable (i.e, it is possible to say upfront in which situations it is falsified)
• This principle leads to the falsificationist challenge to religious belief
A statement has cognitive meaning if and only if it is analytic or
empirically falsifiable (the falsification principle of meaning)
Religious statements are neither analytic nor empirically falsifiable
Therefore, religious statements are cognitively meaningless
Responding to both challenges
• Many philosophers of religion who wanted to hold on to rationally justified
religious beliefs, felt the need to somehow meet these challenges
• Typically they accepted the first premise of both challenges, and tried to
argue that religious discourse is compatible with the verification or
falsification principle. That is, they rejected the second premise
• But then some started to realize that both principles were themselves
problematic. They are according to they own standards meaningless.
Therefore, they are self-referentially incoherent (Plantinga)
• Moreover, it became clear that not even science is able to satisfy these
demands. For many scientific theories it is not possible to specify the
verification or falsification conditions
• The consensus became that both principles are inadequate. A statement
does not have to adhere to these principles in order to be meaningful
– Religious beliefs passes the semantic test. But can they be rationally justified?
The scientific challenge to religious belief
• If religious beliefs are cognitively meaningful, the crucial question becomes
whether they pass the epistemological test. Are these beliefs rational?
Can they be rationally accepted?
• We may then consider the following scientific challenge to the rationality
of religious belief
Religious beliefs must fulfill the same, or at least similar, standards of
rationality as scientific beliefs in order to be considered rational
Religious beliefs do not fulfill the same, or at least similar, standards of
rationality as scientific beliefs do
Therefore, religious beliefs are irrational
• Surely, to give this challenge content one must specifiy the standards of
rationality science one thinks science satisfies
Responding to the scientific challenge
• Religious believers can respond to this challenge in three different ways
The strong response – Accept the first premise and reject the second by arguing
that religion does in fact fulfill the same or similar rationality standards as science
The differentiation response – Reject the first premise by arguing that religious
beliefs are rational but meet (wholly) different standards of rationality as science
The irrationality response – Accept the scientific challenge but claim that this
does not count against religion, since religion has never meant to be rational
The evidentialist challenge to religious belief
• The evidentialist challenge to religious belief consists of an application
of evidentialism to religious beliefs
It is rational to accept religious beliefs only if, and to the extent that,
there are good reasons (or evidence) to believe that they are true
There are no good reasons (or evidence) to believe that religious
beliefs are ture
Therefore, religious beliefs are irrational
• A religious believer might object though that religious beliefs are
properly basic and therefore do not require supporting evidence
The evidentialist challenge to religious belief
• The evidentialist challenge should therefore be stated as follows
Religious beliefs are not properly basic beliefs
It is rational to accept non-basic beliefs only if, and to the extent that,
there are good reasons (or evidence) to believe that they are true
There are no good reasons (or evidence) to believe that religious
beliefs are true
Therefore, religious beliefs are irrational
• The evidentialist challenge to religious belief needs to be distinguished
from the scientific challenge to religious belief. For what counts as good
reasons may differ from what counts as good reasons within science
• The end of rationality advocates of the evidentialist challenge have in mind is
purely epistemic. It is a matter of increasing true and eliminating false beliefs
• It is also formal in the sense that the actual situation of people is irrelevant.
Rationality is ultimately only a property of propositions, not of real agents
Responding to the evidentialist challenge
• Religious believers can respond to this challenge in four different ways
Reject the third premise and claim that there are good reasons for religious belief
Reject the first premise and claim that religious beliefs are properly basic
* Either claim that religious beliefs are properly basic in a strict sense,
i.e. self-evident, incorrigible or evident to the senses
* Or claim that religious beliefs are properly basic but not self-evident,
incorrigible or evident to the senses (Plantinga)
Reject the second premise and opt for a different model of rationality than
Accept the entire evidentialist challenge but claim that this does not count
against religion, since religion has never meant to be rational
A challenge for atheism
• One could also formulate an evidentialist challenge for atheism,
i.e. the belief that God does not exist
The belief that God does not exist is not properly basic
It is rational to accept non-basic beliefs only if, and to the extent that,
there are good reasons (or evidence) to believe that they are true
There are no good reasons (or evidence) to believe that God does not exist
Therefore, the belief that God does not exist is irrational
• Evidentialists can therefore not hold that there should be a presumption
of atheism. If there are no good reasons for or against the existence of God,
then on evidentialism, the rational position is not atheism but agnosticism
• Atheists may respond that atheism is the absence of belief in God. But this
weak definition doesn’t do justice to many self-declared atheists. Moreover,
in that case even baby’s, young children, and those who never considered
the matter should be called atheists, which seems incorrect (if not absurd)
The ‘strong response’ to the evidentialist
challenge to religious belief
• A common response to the evidentialist challenge to religious belief is to
accept evidentialism but to reject premise 3, i.e., to argue that there are good
reasons or evidence to believe that religious beliefs are true (Swinburne)
• One normally identifies a core of religious statements which are central to
the religious worldview in question (e.g., God exists) and defines these claims
as precisely as possible. Everyting else is put aside until this core is justified
• The issue is not whether God’s existence can be conclusively proven, but
whether there is evidence that makes God’s existence probable to such an
extent that it is rational to believe in God’s existence based on the evidence
• This issue needs to be settled by formal methods, e.g. Bayesianism. Theism
is treated here as a large scale theory of reality. Proponents of theism try to
show that the available evidence confirms theism, i.e. renders it probable
• Following Bayes theorem, the epistemic probability of theism given the
available evidence is (just as for any other theory) determined by the
product of the theory’s prior probability and its explanatory power
The ‘strong response’ to the evidentialist
challenge to religious belief
• The prior probability of theism depends on (just as for any other theory) on
its simplicity (hower fewer entities postulated, how simpler), its fit with our
background knowledge and its scope (how many phenomena it covers)
• A Bayesian case for theism would then use Bayes theorem to estimate the
probability that God exists based upon all the available evidence (i.e., the
origin of the cosmos, its fine-tuning, etc.) and its prior probability
P(e|h & k)
P(h|e & k) =
x P(h|k)
h is the hypothesis that theism is true (i.e., the God exists)
k is background knowledge (since theism has maximal scope, k is broadly logical laws)
e is the available evidence appealed to (the origin of the cosmos, its fine-tuning, etc.)
P(h|k) is the prior probability of theism (determined by its simplicity, etc.). It’s 0.25 
P(e|h&k)/P(e|k) is the explanatory power of theism
P(h|e&k) is the epistemic probability of theism given the available evidence. If
this probability is greater than 0.5, then theism can be rationally accepted
The evidence for a formal justification of theism
• The evidence should be laid out carefully and systematically, and should be
shared by believers and unbelievers alike. It should not be controversial
• Evidence is often obtained from common sense or established scientific
beliefs. It includes general features of the world and human experience
• In addition a formal justification of theism will appeal to certain principles
of theory-choice, such as simplicity, explanatory power, efficiency, etc.)
• A formal case for theism is cumulative. It combines different evidence’s ei
e1 The origin of the cosmos (appealed to in cosmological argument for God)
e2 The fine-tuning of the cosmos (appealed to in teleological argument for God)
e3 The experience of objective moral values (appealed to in moral argument for God)
e4 Religous experience (appealed to in argument from religious experience for God)
eN Etc.
• So, the relevant epistemic probability of h is P(h|e1 & e2 & e3 & … & eN & k)
‘The formal justification of theism’ and science
• A presupposition of many formal cases for theism is that theism is a
hypothesis very much like a scientific hypothesis, that explains
why the world exists and why it looks the way it looks
• Thus these cases are very similar to scientific reasoning. They can
therefore be understood as meeting both the evidentialist and
the scientific challenge to religious belief
• Moreover, by showing the close similarities between the assessment of
theism and respectable scientific theories (about things beyond immediate
experience) the Hume-Kant objection (that we are not rationally justified
in believing anything beyond our immediate experience) loses its force
• Finally, for many science is the paradigm of rationality. So, if theistic belief
is supported by the same methodology and evidence as scientific claims,
then the rationality of theism would be firmly and clearly established
The dilemma of religious assent
• The formal cumulative case for theism aims to show that theism is a
rationally justified hypothesis. But this seems to conflict with the fact
that many religious believers accept their beliefs with full commitment
• The firmness of those religious beliefs is therefore not in proportion to the
evidence. The proportionality principle of formal evidentialism is violated
– Insufficient evidence to warrant the certainty with which religious beliefs are hold
• So we have a dilemma of religious assent for the formal evidentialistic
case for religious belief
* Either religious believers are rational in their religious beliefs, but then these
religious beliefs do not seem to function properly (no strong full commitment)
* Or religious believers have the commitment religion seems to demand, but
then these beliefs are irrational (since they violate the proportionality principle)
Religious belief as properly basic
• Another response to the evidentialist challenge to religious belief is to
accept evidentialism but to reject premise 1, that is, to argue that (all
or many) religious beliefs are properly basic (Plantinga)
• Plantinga asserts “that it is entirely right, rational, reasonable, and proper
to believe in God without any evidence or argument at all”
• According to “classical foundationalism” a belief is properly basic if and
only if it is self-evident, incorrigible or evident to the senses
• Plantinga argues that “classical foundationalism” is false, for it would entail
that enormous quantities of what we all in fact believe are irrational
I had breakfast this morning
That person is angry
The earth existed hundred years ago
• On “classical foundationalism” these beliefs are not be properly basic nor
derivable from the basis. But then they are irrational, which is absurd
Religious belief as properly basic
• Moreover, “classical foundationalism” is self-referentially incoherent. Its
standard principle of proper basicality is not basic and does not follow
from the basis. But then this principle cannot be rationally accepted
• So “classical foundationalism” fails. But then we are justified to consider
also other beliefs as properly basic, including the belief that God exists
• No necessary and sufficient conditions for proper basicality can be derived
from evident premises. There remains disagreement about what beliefs
are properly basic. But then the evidentialist challenge doesn’t go through
• Still, Plantinga is a formal evidentialist. It is just that he accepts a weaker
version of foundationalism. The only principle he does not accept is the
rule principle, since he holds that there is no formal logical procedure
to unambiguously determine whether a belief is properly basic
Arguing that belief in God is not properly basic
A characteristic of proper basic beliefs is that any person with normally
functioning faculties, when placed in the appropriate circumstances, will
form these beliefs. And this seems false of the belief that God exists
In fact, many people with normally functioning faculties, when placed in
the right circumstances, will form beliefs that are incompatible with theism.
All these “spontaneous" beliefs are thus unreliable and so not properly basic
• A theist though, might respond that these objections go too quick. For why
should the theist concede that these other people have been placed in the
same circumstances as those who spontaneously became to belief in God?
• For all we know, these other people might not have had access to the crucial
resources in a suitable way, they might not have been sincere seekers, etc.
Thus it doesn’t follow that belief in God is not properly basic. Premise 1 of
the evidentialistic challenge has therefore not been established
• Besides, the fact that others form non-inferentially, say, the moral belief that
women are inferior, doesn’t entail that the non-inferentially formed belief of
equality is irrational. So, why shouldn’t this also hold for the belief in God?
Responding to the scientific challenge to
religious belief
• Swinburne is an advocate of the strong response to the scientific challenge.
He claims that theism can satisfy similar standards of rationality as science
• Plantinga is an advocate of the differentiation response to the challenge.
Theism is rational but meet different standards of rationality as science.
– Belief in God is a non-inferential basic belief and scientific beliefs are non-basic
– Theism is thus not like a scientific hypothesis. It is not an explanatory theory
• Moreover, as Plantinga notes, if we think that all beliefs should satisfy the
demands of science, then many reasonable beliefs, such as that there are
other minds, turn out to be irrational – which is an absurd conclusion
• “Hence”, Plantinga writes, “my tentative conclusion: if my belief in other
minds is rational [although it cannot be established by science], so is my
belief in God. But obviously the former is rational, so, therefore, is the latter”
Chapter 5: The Practice-Oriented
View of Science
• In 1962 Thomas Kuhn published his Structure of Scientific Revolutions in
which he argued against the formalists that philosophers of science must
pay attention to the history of science and to current scientific practice
• Philosophy of science must be ‘true’ to science as practiced. So the
‘descriptive’ (history, practice) is relevant for the ‘normative’ (standards).
Science is not just a matter of ‘logical analysis’ - but also of ‘psychology’
• Since 1962 two camps have developed
Sociologists of science opt for contextualism. They only describe actual practice
from the inside and hold that a normative account from the outside is impossible
Historians of science opt for a practice-oriented approach: a normative account
of science from the outside that is in substantial contact with actual practice
• In contrast to the pre-Kuhnian formal approach, the post-Kuhnian practiceoriented approach is a posteriori and not a priori. Practice matters, but
independent philosophical analysis is important too
History of Science
• Historians of science take it that the history of science (reflecting actual
practices) puts constraints on normative models of scientific rationality
• Historical case studies of actual scientific practices can constitute
evidence for or against a proposed scientific model of rationality
“One must not attempt to prescribe a rationality for an activity as
complex as science without first examining its actual practice” (McMullin)
“For too long philosophers have debated how scientists ought to judge
hypotheses in glaring ignorance of how they in fact judge hypotheses” (Giere)
• The actual cognitive goals of scientists need to be taken into account. The
proposed standards must be scientifically relevant. Relevant for the practice
• But what have been the main consequences of the practice-oriented
approach to scientific rationality? For that we turn to Kuhn’s findings
Kuhn’s philosophy of science
• As Kuhn discovered, the formal view of science “did not at all fit the
enterprise that historical study displays”
• Kuhn (and the practice-oriented philosophers after him) rejected the rule
and stability principle: Scientific theory-choice is not a matter of neutrally
following clear objective logical rules, and methodology changes over time
• Paraphrasing Kuhn: “During the sixty years after Newton’s computations, the
predicted motion of the moon’s perigee violated Newton’s laws. But no one
took proposals to reject Newton’s theory seriously. Physicists patience to
protect Newton’s theory against a major anomaly proved justified, since it
was shown later-on that the mathematics was wronly applied in this case”
• Science isn’t likely to follow a formal Popperian policy of rapid refutation of
theories in the face of counter-evidence. Scientists are conservative and
seeks ways of protecting the theories they are strongly commited to
• Counter-evidence is viewed as anomaly that can be put aside for later work
Kuhn’s philosophy of science
• During periods of normal science, scientists work within an established
paradigm. They want to solve problem in it. Anomalies are put aside
• A paradigm consists of a set of fundamental conceptual and methodological
assumptions that indicate what is to be studied, how it should be studied,
what the questions are that must be answered, and what counts as answer
• When anomalies within a paradigm accumulate and scientists begin to doubt
that the paradigm has the resources to solve the problems, a crisis emerges
• A crisis leads to a scientific revolution. The existing paradigm is replaced by
another paradigm. In physics the Aristotelian paradigm was replaced by the
Cartesian-Newtonian paradigm. The latter was replaced by modern physics
• Contra formalism, there are no clear rules for paradigm shifts. Competing
paradigms are incommensurable. Proponents of conflicting paradigms do not
even agree on what should be the evaluation criteria (deep disagreement)
The revisability thesis
• Further, to the extend that there are some rules acccording to which theorychoices are made, these rules change as science develops (revisability thesis)
• There is no constancy of logic or method in science. Scientific methodologies
have undergone enormous transformations throughout the history of science
• Even the very boundaries between ‘the scientific’ and the ‘non-scientific’,
between scientific problems and merely pseudo-problems, shift over time
• Scientific change is not just about the alternation of factual beliefs, but
extends to the rules and methods of science itself. And this should not
surprise us. For why should we know up-front how to study nature?
• In particular, empirical studies show that, again contra formalism, scientists
are not Bayesian agents. They are not even intuitive Bayesians. They do not
apply explicitly or intuitively Bayesian models as rules for theory-choice
• “No Bayesian model fits the thoughts or actions of real scientists” (Giere)
An alternative model of scientific rationality
• Kuhn argues that a rule-based logic of science does in fact not exist when
fundamental changes occur. Is science thus irrational according to Kuhn?
• No, he doesn’t deny the rationality of science. He rejects the formal view of
science. He holds that rule-following isn’t necessary for scientific rationality
• Kuhn attempts “to show that existing theories of rationality are not quite
right and must be changed to explain why science works as it does”
• Kuhn and his followers didn’t conclude that science is irrational. In fact, they
agreed with the formalists that it is clearly a paradigm example of rationality
– A model of scientific rationality that entails that science is irrational has to go
• That science cannot be formalized is therefore not a shortcoming of science,
as the formalists would have to say, but of the formalists’ logicality thesis
• Hence the Kuhnian practice-oriented philosophers of science wanted,
contra formalism, to develop an alternative model of scientific rationality
An alternative model of scientific rationality
• Kuhn insists that scientific rationality is governed by values instead of rules.
Scientists who share them may make different choices in the same situation
• But the proposed values (accuracy, explanatory scope, simplicity, predictability, etc.)
are also accepted by the formalists. Kuhn only denies that these values can
be mechanically applied as rules to determine theory-choice
• The values thus do not replace the rules. So what is Kuhn really after?
• Many critics gave Kuhn two options. Either science is a matter of impersonal
logical rules (and thus rational) or it is a matter of personal subjective opinions
(and thus irrational). But Kuhn wants to break free from this false dichotomy
• He proposes a third alternative: Science is rational because it is governed by
the informed judgments of the whole community of experienced scientists
• In some cases scientists uses rules for theory-choice. But these algorithms are
developed by exercising judgments. Hence, judgments are more fundamental
So the formalizations of formalism don’t illuminate what scientists actually do
Explicating informed judgments
• Judgment is the ability to make decisions without being able to provide an
explicit justification. It’s a skill that an experienced person can exercise
• Judgments are cognitive acts that involve decisions based on information.
They are not arbitrary, although they aren’t arrived at by following rules
– Evaluate the situation, assess the evidence and consider different viewpoints
• Brown identifies three basic characteristics of informed judgments
It is a cognitive skill that improves with practice. However, it is not possible to
reduce these skills to the application of rules. It’s not about rule following
They are fallible but not arbitrary. It is based on a careful reflection of what is
involved in the issue, understanding of the situation, adequate experience, etc.
They must be exercised by those who are competent in the research field, who
have adequate knowledge and training. These persons must have the relevant
background information and reliable information relevant to the case at hand
• For the rational evaluation of theories, their actual detailed content is crucial.
Evaluation is no longer a matter of following content-agnostic formal rules
The social condition of rational theory-choice
• Rational theory-choice also requires that scientists take into account the
judgments of their peers and re-evaluate their position if necessary
• The community of scientists guarantees the rationality of science. Rational
decisions are based on the informed judgments of the scientific community
• Communial assessment, discussion and consensus is key for rational theorychoice. Individual judgments must be tested by the community (peer review)
• According to practice-oriented philosophers scientific rationality consists of
two demands: informed judgments and proper peer evaluation
• Scientific belief-formation and regulation is rational to the extend that it (a)
is governed by informed judgments and (b) involves intersubjective testing
Social evidentialism
• Practice-oriented philosophers accept the revisability thesis (scientific
methods change over time). Yet one thing is stable: reliance on evidence
• What they have in common with the formalists is that evidentialism or
the evidential principle is a constant feature of scientific rationality
• An informed judgment takes into account an assessment of the evidence or
good reasons for or against a theory. A rational decision needs such reasons
• What counts as evidence or good reasons though, changes over time. And it
typically changes during a scientific revolution (i.e. during a paradigm shift)
• Practice-oriented philosophers of science accept social evidentialism instead
of formal evidentialism. The rule principle is replaced by the social principle
• However, practice-oriented philosophers reject the proportionality principle
(the firmness of belief must be in proportion to the strength of the evidence)
The reason is that the proportionality principle doesn’t fit the history of science
Degree of assent in science
• Science is so successful because scientists do not follow the proportionality
principle. The commitment of scientists to their theories reaches further
than the available evidence. And there are reasons for this commitment
• If scientists would reject their theories at the first falsification, science
would degenerate. For no theory solves all problems it is confronted by
• In fact, scientists do never lay down in advance what would lead them to
give up, say, Quantum Mechanics. This is simply not how science works
• Stronger belief than warranted by the evidence is necessary for progress
in science. A new theory must have the change to develop its potential
• Therefore, all else equal, scientists are rationally entitled to hold beliefs
with a commitment that exceeds the strength of the available evidence
• Strong commitment (convergent thinking) is as important as an open-minded
attitude of challenging the paradigm (divergent thinking) for scientific success
Chapter 6: Social Evidentialism
Social evidentialism as a
general model of rationality
• The practice-oriented view of scientific rationality has been developed
into a general model of rationality, refered to as social evidentialism
• Why a general model? All things being equal, a model that can be applied to
many human practices is to be preferred over a less widely applicable model
• If what is reasonable would completely change as we move from one practice
to another, then this is not desirable. Cognitive integration is to be preferred
• Social evidentialists claim that the following two standards apply to all
areas of life or all human practices where the question of rationality arises:
The evidential principle – A belief is rationally acceptable only if it is arrived
at by a person who exercises informed judgment (rely on expertise or expert advice)
The social principle – A belief is rationally acceptable only if it has been exposed
to or tested against the judgments of a community of relevant expertise
• Judgment is in accordance with our pre-analytical conception of rationality
– Rational persons are those who are able to come to adequate conclusions in
situations without clear rules. Only mechanically following rules isn’t rational
Social evidentialism as a
general model of rationality
• As any model of rationality, social evidentialism must do better than chance
in the long run. It meets this meta-requirement due to its focus on evidence
• Social evidentialists deny that rational beliefs must satisfy proportionality, as
long as believers remain open to evidence and opinions of competent others
– In beliefs a dialectic between commitment and tentativeness must remain
• The social principle replaces ‘formal rules’ with ‘intersubjective tests, peer
review, tests by competent others’ as the non-arbitrary decision procedure
For rational belief revision it isn’t enough that just one or two epistemic peers
disagree, there must be a substantial number of them (e.g., the majority)
Moreover, it must be possible for the person in question to see that these
others who disagree are competent others. They must be recognized as peers
• According to the social principle Robinson Crusoe alone on his island can
exercise judment, but he can’t be rational. For just like playing baseball,
rationality requires other people – and not just any people (but peers)
Social evidentialism as a
general model of rationality
• The evidential principle is necessary but not sufficient for rationality. For
critical evaluation of a belief by the relevant community is needed as well
• The social principle is necessary but not sufficient either. Communial assent
is not enough. It is also required that beliefs are arrived at appropriately
– Consensus by e.g. force isn’t rational. Belief must result from exercise of expertise
• On social evidentialism both principles together are sufficient for rationality
• Who chooses the experts for a field? Normally, the current experts choose
the new ones. For only they have the expertise to train and evaluate them
• But isn’t this process fallible? Aren’t experts who want to preserve their own
opinions going to accept only new candidates who conform to their views?
• Yes, it is fallible, but it is the best process we have. What is crucial is that
there is a continual interaction between the experts and the general public
– This interaction requires that the public is able to recognize when the field ceases
to be reliable and so to question the experts indirectly (e.g., when bridges fail)
The locus of rationality
• Formal evidentialists ground ‘rational person’ in ‘rational belief’. But social
evidentialists, on the other hand, ground ‘rational belief’ in ‘rational person’
• According to formal evidentialists the locus of rationality is a belief. The
notion of a rational person is secondary. It is derived from ‘rational belief’
• Formalists thus take a system of beliefs as the central category of rationality.
A rational person is someone who accepts only rational beliefs
• According to social evidentialists the locus of rationality is a person. The
notion of a rational belief is secondary. It is derived from ‘rational person’
• Social evidentialism takes the agent as basic. So, there are no rational beliefs
per se. Rationality becomes relative to individuals. It is person-dependent
• Taking the notion of rational person as primary implies that the constraints
put on a rational person determine the constraints put on rational belief
• Rationality is ultimately a question under what circumstances a person is
rational in accepting a belief. It is an ability that a person (fails to) exercise(s)
Disagreement on social evidentialism
• Should a person give up his or her belief in case the majority of competent
others disagree? According to Hilary Putnam this is not always the case
• Nevertheless, when the majority of experts disagree, I must try to justify
my belief. For epistemic egoism (I keep my belief because it’s mine) isn’t rational
• Now, according to Kuhn and Gutting, if I try to justify my belief and fail to
convince the majority of the other experts, I have to give up my belief
• Is Kuhn’s and Gutting’s position satisfying? It conflicts with Putnam’s claim
• Brown’s position seems more reasonable. Agreement with the majority is not
required. For in science disagreement is quite common. And it isn’t sufficient
either, since a not-appropriately formed majority belief is still irrational
• Brown holds that persons only have to submit their beliefs for evaluation by
the competent others who are well-informed, and take this review seriously
Strong social evidentialism
• Brown’s position accords with social evidentialism. Kuhn’s and Gutting’s
position though adds a third condition to social evidentialism
Evidential principle – People are rational in their beliefs if they arrive at these
beliefs by exercising an informed judgement
Social principle – People are rational in their beliefs if they submit these beliefs
to evaluation by competent others and their evaluation is taken seriously
Conformity principle – People are rational in their beliefs if they reject them if the
majority (or a sufficient number) of competent others disagree with these beliefs
• This stronger form of social evidentialism is called strong social evidentialism
Social evidentialism and foundationalism
• Does social evidentialism allow for properly basic beliefs, that is, beliefs
that are immediately justified – without informed judgment or inference?
• Beliefs such as “I see a tree in front of me” and “I had breakfast this morning”
are not the result of informed judgment. Nor do I infer them logically from
other beliefs. So, on social evidentialism they are properly basic as well
• Yet, on social evidentialism I must submit my basic beliefs to the evaluation
of competent others – and try to justify them if many other experts disagree
• Brown would agree with the previous point. But Kuhn and Gutting would
add that if my justification attempt fails, I have to reject those basic beliefs
Social evidentialism and truth
• On social evidentialism, rational beliefs are not assured to be true. For our
best informed judgments are tied to our limited capacities and evidence
• Rational beliefs are only true with a high degree of probability in case we
limit ourselves to the acceptance of logical inferences from a self-evident
absolutely certain foundation (strong foundationalism)
• Thus, contrary to strong foundationalism (or other formalisms), there is
no tight connection between rationality and truth. The link is rather weak
• In addition to limited cognition and evidence, there is no formal rule-based
procedure. Hence different experts may reach different informed judgements
• So on social evidentialism truth isn’t a necessary condition for rationality. In
fact it is likely that false beliefs could be rationally acceptable
• Yet, we proceed rationally in attempting to discover truth. We take rational
conclusions as our best estimates of the truth. We also presume that in
proceeding rationally, there will be a convergence towards truth
Social evidentialism and truth
• It is possible to view social evidentialism from a realist or anti-realist point of
view. This has consequences for the connection between rationality and truth
• On a realist account of social evidentialism, rationality is person-dependent
and truth is person-independent. Truth tracks our mind-independent reality
• On an anti-realist account of social evidentialism, both notions are persondependent. For truth is what is rationally justified in the ideal epistemic limit
Epistemic interpretation – We only have access to how the world is for our minds
Ontological interpretation – There is nothing beyond this mind-dependent world
• Some anti-realists are also truths pluralists. According to these pluralists
there is more than one ‘true’ theory of the world
• In what follows realism will be assumed
The scope of social evidentialism
• The main argument of social evidentialists is that – contra formalism – their
model fits science, which is taken to be “the paradigm case” for rationality
• Besides, the scope of social evidentialism is wider than that of formalism
Breadth – Contrary to formalism, it applies to many different practices involving
‘judgment’ and ‘informed others’ (such as everyday life and religion)
Depth – Contrary to formalism, it applies not only to the context of justification
of a practice, but also to the context of discovery of that practice
• The formalists’ limitation of rationality to the context of justification (a
context for which there are supposedly formal rules) doesn’t fit our intuitions
– In the absence of rules for discovery, informed judgment is needed here as well
• With respect to the nature of rationality, social evidentialism can be understood deontologically, i.e., its principles are obligations we ought to fulfill
• And similar to formal evidentialism, it can be combined with a means-end
notion of rationality. Also “its demands are in our power to satisfy” (Brown)
Chapter 7: Social Evidentialism
and Religious Belief
Social Evidentialism and Religious Belief
• Is social evidentialism applicable to religious practice? And if so, can a
person be rational in his or her religious believing on this model?
• In religious matters we can also exercise informed judgment and expose our
beliefs to peer evaluation. But then social evidentialism is applicable to it
• A practice-oriented approach must take into account the actual practice of
religious believers. This has typically not been done in philosophy of religion
– Contrary to science, the initial presumption to religion has not been positive
• Social evidentialists claim that the philosopher of religion should reject such a
negative presumption. Radical suspicion should not be the point of departure
– Philosophy of religion is to be handled similar to philosophy of science, art, etc.
• As with science and art, our initial confidence should be in favor of so central
and comprehensive a practice as religion. Depart from sympathetic attitude
• In short, philosophy of religion should start from a positive presumption
– Not to eliminate criticism, but to refocus it. Critique from outside surely possible
The origin of the negative presumption
• Many philosophers of religion accepted the asymmetry thesis: Believers are
committed to a position, whereas non-believers simply withhold judgment
• This thesis is already reflected in the terms ‘believer’ and ‘non-believer’
• Philosophers of religion who accept the asymmetry thesis find belief rather
than non-belief puzzling. Hence they see religion with skepcis and suspicion
• But the asymmetry thesis is unrealistic. Disagreement between believers
and non-believers involves substantial positive claims on both sides. Nonbelievers have their own ‘view of life’ consisting of guiding secular beliefs
• It’s more adequate to talk about religious believers and secular believers.
Both have beliefs on the nature of humans, the world, the good life, etc.
• Given that a negative presumption is misleading, why should our initial
presumption to religion be positive? Why not simply be neutral?
• Contra general neutrality, philosophy must start somewhere. In this case the
presumed cognitive significance or validity of religion as a central practice
Informed Judgment in Religion
• Social evidentialists agree with formal evidentialists that science is our best
example of a rational practice. But they deny the formal account of science
• Thus the content of the scientific challenge to religious belief (‘religion
must fulfill science’s rationality standards’) changes on social evidentialism
The traditional scientific challenge assumes a formal view of science: Religion
fails to fulfill science’s formal evidentialism and is thus irrational
The contemporary scientific challenge assumes a practice-oriented view of
science: Religion fails to fulfill science’s social evidentialism and is thus irrational
• On the contemporary scientific challenge to religious belief, following
rules is not a necessary condition for obtaining rational religious beliefs
– What is required is the possibility of informed judgments and peer evaluation
• A religious belief is rational only if it is arrived at by informed judgment
• But is it possible? Do we find in experience and thoughts of believers proper
grounds for their beliefs? Adequate assessment, reflection and evaluation?
Informed Judgment in Religion
• An informed judgment for theism could be multi-dimensional or cumulative
(“the apparent power of a theistic framework to explain diverse features of
our world” in combination with “centrality of theistic religious experience”)
• Contra formal evidentialism, these cumulative judgments do not involve
mathematical probability calculus. They are qualitative and not quantitative
• Social evidentialistic cases for religious belief rejects the simple core
approach accepted by formal evidentialistic cases for religious belief
Formal cases for religion isolate core beliefs (e.g., “God exists”) that need to
be rationally justified. Everything else can be put aside until the core is justified
Social evidentialistic cases for religion focus on religious belief as a whole. They
focus on an entire complex large-scale religous world-view (such as Christianity)
• Rule-following formalists are limited to the evaluation of religious beliefs, in
trying to answer the question of whether religion is a rational practice
• For social evidentialists though, this limitation does not hold, if we assume
that judgment can include the evaluation of elements other than beliefs
Informed Judgment in Religion
• Formal evidentialists (necessarily focusing on beliefs) and those social
evidentialists who focus on beliefs make different types of reduction
Formal evidentialists make an epistemic reduction of religion. Only religious
beliefs count for the question of whether religion is epistemically justified
Social evidentialists focusing on religious beliefs make methodological reduction of
religion. For this focus doesn’t entail that only beliefs count for religion’s rationality
Neither make an ontological reduction. For neither claim that religion is nothing
but religious beliefs
• A difficulty for a social evidentialistic account of religious rationality is that
many religious believers do not arrive at their belief by informed judgment
• Religious beliefs are often a natural part of the cultural environment in which
people grow up. They are culturally produced and unreflectively integrated
• So, are most religious believers irrational? We need to ask two questions
Can and should religious believers involved in e.g. philosopy of religion arrive
at their beliefs by making informed judgments?
Can and should “ordinary” religious believers arrive at their beliefs by making
informed judgments?
Informed Judgment in Religion
• However, many beliefs are culturally produced, not only religious beliefs, but
also ordinary beliefs, secular beliefs, moral beliefs and even scientific beliefs
• There are many beliefs that people in a culture naturally accept without solid
assessment, that enjoy widespread acceptance and go with subtle influence
• It is harsh to claim that all these people are irrational. But there is a way out
for the social evidentialist: belief formation differs from belief maintenance
• The evidential principle (‘make judgments’) applies to belief maintenance
(‘how we sustain beliefs’) and not to belief formation (‘how we arrive at beliefs’)
• It is formulated as follows: People must give good (judgmental) reasons
for their beliefs – no matter how they arrived at them in the first place
• It is possible for many more religious believers to meet this version of the
evidential principle than the previous one – let alone the formalists’ one
The social principle and religion
• What implications does the second standard of rationality proposed by
social evidentialists, the social principle, has for the rationality of religion?
• Gutting argues that due to lack of consensus in religion, the social principle
isn’t satisfied, and thus religious belief is not rational on social evidentialism
• But here he invokes a principle that is not even demanded by strong social
evidentialism, i.e. the consensus principle: rational belief requires consensus
• His defence for it seems to be that it is satisfied in science, and since science
is paradigm for rationality, all other practices (incl. religion) should satisfy it
• But in many sciences there is no consensus. Besides, the consensus principle
entails that all Kuhnean periods of scientific revolution are wholly irrational
• The consensus principle is thus too demanding. Yet, if religion wants to have
the same cognitive status as science, then religious experts must at least
sometimes come to consensus about important aspects of religion
• Still, the question is whether religious belief can be rational, not whether
one can reach in religion the same level of cognitive authority as in science
The social principle and religion
• On strong social evidentialism, if there would be a consensus in religion,
then all religious believers must agree with it in order to be rational
• Yet, as Putnam has it, sometimes one seems justified in relying on one’s
own judgment even though the majority of competent others disagree
– We may only infer that prima facie communal judgments outwit personal ones
• Let us thus focus on the social principle (‘expose informed judgments to peer
review’), which is less demanding than conformity and consensus principle
• The fulfillment of the social requirement (‘evaluation by competent others’)
is only possible if we suppose there are experts in religion. Who are they?
• But this doesn’t have to be a problem. There being groups of theologians or
well-informed believers in general, seems sufficient for peer evaluation
• Besides, since scientific evidence is better supported than religious evidence,
it is rational to stick to scientific beliefs in case of a conflict. But then science
also delivers competent others for proper peer review in religious matters
Social evidentialists challenge to religious belief
• The social evidentialists challenge to religious belief is not necessarily the
same as scientific challenge, since one may accept other kinds of evidence
• Typically “when arguing about the epistemic status of belief in God we
should accept more kinds of reasons than those used in science” (Jeffner)
• A social evidentialist challenge to religious belief can be stated as follows
1. Religious beliefs are either non-basic or sufficiently questioned basic beliefs
2. A non-basic belief or a sufficiently questioned basic belief is rationally acceptable
only if it is arrived at by an exercise of informed judgment, or if it is, at least, held
on to only after it has been evaluated by an exercise of informed judgment
3. A belief is rationally acceptable only if it is submitted to evaluation by a
community of relevant expertise, and if it is rejected if a sufficient number
of experts keeps disagreeing (‘conformity principle’ of strong social evidentialism)
4. (Most) religious beliefs do not satisfy (2) and/or (3)
5. Therefore, (most) religious beliefs are irrational
Social evidentialists challenge to religious belief
• Note that the proportionality principle (‘firmness of a belief ought to be in
proportion to the strength of the evidence’) is not part of this challenge
• Contra formalism, social evidentialists reject proportionality. Thus they
do not hold that strong religious commitment is necessarily irrational
– For in science, strong commitment to a theory is essential for progress
• But one might still object to religion that a solid dialectical relation between
the attitudes of tentatively and commitment is not as apparent as in science
• A reasonable demand might be that rationality requires that a tentative
attitude must be more emphasized in religion than currently
• Yet, even if this demand is indeed reasonable, the much stronger claim
that religious belief is irrational does not follow. For that more is needed
• The social evidentialists challenge to religious belief takes science to be the
paradigm case of rationality. It is science that delivers the model. But what if
this model is not suited for other crucial practices, such as everyday life?
Chapter 8: Presumptionism
• Social evidentialism may be our best account of scientific rationality. But
one should doubt whether it is adequate as a general model of rationality
• Is social evidentialism applicable to the practice of everyday belief? And if it
isn’t, can it then be an appropriate model of rationality for religious belief?
• Answer to both will be ‘no’. Thus another model is needed: presumptionism.
It must, better than evidentialism, take into account the human condition
– Human beings are finite beings with limited cognitive resources
– Humans beings live in specific social, cultural and historical situations
• According to the presumptionist, rationality must be realistic. It cannot
reasonably require more from a real agent than what she can possibly do
• Presumptionists thus accept the Axiom of Reasonable Demand. Too idealized
models violate this axiom. They ask more than a real agent can possibly do
• What reasonably can be expected from a real agent depends on that agent’s
concrete situation and resources, on her actual predicament and constitution
• So rationality is not only person-related, but also person- or kind-relative. At
least the standards of rationality are not the same for all (kinds of) persons
– For us (but perhaps not other beings) evidentialism is too demanding in daily life
• Thus for presumptionism it is crucial to ask whose rationality we talk about.
Only when we know who the concrete agent is, can we assess its rationality
• Next to the agent’s constitution or resources (‘the who’) presumptionists
take into account their actual situation or circumstances (‘the where’)
• Presumptionism can be general in the sense that beings of the same kind
and in the same situation face the same standards of rationality
• We focus on human rationality. What is rational for a human being to belief?
For this we need to ask what cognitive resources we have, how we use them,
what are their limitations, and what are our social and cultural situations
• In addition to our (cognitive) abilities and situations, presumptionism as a
realistic model of rationality must also relate to our aims in those situations
• Formal and social evidentialists, and presumptionists accept that an
adequate model of rationality must not entail that science is irrational
– That is a major reason why social evidentialists reject formal evidentialism
• But contrary to evidentialists, presumptionists don’t accept that if a practice
does not meet the standards of rationality of science, then it is irrational
• On evidentialism “science is paradigm case for rationality” means that its
model of rationality is the norm for all other practices (and none meets it?)
• On presumptionism “science is paradigm case for rationality” means that
science is our best example of a practice that is to be considered rational
• A reason for rejecting that science’s tough rationality model is the norm for all
other practices, is that (because of our condition) they would all be irrational
– Such a conception of rationality is still too idealized to apply to the whole of life
• If we focus only on our best example of rationality, we get a distorted view of
what rationality in general is all about. What’s the least we can and should do?
• Everyday life beliefs (real life) is also an important control case of rationality
Everyday belief
• Everyday belief is also a paradigm case of rationality in that it is not the best
but the most we humans do. Daily life constitutes the largest belief domain
and nobody can do without them. They are indispensable for our lives
• Hence, an adequate general model of rationality must be able to make sense
out of everyday belief formation and regulation (not render them irrational)
• The question is not whether we are rationally in daily life or not, but to what
extent we are. We require only minimal rationality, for else we’re all irrational
• Thus formal evidentialism (“perfect”) and social evidentialism (“very good”)
shouldn’t be accepted as general models. For they render daily life irrational
• Only presumptionism (“good enough”) renders most daily life beliefs rational
• These claims are premised on claims about our finite cognitive resources
Internal rationality
• Human memory is a crucial cognitive capacity for our beliefs. It consists of
two units: a small active short-term and a large passive long-term memory
• What we can think about at a moment is what is in short-term memory. But
we can recall beliefs from long-term memory by copying them into short-term
• Our enormously large belief systems are almost entirely in our passive longterm memory. Hence we cannot consider (the implications of) all our beliefs
• So the principle of deductive closure (“Believe all implications of your beliefs”) is
unreasonable. For we only oversee the implications of beliefs in short-term
• We likely only notice inconsistencies between short-term beliefs. Consistency
principle (“Total belief system must be consistent”) is therefore also unreasonable
• What limited ability to detect implications and eliminate inconsistencies can
we reasonably expect? It is for us wholly rational to use sub-optimal heuristics
• And this not only due to our short-term memory limitations, but also because
computing all deductions and full consistency check would take too much time
Internal rationality
• Formalists can respond that principle of deductive closure and consistency are
ideals that we strive to reach. We should at least do our best. But should we?
• Presumptionists hold that this still requires too much. Many inferences are useless. They prevent us from using our limited abilities to do more urgent things
– It is a waste of time to continuously trace irrelevant inferences from our beliefs
• So, presumptionistic principles for internal rationality take into account not only
logic validity, but also feasibility and usefulness for people in a given situation
• Presumptionistic principle of deductive closure: People must make inferences
that are sound and useful for them in their current situation – given that the
inferences are derivable from beliefs in (or close to) their short-term memory
• Presumptionistic principle of consistency: People must eliminate inconsistent
beliefs when elimination is feasible and useful for them in their current situation
– given that the inconsistencies are in (or close to) their short-term memory
• Internation rationality is thus not settled by logical features of beliefs alone. We
need to take into account our actual cognitive resources and concrete situations
External rationality
• What are the consequences of presumptionism for external rationality?
• Evidentialists hold that we must have good reasons for our beliefs. We have
an intellectual duty to hold only beliefs for which we have sufficient evidence
• But in practice we do not always have sufficient evidence for all our beliefs.
Are we thus all irrational? Should we reject the vast majority of our beliefs?
• If we take the evidential principle seriously, then we cannot longer function
properly as human beings. This is not reasonable. We simply cannot in our
short lifetime search for (and think through) the evidence for all our beliefs
– Constantly searching for proper evidence wastes our limited resources and time
• But then evidentialism violates the axiom of reasonable demand. It is not
a realistic principle for external rationality. We often must act out of habit
• Also, many beliefs aren’t usefull enough to spend costly validation-time on
• One may restrict evidentialism to important beliefs only. But even then it isn’t
reasonable. We cannot ground them all. For we would never get around in life
The Principle of Presumption
• Instead of modifying the evidential principle, presumptionists reject it. For
they argue that evidentialists get things the wrong way from the start
• Evidentialists start from high suspicion. We are not allowed to believe something until we have good reasons for it. Beliefs are guilty until proven innocent
• But according to presumptionists a constant questioning of our beliefs is in
fact irrational. For it wastes too much of our limited cognitive resources
• Presumptionists claim instead that the rational initial attitude towards our
beliefs is trust and not distrust. Our beliefs are innocent until proven guilty
– Evidence is first negative (rationality-removing); not positive (rationality-establishing)
• The principle of presumption: It is rationally permitted to accept a belief unless
there are good reasons for thinking that it is not true (sufficient counter evidence)
• On presumptionism, in the absence of defeaters, one does not have to provide
sufficient evidence for ones beliefs since belief-formation is presumed reliable
– We have epistemic right to initially rely on our cognitive equipment and processes
• Principle of presumption is more realistic than the evidential principle. It does
not waste limited time and resources, and meets our need to start somewhere
The Principle of Presumption
• Presumptionism is not fideism because the former does not deny that there
might be good evidence for a belief, or that evidence can count against a belief
• Like social evidentialism, presumptionism rejects the proportionality principle
• On Presumptionism “situations of resistance” are counter-evidence. We reject
our belief or we justify it by arguing that counter-evidence isn’t strong enough
• Do presumptionists allow more as rational than is reasonable? Take John who
just decides to believe there is a parallel wholly blue universe. Is he rational?
• Presumptionism does not imply that John is rational. He knows that the way
he arrived at his belief is unreliable. Hence he has a good reason to give it up
• Presumptionists also reject the social principle of social evidentialism. But
they do accept that judgment and not rules is our primary mode of evaluation
• Presumptionists also accept that the locus of rationality is persons, not beliefs
Rationality, knowledge and truth
• What turns a rationally justified belief into knowledge? On many accounts
of knowledge it is at least the objective truth of the belief in question
• So, rationality has to do only with the subjective pole, i.e. the subject and her
cognitive belief processes. Knowledge refers also to objective truth of beliefs
• Rationality is only about what can be rightly demanded or expected from
people with respect to their beliefs. And ‘truth’ is not one of these demands
• A subtle distinction can be made between rational and epistemic justification
Rational justification has to do with when (under what circumstances)
a person is entitled – has the intellectual right – to believe what she believes
Epistemic justification is much stronger. It has to do with when a person has
provided sufficient grounds for a belief to render it generally acceptable
• If a rationally justified belief also qualifies as an epistemically justified belief,
then on the JTB account of knowledge, only ‘truth’ is needed for knowledge
• If not, then what is needed to turn the rationally justified belief into knowledge is its truth plus what turns the rational into an epistemic justification
• Warrant turns true belief into knowledge. On JTB it is ‘epistemic justification’
Rationality, knowledge and truth
• Given the aforementioned we can distinguish two different ambitions
(a) Rationality ambition – Are we in our intellectual rights in believing a particular belief?
(b) Knowledge or truth ambition – Is a particular belief a piece of knowledge or true?
• On presumptionism (a) does not entail (b). For Ed may rationally believe that it
is the case that John rationally believes P and Mary rationally believes not-P
• The example also holds for social evidentialism but not for formal evidentialism.
For on the latter there are rules that decide which one of both beliefs is rational
• Contrary to social evidentialists, presumptionists do not require that a rational
belief rests on the best available judgment and on the best available evidence
• On presumptionism Ed can be rationally entitled to believe something that the
experts reject. The experts aren’t entitled to it. For they have counter-evidence
• Yet, on presumptionism there can be a strong link with truth. Take scientists
who accept a theory for which they have argued that all proposed defeaters fail
• Thus whether presumptionism is weak or strong depends on the context, i.e.
the cognitive resources, the situation (and the aim) of the person in question
Another objection to presumptionism
Consider Brigitte. She believes something that is quite normal to believe. But
she deliberately avoids all situations in which she could be confronted with
counter-evidence to her belief. She is clearly not behaving rationally. But on
presumptionism her belief would still be rational. Thus presumptionism fails
• This objection leads to a revised version of the principle of presumption: It
is rational to accept an important belief only if one actively tries to find
good reasons to think otherwise, but is in fact unable to find any
• Thus we have to actively put ourselves in situations where we can expect
resistance. We have to actively test our important beliefs with opponents
• But is this revision really required? Aren’t we confusing the rationality ambition
with the truth ambition? The objection seems only an argument for the claim
that we must do more if we want truth and not only rationality. But that’s clear
Presumptionism as a general principle?
• If the main reason for accepting presumptionism is that it is more economical
than evidentialism, why should we accept it then in cases where (i) it is worth
spending more of our resources, or (ii) where the cognitive costs are low?
• In short, if a belief is important and the cognitive costs low, evidence is needed
• Take the following example: Suppose we are mountain climbing and I set the
anchor for our life line, and I form the belief that this anchor point is solid. In
this case I have an intellectual duty to test my belief, to obtain evidence for it
• Yet, examples such as these do not count against presumptionism as a general
principle, that is, how we should normally proceed. It shows only that in some
cases it is rational to spend more of our cognitive resources than usual
Low costs
High costs
Further objections against presumptionism
Take Mike. His cognitive equipment malfunctions. He forms the belief that
people who smile at him hate him. He isn’t able to grasp counter-evidence,
even if presented. On presumptionism he is rational. But that seems false
• Objection fails. For Mike hasn’t generic rationality (‘having capacity to reason’)
which is a necessary condition for normative rationality (‘reasoning properly’)
• If Mike is irrational in generic sense, he cannot be rational in normative sense
• Presumptionism is a model for normative rationality. Thus generic rationality
is a precondition. The model assumes that cognitive equipment isn’t broken
• On presumptionism what is rational to believe for us is obviously situation
relative (“Believing that P might be rational for you in situation S1 but not in S2”)
• But from this it doesn’t follow that the standards of rationality (e.g., principle
of presumption) are situation relative. Let alone rationality’s nature (duty)
Science and Presumptionism
• One could argue that presumptionism rather than social evidentialism is
not only the adequate model for everyday life – but even for science itself
• For the demands of social evidentialism seem reasonable for epistemic justification – but not necessarily for rational justification. After all, fulfillment of
the principle of presumption seems sufficient for being scientifically rational
• Indeed, social evidentialism makes sense for epistemic justification, which
is all about whether a belief can be generally accepted by the community
• But, contrary to presumptionism, it doesn’t make sense for rational justification.
For it entails that I cannot be rational if I’m alone – which is utterly implausible
• Besides, social evidentialism as a standard for rationality wastes far too much
of our limited time and resources. Do I need peer review to be entitled to
believe that I’m tyred or that Brigitte lives in Amsterdam? Clearly not
• But then presumptionism appears to be the best generic model of rationality.
For it satisfies our two paradigm cases of rationality: everyday life and science
• From this one may suspect that it is the most adequate model for religion too
Chapter 9: The Nature and
Function of Religious Belief
Nature and function of religious belief
• In order to answer the question of whether religious belief is rational, we
must first clarify our account of rationality
• Rationality should be understood realistically. It has to do with what a
person can be expected to do, given his or her resources and situation
• Rationality has to take into account the character (of the beliefs) of the
practice in question. What’s its nature? What function does it fulfill?
• Therefore, rationality is both person-relative and practice-relative. We must
focus both on the kind of agents and the kind of (beliefs of the) practices
• Hence, to identify the standards of rationality for evaluating religious belief, we
must discover what’s distinctive of the practice. What’s its nature and function?
• So, what then is characteristic of religious believing?
The function of religion
• Religious beliefs try to answer existential questions. Religious beliefs are formed
as a response to existential experiences. Religious beliefs are existential beliefs
• Religion is broadly speaking concerned with the human search for meaning. It
deals with our existential concerns, constraints and longings as human beings
• As such, religious beliefs are life-orienting beliefs. They deeply affect how we
live and understand our lives. They largely determine the direction of our lives
• Human beings cannot avoid sometimes raising and responding to existential
questions, either explicitly or implicitly in the way we choose to live our lives
• If we consider religion as an existential stance, then a Marxist or a Materialist is
not less religious than a Christian or a Hindu. They all have life-orienting beliefs
• Humans develop a view of life or world view in response to existential questions
and existential experiences. Now, instead of considering all views of life religious,
it seems more adequate to distinguish between religious and secular views of life
A view of life
• Existential responses don’t have to be explicit, complete, integrated or unified.
Life views can thus be (partially) implicit, incomplete, fragmented or plural
• Therefore it is better to talk about a view of life than a world view, since ‘world
view’ has the connotation of an all inclusive comprehensive picture of the world
• As with any view of life, the choice for a fragmented or plural view of life can
be either conscious (deliberate) or unconscious (without being aware of it)
• Given the aforementioned, every human being has a view of life, though not all
humans have a world view. And having a world view entails having a view of life
• Hence, our only existential choice is between having a religious or secular view
of life. Moreover, both types of life views can be either unified or fragmented
• What is important for the question of the rationality of religious believing, is
that the choice thus cannot be between having a view of life or not
A view of life
• According to Jeffner a view of life consists of three elements
Cognitive element – A person’s comprehensive view about the nature of humans,
the world and the relation between humans and world (our place in the world)
Evaluative element – A person’s central values (his or her view of ‘the good life’)
Affective element – A person’s basic mood or attitude (the ‘feeling’ towards life
– e.g. fear, despair, trust or hope – which gives an emotional color to experiences)
• One might object against Jeffner that fragmented views of life does not include
a comprehensive picture of the nature of humans, reality and their interrelation
• Moreover, the first element is called cognitive, which seems to exclude without
warrant cognitive accounts of value. It can better be called the doxastic element
Doxastic element – A person’s beliefs that are of crucial importance for what it
means to exist in the world as a human being
A view of life
• Another problem with Jeffner’s account is that it’s a rather theoretical notion of
a view of life. It’s an emotionally colored conceptual system of beliefs and values
• But a view of life is not merely a vision of life, it is always a vision for life. It leads
us in a particular way. It regulates the way we live our lives. It guides what we do
• Every view of life is practical in the sense of being actually lived. It ultimately is
about what way of life one adopts and not what theories of life one accepts
• Views of life are thus not exhausted by conceptual systems. Rather, conceptual
systems are derived from a view of life. They express its intellectual dimension
• Therefore, we need to add a fourth element to Jeffner’s account. This element
is to capture the life-directing character of a view of life. Its disposition to act
Regulative element – A person’s volitional affirmation of the practical
consequences (action) of the doxastic, evaluative and affective elements
A view of life
• By combining the four elements we get the following definition of a view of life
A view of life consists of the assumptions that constitute or have a decisive importance for a picture of human beings and the world, that form a central system of
values, that express a basic mood, and that have a regulative function in our lives
• From this it follows that every view of life has a twofold function
A theoretical or perceptual function – It structures and makes reality intelligibly. It
determines the place of human beings in it. It also states what is worthwhile in life
A practical or regulative function – It guides the adherents in their concrete lives.
It leads people in how they should live their lives. It is directing or life-orienting
• Views of life are not merely intellectual, but also – and most of all – practical
answers or responses to our existential concerns, constraints and longings
– Existential questions require such answers. For else the question is misunderstood
• Views of life cannot be reduced to metaphysics. But the intellectual (cognitive
and evaluative) dimension of a view of life can be called a metaphysical picture
The conception of religion
• It is reasonable to think that everyone has a view of life. But what is the main
difference between a secular and a religious view of life?
• According to Holte “religion is a view of life marked by a trust in divine being(s),
together with ritual behavior”. But this is too narrow (e.g., versions of Buddhism)
• What seems characteristic of religious believers is that they are convinced that
the sacred provides a way of solving our existential problems
• So if we add the following element to a view of life we have a religion
Transcendent Element – A person’s consciousness of and trust in the sacred
• The sacred or holy transcends the profane or ordinary reality. It can refer to
what goes beyond the universe, to the depth of being or to something else
• Moreover, the sacred can be conceived of as unity or diversity, as personal or
impersonal, and so on. Different religions have different notions of the sacred
• The sacred is felt and viewed as a mystery. It goes beyond our understanding
The conception of religion
• It gives the lives of religious believers substance and meaning. It also gives
strength and wisdom to deal with existential experiences and constraints
• The sacred evokes feelings of joy, grace, awe and fascination. But also, when
religious believers experience the majesty of the holy, they tremble (tremor)
• In his book ‘The Holy’ Rudolf Otto thus characterizes the sacred adequately
as mysterium tremendum or mysterium tremenda majestas et fascinans
• Contra Holte, this notion of religion includes views of life that we intuitively
consider religious, but it is still narrow enough to exclude secular views of life
• Religion is oriented towards (is experience of) the sacred or the numinous, which
is compatible with monotheïsm, Buddhism, Hinduism and many other religions
• Given that a religion is a view of life, it cannot be reduced to a set of beliefs. As
a view of life religion is life-directing. Religion has to be lived. It is a way of life
Existential questions and religious belief
• In general there are five different sorts of questions
Practical questions – Questions on how to do certain things (How to rent a car?)
Factual questions – Questions on what is the case (Do electrons exist?)
Evaluative questions – Questions on what is good or right (Is torture wrong?)
Conceptual questions – Questions on the meaning of concepts (What means ‘factual’?)
Existential questions – Questions on the meaning of life (Why suffering? Is there hope?)
• A life view is an answer to existential questions. What are their characteristics?
• These questions arise because of the kind of beings humans are. We are selfreflective. We are able to reflect on our existence. We take a stance towards it
• These questions have a practical component. They are asked by agents who are
prepared to act, who want to participate in what’s happening. “What can I do?”
Existential questions and religious belief
• These questions also ask what we should do. They also include a normative
component. “What sort of posture do I ought to adopt to myself and life?”
• These questions demand an urgent answer. They are action-obligatory, timebound and irrevocable. They concern problems from which we can’t withdraw
– If we neglect to act, we also act. We must do something. We can’t withold judgment
• These questions are also of utmost importance for us. They deeply affect how
we understand ourselves and the world we inhabit. And how we live our lives
• The normative component of existential questions (“What should I do?”)
implies that an answer to them is not purely theoretical, but life-orienting
• But this does not mean that existential questions are only expressive or nonfactual. An answer must contain both a practical and theoretical dimension
Existential questions and religious belief
• Existential questions can be about the same subject as factual or evaluative
questions, but an answer to the former must include what the agent has to do
• An answer to an existential question is how to respond to existential constraints
and existential experiences. The religious answer is that the only way to respond
is to let the sacred (the transcendental dimension of reality) transform our lives
• For religious believers the sacred is of ultimate concern (Paul Tillich). The holy is
more real and of more value than anything else. It’s what they care most about
– It often takes the form of worship and then involves praise, love and gratitude
• For something to be a completely appropriate object of ultimate concern and
fully worthy of worship, it must be the greatest of all possible reality (Anselmus)
– The sacred is thus very different from the objects of scientific or everyday belief. This
is to be taken into account when proposing standards of rationality for believing in it
• A choice for a view of life is an unavoidable existential choice. It’s not purely theoretical, but still cognitive. It’s a choice of what to live for. It’s not only a choice of a
vision of life but a vision for life (regulative). Is that kind of choice rational or not?
Chapter 10: Religious Rationality
Religious Rationality
• Standards for rationality are relative. The question is what is rational to believe
for person N, in situation S, in practice (mode of form of life) P, having goal G
• For determining standards, it is important to consider the aim of a practice.
So, what is the aim of religion and does it differ from science or daily life?
• We must also consider the axiological dimension of rationality: what kinds
of aims is it rational to try to satisfy in our life? What should we value?
• Who we are and our situation affects what aims we can reasonably have
– We are finite beings with limited cognitive resources
– We are situated in an already existing world, with a certain structure
– We have needs that we cannot avoid (e.g., daily life needs and existential needs)
• Surely, people do not need everything they desire, and people do not desire
everything they need. A need is what is important to one’s human flourishing
• The kind of being we are determines a range of ends that are in our interest,
that are good for us (e.g., nourishment, protection, freedom, friendship, love)
Religious Rationality
• On presumptionism a truly rational person is one who takes into account her
own limitations, situation and real interests in deciding what to do and belief
• We are in-the-world and need our beliefs and activities to do certain jobs for
us. In science to predict and control, in views of life to respond to existential
demands, and in daily life to function properly (survival, well-being, etc.)
• But then it is incorrect to claim that the appropriate ends of believing are only
epistemic. Epistemic rationality is not the only form of theoretical rationality
• In fact, the aim to acquire as many true beliefs as possible, inclusing useless
beliefs as the number of grains of sand, is irrational given our predicament
• Even in science, epistemic goals are not the only goals for accepting theories.
Theories are also accepted because of their usefulness, predictive power, etc.
Contemporary Philosophy of Religion Debate
• An inadequate reduction of rationality to epistemic rationality is often made
with respect to religion. For often only epistemic ends are considered relevant
• But real religious believers are not fictive abstract pure epistemic agents who
only evaluate beliefs epistemically. Humans also have non-epistemic interests
– A choice of religion is not purely a theoretical-epistemic choice
• A purely epistemic approach to religion isn’t religiously relevant. It fails just like
formal approaches that do not take into account the actual practice of science
• Acceptance of religion for the actual religious believer is an existential (yet still
cognitive) choice. It is about chosing a particular view of life for living one´s life
• Thus an evaluation of rationality that treats the matter purely theoretically, that
does not take into account that we all have to choose a way of life, is inadequate
• Contrary to idealized agents, actual human agents cannot postpone their choice
for a view of life until sufficient evidence has been gathered. We must live now
Contemporary Philosophy of Religion Debate
• What then is the aim of religious practice? Views of life not only aim to make
our lives theoretically intelligible, but primarily also existentially intelligible
• Views of life make sense out of our existential experiences, and help us to
find a way through existential constraints to fulfill our existential needs
• Since this end is rational (based on intrinsic human needs) it is with respect to
that that religious beliefs must be evaluated (by evidentialists or presumptionists)
• To avoid a purely epistemic interpretation both principles must be reformulated
The revised evidential principle – It is rational to accept a non-basic belief only if
there are good reasons for it
The revised principle of presumption – It is rational to accept a belief unless there
are good reasons to cease from holding it
• Both principles are open to epistemic and non-epistemic goals. In the case of
religion: making reality intelligible and guiding adherents in their concrete lives
• So, only standards of life-view rationality that take into account the twofold
function of views of life (theoretical and existential) can be adequate
Do Science and Religion have different rationalities?
• Science and religion have different ends. But this does not entail that they
have different conceptions of what rationality is (the nature of rationality)
• Science and religion having different aims also does not entail that they have
different generic standards of rationality (evidentialism or presumptionism)
• But it entails that the detailed standards of rationality for both are different.
For what count as good reasons depends on what the end in question is
– “No predictive power” counts against scientific theory – but not against view of life
– “No moral insights” counts against view of life – but not against scientific theory
• Still, some reasons are relevant both for religion and science (e.g., consistency)
• We should distinguish between reasons that are relevant for all practices (basic
reasons) and reasons that are dependent on a specific practice (derived reasons)
Good reasons in science
A good reason for (or against) the rational acceptance of a set of beliefs might be
(1) It is logically consistent, that is, it avoids self-contradiction
(2) It is consistent with other already accepted sets of beliefs
(3) It is coherent, that is, its components hang together
(4) It is coherent with other theories
(5) It is less complex than rival theories
(6) It makes possible the prediction of new phenomena
(7) It provides illuminating explanations of puzzling phenomena
(8) It is more comprehensive than other rival theories
(9) It is easier to apply than its rivals, that is, it is practically more useful
Good reasons in science and religion
A good reason for (or against) the rational acceptance of a set of beliefs might be
(1) It is logically consistent, that is, it avoids self-contradiction
(2) It is consistent with other already accepted sets of beliefs
(3) It is coherent, that is, its components hang together
(4) It is coherent with other theories
(5) It is less complex than rival theories
(6) It makes possible the prediction of new phenomena
(7) It provides illuminating explanations of puzzling phenomena
(8) It is more comprehensive than other rival theories
(9) It is easier to apply than its rivals, that is, it is practically more useful
Good reasons in science and religion
A good reason for (or against) the rational acceptance of a set of beliefs might be
(1) It is logically consistent, that is, it avoids self-contradiction
(2) It is consistent with other already accepted sets of beliefs
(3) It is coherent, that is, its components hang together
(4) It is coherent with other theories
(5) It is less complex than rival theories
(6) It makes possible the prediction of new phenomena
(7) It provides illuminating explanations of puzzling phenomena
(8) It is more comprehensive than other rival theories
(9) It is easier to apply than its rivals, that is, it is practically more useful
(10) It helps satisfy intrinsic human needs, that is, it is adequate for human lives
(11) It answers or helps deal with existential matters better than other accounts
(12) It helps people better than other accounts to life a morally good life or the like
The scientific challenge reconsidered
• According to the scientific challenge, religious belief is irrational since it
does not meet the same standards of rationality as scientific beliefs
• If this challenge succeeds, than the following challenge succeeds as well
(1) Secular or religious views of life must fulfill scientific standards
(2) Secular or religious views of life do not fulfull these standards
(3) Therefore, secular or religious views of life are irrational
• Thus, it would follow that all views of life are irrational. But since every human
necessarily has some life view (to function properly and to meet our existenial
concerns) it follows that all humans are irrational – which is absurd
• Advocates of the challenge may respond as follows. It might be necessary for us
to choose a view of life, but that has nothing to do with rationality. It is merely a
matter of taste (just as some like pizza and others not) and thus a-rational
• But life views are sufficiently similar to scientific theories and everyday life beliefs.
So we can rationally compare life views, although the standards differ somewhat
The scientific challenge reconsidered
• And given that we have to choose some view of life, the crucial question is not
whether religion is or is not scientifically acceptable, but whether religious life
views are or are not less rational than secular life views
• For evidentialists it is important to note that the rational evaluation of views of
life is not a two-way confrontation between a view of life and the evidence. It
is at least a three-way confrontation between rival views and the evidence
• Given that we cannot withhold judgment, a view of life shouldn’t be abandoned,
even if there exists evidence against it, except in favor of a better view of life
• To assess the rationality of life-views, the focus should not be on arguments for or
against the existence of the sacred (propositions), but on arguments for or against
the rationality of believing in the existence of the sacred (real people in situations)
– “When is a belief P rational?” differs from “When is an agent rational in believing P?”
• The former has to do with whether the belief that the sacred exists is true or
should be epistemically accepted. The latter is about whether specific people in
their concrete situations are rational to believe in the existence of the sacred
The evidentialist challenge reconsidered
• According to the evidentialist challenge, most religious people are irrational in
their religious beliefs, since they do not base these beliefs on adequate evidence
• But if that challenge suceeds, this one also suceeds: “Most secular people are
irrational in their secular life-views (e.g., “man is the measure of all things”, “reality
is ultimately impersonal”, etc.) since they do not base it on proper evidence
• Thus, most religious and secular people would be irrational, which is absurd
• The evidential principle is thus problematic. Indeed, even in daily life most people
do not follow it. It would in fact be irrational given our limited time and resources
• That is why the presumptionist ‘turns the tables’. It is rational to trust the deliverances of our cognitive resources as long as there is no sufficient counter-evidence
Evidence, grounds and skepticism
• To say that one needs no evidence for one’s beliefs does not mean that the
beliefs are groundless. They are grounded in the source from which they arise
• The presumptionist merely argues that we have the right to initially trust our
cognitive faculties and their deliverances (our natural belief-forming processes)
• We may initially rely on the deliverances of our belief-producing mechanisms,
unless we obtain good reasons to think that these deliverances are not reliable
– What else can we rationally do except initially trust the only light we have?
• The evidentialist cannot even give adequate evidence for the (un)reliability of
our belief-forming processes. For any argument relies on our cognitive faculties
• The presumptionist claims that we do not need to justify these processes to be
rational in believing them. We trust our beliefs until counter-evidence emerges
• Of course foundational evidentialists partially agree, namely with respect to
the proper basic beliefs. But this class is typically taken as being very limited
Evidence, grounds and skepticism
• Presumptionists claim that it is first in situations of resistance (‘defeaters’) that
a rational person should start to challenge his or her initial trust – and only then
• Presumptionists treat all our naturally formed beliefs as properly basic, not out
of laziness, but because rationality must be realistic given our time and resources
• But how then can one be irrational on presumptionism? One answer is that one
is irrational if one fails to take into account things that count against ones view
• More specifically, one can be irrational in respect of ends, means and situation
With respect to situation – Rational individuals must take into account their resources
and the environment they live in. “Be realistic. Do not overestimate time or abilities”
With respect to means – Rational individuals must select means that are sufficient and
appropriate. “Do not select means that do not satisfy your ends or are destructive”
With respect to ends – Rational individuals must select ends that do not go against
their real (daily and existential) human interests
Presumptionist challenge to religious believing
• On presumptionism it is rational to accept a life-view only if there are no good
reasons not to do so (such as there being a rival life-view one is better off believing)
• A religious believer is irrational if she is aware of good reasons against her belief,
(‘defeaters’) or if she is not open to potential defeaters that are accessible for her
• What this all amounts to is that in the end the unfriendly religious believer (“No
one is rationally justified to adopt a secular life-view”) and the unfriendly secular
believer (“No one is rationally justified to adopt a religious life-view”) are wrong
• The most reasonable position to adopt is either to be a friendly secular believer
(“some religious believers are rational in their life view”) or to be a friendly
religious believer (“some secular believers are rational in their life view”)
– Since rational belief does not entail true belief, friendly believers aren’t inconsistent
• The traditional issue concerning whether people are rational in their religious or
secular beliefs has come to an end. But it is one thing to be entitled to accept a
belief, quite another to be entitled to believe it with a particular strength
Religious commitment and rationality
• In addition to what (the content) and why (the ends) we should also look at
how one believes (‘with what strength or the degree of commitment’)
• What bothers non-religious people is that religious believers typically believe
in the sacred “no matter what happens”. Here lies a real problem for religion
• The tentative challenge to religious belief is that these beliefs ought not to be
held as firmly as is normally done. The assent is to be interim and not decisive
• In the case of interim assent a belief is tentatively and not fully accepted. That
is to say, one accepts it but also thinks there is a need to investigate it further
• On the principle of tentativity all beliefs should be tentatively accepted. No
belief is beyond doubt. One should always be open to possible defeaters
• The tentative challenge has it that holding religious beliefs tentatively may be
rational for presumptionists, but holding them in a religious way (“fully”) is not
• On the tentative challenge the only reasonable version of presumptionism is
tentative presumptionism: One is rational in tentatively accepting a belief
unless there are good reasons to cease from believing it
In defense of full presumptionism
• According to full presumptionism one is in fact rational in fully accepting
a belief unless there are good reasons to cease from accepting it fully
• Stenmark defends full presumptionism against the tentative challenge
• Tentative acceptance is hard. Cognitive costs are high. One must keep a detailed record of all tentatively accepted beliefs, justifies and entailment relations
– Constant questioning is not rational. It wastes too much of our resources
• There are examples where full acceptance fits well with presumptionism, e.g.
asking the direction to the train station or believing that your partner loves you
– It is irrational to constantly question whether she loves me. It is even destructive
• Full acceptance is not the same as dogmatic acceptance. For the belief that no
further inquiry is needed does not entail disregard for future possible defeaters
– Full acceptance of P is not the same as dogmatically believing in P no matter what
• Given function and end of religious practice, full belief (trust) is more adequate
than tentative belief. Although it isn’t necessary. There are tentative believers

similar documents