Peterson&Vanarragon (eds) - Contemporary Debates in Philosophy of Religion

CHAPTER

THREE

Does Science

Discredit Religion?

In this debate, John Worrall argues that there are irreconcilable differences between

science and religion. A fundamental difference pertains to methodology: science

evaluates candidates for belief only on the basis of evidence, while religion patently

does not. So, Worrall concludes that a properly scientiﬁcally minded person cannot

give credence to religious belief. Del Ratzsch argues that the relation between

religion and science is much less clear than that, and that the arguments that

science discredits religion are not nearly so powerful as some have believed.

Science Discredits Religion

John Worrall

We get the ages of rock, and they get the rock of ages; we work out how the heavens

go and they work out how to get to heaven.

– Old saying amongst some scientists

Strong Son of God, immortal love,

Whom we, who have not seen Thy face,

By faith, and faith alone, embrace,

Believing where we cannot prove.

– Tennyson, In Memoriam

1Introduction

Science and religion are in irreconcilable conﬂict – or so I shall argue in this chapter.

There is no way in which you can be both properly scientiﬁcally minded and a true

religious believer.

This might seem a surprising thesis in view of the undoubted fact that many

scientists (even some of the most eminent ones) were or are also religious believers

of one sort or another. But this results, I hold, from a mixture of three factors: (i) a

simple failure to think things through fully; (ii) a failure to be properly scientiﬁc (as

I shall explain, this involves more than simply giving due weight to well-accredited

scientiﬁc results and theories, it also involves bringing a scientiﬁc attitude to the

appraisal of claims and the weighing of evidence in general); (iii) adopting the atti-

tude hinted at in the ﬁrst of my epigraphs – one that holds that science and religion

not only do not conﬂict, they cannot conﬂict, because they cover quite different

domains.

The structure of my argument is very simple. I begin by analyzing the attitude just

mentioned in (iii), arguing that, when properly understood, that attitude is (essen-

tially) untenable for a religious person – the cost of adopting it is too high. This entails

that both the scientist and the religious believer are playing the same game; they are

both making substantive, descriptive, “explanatory” claims about the way the world

is. But then they must surely also play by the same rules – all such claims must be

judged by how well they stand up to the evidence. This, of course, is indeed how

claims in science are appraised and accredited. When religious claims are appraised

in this way, however, they all turn out to be untenable. Science, or rather a scientiﬁc

attitude, is incompatible with religious belief.

2Two Separate Domains?

The view is perhaps increasing in popularity that science and religion are about

different domains, are two different “non-overlapping magisteria” (NOMA), as the

eminent biologist Stephen Jay Gould recently put it,

and hence that, when properly

understood, there can be no conﬂict between them. I can see three ways in which

that view might be interpreted.

On one interpretation the view simply attempts to legislate away any clash by

creating a separate, “spiritual reality,” alongside ordinary “material reality” – science

teaches us about the latter, religion about the former. But this is based on a confu-

sion – it elevates a (perhaps natural, but clearly sloppy) way of speaking into an obvi-

ously untenable ontological doctrine. There is only one reality; that reality either does

or does not contain a god, an afterlife, or whatever, just as it either does or does not

contain quarks or superstrings or whatever; and the question that needs to be

addressed about both sets of equally unobservable (alleged) entities is what evidence

Chapter Three

1See S. J. Gould, Rocks of Ages: Science and Religion in the Fullness of Life (London: Jonathan Cape,

2001).

we have for their existence.

To take an analogy: defenders of the paranormal, like

Uri Geller, may speak loosely of “another reality” beyond the mundane one, but what

they really mean of course is that extrasensory perception and psychokinetic powers

are aspects of this reality – what else could they mean? If what they claim to be true

is true, then ESP and psychokinetic powers are parts of this reality (what else?). The

question is simply whether what they say is true (or rather, since they are talking

about (alleged) things that are not directly observable, whether there is good evidence

that what they say is true). If religion is committed to making allegedly factual asser-

tions about the world, whether about its material or (supposed) spiritual aspects, then

science and religion seem to be two competing “magisteria,” not distinct ones.

Or are they? On a second way of understanding the NOMA view (perhaps to be

thought of as a reﬁnement of the ﬁrst), religion does indeed make descriptive, factual

assertions about the universe – even about some of its “material” aspects – but there

is no conﬂict because religion kicks in only once science has gone as far as it can.

Newton’s theory may have given a perfect explanation of the movements of the

planets (let’s suppose), but of course it gives no explanation of how those planets

were created. Religion should avoid interfering in the law-governed “mundane” reality

of planetary motions (religious sentiments led even Newton himself to make a mistake

here), but it comes into its own at the level of creation. One problem with this version

of the view, as the example illustrates, is that the line between what science can

explain and what it cannot has a habit of shifting – we do now have well-accredited

theories of the formation of the solar system, ones radically at odds with those

invented earlier by theists.

But, shifting or not, there always is a line – at any stage in science, there will be

features of the universe (those described by the most fundamental theories then avail-

able) that science treats as “bottom line”: being most fundamental exactly means that

those theories cannot (to repeat: cannot at that stage in science) themselves be

explained. The suggestion, then, might be that religion can penetrate to a deeper level,

by explaining why those scientiﬁcally basic theories are true; and that there is no

clash because, by deﬁnition, science has nothing to say at that deeper level. But think

what such a claim would entail. Surely, in order to count not as mere speculative

assertions, but as genuine explanations, any such ideas presented by religion would

have to carry some rational warrant. But, on the assumption that there is only one

set of standards for appraising substantive explanatory claims about the world in the

light of evidence, this makes this second interpretation of the NOMA view incoher-

ent. If all explanations involving substantive, synthetic claims about the world must

satisfy the same criteria, then it is simply nonsense to claim that religion can explain

the scientiﬁcally inexplicable. To deny the assumption and assert instead that there

are different standards of explanation in the different ﬁelds is simple relativism. And

I assume that this is acceptable to no one in this debate. For one thing, if we allow

Science and Religion

2Those of you who may have heard of “multi-universe” interpretations of quantum mechanics should

not be confused. According to such interpretations (which, by the way, have precious little to recommend

them), the universe has many (causally non-connected) sub-parts. (Again, there is nothing else they could

mean!)

different standards for explanations in religion, why not also in the study of the para-

normal, or voodoo or scientology or...and so the list goes on? (I will address this

view again, from a somewhat different angle, later.)

I take it, then, that the only coherent version of the NOMA doctrine, the only one

that thoughtful commentators might really want to defend, is quite different from the

two so far considered. This third version sees religion as advancing no descriptive

doctrine at all, as making no real claims about the way the world really is, but “only”

as making claims about what is and what is not valuable, what is and what is not a

worthwhile life. This view concedes to science exclusive rights to inform us about the

world of fact (understood in the broad sense to include general structural features)

and accepts that religion is restricted to the world of value. This is certainly the version

advocated by Gould:

Science tries to document the factual character of the natural world, and to develop

theories that coordinate and explain these facts. Religion, on the other hand, operates

in the equally important, but utterly different realm of human purposes, meanings and

values – subjects that the factual domain of science might illuminate, but can never

resolve.

Obviously, this view of religion does indeed eliminate any possibility of a clash

between it and science.

But at what a price! (My advice here to religious people would

be to avoid scientists bearing gifts.) For one thing, the religious person who adopts

this view cannot call on any of the usual justiﬁcations for whatever value system she

endorses. She cannot, for example, claim that one ought to “love thy neighbor as

thyself” because this is what pleases our loving creator. This justiﬁcation, of course,

along with any of its ilk, involves a claim of exactly the sort from which she must

now abstain. Avoiding the conﬂict in this way means abstaining from asserting any

descriptive claim – not just speciﬁc claims about Adam and Eve or the Virgin Birth

or the like that thoughtful religious people often have difﬁculty with in any case, but

also more general ones about our possessing souls, or even about the universe being

the creation of a superhuman “entity” or “force” – and treating all such claims as, at

best, merely metaphorical.

I concede to no one in my appreciation of the importance of issues about what

sorts of lives are valuable, and about ethical issues more generally. But coming to a

view on such issues is surely not the exclusive prerogative of religious people. (Gould

acknowledges this and in fact quietly takes the “magisterium” of “religion” to consist

in the discussion of ethical issues, whether or not based on religion in the more usual

sense.

) More centrally for current purposes, it seems very doubtful that a religious

faith stripped of any substantive descriptive claim about the universe, its history, and

its creation can really count as a religious belief. The theologian Ian Barbour surely

has it right:

Chapter Three

3. Gould, Rocks of Ages, p. 4.

4Actually it isn’t so obvious – there is a substantial literature examining the issue of whether the domains

of fact and value are logically distinct. Nonetheless, it is true.

5“[I] construe as fundamentally religious (literally, binding us together) all moral discourse on principles

that might activate the ideal of universal fellowship among people” (Gould, Rocks of Ages, p. 62).

[R]eligious language does indeed express and evoke distinctive attitudes. It does encour-

age self-commitment to a way of life; it acknowledges allegiance to ethical principles

and afﬁrms the intention to act in particular ways. But...these non-cognitive uses pre-

suppose cognitive beliefs....[R]eligious faith is not simply assent to the truth of propo-

sitions; but it does require the assumption that certain propositions are true. It would be

unreasonable to adopt or recommend a way of life unless one believes that the universe

is of such a character that this way of life is appropriate.

And once a religious faith “requires” such beliefs about the universe, then the clash

with science (or more accurately the scientiﬁc attitude) is inevitable. Or so I now go

on to argue.

3Three Types of Religious Belief, Three Types of Clash

with Science

Many beliefs about the world, its origins, and structure have been, and are, held in

the name of religion. In order to examine carefully the issue of whether science and

religion clash, we need to differentiate at least three types.

Into the ﬁrst category fall quite speciﬁc beliefs about the universe and its history

that some believers have certainly held (and in some cases presently hold) on the basis

of their religion, but which are directly inconsistent with well-accredited scientiﬁc

theories. One example is the claim that the earth is stationary in absolute space and

that the sun and other planets orbit it; another is the claim that there were two

humans, Adam and Eve, who had no ancestors, either human or human-like, and of

whom all humans are descendants (or, more generally, the claim that the universe

was created with essentially the same ﬂora and fauna it presently exhibits in 4004

BC). I shall take it that no one seriously disputes that such claims are indeed incon-

sistent with well-accredited scientiﬁc theories.

However, no serious thinker any longer

feels the need to defend the ﬁrst claim on biblical grounds – even the Vatican now

thinks that its attack on Galileo, and his Copernican allies, was a mistake (though it

did take it until 1820 to remove Copernicus’s De Revolutionibus from its Index of

forbidden books). And none but a few (though very noisy) fundamentalists still feel

the need to defend the second claim on biblical grounds. Again, even the Vatican

seems to have reconciled itself to the idea that evolutionary theory is more than a

mere “hypothesis.”

I shall not go into details about either of these particular claims or others of similar

status. There is, of course, an enormous literature on such matters. I shall simply

assume that the upshot of this literature is that,

(1) on the one hand, such claims are directly inconsistent with well-accredited

scientiﬁc theories (and indeed that where this is true it is the erstwhile religious claim

that must, from a rational point of view, give way);

Science and Religion

6Ian G. Barbour, Myths, Models and Paradigms: A Comparative Study in Science and Religion (San

Francisco: Harper and Ran, 1974), p. 58.

7Anyone in any remaining doubt about the ineradicable clash between so-called scientiﬁc creationism

and real science should consult Philip Kitcher, Abusing Science (Cambridge, Mass.: MIT Press, 1982).

(2) while on the other hand, there is no need for a religious person to commit

herself to any such precise claim. A person may remain, in a clear sense, a religious

believer without committing herself to any such claim that is (or at any rate is so

obviously) inconsistent with well-accredited scientiﬁc theories.

While having seemingly reconciled itself to the (likely) truth of Darwinism, the

Vatican, if I have understood its position correctly, continues to insist that there is

some point along the branch of the evolutionary tree from chemical molecules to

current humans at which “souls” were “infused” into some organism.

Such a belief,

along with other relatively general beliefs about souls and like “entities,” falls into a

second category. This, I suggest, is the category of beliefs that while not, perhaps,

directly inconsistent with any well-accredited theory in science nonetheless seem to

be in a clear and strong sense contraindicated by science.

Nothing in neurophysiology is directly inconsistent with the claim that alongside

the 10

or so neurons in the human central nervous system with their chemically

governed activity there is another entity called the “soul.” (Any more than that New-

tonian gravitational theory is directly inconsistent with the claim that the reason why

every particle of matter in the universe attracts every other with a force proportional

to the product of their masses and inversely proportional to the square of the dis-

tance between them is that each such particle possesses an immaterial mind which

happens to will this to be the case.) It is just that neurophysiology has no need for

such a hypothesis – the “soul” simply and increasingly has no role to play (any more

than those particulate “minds” would have any role to play in gravitation theory).

Neurophysiology is, of course, a science still very much in its infancy, but it has

already made impressive strides in explaining, in its own – “purely material” – terms,

phenomena such as memories and pains that earlier thinkers held were in some

irreducible sense “mental” and hence required some sort of “mind” or “spirit” whose

properties and states they were. The religiously inclined could, then, readily identify

the “soul” with such a “mind” or “spirit.” But minds or spirits separate from bodies

are no longer seriously available. There remain interesting and challenging issues

about whether or not human psychology is fully “reducible” to the laws of nature

governing matter. But these are to do with whether or not mental properties can be

reduced to material properties – no serious scientist (even if religiously inclined) still

holds any version of the Cartesian dualist view of a mental substance that is sepa-

rate from the matter of the brain and central nervous system.

The reason why the soul can play no role, and why the idea of it gets in the way

of proper scientiﬁc theorizing, is that the idea not only has no empirical support, it

inevitably – in principle – can have no such support. There is reason to think (from

the way that progressive science is going) that all the observable effects here are pro-

duced by neurophysiology, and that the “soul” can therefore itself have no observ-

Chapter Three

8See again the sympathetic treatment in Gould, Rocks of Ages.

9Of course, I take it here that talk of souls and the like is to be understood in some “literal” sense – one

that, if true, would make a factual difference to the way the world is – and not in any “metaphorical”

sense, or sense that makes claims about the existence of souls some sort of hidden moral injunction. As

already indicated, I readily concede that all “conﬂict” between science and religion can be eliminated by

going fully “metaphorical” or restricting religious claims entirely to claims about values.

able causal effects. There is a principle of good scientiﬁc reasoning, sometimes called

“Occam’s razor,” that is incorporated in one form or other into every sensible system

of scientiﬁc conﬁrmation, and which states that if some notion plays no role in – if

it can be excised without cognitive loss from – our system of knowledge, then it

should be so excised. This principle is uniformly applied within science itself. The

nineteenth-century idea of a space-ﬁlling aether – a mechanical medium that was

alleged to be the seat of the electromagnetic ﬁeld (and earlier the carrier of the

disturbances that constitute light) – is now rejected by physics. But the notion was

rejected not because it is actually inconsistent with any new, well-accredited scien-

tiﬁc theory, but rather because it is – provably – otiose. Once we have the real science,

in this case the special and general theories of relativity, then further postulating the

aether makes no empirical difference – not only is there no empirical evidence of its

existence, there cannot be any such evidence. The same applies to this sort of inter-

mediate second category of religious claim, exempliﬁed by the idea of the soul.

This second category of religious belief slides over into a third. Beliefs in this third

category are the most general of all. One example would be a general belief that the

universe was created by (whatever that might actually mean) a superhuman power

(whatever that might actually be). A retreat to claims in this third category may bring

with it, for the religious person, the advantage of avoiding the need to account for

why two religious persons selected at random from the world’s population are likely

to have apparently quite different speciﬁc beliefs – those differences being clearly

correlated with accidents of birth, culture, and geography. What you are likely to

believe about the speed of light or the half-life of particular isotopes of uranium, if

you hold any such beliefs at all, is unlikely to depend on whether you were born and

educated in Shanghai, Sydney, or Suez. But what speciﬁc religious beliefs you are

likely to hold, assuming you hold any, are highly dependent on where you were born

and educated. This fact, which surely ought to be disturbing for the thinking believer,

may, perhaps, be nulliﬁed if one resorts to the very general level of belief – perhaps

all religious people agree that the universe is the creation of some sort of super-

human power, and perhaps all the more speciﬁc claims should be thought of as merely

metaphorical (and it must be said, then, pretty misleading) ways of endorsing that

general one. (I take it that something like this is what J. S. Haldane had in mind when

he claimed that “behind the recognized churches, there is an unrecognised church to

which all may belong.”

)

A general claim of this kind not only fails to be inconsistent with any scientiﬁc

claim, the structure of science itself guarantees its consistency. Let me ﬁrst explain

why this is so, and then why I, nonetheless, hold that belief in even such general

claims is unscientiﬁc.

Explanation in science is essentially derivative. In order to avoid unnecessary com-

plexities, assume we are back in the nineteenth century before relativity theory super-

seded Newtonian “classical” physics. If you had asked a scientist at that time why it

is that the planets move in the way they do – why, for example, they move in (some-

what perturbed) elliptical orbits around the sun – he would have had a ready answer.

He could show that the assertion that the planets move in that way follows logically

Science and Religion

10 Gifford Lecture, 1927; quoted from Gould, Rocks of Ages, p. 92.

from Newton’s theory of mechanics plus universal gravitation (actually together with

an “initial condition” about the planet’s velocity). Given that every material particle

in the universe attracts every other with a force proportional to the product of their

masses and inversely proportional to the square of the distance between them, then

it follows that a planet must move in a (roughly) elliptical orbit around the sun. But

suppose you asked such a nineteenth century scientist to explain in turn the “given”

in that initial explanation – to explain why Newton’s theory itself is true, not why

he thought it was true (a question about evidence) but, assuming for the sake of argu-

ment that it is true, why the universe obeys Newton’s theory rather than any other.

Why, for example, is the gravitational force inversely proportional to the square of

the distance rather than, say, to the cube of the distance? Our nineteenth-century

physicist would be nonplussed by this question. Again, he can readily explain how

he “knows” that it’s the square of the distance – that assumption and only that

assumption yields the right observational results – but we are now considering an

ontological question, not an epistemological one: that of why the universe happens

to obey this particular law (assuming that it does) rather than any other. And the only

answer to that question that our physicist could give would be some variation on the

theme of “that’s the way the cookie crumbles.” Relative to the state of science at the

time, it had to be taken as just a “brute fact” that the universe instantiates Newton’s

theory.

Nothing, of course, prevents a scientist from attempting to go deeper, from attempt-

ing to explain why Newton’s theory holds. Indeed, one eminent scientist who endorsed

that attempt was Isaac Newton himself – he famously denied that gravity could be

an “essential property” of matter, and hence denied that his theory could be the ulti-

mate explanatory “bottom line.” Newton was tempted by a Cartesian-style explana-

tion of gravity in terms of some pressure gradient in an all-pervading elastic aether

(though he himself established that Descartes’s own particular explanation along these

lines was hopeless). Suppose that Newton had succeeded, that he had produced a

theory about the constitution of a space-ﬁlling plenum, pressures in which gave bodies

a tendency to move towards one another in accordance with his principle of univer-

sal gravitation. The logical point would of course remain: while we would then have

an explanation for the state of affairs described by Newton’s theory, the facts about

this plenum and its properties that did the explaining would then themselves be unex-

plained; those facts would be – as science would then have stood – unexplained

explainers. Explanation must always start somewhere, no matter what stage science

has achieved. Yesterday’s “brute facts” may indeed become today’s explained facts –

if so, then science has made progress; but the logic of scientiﬁc explanation makes

the existence of some “brute facts” inevitable at any stage.

This is what provides the (logically inevitable) latitude for the religiously inclined.

Since the attempt to reduce gravity to the actions of an aether failed, the poor

scientist cannot explain why the force of gravity between two bodies happens to be

inversely proportional to the square of the distance between them; but the religious

person, it seems, has no problem – that’s how the creator willed it to be. Or, to take

a more up-to-date example, the scientiﬁc cosmologist cannot explain why it was that

the so-called escape velocity of matter at “Planck time” shortly after the Big Bang

had the value it did – she must just take it as a “brute fact” (in that case a brute fact

Chapter Three

reﬂecting an “initial condition” rather than a law of nature). The religious person can,

as always, “explain” that value by invoking a creator and his wishes – and indeed

can, in that case, add a little more to the story, as we “know” on scientiﬁc grounds

that if the value of that escape velocity had been just a little different than it in fact

was, then galaxies (and hence humans) could never have formed. God ﬁxed the value

of the escape velocity because he wanted it to be possible for humans to evolve.

Although the structure of science inevitably leaves religion free to claim it can

give “deeper” explanations than science, what could warrant such claims? As I

explained earlier, a version of the “no conﬂict” (NOMA) account can be developed by

allowing different standards for explanations in the two ﬁelds. But only, as we saw,

at the surely unacceptable cost of adopting a purely relativistic viewpoint. If, as I

urge, we refuse to pay that cost, then the credentials of these alleged religious expla-

nations must be examined in the same way, by the same standards, as are scientiﬁc

explanations; but if judged in that way, then those alleged explanations fail.

Notice, ﬁrst, that once everyone is playing by the same (exacting) rules, then any

claim that religion is in a superior position from the point of view of explaining the

world is logically misplaced. A religious explanation in terms of a creator and his

intentions is just another (attempted) explanation, and, even were it accepted, then,

exactly as the request for an explanation of the latest scientiﬁc theory can always be

made, so we can request an explanation of the religious claim: Why did the creator

choose an inverse square law rather than, say, an inverse cube one? Why was the

possibility of human evolution part of the creator’s plan? The idea that religion can

do what science cannot by “explaining everything” is an illusion.

This shows that the religious “explanation” can, at best, achieve parity – in fact,

parity is far beyond its reach. Let’s retreat just a little and ask why it was, for example,

that the attempt to explain the law of gravity in terms of pressure gradients in a

plenum was eventually deemed to be a failure. It could obviously and trivially simply

have been claimed that there is an all-pervading medium and that – without speci-

fying exactly how – there just happen to be pressure gradients set up in it that account

for the gravitational attraction. But such an “explanation” would never be accepted

in science because it is entirely ad hoc (in the pejorative sense) – it permits no inde-

pendent test. All the alleged explanation does is to deliver what we already know –

indeed, in the form I gave it, it was precisely designed so as to deliver that and only

that.

A successful explanation, one that will be accepted in science, on the contrary, is

independently testable and passes independent tests – that is, it not only entails the

results it set out to entail, it also makes, often surprising, and hitherto unsuspected,

empirical predictions which turn out to be correct. Independent testability and success

in independent tests are the key to scientiﬁc progress.

The explanation of Kepler’s

laws of planetary motion by Newton’s theory was a scientiﬁc success (and the theory

was correspondingly regarded as empirically highly conﬁrmed), because that theory

turned out to entail not only those laws (or rather, in fact, a modiﬁed version of them)

but also a range of other testable predictions – about the precession of the equinoxes,

Science and Religion

11 Se e, e. g. , L ak a t o s , “ Fa l s i ﬁc at io n a nd th e M et ho do l o g y o f S ci en ti ﬁc Re se ar ch P r o g r am me s, ” r ep r. i n

Philosophical Papers, vol. 1 (Cambridge: Cambridge University Press, 1974).

the return of Halley’s comet, and so on (and later of the existence of a hitherto unsus-

pected planet), all of which turned out to be correct. The wave theory of light, devel-

oped by the French physicist Fresnel in the early nineteenth century, not only

explained known optical effects, like reﬂection and refraction, it also turned out cor-

rectly to predict the existence of hitherto unsuspected and surprising phenomena –

such as that the center of the (geometrical) shadow of a small opaque disc held in

light diverging from a point source is illuminated, and illuminated just as strongly as

if no obstacle were held in the light’s path.

The “explanation” of, say, the facts revealed by Newton’s theory or the value of

the “escape velocity” of matter (or, of course, of any other feature of the universe) by

the postulation that those facts reﬂect the wishes of a creator is, on the contrary,

essentially non-independently testable. It is not just that such postulations happen to

fail to be independently testable as yet. They can never in principle be subjected to

independent tests – precisely because, unlike real successful explanations, they are

explicitly designed to yield the already known facts (inverse square attraction, value

of the “escape velocity”) and nothing more.

Science not only declines to accept theories that fail to be independently testable,

it positively rejects them. When a whole series of investigators, including, as I men-

tioned, Newton himself, consistently failed to produce any deeper account of gravity

in terms of pressure gradients in a mechanical medium that was independently

testable, then science adopted the view that there was no such deeper account to be

had, and, despite initial reservations, accepted that matter just does act on other matter

at a distance. Similarly, the initial reaction to Maxwell’s postulation of the electro-

magnetic ﬁeld was that such a ﬁeld could not simply be a mysterious, primitive feature

of the universe; the electric and magnetic ﬁeld strengths at each point of space had

to reﬂect the contortions of some underlying mechanical space-ﬁlling medium (our

old friend, the aether). But when a whole series of investigators (again including,

interestingly enough, the chief scientiﬁc innovator, in this case Maxwell) tried and

failed to produce “mechanical models” of the ﬁeld that were independently testable

and independently conﬁrmed, science came reluctantly to the view that the ﬁeld is

indeed a sui generis, independent, primitive part of the furniture of the universe –

that is, the mechanical aether, at least as an underpinning of the electromagnetic ﬁeld,

was rejected.

I conclude that this third and most general type of religious belief, although not

actually inconsistent with any substantive scientiﬁc theories, nonetheless runs counter

to the practices that have informed successful science; and hence such beliefs,

too, are unscientiﬁc. Some particular religious beliefs are inconsistent with well-

accredited scientiﬁc theories, but all are inconsistent with a scientiﬁc attitude.

Religious belief must, as Tennyson so eloquently reminds us in the second of my two

epigraphs, rely on faith; and faith is unscientiﬁc.

4Objections, Complexities, and Some Food for Further

Thought

Not everyone who has contributed to the – very extensive – literature on the

relationship between science and religion is likely to be convinced by the above

Chapter Three

argument. (Indeed, this may count as one of the all-time understatements, even by

customary English standards!) In a longer treatment, objections would need to be met,

complexities unraveled, and, above all, further confusions exposed and clariﬁed. Let

me end by indicating – in rough outline – some of the necessary elaborations, if only

in an attempt to facilitate further thinking about the issue.

4.1 Belief in science

I have talked so far as if science and religion were two (conﬂicting) ways of gener-

ating beliefs about the world. In fact, however, the relationship between science and

(outright) belief is not at all straightforward. As recent studies have underlined, it

would be a bold thinker who, in view of the history of radical theory change in science,

believed that our currently accepted best fundamental theories are true. If any sort of

belief concerning those fundamental theories is rationally mandated, it is at best a

belief in their approximate truth, which really amounts to the meta-level belief that

those theories will be retained in “limiting-case form” in any future replacement

theories. (Einstein’s relativity theory is logically inconsistent with Newtonian theory,

but yields the latter as a fully adequate approximation for cases of bodies moving at

velocities small compared to that of light.) Outright belief – if reasonable at all –

would be reserved for statements of evidence and, perhaps, for lower-level theories

(such as that matter has some sort of atomic structure) that seem so ﬁrmly entrenched

that their replacement is inconceivable. (This is reﬂected in the currently most popular

formal account of the relationship between theory and evidence in science – person-

alist Bayesianism. This sees rational agents as assigning probabilities (short of 1) to

explanatory theories – probability 1 (effective certainty) being reserved for statements

of evidence and of “background knowledge.”

) I need hardly say perhaps that this,

if correct, sharpens the clash between science and religion: if outright belief at least

in fundamental, explanatory theories is not rational – that is, not scientiﬁc – even

in science, despite their enormous empirical success, then the same must apply a

fortiori to religious explanatory claims, which have no empirical success at all.

4.2 Kuhnian “commitment” in science

In his interesting book Myths, Models and Paradigms, Ian Barbour suggests that devel-

opments in the philosophy of science – notably in Thomas Kuhn’s The Structure

of Scientiﬁc Revolutions (1962) – have reduced the differences between science and

religion to differences of degree, rather than kind. Barbour’s argument merits a more

systematic rebuttal than can be given here. My response, however, is that it relies on

Science and Religion

12 See, e.g., C. Howson and P. M. Urbach, Scientiﬁc Reasoning: The Bayesian Approach, 2nd edn (La

Salle, Ill.: Open Court, 1993). I should add that there are a number of Bayesian, probabilistic arguments

for religious claims that I lack space to consider here – see, e.g., R. Swinburne, “Argument from the Fine-

tuning of the Universe,” in J. Leslie (ed.), Physical Cosmology and Philosophy (New York: Macmillan, 1990).

(Swinburne’s argument is subjected to heavyweight criticism by Adolf Grünbaum in “A New Critique of

Theological Interpretations of Cosmology,” British Journal for the Philosophy Science, 51 (2000), pp. 1–43.

See also Swinburne’s reply in “Comments on Grünbaum,” British Journal for the Philosophy of Science,

51 (2000), pp. 481–5. The arguments are tied to ideas about what can, and cannot, possibly count as an

unexplained “brute fact.” Those ideas are considered brieﬂy below.

overinterpretation of Kuhn’s views. Kuhn does suggest that successful, mature science

requires “faith” in basic, paradigm-forming theories. Scientists must have faith in

those theories in order, for example, not to promote “anomalies” into outright falsi-

ﬁcations – holding that work within the paradigm will eventually solve them.

However, nothing like religious faith is necessary here. Scientists’ “commitments” are

temporary, pragmatic, and defeasible. One needn’t have believed in the absolute truth

of Newton’s theory in the nineteenth century to see that the anomalies for it posed

by observations of Uranus’s orbit were probably best dealt with within the Newton-

ian paradigm – ideas associated with that paradigm provided ways of approaching

the problems with Uranus’s orbit (perhaps, for example, there was another planet in

the heavens, so far misidentiﬁed, and once its gravitational action on Uranus was

taken into account, the anomaly would disappear). Whereas had a mid-nineteenth

century scientist proposed to “abandon” Newtonian theory, he would have been left

with absolutely no idea about how to proceed. Moreover, anomalies must eventually

be resolved, and what counts as a resolution is clear and a ﬁxed feature of science –

scientiﬁc “faith” is temporary and eventually called to account (in this life!).

4.3 Worries about independent testability

My argument is oversimpliﬁed as it currently stands: not all accepted scientiﬁc expla-

nations are independently testable. For example (there are many), the explanation of

the failure to observe any stellar parallax that was (surely correctly) accepted in the

seventeenth and eighteenth centuries was that there is indeed parallactic motion but

available telescopes were not sufﬁciently powerful to observe it. (If we are on a moving

observatory, the earth, then pairs of “ﬁxed stars” ought to seem at least slightly further

apart when we are at our nearest point to them than they do when we are furthest

away.) This explanation itself was certainly not independently testable – it simply

explained away the problem it was introduced to deal with. Is this not exactly like

the religious “explanations” I have castigated as unscientiﬁc? But notice two things.

First, the basic theories involved here – the Newtonian version of the Copernican view

– were (massively) independently conﬁrmed in other areas, by other phenomena. So

the “faith” is underwritten. Moreover, the lack of independent testability of even the

speciﬁc theory is again of temporary duration – if telescopes had continued to improve

in accuracy and still no parallax was observed, then the Newtonian/Copernican view

would have been in unambiguous trouble. The point about the difference between

scientiﬁc and religious explanations therefore remains in tact.

4.4 Explanation as “understanding” or “making sense”

The discussion of the idea of scientiﬁc explanation has, in my view, been dogged from

the beginning by certain associations of the word “explanation” that ought to be

excised. It is natural to think that explanation has something to do with (human)

understanding, or “making sense” of, the universe. The whole structure of scientiﬁc

explanation surely shows, however, that this is a mistake – all such explanation is

derivative, and that means that we don’t ever really understand anything about the

universe (why should we?). Instead, we simply attempt to describe it – eliminating

Chapter Three

minor mysteries (why do the planets move in ellipses?) in favor of major mysteries

(why do all bodies attract one another in a certain way?).

An objection to this that certainly merits consideration is that even scientists allow

that some theoretical claims “make sense” (reﬂect what can reasonably be taken as

“natural” states of the universe), while others cannot be simply accepted as “brute

facts” but demand explanation. It can then be argued that there are certain claims

that must always remain brute facts on any scientiﬁc account (one much discussed

contender is the fact that “there is something rather than nothing”

), but which cannot

rationally be taken as brute, and that this therefore gives rational credit to religious

claims which can explain them (and hence remove their erstwhile brutish character).

I can in response here only state my own view: namely, that all conceptions about

what “makes sense” or what are plausibly “natural” states of the universe are histor-

ically conditioned by the successful research programs at the time, and hence are

themselves subject to change in the light of the always dominant criterion of inde-

pendent empirical support. Aristotelians demanded an explanation for any motion,

Newtonians only for any change in motion; pre-quantum theory explanation in

physics demanded a determinist theory, now that is no longer taken for granted; once

Newtonian theory was established (and the aether-reductionist approach had failed),

scientists were happy (for a while!) to take hitherto barely thinkable, action-at-a-

distance as a brute fact. I hold, then, that there is no such thing as a fact that “cannot

be taken as brute” – the sort of assumption that a scientist is happy to take as reﬂect-

ing a brute fact is historically conditioned and historically variable.

4.5 Am I the victim of an evidentialist prejudice?

Finally, I have made it clear that my whole argument rests on the assumption that a

rational, scientiﬁc person needs good evidence before admitting God into her world

view, just as she would before admitting, say, electrons into it. Alvin Plantinga has

mounted a well-known defense of the striking claim that belief in God can be “prop-

erly basic” – that is, taken to require no evidence.

Although again it requires detailed

treatment which I cannot give here, I should at least indicate my response. This is

that, on analysis, Plantinga’s view amounts to no more than the obviously true

descriptive claim that some people as a matter of fact take belief in God as basic. But

this is no news, the question of course is whether or not they are justiﬁed in doing

so; and, insofar as Plantinga has anything to say about this issue, it seems to rest on

the sort of simple-minded relativism that I have throughout taken to be eschewed.

His response, for example, to the obvious question of why in that case one couldn’t

take belief in a ﬂat earth (or come to that, the innate superiority of the “Aryan” race)

as “properly basic” seems to be simply that no Christian would in fact take – or is

under any obligation to take – such beliefs as “properly basic.” This, however, is

plainly not the issue. The question is what such a Christian would say to someone

Science and Religion

13 I cannot resist here citing Adolf Grunbaum’s response to Richard Swinburne on this issue: “Surpris-

ingly, Swinburne deems the existence of something or other to be ‘extraordinary’, i.e., literally out of the

ordinary. To the contrary, surely, the most pervasively ordinary feature of our experience is that we are

immersed in an ambiance of existence.” (“New Critique,” p. 3)

14 See, e.g., A. Plantinga, “Is Belief in God Properly Basic?”, Nous, 15 (1981), pp. 41–51.

who did assert as “properly basic” (that is, on no basis at all) a claim that she, the

Christian, found abhorrent – and, assuming that she would want to challenge that

claim, how she would deal with the tu quoque objection. Long live evidentialism!

The Demise of Religion: Greatly

Exaggerated Reports from the

Science/Religion “Wars”

Del Ratzsch

The supernatural is being swept out of the universe in the ﬂood of new knowledge

of what is natural. It will soon be as impossible for an intelligent, educated man

or woman to believe in a god as it is now to believe that the earth is ﬂat, that

ﬂies can be spontaneously generated . . . or that death is always due to witchcraft.

– Julian Huxley

This statement reﬂects a widespread intuition that the continuing triumphal march of

science has resulted in religion gradually dissolving toward rational oblivion. In one

of my earlier questing phases, that intuition seemed a welcome weapon against the

demands of the particular religious beliefs I had been raised with. After a (sometimes

rocky) reconciliation with the core of those beliefs, that intuition constituted a con-

stant threat stalking the periphery of my world view. But neither weapon nor threat

is genuine unless the picture of science inexorably destabilizing religious rationality

is accurate. In what follows, I shall examine some of the justiﬁcations for that picture

and will argue that they are inadequate. I come to this exploration with convictions

contrary to, but not deeply hostile toward, that picture. I too have felt its force.

Chapter Three

15 S. Wykstra, “Toward a Sensible Evidentialism: On the Notion of ‘Needing Evidence’,” in W. L. Rave

and W. J. Wainwright (eds), Philosophy of Religion: Selected Readings, 3rd edn (Fort Worth, Tex.: Harcourt

Brace, 1998), argues in response that Plantinga at least shows that we need a modiﬁed, more “sensible”

evidentialism, because while everyone accepts that our access to electrons (if indeed they exist) is neces-

sarily via inference, believers claim to have direct access to God. The important word here is “claim”: what

they are really saying is that, given their – clearly theoretical beliefs – they take themselves to be, in certain

circumstances, in direct contact with God. But, contrary to Wykstra, the situation is precisely analogous

in science, at least with respect to some theoretical entities: because of theories we accept (both about the

nature of light and about our physiology), we take ourselves to be in regular (pretty well) direct contact

with photons of various frequencies. In both cases the “access” is, whatever someone might believe, infer-

ential; it relies on accepting a theoretical premise, and reasonable acceptance of such a premise requires

an inference from evidence.

16 Julian Huxley, Religion wihout Revelation (New York: Mentor, 1957), p. 62.

1Refutation: Some Preliminaries

The direst difﬁculty science could pose for religious belief would be direct scientiﬁc

refutation of essential religious principles.

But refutation can emerge only out of

genuine conﬂict, and that fact imposes some boundaries. For instance, many believe

that science and religion operate in different domains or levels. If such positions are

correct, there can be no genuine conﬂict. Any apparent conﬂict would represent

trespassing or confusion. Furthermore, serious conﬂict between science and religious

belief is possible only if both purport to be true. Consequently, if religious commit-

ment is noncognitive or nonpropositional, genuine conﬂict seems impossible. It also

follows that science taken anti-realistically (as in instrumentalism, social construc-

tivism, etc.) poses minimal challenge. If the ultimate intent shaping science is mere

empirical adequacy, or if the underlying engine of science is sexual dominance or the

suppression of one’s social competitors (as some allege), then whatever science says

will have little rational force.

Suppose, however, that science can conﬂict with religion. Where would conﬂicts

occur? For present purposes, we can separate religious claims into two rough cate-

gories: core beliefs shared by nearly every religion, and the more varied outlying

beliefs constituting the specialized, characteristic beliefs of particular religious groups.

The core usually includes the following:

1A supernatural person – God – created the cosmos.

2God cares about humans.

3God ultimately controls cosmic and human history.

4God can intervene in earthly events.

5There is objective meaning/signiﬁcance to human life, both now and after death.

There are only limited prospects of science contradicting that core. Plate tectonics,

stellar and biological evolution, the periodic table, relativity, quantum mechanics, or

other such results of science do not have even the appearance of contradicting any

of the above. Evolution, for example, could be the means God used to achieve certain

desired results, or the world could be quantum-mechanical because that is the way

God wanted the cosmos to operate. If those claims are coherent (whether scientiﬁc or

true), then the theories in question do not contradict (1)–(5).

The typically cited “conﬂict” episodes nearly always involve specialized beliefs

outside the core, such beliefs often being ascribed to special sources of information

(e.g., revelation). For instance, many believers historically took the earth to be

stationary, at the center of the cosmos. Some contemporary religious groups see the

Science and Religion

17 Additional discussion of several relevant points can be found in my Science and its Limits (Downers

Grove, Ill.: InterVarsity Press, 2000), and “Space Exploration and Challenges to Religion,” Monist, 70/4

(Oct. 1987), pp. 101–13.

18 It might be countered that this sort of reference to God is empirically empty and adds no content to

the scientiﬁc claims in question. That may or may not be true, but if it is true, then the proposition that

God made the world to be quantum-mechanical is logically consistent if the claim that the world is quantum-

mechanical is consistent. Religious claims cannot be simultaneously empirically empty and empirically

refuted.

earth as quite young. Others take the basic kinds of organisms to be unchanged since

the creation. These and other specialized religion-inspired claims are widely perceived

as having been discredited by science. But would that be problematic for deeper reli-

gious belief? How would an anti-religious argument proceed from there? One line of

thought is that such refutations undermine the claims to “revealed truth,” and that

since even the deeper core religious principles rest on that same source, they are thus

rationally unsupported. A different line is that multiple consistent failures even of

peripheral religious beliefs support an inductive case for the falsehood of religious

principles in general, including the more fundamental core beliefs. Both will be dis-

cussed later.

2Foundations: Deep Conﬂict?

Science would challenge religious belief were there principles essential to scientiﬁc

method, scientiﬁc explanations, etc. which were thus presupposed in the very exis-

tence of science, and which conﬂicted with essential components of religion. For

instance, Norman and Lucia Hall claim that there is a “fundamental incompatibility

between the supernaturalism of traditional religion and the experimental method of

science.”

Is that correct?

2.1 The larger web

First, some cautions. Since most scientists historically were religious believers, we

have to attribute intellectual blindness, self-deception, or hypocrisy to those scien-

tists who missed this “fundamental incompatibility.” But classifying Copernicus,

Galileo, Newton, Kepler, Boyle, Maxwell, Faraday, Herschel, etc. as imperceptive or as

religious hypocrites violates substantial historical evidence.

And since about 40

percent of current scientists classify themselves as believers, and even many who do

not nonetheless see no fundamental conﬂict here, the present charge would indict the

majority of scientists who ever lived as not fully grasping what they were doing. That

seems implausible.

Furthermore, any science-based case against the rationality of religion must

presume the rational justiﬁcation of science itself, including its foundational presup-

positions – the uniformity of nature, the basic reliability of human observation, the

appropriateness of human conceptual and cognitive resources, etc. But science cannot

straightforwardly establish the legitimacy of the foundations upon which it itself rests.

If science is the only source of rational justiﬁcation, then the foundational principles

upon which science itself rests must be simply accepted on brute faith – effectively

undermining a key purported distinction between science and religion. Otherwise,

science’s foundational presuppositions must obtain rational legitimation elsewhere.

Chapter Three

19 Norman F. and Lucia K. B. Hall, “Is the War between Science and Religion Over?” The Humanist,

May/June 1986, pp. 26–8: p. 27.

20 We cannot just let these scientists off the hook by claiming that the speciﬁc scientiﬁc facts and

theories which generated problems were not yet known in their day, if the conﬂict ﬂowed out of the very

structure and necessary presuppositions of the scientiﬁc project itself in which they were intensely engaged.

Two things ensue. First, science could not be the only source of rational justiﬁcation.

Second, one question becomes inescapable: How does one give a non-circular natu-

ralistic justiﬁcation for the cognitive faculties we employ in science – that is, a

justiﬁcation, recognition of the rational adequacy of which does not itself rely on pre-

cisely the cognitive faculties whose justiﬁcation is at issue?

Christian scientists in

the past proposed religiously based solutions to justiﬁcation problems. For instance,

human observation and intellection could be trusted if properly employed, given that

those faculties had been deliberately created in us for cognition of this cosmos.

Although there are debates over details, historians of science no longer question the

foundational role which religion – speciﬁcally, the doctrines of divine creation and

divine voluntarism – played in the birth of modern science itself.

Those foundations may be of more than merely historical interest. Despite cen-

turies of development, science may not be disengaged from those roots even now.

Physicist Paul Davies remarks that “Science began as an outgrowth of theology,

and all scientists, whether atheists or theists...accept an essentially theological

worldview.”

If Davies is right, prospects for anti-religious cases of the present sort are not

promising, since science still depends upon foundational structures appropriated from

its religious world view heritage, and thus seems unlikely to constitute a refutation

of them.

2.2 Cases

Let us nonetheless look at two examples.

2.2.1 Naturalism: philosophical and methodological

Again, the Halls: “Science...assumes that there are no transcendent, immaterial

forces and that all forces which do exist within the universe behave in an ultimately

objective or random fashion....[N]aturalism is the unifying theory for all of

science.”

But does science require philosophical naturalism? Many scientists – believ-

ers and nonbelievers – argue that science requires only methodological naturalism

(sometimes called “methodological materialism,” or “methodological atheism”). Advo-

cates of this position argue that whether or not reality includes more than the natural

realm, science by its very nature can deal only with the purely natural and must

rigidly restrict itself to that realm. For instance, Eugenie Scott: “Science has made a

little deal with itself; because you can’t put God in a test tube (or keep it [sic] out of

one) science acts as if the supernatural did not exist. This methodological material-

Science and Religion

21 Darwin himself worried about this. See his July 3, 1881, letter to William Graham, in The Life and

Letters of Charles Darwin, ed. Francis Darwin (New York: n.p., 1889).

22 In fact, some historians of science believe that scientiﬁc method was developed as empirical and

experimental precisely because, as early scientists saw it, given the doctrine that God had created freely,

unhampered by substantive constraints, science had to actually look to see how the cosmos was structured

and governed.

23 Paul Davies, Are We Alone? (New York: Basic Books, 1995), p. 138.

24 Hall and Hall, “Is the War between Science and Religion Over,” pp. 26–7 (lengthy ellipsis).

ism is the cornerstone of modern science.”

On this version of methodological

naturalism, science must pretend that what it cannot control does not exist, and so

must operate as if there is no supernatural realm.

It is evident that whether or not there is a nonnatural realm, the methods that

science would employ and the results that science would obtain presupposing philo-

sophical naturalism would be identical to those it would obtain employing method-

ological naturalism of this sort.

There is thus no scientiﬁc reason for insisting on

philosophical, as opposed to this methodological, naturalism – whatever the philo-

sophical rewards might be.

The critic might take a different tack here, claiming that science does require

methodological naturalism, but that the continued success that science has achieved

by thus insistently ignoring any alleged nonnatural realm constitutes indirect conﬁr-

mation that the natural realm is, after all, the only reality – that is, that philosophi-

cal naturalism is true.

That move is sensible but not completely simple. First, science may not require

even methodological naturalism. Science historically sometimes employed a non-

naturalistic conception of law as regularities in God’s immediate governance of the

cosmos. Such views may even offer the only available explanation of unique logical

characteristics of “natural laws.”

A second reason involves what success means in

this context. Most scientists do take methodological naturalism as a working pre-

scription. This means that methodological naturalism deﬁnes the terms in which

acceptability of scientiﬁc theories is assessed. Thus, if a theory is inadequate, it will

as a matter of methodological policy be replaced only by some alternative theory

which also meets methodological naturalistic criteria. Nonnaturalistic theories –

regardless of how explanatorily powerful – will simply be ruled out of consideration

by ﬁat. Given this procedure, only “naturalistic” theories – whatever their problems

Chapter Three

25 Engenie Scott, “Darwin Prosecuted,” Creation/Evolution, 13/2 (Winter 1993), p. 43. It is interesting

that the deal that science purportedly makes is with itself. Shouldn’t science be making deals with nature?

A particularly forceful statement of methodological (at least) naturalism as a faith claim comes from

Harvard biologist Richard Lewontin:

Our willingness to accept scientiﬁc claims that are against common sense is the key to an understanding

of the real struggle between science and the supernatural. We take the side of science...because we have

a prior commitment, a commitment to materialism. It is not that the methods and institutions of science

somehow compel us to accept a material explanation of the phenomenal world, but, on the contrary, that

we are forced by our a priori adherence to material causes to create an apparatus of investigation and

a set of concepts that produce material explanations, no matter how counter-intuitive, no matter how

mystifying to the uninitiated. Moreover, that materialism is absolute, for we cannot allow a Divine Foot

in the door. (“Billions and Billions of Demons,” New York Review of Books, January 9, 1997, p. 44 (1))

26 As Steven Wykstra has pointed out to me, the range of what is considered “natural” could be differ-

ent in theistic and nontheistic universes. That would imply that a principle that science can involve only

what is natural is not equivalent to the principle that science must proceed as if the natural is all that

exists, as Scott and others seem to believe. Wykstra is developing this point in a manuscript currently in

progress.

27 Of course, if philosophical naturalism is not true, then assuming either philosophical or methodologi-

cal naturalism in science may well lead science irretrievably off track, but that is a different issue.

28 See my “Nomo(theo)logical Necessity,” Faith and Philosophy, 4 (1987), pp. 383–402. Those unique

characteristics include their being located between material generalizations and necessities, their support

of counterfactuals, etc.

– can ever be candidates for “success.” The claim, then, that naturalism has a mo-

nopoly on scientiﬁc success is both unsurprising and of restricted evidential force.

That is not to say that it has no force. But the situation resembles that of the ruling

party in a one-party country citing its unbroken history of electoral success – where

only party members are even eligible to appear on a ballot – as evidence of the voters’

high regard.

In any case, conﬂict does not automatically entail that religion is in trouble without

additional principles – for example, that science and its presuppositions take prece-

dence over religion and its presuppositions. That is a philosophical – not a scientiﬁc

– assertion and will be discussed later.

2.2.2 The larger conceptual matrix

Richard Dawkins believes that religion’s foundational outlook is profoundly mis-

oriented. Religion does have empirical content:

[Y]ou can’t escape the scientiﬁc implications of religion. A universe with a God would

look quite different from a universe without one. A physics, a biology where there is

a God is bound to look different. So the most basic claims of religion are scientiﬁc.

Religion is a scientiﬁc theory.

But, unfortunately for religion, the empirical expectations it generates are precisely

wrong:

[I]f the universe were just electrons and selﬁsh genes, meaningless tragedies . . . are

exactly what we should expect, along with equally meaningless good fortune. . . . In a

universe of blind physical forces and genetic replication, some people are going to get

hurt, other people are going to get lucky, and you won’t ﬁnd any rhyme or reason in it,

nor any justice. The universe we observe has precisely the properties we should expect

if there is, at bottom, no design, no purpose, no evil and no good, nothing but blind,

pitiless indifference.

Science and Religion

29 It is worth keeping in mind that the evidence in question (the claimed success of naturalism in science)

is not only contingent, historical, and problematic, but that the conclusion it is supposed to support –

philosophical naturalism – is philosophical. Such cross-categorial moves are not straightforward even under

the best of circumstances.

30 It is often argued that methodological naturalism acts as an important safeguard against scientiﬁc

investigation being short-circuited by scientists being too ready to take the easy way out by citing super-

natural explanations for phenomena for which genuine scientiﬁc explanations could be found were inves-

tigations to continue. This might be true, but even if it is, that justiﬁes methodological naturalism only as

a pragmatic strategy, which has no substantive implications in the present context.

31 Richard Dawkins, The Nulliﬁdian, 1/8 (Dec 1994). (The Nulliﬁdian is an e-journal). And Julian Huxley

again:

The supernatural hypothesis, taken as involving both the god hypothesis and the spirit hypothesis and the

various consequences drawn from them, appears to have reached the limits of its usefulness as an inter-

pretation of the universe and of human destiny, and as a satisfactory basis for religion. It is no longer

adequate to deal with phenomena, as disclosed by the advance of knowledge and discovery. (Religion

without Revelation, p. 185)

32 Richard Dawkins, River out of Eden (Cambridge, Mass.: Perseus Publishing, 1996), pp. 132–3. See also

Peter Atkins, The Creation (San Francisco: W. H. Freeman, 1981), pp. 17, 23.

The problem is not just that some speciﬁc scientiﬁc theory associated with religion is

mistaken, but that the whole base orientation of religion is orthogonal to scientiﬁ-

cally revealed reality.

It is not obvious that Dawkins is right. (Even many scientists disagree with him.)

Some would deny that empirical results bear upon issues of meaning at all. In any

case, we have experienced exactly one universe. Is this universe precisely what we’d

expect of an undesigned, purposeless, blind, and pitiless one? I doubt that we know.

Could we even reliably distinguish purposeless universes from designed, purposeful

universes in which something has gone badly wrong (a familiar religious claim)?

While Dawkins’s intuitions here may be understandable, they are not rationally

obligatory.

In any case, if we do form such expectations, and if we observe aspects of the

world which clash with those expectations, the problems may lie in our expectations.

It is worth noting that nearly every scientiﬁc revolution has involved reality itself

violating our previous best scientiﬁc expectations concerning the natural. Our human

expectations concerning the supernatural may be far off the mark. But if our expec-

tations do bear some weight here, it must be kept in mind that the world also exhibits

characteristics we would not expect unless it were supernaturally created. Science

itself may have something to say on this side of the issue. Although controversial,

cosmological ﬁne-tuning is at least suggestive.

3Epistemic Undertows: Dissolving Rationality

A number of “scientiﬁc” critiques of religious belief consist of

(a) citing purported causes of such belief,

then

(b) claiming that those causes are not rationally legitimate.

While such considerations would not show the falsehood of religious belief, wouldn’t

they undercut the rationality of religious belief?

Let us look brieﬂy at two popular

versions of this critique – roughly Freudian and Marxist respectively.

According to the Freudian proposal, religious belief represents wish fulﬁllment. We

have deep, hidden psychological needs and terrors, and we construct emotionally com-

forting religious beliefs in response to them. But nonconscious, need-driven processes

of forming beliefs are nonrational procedures for generating beliefs of any sort, rep-

resent profound immaturity (indeed, neurosis), and have no prospect of generating

beliefs connected to the actualities of the world and ourselves. For Marx, religion is

a reality-ﬂeeing, empty promise of future compensation for present suffering whose

Chapter Three

33 None of this is to say that there is not a problem here for religious believers. In any case, religious

believers, far from ducking this and related problems historically, have been among those most insistent

on coming to grips with them.

34 There are a number of technical qualiﬁcations that would be required even for that more modest

project, but I shall bypass most of them.

true (societal) causes it deliberately conceals – an “opium,” as he famously labeled it.

In both cases, the belief-generating process is oriented toward something other than

truth. The governing aim of Freud’s wish fulﬁllment is psychological insulation. The

governing aim of Marx’s opium is psychological compensation. Both processes involve

belief misorientation, and consequently do not deliver rational justiﬁcation.

Success, for this critique, requires two things. First, there must be a plausible case

that the proposed source of belief is in fact its actual source. Such cases are not easy

to come by. Freud produced speculative stories involving a hypothetical domineering

prehistoric father and his conscience-ridden cannibal sons.

It is not clear that such

unveriﬁable speculations constitute a scientiﬁc threat to religious rationality.

Second, such critiques require a case for thinking that the proposed religion-

producing tendencies are indeed unreliable – that they were not, for instance, placed

in us by God exactly for the purpose of alerting us to spiritual matters. Showing that

is not trivial.

Several additional points are worth noting. This criticism cuts in both directions.

Some people may embrace religion because of fear of death, etc. But it is equally pos-

sible that some people embrace anti-religion because they fear ultimate accountabil-

ity, have difﬁculty dealing with the idea of some Being immeasurably superior to

them, cannot cope with being mere dependent creatures, etc.

Furthermore, if Darwin is right – as most critics believe – then natural selection

produced the faculties and cognitive structures with which we form beliefs and pursue

science. The governing aim of natural selection is reproductive success – not theo-

retical truth.

But if the governing aim of a belief-production mechanism being other

than truth undercuts the rational legitimacy of the beliefs so produced – as Marxist

and Freudian critiques presuppose – then exactly the same principle poses potential

problems for scientiﬁc beliefs and for anti-religious arguments produced by cognitive

faculties developed by Darwinian processes ultimately directed toward enhancing

reproductive ﬁtness.

Finally, the present criticism categorizes religion as an explanatory hypothesis

competing with other hypotheses, answering to scientiﬁc criteria appropriate to such

hypotheses. Later I will discuss one ground for questioning that.

Science and Religion

35 Freud, Moses and Monotheism (New York: Vintage, 1958), pp. 102ff.

36 Thus, Patricia Churchland:

There is a fatal tendency to think of the brain . . . as a device whose primary function is to acquire propo-

sitional knowledge....From a biological perspective, however, this does not make much sense. Looked at

from an evolutionary point of view...[t]he principle chore of nervous systems is to get the body parts

where they should be in order that the organism may survive. . . . Improvements in sensorimotor control

confer an evolutionary advantage: a fancier style of representing is advantageous so long as it is geared

to the organism’s way of life and enhances the organism’s chances of survival. Truth, whatever that is,

deﬁnitely takes the hindmost. (“Epistemology in the Age of Neuroscience,” Journal of Philosophy, 84/10

(Oct. 1987), pp. 544–53: pp. 548–9, emphasis original.)

37 Indeed, the implications apply even to the faculties Darwin employed in forming his own beliefs that

evolution explained the existence of those same faculties. One might claim that the successful track record

of the Darwinian-produced faculties has established their reliability, but that is not completely unprob-

lematic, given that the judgment of “successfulness” essentially employs and depends upon precisely the

cognitive faculties in question. Also, the recent dismal fates of both Freudian and Marxist systems might

be worth pondering here as well. Other similar deconstructive critiques – e.g., postmodernism – also seem

well on their way to dismal fates.

4Conﬂicting Mind-Sets

Some cases involve psychological contrasts. For instance, Darwin’s cousin Francis

Galton remarked that “the pursuit of science is uncongenial to the priestly charac-

ter.”

The claim is not obviously true – at least, there are important exceptions (Coper-

nicus and LeMaître, for instance). But what is it about scientiﬁc and religious

mind-sets which is supposed to generate tension? The usual claim is that science

requires an open, tentative, inquiring – even skeptical – mind-set, whereas religious

belief requires a closed, dogmatic, authority-accepting, blinkered mind-set. One mind,

it is claimed, cannot easily be of both sorts. Nobel physicist Richard Feynman describes

as “a kind of conﬂict between science and religion” the “human difﬁculty that happens

when you are educated two ways.”

If different mind-sets operated in different areas, there would need be no conﬂict.

The objection thus suggests that, ideally, one consistent mind-set should dominate

one’s character – especially if one is not to be riddled with inner tensions. But obvi-

ously, rational people have different traits in different areas. The ﬁrst collection of

traits might be inappropriate and even irrational in the interpersonal relationships

even of a scientist. The second set might be inappropriate and even irrational in the

practical pursuits even of a fervent religious believer. The exaggerated dichotomy may

thus misrepresent both sides of the discussion.

Furthermore, the mind-sets are not as distinct as critics would have it. Kuhn and

others have taught us that certain degrees of dogmatism and similar traits are

absolutely essential to the effective operation of science itself. Many scientists accept

what they see as scientiﬁcally essential presuppositions as virtually nonnegotiable

faith commitments. Some hold speciﬁc theories pretty dogmatically – e.g., Richard

Dawkins: “The theory of evolution by cumulative natural selection is the only theory

we know that is in principle capable of explaining the existence of organized

complexity. Even if the evidence did not favor it, it would still be the best theory

available.”

On the other hand, numerous religious traditions have valued – and devoted enor-

mous effort to – reasoned justiﬁcations for their beliefs. In fact, basic theistic belief

seems to be perfectly consistent with a “scientiﬁc” mind-set. There were those in the

natural theology movement who refused to accept religious authority and revelation

as legitimate and who undertook to accept beliefs about the supernatural only to the

extent that such beliefs could be empirically substantiated, but who were convinced

that God’s existence and some of his properties could be so discovered. Perhaps their

arguments were defective, but they were believers with no evident inconsistency

within their overall mind-set.

Chapter Three

38 Francis Galton, English Men of Science: Their Nurture and Nature (London: Macmillan, 1874), p. 24.

39 Richard Feynman, The Meaning of it All (Cambridge, Mass.: Perseus Publishing, 1999), p. 38. Even

John Wesley noted this sort of difﬁculty (at least for some, including himself) in his sermon, “The Use of

Money”: “I am convinced, from many experiments, I could not study, to any degree of perfection, either

mathematics, arithmetic, or algebra, without being a Deist, if not an Atheist.”

40 Richard Dawkins, The Blind Watchmaker (New York: W. W. Norton, 1996), p. 317, emphasis original.

5Historical Erosion

The problem which science purportedly presents for religion is perhaps most often

seen not in terms of episodes of decisive confrontation between the two, but as a

gradual historical erosion. Religion, the story goes, supplied prescientiﬁc explanations

for otherwise inexplicable aspects of human experiences. Such purported explana-

tions rested on religious conceptual foundations, constructed religious explanatory

resources, and applied them to phenomena – both real and otherwise – countenanced

within religious outlooks. These explanations might in their time have been ration-

ally defensible, but, the story continues, science has progressively undermined all of

those areas – religious conceptual foundations, explanatory resources, proposed expla-

nations, and in some cases even the alleged phenomena “explained.” Science has irre-

versibly eroded the conceptual foundations by revealing a cosmos ever more strikingly

out of sync with the expectations those religious conceptual foundations generate.

The “substance” of explanatory resources harvested from such foundations has pro-

gressively evaporated, and there is increasingly little within scientiﬁcally revealed

reality for them to apply to. The alleged explanations of even those phenomena which

are real (e.g., diseases) have exhibited a growing causal irrelevance. And some of

the “explained” phenomena themselves have simply melted away under objective

scientiﬁc scrutiny.

5.1 Dissolution by induction

One continuation of this story goes as follows. The various bits and pieces which have

crumbled away from the religious conceptual scheme might not, individually, have

been essential to religious belief, but, added together over the longer haul, they con-

stitute a track record of serial failure substantive, persistent, and consistent enough

to establish that whatever religious beliefs in general might be alleged to reﬂect, they

do not reﬂect anything remotely resembling reality or truth. As such, religion is now

at best a free-ﬂoating irrelevance, at worst, worse.

Such objections are quite popular. Stephen Hawking suggests that explanatory

appeal to God may no longer be appropriate even for the bare existence of the cosmos:

[T]he quantum theory of gravity has opened up a new possibility, in which there would

be no boundary to space-time...The universe would be completely self-contained . . .

But if the universe is really completely self-contained, having no boundary or edge, it

would have neither beginning nor end; it would simply be. What place, then, for a

creator?

5.1.1 Imploding the gaps

One common way of packaging this “What place?” challenge is the “God-of-the-gaps”

picture. According to this picture, religious explanations ﬂourished before science

Science and Religion

41 Stephen Hawking, A Brief History of Time (New York: Bantam Books, 1988), pp. 136, 141. See also

Atkins, Creation, p. 17.

began to conquer the vast plains of human ignorance. But, as that conquest got under

way, religious “explanations” simply could not compete with the conﬁrmable expla-

nations produced by science and were increasingly displaced by them. The available

ﬁeld of operations for religious explanations inevitably shrank, reducing religion

to ﬁghting for its life in doomed rearguard actions from within whatever gaps in

the broader scientiﬁc picture happened to be (as yet) unclosed. But even those tem-

porary shelters are, if not exhausted, well within the current gun sights of scientiﬁc

inevitability.

Such objections are not without surface plausibility, but they may be less power-

ful than their advocates believe. First, there still are gaps in our scientiﬁc pictures.

Although there are recurrent claims that we ﬁnally have in hand all the necessary

materials for completing the scientiﬁc picture (and such claims have a long and, to

this point, unsuccessful history

), such promises rest upon optimistic induction at best

and prior philosophical commitments at worst, and share the hazards such processes

embody. Second, Kuhn argued that revolutions and advances sometimes reopen

scientiﬁc issues previously thought to be settled. Since there is no guarantee that

closed gaps will stay closed, closed gaps may be an unstable launching platform for

critiques.

Third, it is not clear that religion exhibits an unbroken record of being driven into

gaps. For nearly two centuries, some scientists have argued that the fundamental

empirically discovered structure of nature (governing laws, etc.) can best – maybe

only – be sensibly explained in terms of deliberate design and adjustment. Indeed,

theistic sympathies raised by “ﬁne-tuning” considerations get progressively more

pronounced, the more we know about nomic structures, incredibly tightly constrained

natural constants, and so forth. Such positions have become noticeably stronger

recently, and have attracted even some scientists who cannot by any stretch be clas-

siﬁed as traditional religious believers.

There are, of course, proposed cosmologies

which militate in the opposite direction – Hawking’s view above, many-worlds

theories, etc. But such cosmologies are currently largely speculative, and speculative

cosmologies have been quite unstable historically.

And it should not be overlooked

that some cosmologies – perhaps including some many-world theories – have been

embraced by some precisely for the purpose of escaping what otherwise look like

broadly religious implications of competing cosmologies.

5.1.2 Creeping marginalization

I am all in favor of a dialogue between science and religion, but not a construc-

tive dialogue. One of the great achievements of science has been, if not to make it

Chapter Three

42 e.g., Atkins, Creation, p. 127, says: “Complete knowledge is just within our grasp.”

43 Furthermore, the initial emergence of this focus not on objects in nature but on design-friendly law

structures underlying nature was not part of a retreat from Darwin (as it is frequently characterized), but

predated Darwin by a number of decades.

44 The physicist John Archibald Wheeler reports one of his colleagues as advising people not to chase

after a bus, a member of the opposite sex, or a cosmological theory – since, after all, in each case there

will be another one along in about three minutes.

impossible for intelligent people to be religious, then at least to make it possible

for them not to be religious. We should not retreat from this accomplishment.

– Steven Weinberg

The perception expressed here by Nobel physicist Steven Weinberg represents a variant

erosion theme. Perhaps science has not rendered religious belief epistemologically

pathological, but it does make explanatory appeal to any supernatural agency unnec-

essary if not downright pointless. Religion is superﬂuous because (a) its primary pur-

ported function is explanatory, and (b) any domain of actual reality it might lay claim

to will ultimately be covered by science. That is guaranteed by science’s potential

explanatory completeness – a potential inductively supported by its continually

expanding track record to this point. So perhaps one can still believe if one wants,

but one should not be under the illusion that religious belief does any essential

explanatory work or is otherwise rationally indispensable.

Contention (a) will be addressed shortly. Concerning (b), we may need to be a bit

circumspect. We must appreciate the power which science has and the incredible

things it has achieved. But we must not overlook the fact that perceptions of science’s

potential completeness are at this point projections involving surprising leaps, and

some of the remaining gaps are worth reﬂection. For instance, few things are more

familiar than the fall of a raindrop. Yet no scientist anywhere has ever accurately

predicted – let alone observed, measured, and conﬁrmed – the path of a single

descending raindrop. There are, of course, readily proposed, perhaps perfectly correct,

explanations for that. But the fact is that assertions of global scientiﬁc capabilities

constitute an unfulﬁlled promissory note even for vast stretches of the most familiar,

directly observable physical events.

One other point for reﬂection here. Those who advance an “erosion” picture assume

that science drove the process. But that is a historical claim and may not be totally

accurate. On a closely related issue, historian John Henry says: “Far from being the

predominant driving force in secularization, the practice of science itself was secu-

larized as a result of inﬂuence from the wider culture.”

The “credit” for the erosion

may not be science’s to claim – other, perhaps less than epistemically upright factors

may have been driving the evolution of science itself.

5.2 Revelation

The claimed erosion also factors into another popular critique. Suppose that some

person claims to have a special source of supernaturally revealed truth, but that the

alleged revelation has a track record of frequent mistakes in readily testable areas.

Suspicion of the claim to special access would certainly be understandable – maybe

even proper. (There is, of course, the possibility that the mistakes result from mis-

interpretation of genuine revelation – not to mention the possibility that the alleged

refuting scientiﬁc results are where the mistake lies.) But suppose that the revelation

Science and Religion

45 Steven Weinberg, “A Designer Universe?” New York Review of Books, Oct. 21, 1999, p. 48.

46 John Henry, “Atheism,” in Gary B. Ferngren (ed.), The History of Science and Religion in the Western

Tradition (New York: Garland, 2000) pp. 182–8: p. 186.

– even all revelation – were shown to be spurious. Would not religion have to abandon

all pretense of rational justiﬁcation?

That follows only if the core rests solely on purported revelation – now discred-

ited – and is devoid of other possible rational justiﬁcation. By “rational justiﬁcation,”

critics typically mean argumentation, evidence, experiences, observations, explana-

tory hypotheses, and the like. Of course, some believers (e.g., mystics) have cited

special sorts of direct experience. Others, during nearly every historical period, have

constructed formal arguments. Such attempts are generally rejected by critics as being

either logically inadequate or somehow of the wrong sort – the experiences are

non-reproducible; the arguments are contrived, fallacious, or rest upon empirically

ambiguous foundations. The usual tacit presupposition is that any rational justiﬁca-

tion deserving of the title must conform to something like a scientiﬁc, explanatory,

predictive model of rationality.

But is that presupposition correct? That question will

be addressed momentarily.

6Conﬂict and Rational Justiﬁcation

6.1 Conﬂict: limited signiﬁcance

As noted earlier, some argue that science and religion operate in different arenas and

cannot even in principle conﬂict, meaning that science cannot possibly undercut

religion. Perhaps that is true. But suppose that there is genuine, irresolvable conﬂict.

What would the signiﬁcance of that be? As noted, conﬂict would inexorably undo

religious belief only given the epistemic priority of science – and establishing that

may not be straightforward, for several reasons.

(1) Although we often speak of “scientiﬁc proof,” the logical structure of scientiﬁc

investigation and conﬁrmation precludes rigorous proof of scientiﬁc theories and

results. That well-known fact is one reason why science is typically described as pro-

visional and is always willing to give up speciﬁc positions when necessary. The pos-

sibility that speciﬁc scientiﬁc beliefs involved in science/religion conﬂict are mistaken

should not be overlooked.

(2) Science is done by humans and reﬂects human limitations. For instance, sci-

entiﬁc theories and results are of necessity limited to reasonings, concepts, observa-

tions, measurements, and other resources which native human faculties can, either

directly or indirectly, connect to. Regardless of how far instruments, computers, and

other aids can extend humanity’s scientiﬁc reach, that reach must have traceable ties

ultimately to a fundamentally human bedrock. The anchors of human science must

catch there, because our human faculties and capabilities are the only ones we have,

and without such connections, we could neither grasp nor pursue our own science.

Although we try – via the “scientiﬁc method” – rigorously to govern, correct, and

Chapter Three

47 Some – e.g., Nancey Murphy and Michael Banner – have argued that there are signiﬁcant formal

parallels between science and religious belief.

test our faculties, reasonings, concepts, and theories, we have no alternative but to

employ – and ultimately to trust – inescapably human capacities and insights. Claim-

ing that we subject them to nature’s verdicts does not circumvent the loop. After all,

we must ultimately rely on our convictions concerning what constitutes a test, what

constitutes passing such a test, what the evidence of such passing does or does not

include, what proper evaluation of such an attempt involves – and none of these

matters are just dictated to us by nature. Careful, methodical, and responsible as we

might try to be, those matters inevitably have human ﬁngerprints all over them.

(3) Ultimately, then, we have no choice but to accept some deliverances of some

basic human faculties as cognitive bedrock for any human enterprise. But the cogni-

tive faculties and intuitions underlying science are not the only ones we humans have

– or if they are, they also underlie other characteristic human projects which may

thus have ultimate foundations just as legitimate as science’s. If, then, there is any

deep conﬂict between science and religion, each side may rest upon fundamental

aspects of human nature, and each may have equally legitimate claims on us. In that

case, science would have no more inherent claim to deeper allegiance than does

religious belief. It may be that other dimensions of broader human existence – faith,

loyalty, perseverance, love, religious commitment – should sometimes outweigh

commitment to the abstract, provisional, inductive, theoretical, highly indirect, only

partially conﬁrmed theories and hypotheses of the scientiﬁc dimension of human exis-

tence – and on precisely the same ultimate grounds that science ought sometimes to

overrule speciﬁc religious beliefs. That seems at least possible, and if so, then the mere

fact of even genuine and irresolvable conﬂict would not automatically imply that reli-

gious belief should always give way to science.

(4) The history of science itself provides a caution here. Historically, the bulk of

all scientiﬁc theories ever proposed or accepted by scientists have turned out to be

incorrect, at least in detail. Anyone who risked all for nearly any scientiﬁc theory

in the past would have lost all. Future generations may well say exactly the same

about our present science. If so, then those who advocate the absolute primacy of

science must either argue that at last, fortunately, we happen to be the lucky gener-

ation that ﬁnally got things right (which seems both unlikely and overly self-

congratulatory), or else must base that claim on the prediction that although science

has not yet gotten things quite right, it will someday. This means that theory-based

scientiﬁc cases against religious belief rest partially upon faith in a promise for the

future.

(5) The implications of inconsistency are not always straightforward even within

science itself. (i) Theory/data conﬂict: nearly every successful scientiﬁc theory is

proposed, developed, and accepted in the context of known contradicting data. That

has been so prevalent historically that one historian of science remarked that every

theory is born refuted. Were problematic data an automatic reason to reject scientiﬁc

theories, we’d have to reject them all. (ii) Theory/theory conﬂict: General relativity

and quantum mechanics – two of the best theories contemporary science owns – are

mathematically inconsistent with each other. But science has so far – perfectly

properly – refused to part with either. Logical ambiguity is where we ﬂesh-and-blood

humans – scientists included – must sometimes live.

Science and Religion

6.2 Rational justiﬁcation: sources

The conception of religious belief as constituting an explanatory hypothesis compet-

ing with science, whose rational credentials depend upon successfully meeting the

criteria for such hypotheses, has arisen repeatedly in the foregoing. Is that concep-

tion correct? Very recently, some scholars have been struck by the fact that virtually

none of our truly fundamental, commonsense, life-governing beliefs are generated by

argumentation; nor are they merely provisional explanatory hypotheses or anything

of the sort. Nor do we acquire or justify beliefs that those around us have minds, that

there has been a past, that our reason applies to reality, or that there is an external

world, on some ‘scientiﬁc’ model of argumentation, hypothesis, testing, and con-

ﬁrmation. We could not do so were we to try – as the long history of failed philoso-

phical attempts amply attests. And belief that one’s spouse loves one is neither a

hypothesis to explain otherwise puzzling behavior, some sort of induction, nor an

empty irrelevancy. Furthermore, centuries of serious skeptical arguments from various

philosophical movements have failed to make the slightest dent in commitment to

such pervasive and persistent beliefs as that we know that there is a real external

world. Yet, despite their lack of formal support and their unrufﬂed immunity to con-

trary argumentation, such fundamental beliefs are surely rational if any of our beliefs

are. Rational justiﬁcation here must thus have some different and deeper source, and

artiﬁcially limiting acceptable grounds to explanatory hypotheses, argumentation, and

the like does serious violence to human rationality. That has led some “reformed epis-

temologists” to suggest analogously that artiﬁcially limiting possible grounds for

rational justiﬁcation of religious beliefs to a set of “scientiﬁc” procedures which is

demonstrably inadequate, even for the world of ordinary experience, may be equally

misguided. We may, they argue, have deep inbuilt faculties which generate religious

belief, just as we apparently have deep inbuilt faculties which generate the bulk of

our commonsense beliefs. (Indeed, the indispensable presuppositions of science itself

may ultimately rest upon an exactly similar foundation.) Relevant assessments of

rationality may in each case require an approach very different from more formal

‘scientiﬁc’ ones.

If that proves correct, then some common demands – e.g., that religious believers

produce science-shaped arguments or hypotheses, or identify empirical arenas in

which religion withstands all empirical comers – may simply be inappropriate. Any

inability of believers to meet such demands may be of as little rational signiﬁcance

as the inability of everyone to produce arguments for the existence of the external

world, the minds of their colleagues, the reality of the past – or for the legitimacy of

bedrock presuppositions of science. In none of these areas would any such inability

alone establish lack of rational justiﬁcation.

7Conclusion

I have not tried to show that a theistic world view is correct, that science supports

such a world view, that there have not been historical tensions, that there is no present

ferment, or anything of the sort. In fact, ferment seems to be the natural condition

Chapter Three

of most truly open human projects – intellectual or otherwise. But making the case

that science destabilizes fundamental religious belief is not so easy as some claims

would make it seem. Furthermore, the force even of those critiques which may have

substance is less than is often attributed to them. And if science really is to be our

intellectual model here, as some critics suggest, then, given standard claims that

science differs from blind religious dogma in being tentative, provisional, and always

prepared to revise in the face of new information and insight, it would appear that

some critics of religious belief are engaged in an existential inconsistency.

Reply to Ratzsch

Del Ratzsch alludes to any number of arguments. I think it is fair, however, to take

his paper’s central claims to be: (i) the clash between science and religion is by no

means as clear-cut as is often supposed, and (ii) even where there is, let’s say, tension

between the two ﬁelds, it is not obvious that epistemic considerations unambiguously

dictate that it is religion that should give way.

Since I already conceded that religious believers can – though with more difﬁculty

than Ratzsch’s treatment suggests – avoid outright inconsistency with the “results” of

science, I concentrate on part (ii) of his argument. His most challenging points seem

to be these. First, there is not even the appearance of a clash between science and

religion unless we take it that science provides rational warrant for belief in the truth,

or approximate truth, of its theories. And there are widely defended views of scien-

tiﬁc theories that deny this. Secondly, the thesis that our beliefs (or degrees of belief)

should be governed by the scientiﬁc method is itself a philosophical, rather than a

scientiﬁc, claim. This implies – point three – that even those who support the epis-

temic priority of science cannot hold that science is the only source of knowledge.

Are the nonscientiﬁcally endorsed beliefs that underpin the scientiﬁc method really

any different from the nonscientiﬁcally endorsed beliefs that underpin religion?

Fourth, and ﬁnally, he claims that there are things that everyone would accept that

we know – such as that there exists a world external to our consciousness – that we

do not know via the scientiﬁc method; but then we may know religious truths in this

same “alternative” way.

1Anti-Realism and Religious Belief

Even the standard conﬂicts between science and particular commitments inspired by

religious views – for example, over Copernican or Darwinian theory – undoubtedly

depend on taking a “realist” view of those theories. If, instead, theories are regarded

simply as tools for organizing empirical data, then there can be no conﬂict.

Science and Religion

48 I wish to thank my colleagues in the Calvin College Philosophy Department, especially Steve Wykstra

and Kelly Clark, and David Van Baak.

Now, whatever may be the case concerning “fundamental” theories, it is difﬁcult

to take seriously an anti-realist attitude to less-frontier, more well-entrenched theo-

retical claims such as that matter has some sort of atomic structure, the earth is not

stationary, the fossil record is indeed a collection of the imprints or bones of now

extinct species, and so on. But aside from this, it surely seems odd, to say the least,

for someone defending religious faith to appeal to anti-realism. (Though it is, I admit,

an oddity that frequently occurs.) After all, the motivation for that anti-realist view

is the desire to go as little as possible beyond the evidence, beyond what we can rea-

sonably take as certain. No matter how strongly our evidence from, say, cloud-

chamber photographs and the like may seem to indicate the existence of electrons,

that evidence can never be logically conclusive – so, the suggestion goes, stick to

what is conclusive (namely, that electron theory makes true empirical predictions) and

avoid commitment to the truth of electron theory. But this hard-headed, skeptical atti-

tude is entirely at odds with the sort of credulousness required of religious believers,

who advocate belief in substantive claims about the universe on the basis of no evi-

dence at all. The anti-realist deist, having strained at a gnat, promptly swallows a

camel with gusto.

2 ? No Thank You

Believers through the ages have used the “tu quoque” response to those who try to

use scientiﬁc rationality against them. The claim is that, when carefully analyzed, the

principles that underwrite the canons of scientiﬁc rationality cannot themselves be

defended rationally or scientiﬁcally, and hence that the defender of scientiﬁc ration-

ality herself appeals to principles which must ultimately have the status of unproven

(because basic) dogmas.

I agree with Ratzsch that the principles of scientiﬁc method cannot themselves be

defended scientiﬁcally – all attempts to “naturalize epistemology” are circular.

Indeed, as Lewis Carroll showed,

– any attempt to justify even the simple principles

of deductive logic, such as the rule of modus ponens, will inevitably presuppose those

principles themselves. It follows that the principles of logic and of evidence under-

pinning science can ultimately only be defended “dogmatically”: we just know that

the rule of modus ponens, for example, transmits truth; we just know that it’s a good

idea to test claims against plausible rivals before accepting them, and so on.

This does not, however, entail, as Ratzsch asserts it does, that scientiﬁc “natural-

ism” is self-certifying. First, I cannot see, as explained in my essay, that it is coher-

ent to contrast “naturalism” and “supernaturalism”: there is what there is, and if what

there is includes so-called supernatural entities, then those entities are really natural

– they are parts of the universe as a whole (again, what else could they be?).

Secondly, science has no inbuilt prejudice against God (or against psychic forces or

teleportation or whatever). The scientiﬁc approach is not inherently “naturalist,” but

Tu quoque

Chapter Three

49 John Worrall, “Two Cheers for Naturalised Philosophy of Science – or Why Naturalised Philosophy of

Science Is Not the Cat’s Whiskers,” Science and Education, 8 (1999), pp. 339–61.

50 Lewis Carroll, “What the Tortoise Said to Achilles,” Mind, New Series, 4/14 (1895), pp. 278–80.

rather inherently evidentialist: it accredits claims about features of the structure of

reality exactly to the extent that there is evidence for the existence of those features.

The problem with God is not that there is no room for her in physics – the problem

is that there is no room for her in physics because there is no evidence of her

physical (what else?) reality.

I do accept, though, that the principles of evidentialism have to be “taken as read.”

There is no way that someone could be rationally convinced that logic and evidence

are essential parts of rationality unless they had already accepted logic and evidence!

Does that mean, as Ratzsch intimates (without quite saying so), that in the end the

pro-science and the pro-religion person are “equally dogmatic”? This would be like

arguing that, because they both acted immorally, the hungry little old lady who steals

a lamb chop from the supermarket stands on a par with Adolf Hitler. There are grades

of wickedness, and there are grades of dogmatism – not all sins are equal, and neither

are all “dogmas.” There is an obvious difference between, on the one hand, asserting

that it is just a basic truth, one that cannot be justiﬁed on the basis of anything

simpler, that if, if p then q, and p are both true, then so also must be q, and, on the

other hand, asserting that it is just a basic truth, not one that can be justiﬁed on

the basis of anything simpler, that there exists an omnipotent, omniscient, omni-

benevolent god who “created” the universe. The “dogmas” that underpin the scien-

tiﬁc method are formal, intuitive, of minimal content, and, once understood,

universally accepted by all serious parties; the dogmas of religion make enormously

contentful claims and are matters of heated and continued dispute between serious

parties – hence they surely cry out for (but fail) evidential assessment. As it says in

“the good book,” “Why beholdest thou the mote that is in thy brother’s eye, but con-

siderest not the beam that is in thine own eye?” (Matthew 7:3, KJV).

3Different Ways of Knowing?

Del Ratzsch echoes the view of the ‘reformed epistemologists’ that there are features

of the universe (such as its existence independently of human consciousness) that we

(i) know as ﬁrmly as we could possibly know anything and (ii) do not know on the

basis of any scientiﬁc reasoning. Hence there are nonscientiﬁc ways of knowing, and

who is to say that we cannot know religious truths in some such nonscientiﬁc way?

One way in which this sort of thesis can be argued relies on the sort of outright

relativism that I am taking to be rejected on all sides: if there is a “religious way of

knowing” alongside the scientiﬁc, why not a parapsychological or a scientological or

a . . . way of knowing? Ratzsch is not endorsing such easy (and clearly unacceptable)

relativism, but rather is claiming that anyone serious must acknowledge that we know

certain things in a way not dependent on the procedures of science, as well (or better)

than we know things through science.

Surely, however, we do know that there exists a world independent of our con-

sciousnesses precisely on scientiﬁc grounds. That claim is so fundamental, such a core

assumption in all our scientiﬁc theories, that we may not think of it this way, but we

accept it because of its (immense) predictive success (unlike its idealist rivals, it

successfully predicts that the tree that appeared to be there before we closed our eyes

Science and Religion

will still appear to be there when we reopen them; we accept it because it is the best

explanation of the evidence). Similar considerations of course apply to the case of

the existence of other minds. Indeed, I would argue that all of our so-called com-

monsense knowledge is either ﬂawed (and therefore not really knowledge at all) or

based on scientiﬁc reasoning, very broadly construed (and hence in effect consists of

well-conﬁrmed hypotheses). I would certainly include here Ratzsch’s example of the

person who “knows” that his spouse loves him: if he knows this at all – and, sadly,

the divorce courts churn out thousands of counterexamples a day – he knows it as a

well-conﬁrmed hypothesis (without of course formally thinking of it this way).

Notice that Ratzsch’s argument for our not knowing of the external world or other

minds in a scientiﬁc way is confused. He correctly remarks that philosophical attempts

to prove the existence of the real world (and of other minds) fail. But the point is that

we should never have expected a proof (science never offers these) even though the

claims seem so evidently true. Exactly when we replace philosophy with science, we

get the right answer: these claims are extremely well-conﬁrmed hypotheses. Ratzsch’s

suggestion that all ﬁndings of science are “provisional,” while the external world

claim is not, again rests on concentrating – as much recent philosophy of science has

done – on fundamental scientiﬁc theories. These, as cases of “scientiﬁc revolutions”

have shown, are seriously fallible; but the same cannot be said of “lower-level” claims

such as that the heart pumps blood around the body, that the earth is in motion, that

DNA has a helical structure, and so on. The “external world” claim is even lower-

level, even more deeply entrenched. Of course, being a synthetic claim, it is fallible

in principle, but there is no reason to take its fallibility seriously.

The central thesis remains in tact: everything that we can legitimately claim to

know about the world (as opposed to our methods) is based on the methods and rules

of evidence employed in science, and there is no evidence for the existence of a god.

Reply to Worrall

Worrall has developed very clearly some intuitions which often underlie criticisms of

religious belief but which are seldom explicated in such perspicuous form. Some of

his views I agree with – for example, suspicion of NOMA, and that religious belief

doesn’t come to much if metaphorized out of all genuine substance. Furthermore,

some of our disagreements do not involve matters of principle and could probably be

resolved through further discussion.

But we do have signiﬁcant differences.

Fundamental to Worrall’s case is the view that one single set of criteria deﬁnes

rational justiﬁcation and that (on pain of an invidious relativism) all substantive

beliefs must meet those same criteria to be rationally justiﬁed. But it is surely pos-

sible that sufﬁciently different substantive facets of the world require appropriately

Chapter Three

51 For instance, I read Alvin Plantinga quite differently than does Worrall, and I am not convinced that

the demographics of religious belief have the negative implications that Worrall suggests.

distinct approaches.

Were that the case, the demand that all evaluation be in iden-

tical lockstep terms might avoid relativism but produce conceptual chaos (or con-

ceptual impoverishment). However, Worrall’s initial candidate for this single

evaluative structure is the plausible-sounding principle that substantive claims must

be evaluated ultimately by reference to evidence.

Although that principle is actually

controversial, one might accept it as a general evaluative requirement, then argue that

different types of beliefs might be rationally justiﬁed in terms of appropriately dif-

ferent types of evidence. That would give a unitary underlying structure to rational

justiﬁcation (certainly plausible) while allowing speciﬁc sub-types of justiﬁcation to

reﬂect (sensibly enough) characteristic uniquenesses of distinct facets of reality. Such

recognition of uniquenesses seems potentially crucial in the present case, since reli-

gious explanations are ultimately agent (person) explanations – involving God’s

actions, wishes, purposes, commands. Agent explanations are widely seen as having

unique characteristics, and evidence concerning the wishes and purposes of other

agents is often of a special sort – deliberate agent communication.

However, midway through his essay, Worrall’s principle, with neither notice nor

stated justiﬁcation, mutates to the principle that such evaluation must be in terms

not just of evidence, but of empirical or observable evidence. So not only is there only

one admissible structure for rationality (that being a broad evidentialism), there is

only one category of admissible evidence – the empirical. (Direct supernatural agent

communication – e.g., revelation – is evidently ruled out by stipulation.) As Worrall

sees it, then, all rational evaluation of substantive claims about the world becomes

scientiﬁc evaluation

– the norms of empirical science simply become the norms of

rationality, which substantive religious belief must then meet to be rational.

But extreme care is required here. Most scientists have believed that simplicity,

mathematical beauty, explanatory power, systematizing capability, etc. – none of

which constitute empirical aspects of nature – were essential evaluative criteria for

scientiﬁc theories. Indeed, given the underdetermination of theory by empirical data,

no realist conception of science can avoid nonempirical evaluative criteria, and thus

evaluation could not be just in terms of empirical evidence, in any straightforward

sense.

Science and Religion

52 By “substantive,” Worrall may just mean “empirical.” If so, that would seem to involve an ontologi-

cal presupposition which would require support.

53 Although I will not pursue the matter here, that principle requires great care. My belief that I have

consciousness is a substantive, factual matter – the world would be importantly (I think) different were it

not true. But that belief does not rest upon evidence which I note, evaluate, and test, only then accepting

that belief.

54 That equating can be seen in such statements as: “If all explanations involving substantive, synthetic

claims about the world must satisfy the same criteria, then it is simply nonsense to claim that religion can

explain the scientiﬁcally inexplicable” and the phrase “not rational, that is, not scientiﬁc.”

55 Simplicity, for instance, is often taken as evidence, in some broad sense, for the simpler of two em-

pirically equivalent theories. I am not sure whether Worrall would call that “evidence” – it surely isn’t

empirical evidence. Anyway, I am not at all sure that Worrall would disagree with the overall point here.

Much would depend upon details of his view of the nature and exact role of the empirical, which, given

length constraints, were understandably not covered in his main essay.

In any case, is this sort of evidentialism rationally fatal for religion? Couldn’t a

believer propose God’s creative activity as explaining some observable phenomenon,

then cite that as empirical evidence for God’s existence and activity? Worrall’s reply

is that an invariable requirement of scientiﬁc method is the rejection of any explana-

tory proposal having no (possible) ‘independent empirical support.’

A principle,

purported fact, or theory proposed or hypothesized to explain some empirical phe-

nomenon must not only successfully explain it, but must also have additional empir-

ical implications beyond those it was speciﬁcally tailored to handle. Otherwise, it will

be merely ad hoc. Thus, a proposed explanation offering no prospect of further cor-

roboration of the hypothesized matters (e.g., novel predictions) is not genuinely sci-

entiﬁc. Religious “explanations” of empirical matters, Worrall holds, invariably fail

this invariable requirement of scientiﬁc practice.

Is that true? Long before the Big Bang theory, the doctrine of creation was taken

as implying a beginning of the cosmos at a ﬁnite time in the past. That empirical

implication is apparently right. (Some scientists resisted Big Bang cosmology precisely

because it looked suspiciously like creation.) Furthermore, prominent early scientists

insisted that investigation of nature must be empirical, because nature was freely

created by God, and the only way to determine what God had done was to look.

Scientiﬁcally essential conﬁdence in the intelligibility of nature, and in the unifor-

mity of nature, were historically given theological justiﬁcations. Does the subsequent

success of science, and science’s continuing corroboration of those implications of the

doctrine of creation (which, long predating science, manifestly were not proposed just

to explain science and its success) constitute independent support? Science as we con-

ceive it certainly did not have to succeed – its success is a decidedly contingent, decid-

edly substantive fact. (As these issues suggest, religious explanations often focus not

merely upon internal scientiﬁc matters, but on conditions essential for the very exis-

tence of science – matters which science itself is unequipped to address.

Worrall’s

examples of religious explanations of substantive matters (e.g., gravity operates by

an inverse square because that’s the way God wanted it) give a far too constricted

view of the varying types of explanations involved – explanations which, again, are

agent explanations.)

But is the stipulated “independent empirical support” requirement really invariable

anyway? Contrary to Worrall’s contention, even in broadly “scientiﬁc” contexts, it is

not absolute. Suppose that you were the ﬁrst human on Mars, and discovered an

exact, full-scale replica of the Eiffel Tower, or a sign saying “Welcome, Delicious

Earthling.” The most rational explanation would be an agent explanation – aliens.

And the absence of any additional “independent empirical support” for alien exis-

tence, activity, etc., would not have the slightest effect on the cogency of that expla-

nation. Nor would our inability to tell – or even understand – why the aliens had

done it. Nor would the fact that aliens were proposed speciﬁcally to explain only that

Chapter Three

56 Whether or not there is such a thing as ‘the’ scientiﬁc method which is historically invariant is a

matter of dispute among historians and philosophers of science. Prof. Worrall’s reference to “the always

dominant criterion of independent empirical support” suggests he believes there to be at least a core of

invariability.

57 See my initial essay, especially the section entitled “The Larger Web.”

one bit of empirical evidence render the proposal ad hoc.

The alien explanation

would be an agent explanation designed to explain something made by that agent.

Religious explanations are frequently, of course, of exactly that general type as well.

But suppose that Worrall had shown that religious belief did not conform to “the

scientiﬁc attitude.” If religion is religion and not science, wouldn’t one expect it not

to conform to norms deﬁnitive of something else – science? More importantly, why

should that nonconformity be worrisome for religious believers? Should scientists

worry if told that science does not conform to “the religious attitude”? Recall, fur-

thermore, that Worrall warns us against thinking that even our best fundamental sci-

entiﬁc theories are true, and says that not only does science not logically refute basic

theistic beliefs, but that “outright belief at least in fundamental explanatory theories

is not rational, that is, not scientiﬁc, even in science” (my emphasis). So why, we

might ask Worrall, should science epistemologically rule all?

Superﬁcially, Worrall’s answer might seem to be that, regardless of its limitations,

science is still the best we’ve got – it has had “enormous empirical success,” whereas

religion has had “no empirical success at all.” But his reason goes deeper than that.

Recall that Worrall’s position constitutes an identiﬁcation of the rational with the

scientiﬁc – a form of scientism. (Note Worrall’s telling phrase: “not rational, that is,

not scientiﬁc” (my emphasis).) Thus, if religious belief clashes either with well-accred-

ited results of science or with the scientiﬁc attitude, it is thereby automatically in

rational difﬁculty. In effect, if religious belief has any distinctive character different

from that of science, it virtually thereby fails to conform to the requirements of ration-

ality – making genuine religion rationally unacceptable almost by deﬁnition. In par-

ticular, “religious belief must...rely on faith; and faith is unscientiﬁc.”

It must rely

on faith – that is purportedly essential to its character as religious belief. And faith

evidently cannot demand independent empirical support – that is presumably essen-

tial to its character as faith.

And belief despite failure of independent empirical

support constitutes failure to be scientiﬁc – that is, failure to be rational. Science

trumps and triumphs because it very nearly just is rationality, and anything purport-

edly substantive but not scientiﬁc is afortiorinot rational. With the boundaries thus

drawn, religion’s “clashes with science (or more accurately the scientiﬁc attitude) is

inevitable,” and (at least when in conﬂict with established scientiﬁc theory) religious

claims “must, from a rational point of view, give way” (my emphasis).

Worrall’s argument, then, rests upon reduction of all evaluation of substantive

matters to one preferred structure of evaluation (evidentialism), all legitimate evidence

to one preferred sort of evidence (empirical), and all rationality to one preferred model

of rationality (science). It seems to me that not only has Worrall not given adequate

Science and Religion

58 This example might be far-fetched, but, as David Van Baak remarked to me, the scientiﬁc analysis of

the ﬁrst radio message recovered by a SETI program would have precisely this logical structure.

59 The reference here is to Tennyson, but Worrall clearly endorses the position.

60 This conception of faith would be deeply disputed by many believers.

61 So, as Worrall structures the issue, evaluation is stipulated as empirical, but religion will not get credit

even when it happens to be empirically right. If a religious belief does have the proper empirical creden-

tials, then it is ipso facto science, and not religion after all, and religion, again, does not get credit. One

upshot here is that belief based just on, e.g., revelation is, by deﬁnition, not rationally justiﬁed – natural

theology is in effect the only even possible justiﬁcation for substantive religious belief.

support for that series of reductions, but that the speciﬁc picture of science presup-

posed may not be rich enough to accommodate either the history or the structure of

science itself. And rather than declaring all that to characterize ‘the scientiﬁc atti-

tude,’ the true scientiﬁc attitude might be to treat scientiﬁcally things discovered to

fall within the proper domain of science’s competence and qua scientist to remain

silent (versus blanket rejection) on issues (if any) beyond that domain – leaving the

extent of that domain as a matter for discovery, rather than stipulation.

Worrall began by describing religious scientists as muddled in various ways. But

that perception rests at least in part on Worrall’s philosophical preferences. Those

philosophical preferences may or may not be defensible – but they are surely not

rationally obligatory.

Chapter Three

62 Some of Worrall’s own views may generate difﬁculties here. For instance, his key principle that any

acceptable scientiﬁc explanation must be independently empirically testable seems substantive – neither

empty nor necessary – but it is unclear what observable evidence that principle rests upon. For that matter,

what is the empirical evidence for Occam’s razor?

63 I am indebted to David Van Baak and Stephen Wykstra for extremely helpful discussion of a number

of issues raised here.