The Rationality of Theism

Paul Copan, Paul Moser

• Publisher: Routledge
• Year: 2004
• Print ISBN: 9780415263313
• eBook ISBN: 9781134574872
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

The Rationality of Theism is a controversial collection of brand new papers by thirteen outstanding philosophers and scholars. Its aim is to offer comprehensive theistic replies to the traditional arguments against the existence of God, offering a positive case for theism as well as rebuttals of recent influential criticisms of theism.

Part I
FOUNDATIONAL
CONSIDERATIONS

1
RELIGIOUS LANGUAGE AND VERIFICATIONISM
William P. Alston

In this chapter I shall consider the challenge to the factual meaningfulness of attempted statements about God posed by the verifiability criterion of factual meaningfulness. I shall first lay out the criterion and something of its history, and then shape it into its most coherent form. I shall consider what can be said for and against it. Then I shall examine its application to talk about God. This will involve considering both whether attempted statements about God must satisfy the criterion in order to be factually meaningful, and also whether (some) statements about God do satisfy it. The overall conclusion will be that the verifiability criterion poses no serious threat to the factual meaningfulness of what is said about God.

The criterion, its history, and the most coherent form of the theory

In the early decades of the twentieth century, a number of philosophers in Vienna began a group for the discussion of issues in the philosophy of science, a group that became known as the Vienna Circle. Among its leading members were Moritz Schlick, Rudolf Carnap, Otto Neurath, Friedrich Waismann, and Herbert Feigl. The group was concerned with the logic of mathematics and science and with giving philosophy a scientific orientation. The movement was labeled “logical positivism.” Its influence spread to Britain, especially from the influence of A. J. Ayer’s Language, Truth, and Logic, and to the United States, to which several of the Vienna Circle, including Carnap and Feigl, emigrated after the rise of Hitler.

The logical positivists felt that traditional philosophy had reached a dead end and that it needed a fresh start. The main weapon they wielded in the negative part of their endeavor – to discredit traditional metaphysics, epistemology, and ethics – was verificationism, the “verifiability theory of meaning,” according to which a statement is genuinely “factually meaningful” only if it is capable of empirical verification or falsification. Since, in their view, most traditional philosophical theses failed this test, they were deemed not to be what they professed to be – true (or even false) claims about objective reality. This chapter is specifically concerned with the application of verificationism to religious statements. But to come to grips with this, we must do some fine-tuning of the verifiability theory itself.

First, there is the difference between verifiability as a theory of meaning and as a criterion of meaningfulness. The former can be formulated thus: “the meaning of a proposition is the method of its verification.” In other words, “the meaning of a proposition can be given only by giving the rules of its verification in experience.”¹ This is a reductively empiricist account of meaning: there is no meaning to a factual assertion except what would be involved in verifying it. It amounts to an unrestricted generalization of the position in philosophy of science known as “operationalism,” according to which, say, the meaning of “temperature” is exhaustively given by a specification of how temperature is measured. But as a criterion of meaningfulness, the principle makes no attempt to say what the meaning of a factual assertion is. Instead, it confines itself to laying down a necessary condition of its meaningfulness – that it is, in principle, capable of being empirically verified or falsified. The concern here is with the criterion form of the view.

There are still a number of loose ends to tie down before we are in a position to discuss what happens when we submit attempted religious assertions to this criterion.

1. If the claim is that possible empirical verifiability is required for any kind of linguistic meaning, the position is wrecked before it leaves port. It makes no sense to speak of “verifying” interrogative, imperative, or expressive utterances. What would it be to verify “What time is it?” or “Please go to the bank for me” or “Thanks a lot”? To have a chance of being acceptable, the criterion must be limited to declarative sentences that are usable to make statements that have truth-values. This point is recognized by the more sophisticated versions of the position, in which this kind of meaning is called “cognitive meaning,” or “factual meaning.”
2. But this raises the more difficult question of what sort of entity is declared to have cognitive meaning only if it is empirically verifiable. The problem is that it is sentences (and their meaningful constituents) that are said to have or lack cognitive, or any kind, of meaning, whereas it is not sentences, but the statements they are used to make, or the propositions they express, that have truth-values and are subject to verification or falsification. We can see that truth-values and empirical testing do not apply, in general, to sentences by considering sentences containing indexical terms or proper names. Is the sentence “I’m hungry” true or false? What would count as empirical evidence for or against it? It all depends on who utters it and at what time. This very same sentence can be uttered by many people at many different times, and the truth-value of what they say is sometimes true and sometimes false. The same is true with “Jim is at home.” Many people are named “Jim,” and so some utterances of that sentence will be saying something true and some something false. Such examples make it clear that what is true or false, verified or falsified (in the primary, straightforward sense), is not a sentence but what is stated or asserted by uttering the sentence by a certain speaker on a certain occasion, a certain place, etc. Not all sentences exhibit this kind of variation. “Lemons are sour” does not. But for a general formulation the point holds that it is statements – not sentences – that have truth-values and are susceptible to empirical testing. But then we have to reconstrue what empirical verifiability is a criterion for. If we are still to think of it as a criterion for a certain kind of sentence meaning, we have to relate verifiability of statements to a semantic status of sentences. This is not difficult to do, but we lack a snappy way of putting it. Non-snappily, the point is that a sentence must have the potentiality for being used to make statements that are verifiable or falsifiable in order to be factually meaningful. By switching the semantic status to a potentiality for making statements that are verifiable or falsifiable, we allow for sentences that have this status to be usable for making many different statements that vary in their truth-values. The verifiability criterion will then read:
   
   (a) A sentence is cognitively meaningful only if the statements it can be (standardly) used to make are, in principle, empirically verifiable or falsifiable.
   
3. It will not have escaped notice that in (a) the term “in principle” preceded “verifiable or falsifiable.” That is meant to contrast with “in practice.” To have factual meaning, a sentence need not be usable to make statements that we are, at present, able to subject to empirical testing. In the early days of logical positivism, the standard example of a genuine factual statement that could not be tested was: “There are mountains on the other side of the moon.” As with many philosophical illustrations, this one has succumbed to the progress of science and technology. But it is easy to find more stable examples. Take “It was raining at this spot on the earth precisely 2 million years ago.” That is verifiable or falsifiable in principle, e.g. by an observer at that spot at that time, but not by anyone now. The “in principle” requirement is sometimes put by saying that we can describe a verification or falsification. The only trouble with that restatement is that it raises the question of whether that description is factually meaningful, thus raising the specter of an infinite regress. But many such descriptions, like the one above, cannot sensibly be suspected of factual meaninglessness.
4. The early formulations of the criterion made serious use of the terms “verifiability” and “falsifiability,” thus requiring the possibility of a decisive, maximally conclusive establishing of truth or falsity. But it was soon realized that this would rule out all or most of the scientific hypotheses and theories that the logical positivists took as paradigms of factual meaningfulness. For scientific evidence is never so conclusive as to rule out the possibility that the hypothesis supported is false. Enumerative induction is universally recognized to establish only a higher or lesser probability. As for high-level theories, their support comes from their ability to explain and predict empirical data. But for any such theory there is always a potentially indefinitely large number of alternative theories that can do that same job. Moreover, we can never be sure that new data will not pop up which our favored theory cannot handle satisfactorily, as happened in the last 150 years with Newtonian mechanics, which for a very long time was a paradigm of a scientific position that was conclusively established. Hence, there has been a retreat to a more modest form of the criterion that requires only the possibility of empirical data that confirm or disconfirm – support or weaken – the statement in question to a certain extent. In other words, the requirement is that it is possible for some empirical data to tell for or against the statement. This entails another revision of the criterion.
   
   

(b) A sentence is factually meaningful only if the statements it can be (standardly) used to make are, in principle, empirically confirmably or disconfirmable.

Here, “confirmable” and “disconfirmable” are not to be read with the qualifier “conclusively.”

5 The next step in firming up the criterion is to probe into just what it takes for empirical data to tell for or against the acceptability of a statement. This divides up into (a) what counts as empirical data and (b) how such data have to be related to a statement to tell for or against it. These questions get us into all the problems of confirmation theory and, more generally, the epistemology of perception and the epistemology of beliefs and statements that go beyond the directly perceivable. Here I can just skim the surface of all this.
The first point is that to introduce a reasonable degree of precision into talk of empirical confirmation and prevent it from wallowing in a morass of purely intuitive judgments about when the results of observation bear on the acceptability of a hypothesis, we have to explicate the notion in terms of logical relations between the hypothesis and statements that report the observations. As a first approximation, for a hypothesis to be supported by a possible observation is for it, usually together with other premises, to entail a statement reporting the observation. And for a given hypothesis to be weakened by a possible observation is for that hypothesis, together with other premises, to entail the negation of a statement reporting that observation. Thus if the hypothesis (with other premises) entails that a certain liquid will turn cloudy within 10 minutes, an observation of its doing so will support the hypothesis, whereas an observation of its not doing so will weaken it. In neither case does this amount to conclusive confirmation or disconfirmation: not the former, for there are other possible explanations of the observed event; not the latter, for since there are generally other premises involved in the deduction, the trouble might be with some of them rather than with the hypothesis under investigation.
Note that this account presupposes that we can separate out a class of “observation statements” that are confirmed or disconfirmed directly by observation, not indirectly as with the hypothesis just mentioned. There has been much criticism of this assumption in recent times. It is claimed that all perception is shaped by “theory” that we bring to it and that nothing untouched by that is “given” to us in perception. This leads to “reports of observation” themselves being treated as hypotheses that are to be evaluated by how well they explain the data. If those data also turn out to be hypotheses, we are off on an infinite regress or we are forced into a coherentist epistemology, in which the only standard of acceptability is the way a given belief or statement fits into a background system, which itself is to be evaluated solely by its internal coherence.
However all this may be, it is not at all clear that a distinction between what is directly justified by experience and what is, at most, indirectly so justified cannot be maintained, provided we content ourselves with relatively modest claims for the former. This would involve a disavowal of certainty and conclusive verification for perceptual beliefs as well as the higher-level ones. This means that perceptual experience will be taken to give only a prima facie justification of perceptual reports – justification that is, in principle, vulnerable to being canceled by sufficient contrary reasons. With that caution, we can still take such statements as “The liquid is cloudy” to be directly justified by a visual presentation, and take a hypothesis of chemical constitution of the liquid to be assessed in terms of how it, along with the larger system in which it figures, explains and predicts correct observational reports. This is the way the matter is generally construed by advocates of the verifiability criterion, and I shall continue to discuss the matter in these terms.
The exposition just given reflects the fact that the verifiability criterion restricts “experience” to sensory experiences and, correspondingly, restricts “observation” to sense perception and empirical data to sense perceptual reports. These restrictions will turn out to have an important bearing on the application of the criterion to talk about God.

An early objection to the verifiability criterion is that it failed its own requirement for factual meaningfulness and thus was self-refuting. For it does not seem to be ...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Contributors
• Introduction
• Part I: Foundational Considerations
• Part II: Arguments for God’s Existence
• Part III: Potential Defeaters for Theism
• Bibliography