Suppose a person upon reading scripture and hearing the scriptures expounded through preaching from the pulpit believes that God has prohibited a certain class of actions. Suppose further that the believer is aware of no conclusive argument either for God’s existence or for the affirmation of the command in question. Nor is he or she aware of any cogent arguments for the reliability of scripture. The believer on discerning that God affirms this in scripture simply takes it at its word and finds himself or herself deeply convinced or persuaded that the action is wrong. Does this person believe in an intellectual acceptable manner?
Many people think the obvious answer is no, here I want to reflect on one particular argument for this negative assesment.
In an lecture expounding Calvin’s view of faith, Greg Dawes has argued that a person who believes in this fashion does act irrationally. A claim he repeats in his text book Philosophy of Religion which was , at least couple of years ago, the proscribed text for Philosophy of Religion at Otago University. ( I should add that Dawes was one of my thesis examiners) Dawes states
One can concede that it would be perfectly reasonable to believe something on the authority of God, even if one had no other evidence for its truth (what could be more reasonable than to believe something told me by an omniscient and morally perfect being). But on the face of it one would still need evidence in
support of one’s belief that (a) that these propositions are revealed by God; and, (b) that God is a reliable source of knowledge. We’ll see in a moment how believers have responded to this demand.
Here Dawes claims that although it is true that one is warranted in accepting things on a reliable authority such as God, before one does so one needs evidence that the authority in question is, in fact, reliable, that the testimony is from God and that that God is a reliable testifier.
Dawes goes on however to argue that in the case of faith in scriptural testimony, no non-circular evidence is forthcoming. In questioning what makes believers think God speaks authoritatively through Scripture he states,
The more common response … is to ‘bootstrap’ the believer’s sense of certainty: to base the certainty of his belief on the very revelation in which he believes. … Religious faith believes certain propositions on the authority of God on the authority of God. (This is not a typographical error.) The authority of God is simultaneously that which (id quod) and that by virtue of which (id quo) one believes.
The circularity in this position might seem to be the Achilles’ heel, not just of the Protestant system, as David Friedrich Strauss suggested, but of this traditional, ‘bootstrapping’ view of faith in whatever form it is expressed.
Dawes’s argument rests on two assumptions. Firstly, that the believer has no non-circular reasons for thinking that the scripture is a medium of divine discourse. Secondly, that the believer must have such reasons if he or she is to be warranted in accepting anything on these grounds. I think both assumptions are false.
1. Does the Believer Need Reasons for Thinking his Source Reliable?
I suspect Dawes conflates two separate questions in the second of his assumptions; whether a given ground is, in fact, reliable and whether one has grounds or reasons for thinking it is reliable. These are not the same question. It is possible for a ground to be reliable without knowing or having any reason for thinking it is. Likewise, the fact that a person has no reason for thinking something is reliable does not entail that it is not reliable.
In defending Calvin’s view of faith Alvin Plantinga and those who follow a similar tack such as William Alston and Nicholas Wolterstorff utilise a highly-influential theory of knowledge known as Reliabilism. Reliabilism holds that a belief is warranted if, and only if, the source, grounds or mechanism that produces the belief, or that it is based on, is reliable, i.e. likely to produce true beliefs on the topic in question. Robert Nozick, Alvin Goldman, John Armstrong, Alvin Plantinga and William Alston have advanced versions of this theory.
The exact version of Reliabilism does not matter too much but for clarity I will use Alston’s version. Alston suggests that a belief is warranted if it is based on a reliable ground or grounds. A person’s belief is warranted if the grounds on which the belief is held are reliable grounds for holding a belief of that sort. He also adds that the person who holds the belief must not have a reason for thinking that the belief is false nor a reason for thinking that the belief is not reliable.
On a reliabilist account, it is the former and not the latter condition that must be met for a belief to be warranted. Alston’s Reliabilism entails that a belief is warranted if it is based on a reliable ground; it does not hold that it must be based on what one has reason for thinking is a reliable ground. The same can be said for most Externalism and Reliabilism theories in general. Noting that a person has no reason for thinking that something is reliable does not show that this condition is not met and hence, does not show that the belief in question lacks warrant.
An obvious response to this line of argument would be to claim that there is some kind of epistemological principle or requirement to avoid believing something based on a given ground or source until one has reasons for thinking that such sources are reliable and that the believer violates this requirement. However, this position has at least two problems.
Firstly, it leads to absurd conclusions. If I cannot believe any thing on a given ground until I have good reasons for thinking the ground reliable, then I cannot believe anything because I remember it happening. To do so I would need good reasons for thinking my memory is reliable. Clearly, such reasons are not forthcoming as any argument I use to try to demonstrate my memory would be circular. I could attempt to show that most of the times I used my memory in the past it was correct but then I would need to remember how I had used my memory in the past and remember whether or not it was accurate. However, I am not permitted to utilise memory in this way until I have reason for trusting it and hence, any such argument could not get off the ground. Similarly, I would be unable to rationally rely on the deliverance of reason. After all, how can I show that a reason is reliable? This can only be done by offering reasons.
The problems do not stop here because memory is essential to any given line of reasoning one engages in. Reasoning is a temporal process; one begins with the first premise and follows an inference through to a conclusion. One does not hold every step of an argument in one’s mind at the same time. Rather one relies on one’s memory to remember the first steps while the second is assessed and then remembers this step while the third is examined and so on. Hence, without memory one cannot reason at all. This creates an unsolvable sceptical situation. One cannot trust memory if one does not have a reason for thinking it is reliable but one cannot get any such reason unless one trusts one’s own memory. Moreover, one cannot engage in reasoning at all if one does not trust it. This position leads to the destruction of all reason.
The second problem created by this kind of stance is that it leads to an infinite regression. To show that the source were reliable, I would have to appeal to certain premises that are based on some other source but then I would have to show that this source was reliable and so on. The claim that one needs reasons for thinking a ground is reliable before one can be warranted in believing anything on the basis of that ground appears problematic.
A further rejoinder suggests that it is in general true that I do not need to have reasons for thinking a ground reliable before I am warranted in accepting a given proposition on this ground. However, I do need such reasons if the ground of my belief is testimony, that is, the say-so of some other person. Dawes suggests something like this in his paper. He distinguishes beliefs based on testimony from basic beliefs. He states that basic beliefs are such that one is warranted in believing them independently of any argument for them, whereas beliefs based on authority are warranted by “indirect evidence”. By indirect evidence, he means evidence that the testimony in question is reliable. A person who believes something based on testimony will, if they are rational, have “reasons to believe the trustworthiness of the source” from which he or she “gained the information”.
However, this view of testimony is mistaken. Coady summarises the problem. If one is going to have grounds for the reliability of a given authority or testimony then these grounds will be either some other testimony or authority, in which case there is a problem of circularity, or it will be based upon sources apart from testimony.  The problem with this second horn of the dilemma is that if we exclude what we know by way of testimony we will have so little to go on that such grounds will be almost impossible to come by.
To demonstrate this, consider an example Dawes himself provides, the belief that E=MC2. Dawes writes,
Very many of our beliefs are held on the basis of testimony. (In this context I shall sometimes refer to these as beliefs held on the basis of authority.) Does e=mc2 represent the rate at which matter can be transformed into energy? I believe so, although I would not have the faintest idea how to demonstrate its truth I have it on good authority that it is true…Of course, there is a sense in which I do believe this on the basis of evidence. I have reasons to believe in the trustworthiness of the sources from which I gained the information.
Dawes suggests that a non-physicist can rationally believe e=mc2 because he or she has reasons to believe that his or her sources are trustworthy. I believe this last comment is incorrect. Consider, for example, what reasons he could offer for believing that the source of his information was reliable. Presumably, it would be because the author of the book in which he read it or the person who told him it was a physicist. Nevertheless, how does he know this? He could have read the person’s qualifications off a faculty list, off the dust-jacket of the book or been told them by the person himself but in each case he is relying on testimony and so, in the absence of further reasons he cannot believe these sources. Suppose, however, Dawes was to investigate thoroughly and locate the address of the university where the degree in physics was awarded in order to check its original records. Yet again, he will be relying on testimony in the form of an address list and records. He would also have to have trusted the testimony of maps and road signs in getting to the university in question.
Consider then what Dawes would have left to go on if he did not use testimony. He could not rely on any information which he himself did not observe first-hand. This would exclude any information about events prior to his own lifetime, any events in his own lifetime that he did not remember witnessing first hand and any event that happened in a place other than where he was at the time. Nothing read in journals, books, heard in lectures, taught to him by his parents or teachers could be used. Nothing heard on the news, read on the computer, told over the phone or reported on would be included. Almost everything he had learnt through his entire education would be excluded because nearly all of it is based on testimony. It seems, then, that if Dawes were really to comply with the epistemic standards he laid down, he could not rationally believe in e=mc2. It appears he is mistaken in thinking that one needs to have reasons for thinking a given authority is reliable to be warranted in believing in testimony.
I think this example shows that this is not isolated. What we know by way of being told by others accounts for a huge and pervasive amount of what we believe. Everything I know about other places, other times, everything learnt at school, university, from parents, friends, books, newspapers, television, etc. is based on testimony. If I were to try to verify any of these beliefs without first relying on some other piece of testimony, I would be unable to.
2. Does the Believer have Non-Circular Reasons for Believing in Scripture
These observations also give us grounds for calling into question the first contention Dawes makes in his criticism of believing in scriptural testimony. Dawes assumes that the believer has no non-circular grounds for thinking that scripture is a medium of divine discourse.
Here his argument appears to be as follows. Dawes grants that if scripture is a medium of divine discourse then one is warranted in accepting theological beliefs on scriptural testimony. He then notes that the only way one could get to the conclusion that such beliefs are warranted is by affirming the antecedent of this conditional and affirming that scripture is in fact a medium of divine discourse.
However, he goes on to argue that this latter belief is typically believed based on scriptural testimony and hence the argument is circular.
What is mistaken here is Dawes’s assumption that this practice involves some form of argument in which the proposition that scripture is a medium of divine discourse serves as a premise. On the above model a person believes propositions affirmed in scripture not by inferring them via argument but by simply taking scripture’s word for it. Consequently, these propositions are not based on any argument at all and cannot be based upon a circular argument as Dawes suggests.
Here his argument appears to be as follows. First he conceeds that if scripture is a medium of divine discourse then one is warranted in accepting theological beliefs on the basis that scripture affirms them. He then suggests that this entails that one can believe on the basis of scripture only if one has good reasons for thinking scripture is the word of God. However, he goes on to argue that this latter belief is typically believed based on scriptural testimony and hence the argument is circular.
What is mistaken here is Dawes’s assumption that this practice involves some form of argument in which the proposition that scripture is a medium of divine discourse serves as a premise. On the above model a person believes propositions affirmed in scripture not by inferring them via argument but by simply taking scripture’s word for it and hence believes such propositions as basic. Consequently, these propositions are not based on any argument at all and cannot be based upon a circular argument as Dawes suggests.
Perhaps what Dawes is driving at is not that this claim that scripture mediates divine discourse is based on a circular argument but rather, it is circular in some other fashion. The distinction between logical and epistemic circularity is helpful here. Logical circularity occurs when a person affirms in the premise of the argument what he or she is attempting to establish in the conclusion. Such circularity can then only apply to arguments and not to basic beliefs. However, William Alston has pointed out that there is also such a thing as epistemic circularity. This occurs when one in practice relies upon a particular source or type of ground in order to establish the reliability of the type of ground in question. A person who relied on perceptual judgements to argue for the reliability of sense perception would be an example. This approach is not logically circular; the person need not argue from premises affirming the reliability of perception. However, it is circular nevertheless.
It is clear, I think, that the model is epistemically circular. The real question is whether there is anything wrong with such circularity. This is borne out by another point Alston stresses, that every, major, doxastic practice, even ones that are paradigmatically rational, are epistemically circular.  I noted this with memory above; one can only establish that memory is reliable by relying on the deliverance of memory as premises in a deductive argument. Similarly, with beliefs based upon a sound, deductive argument. Such arguments can be shown to be reliable only with other arguments and so on. Even an omniscient being could not demonstrate that his cognitive faculties are reliable without appealing to those faculties. Hence, if the practice of believing in divine commands because they are affirmed in scripture is problematic because one cannot believe the reliability of scripture without engaging in epistemic circularity, then various paradigms of rational belief are also problematic. In fact, rationality is impossible. This is, of course, absurd.
 Greg Dawes, “Faith and Reason”, a paper presented to the University of Otago Theology and Religious Studies Faculty. This is contained in Dawes, Philosophy of Religion, (so far unpublished) 46.
 William Alston, “The Concept of Epistemic Justification,” in Epistemic Justification: Essays in the Theory of Knowledge, ed. William Alston (Ithaca, NY: Cornell University Press, 1989), 77.
 Ibid., 38.
 C.A.J. Coady, Testimony: A Philosophical Study (New York: Oxford University Press, 1995).
 Dawes, “Faith and Reason,” 34.
 Ibid., 9.
 William Alston, “Epistemic Circularity,” in Epistemic Justification: Essays in the Theory of Knowledge, ed. William Alston (Ithaca, NY: Cornell University Press, 1989), 319-349.
 By doxastic practice, I refer to the practice of forming beliefs in response to certain grounds, whether the ground is argument, some form of experience or the assertion of some authority.