I. God, as an explanation of the world and any meaning in life, must be derived by scientific methods and must constitute a scientific explanation. Only scientific answers are real proof and can be checked by others, This is reduction of all knowledge to the only valid kind of knowledge: science.:
II. He uses that position to leverage the battle on to his own turf, The proofs of God must be scientific proofs, therefore he totally discounts revelation.:
III. Having moved the battle to his own turf he argues against Swinburne's use of Bayes so that he can use it as the scientific evidence. Having established that only scientific evidence will do then he's the only one in the debate with scientific evidence.:
The way he accomplishes move I is by dispensing with revelation and personal experience. His first chapter argues for the priority of natural theology.  The establishment of that point is really based upon the supposed weakness of revelatory knowledge.Essentially he asserts that only natural theology can be checked by others, revelation and personal experience are unreliable and subjective. He constructs four dilemmas for the believer designed to leverage her into fighting the battle on his own turf. The first such dilemma is between what he calls “cognitive and non-cognitive.” By non cognitive he may include experiences, but he's speaking more directly of liberal theology, that tends to be metaphorical vs. a factual account. He says, “according to...cognitive interpretations a believer who says 'God exists' is making a factual claim which is either true or false, whereas according to non-cognitive interpretation such sayings have a very different function. Evaluation of these statements in terms of truth or falsity is inappropriate.His examples of non-cognitive are Wittgenstein and Karen Armstrong. The non-cognitive makes belief immune to criticism but also leaves it cut off from non-believers as unprovable.
For the “cognitive believer” there's a second dilemma. Either belief must be backed by evidence or it's not needed. An example of the latter is the work of Alvin Plantinga and his idea of “properly basic beliefs.” In that view belief is rationally warranted though not proven to others. Philipse answers Plantinga, “I shall argue that religious beliefs, and indeed most religious believers, not be called justified or reasonable or warranted unless the latter...can support their beliefs by abducinng good positive reasons to the effect that these beliefs are true.”Here's where he starts leveraging the believer off the ground that makes strong faith (experience) and onto his turf where he controls the landscape (that of scientific knowledge). He says, “...these reasons [the believer's justification for belief] shall not derive from, or not derive merely from, a revelation or from faith. In other words rational or natural theology is indispensable to modern religious believers, if at least they want to be called 'reasonable,'”  Either bring religion into compliance with the priesthood of knowledge (science) or the gate keepers will slam the door in your face and you will be obsolete. Either way religion loses.
He brings out the next dilemma, either the methods used to defend belief will be scientific or not scientific. He says “both horns imply great peril for the believer.” In debate we avoided that position by exposing the false premises upon which the dilemma was based. Most dilemmas are like a long horn steer, a point here, a point there, and a lot of bull in between. First of all abductive reasoning means inference from the best explanation. That does not require proof and it may not involve establishing facts. He's going to have to prove that the best explanations are always and only the scientific ones. To assert that only the scientific answers count means that he reduces all knowledge to science. Secondly he's done a bait and switch in that he's demanded science but clearly substitutes the fortress of facts. Moreover, there is a distinction between warrant for the belief of the believer and warrant for an argument that will persuade others. The former is what Plantinga speaks to, the latter is Philipse's demand. Obviously revelatory proofs will mean more to the one who had the revelation. We can't let others in on it. So personal experience could be valid and warrant belief even if it does not function as proof for others. Personal experience, however, can function as warrant even for others, we do have scientific evidence, I will get to that in a bit.
Let's look at his three dilemmas. The first is factual and non factual. In the non factual camp he really includes liberalism and non literal views of theology. That's important because he has to get the believer on his turf, the fortress of facts. Apparently that's what he means by “science,” the fact producing machine. It is possible to employ metaphor in a belief system without reducing belief in God as a stark choice of either fact or metaphor. The second dilemma is for the so called “cognitive side” (for that I read “literalistic,”) it must be backed up or not. I say it only needs backing up if one needs to prove to others. Otherwise we all have reasons for believing, and if faith is sincere those reasons can only be life transforming. There's nothing irrational about that. If something is so overpowering it can't be denied and if it changes your life dramatically for the better it would be irrational to refuse it merely because it's not part of the truth regime. The third I just discussed, the demand for scientific evidence.
A major problem here is that he has these neat segmented little compartments, where reason and faith are kept apart; where natural theology and reformed theology are kept apart. Real faith and real religious thought is not like that. People have both faith and reason, revelation and objective proofs. It is perfectly possible to mix the two sides and most of us do to some extent (with a thousand variations of degree). We really need to take to heart the distinction between personal faith and the desire to convey one's reasons to others. No one believes for no reason. Belief is an existential/phenomenological matter. It's not a test for a driver's license. The scrutiny I have to care about is my own and that of those I love. When Philipse says things like, “...for modern believers faith is not an epistemic virtue but a vice, unless there are convincing arguments for the truth of its contents,” there is no reason to credit that statement. Convincing to whom? The only person I have to convince is me. I do have reasons I find powerful and if others would not find them so more is the pity for them.
He says modern believers are in a methodological dilemma. “Do They have at their disposal reliable and validated methods of religious research? If so it seems that the content of their faith can be refuted...” If it can be validated it can be refuted. That goes for all positions. I was a college debater I can refute anything. Actually that's not a dilemma it's a double bind. He's saying if it's validated that's bad, then he blames it because it's not validated. “...If not, [validated] how can a religious creed be credible at all in our scientific era?” That depends upon who is doing the validating and upon what basis.. The philosopher kings have been in the ivory tower too long. That statement in and of itself screams ideology. Why should religion not be credible merely because we are in an age of science? We all believe for a reason, modern people have modern reasons.
Of course Philipse, brilliant and rigorous though he is, seems blissfully unaware of the fact that there is good scientific evidence backing up many aspects of religious faith. In the chapter on supernatural I will discuss the vast body of empirical work backing the veracity of religious experience; the skeptic really has no justification for dismissing experience as evidence, they way Philipse does. The “M scale” gives us a scientific means of control for understanding what is a valid experience and what is not (see chapter on supernatural). By the same token the same chapter also documents unexplained events well documented and factual form the basis for modern miracle claims. These both experiences and miracles are controlled for and validated by double blind experiments and medical diagnostics. Even so its only if one intends to persuade the skeptic that one needs to meet the demand for evidence. Then evidence need not be scientific. Yet we do have the validated scientific methods, I wrote a book about them.
(to understand the studies proving scientific methods for a God Argument see the summary of my book on my blog "The Trace of God).
He continues more of the water tight compartmentalizing by stating that there are so many contradictions in the “traditional concept of God.” Then the most troubling part, “without introducing analogy or metaphor...” Why should we want to avoid analogy and metaphor? That's rather irrational since we can't talk without them. It's not the literalistic aspect that makes language work as communication but the space afforded understanding in the use of analogical speech. Not only is all religious language analogical but all language is to some extent.That is a major method in theology, use what we know in analogy to communicate in the face of the unknown. Yes, it can be inaccurate, but how accurate does it have to be? Why doesn't the fact that we have no empirical proof of string theory or multiverse stop scientists from talking about those things? Personal belief is a phenomenological matter. The short cut to defining that term: metaphysics imposes pre set categories upon sense data, phenomenology is allowing the sense data to suggest its own categories. In equating metaphor with untruth he is imposing many preset categories (ideology) upon the sense data.
Philipse associates literal language (no analogy or metaphor) with being reasonable defined as “objective diachronical rationality”). I disagree that “objective” opposes metaphor. One can be objective about the use of metaphor. I certainly object to the implication that metaphor is irrational. To use metaphor properly one must reason about meaning and implication. Since belief is a phenomenological pursuit it requires an individual understanding. It's not science nor can it be. We just have to accept that science is not the only avenue for reason. One can be rational and reason in the continental style of philosophy, using phenomenology for example. The compartmentalizing is a hindrance. That seems to be the ideology of scientism. That's an upshot of the reduction of all knowledge to science.
He equates belief in God with a scientific hypothesis and demands of it the kind of explanatory power one finds in science. Rather he asks (rhetorically) “...whether theism can be an explanatory theory or hypothesis at all.” That rather depends upon what one means by “explanatory.” What kind of explanation are we seeking? Why should it be a scientific hypothesis? It's not meant to be one. One does not dedicate one's life to a mere hypothesis. Hypotheses are meant to be tested. We can't test God. Take the Lord out for a test prayer! Science works by testing ideas and culling bad one's until the last one not culled is assumed true. It's not proved its just assumed (see chapter 3, Popper). To that extent Paul Tillich's notion of theology is scientific.Tillich's “method of correlation” in which he “attempts to correlate the various analyses of the human predicament produced in modern culture with the answers provided by the symbols used in the Christian message.” This a sort of hypothesis testing where one is comparing empirical experience to “theory” (doctrine). This is not to say, however, that doctrine is at all like an hypothesis. Nor is it intended to be. Philipse's demand is an example of the reduction of all knowledge to science. Yet Philipse demands that an understanding history be included to be rational (“diachronical”). Tillich said that the sources for doing theology are not only the Bible but history and culture as well. Moreover, he spoke of theology as explaining the meaning in the human condition. That is a kind of explanatory power science does not have.
In order to counter the realization that all language is analogical he quotes Swinburne saying “...if theists cannot articulate their religious view except by using the key terms in an irreducibly analogical manner theism cannot be a theory of hypothesis.”  Well its not. He still has not bothered to say why it should be. In fact this line of argument seems to be a smokescreen. It's irrelevant what Swinburne makes of analogy because that doesn't answer the point that all language is analogical. The issue there revolves around the term “irreducible.” Philipse acknowledges that all language is somewhat analogical but not irreducibly so. If the term means that there is meaning present to the signifier then he's wrong, all language is irreducibly analogical. There is a literal meaning to all utterance but that doesn't mean that meaning is ever totally present in any signifier. In speaking of God we can say “God is the basis of all that is.” That is a literally true statement and yet it's fraught with ambiguity. God is beyond our understanding, we don't need words to feel love, or to know that love is real. Belief is personal and doesn't stop when you leave the lab. It's given a deep hidden personal slant multiplied times all believers, Naturally, it's not going to have the kind of precision a scientific hypothesis will have. Nor will it have the kind of explanatory power.
He does point out metaphor is used in science. He gives several examples. The mechanical model, seeing universe as a machine (eighteenth century) gives way to the organic model. These “models” are really analogies. He argues that science has clearly defined limits for such ideas but God is transcendent of our understanding so all our speech about God is completely analogy.xxi  The upshot is religion is not science, so religion is false because science is all that is. He argues that some people think they experience God's presence but of curse they don't. It's true that God is not given in sense data. A thing need not be amenable to our sense data in order to be real.:
Although some believers claim to have perceived God, the speakers of the language cannot be perceptually aware of God in the same unambiguous public manner in which human beings, cities, dogs, or mountains can be perceived. Hence no intersubjective referential use of 'God' can be established by such deictic methods. As a consequence one is only able to provide the proper name 'God' with a possible referent by giving a description of what the name is supposed to refer to. If no literal description is possible of an entity to which the word God allegedly refers...since that entity can only be hinted at by irreducible metaphors...we could never succeed in providing the word God with a referent.First of all, speaking a language does not in and of itself assure us a signified (referent) for our signification. It's irrelevant what the speakers perceive qua speakers. Secondly of course God is not going to be as well perceived as a tangible object, at that point he's merely saying God is not tangible, that cannot be equated with “there is no God.” Moreover, not to be pedantic but proper name is not God. That's his job description.
It is basically true that we can't communicate what God is in words because its beyond our understanding. That's why there are mystics. Lots of other things are known to exist but can't be communicated in words. Some examples include love, the colors, the feelings (try describing a feeling without referring to other feelings). Some philosophers try to deny the existence of feelings, those of us who don't need therapy know they exist. Now if it be objected that these things are not physical objects and God must be a physical object, that is an ideological assumption. There is no proof that there can't be real objects that are not physical. The disproof of Philipse's point, however, is the scientific data to which I refer in the chapter on supernatural, in connection with the M scale and Ralph Hood Jr. The data proves that mystics around the world in all faiths are having the same kinds of experiences. These can be separated from physical causes of brain chemistry, drugs, perceptual problems and the like. Not only do they have an intersubjetive basis for referent but that basis furnishes good evidence of a reality external to their own minds which they all experience. The differences in those experiences are the doctrines used to explain them, that acts as a cultural filter, not the experiences themselves.
1 Herman Philipse, God In The Age of Science, Oxford, London: Oxford University Press, 2012, 3
2 Ibid 23
3 Ibid., xiv
9 Way back in the Trojan war, when I was in high school debate the Dallas Jesuit debators used that saying, that's where I got it.1973-74. Jesuit was the top school having won NFL district championship 3 years in a row (1972, 73, 74).
10 Ibid xvi
12 Joseph Hinman, The Trace of God: a Rational Warrant doe Belief, Colorado Springs: Grand Viaduct, 2014. no page indicated. This book is only available on amazon.com. It covers a host of methodologies in over 200 studies on religious experience forming a huge body of work going back to the 19960s. Of these the chief study instrument is the “M Scale.”
13 Philipse xvi
14 Phiipse acknowledges this, 95
15 Ibid., 95
17 Guyton B Hamond, “An Examination of Tillich's Method of Correlation,” Journal of Bible and Religion, 1964,Vol. 32, No. 3 (Jul., 1964), pp. 248-251, 24
19 Philipse, 96
21 Ibid., 97
22 Find chapter and fn on Hood article.