Among the many books written in defense of atheism, perhaps the two best are J. L. Mackie's The Miracle of Theism (Oxford 1982) and Robin Le Poidevin's Arguing for Atheism (Routledge 1996, 2000).
Mackie's magnum opus received a very full and critical review by Alvin Plantinga, "Is theism really a miracle?" Faith & Philosophy (1986) 3:109-34. But, to my knowledge, Le Poidevin's entry into the field has not received the same level of critical scrutiny.
Arguing For Atheism is divided into the three sections. Part 1 is an attack on the "big three" traditional arguments for the existence of God (ontological, cosmological, teleological). Part 2 is an attack on the moral argument, with special reference to the Euthyphro dilemma and the problem of evil. Part 3 is, in part, a defense of theological antirealism, and, in part, a defense of life without an afterlife as meaningful.
In chapter 1, Le Poidevin attacks several versions of the cosmological argument. One problem is that he fails to grasp the Kalaam cosmological argument. This is predicated on the distinction between an actual abstract infinite and a potential concrete infinite. By definition, an actual infinite is a given totality. As such, it must be an abstract object. A concrete object can be a completely given as long as it constitutes a finite totality.
He also has an inadequate notion of ontological or causal priority, which he equates with temporal priority. "Has he [God], for all time, willed the universe to exist. Why, then, did it not come into existence sooner…we are led back to the regress…which the temporal argument was supposed to avoid" (13). No, we are not.
i) God has not, for all time, willed the universe to exist. If the divine fiat is a timeless fiat, then this means that the world began to exist, but that there was never a "time" when the world did not exist. And this first moment is the temporal effect of a timeless cause.
ii) Causation is a subset of constitutive ontological relations. For example, a triangle is constituted by its angles and sides. Yet this is an atemporal and noncausal mode of ontological relations. That is a stronger thesis than we need for the origin of the world, but we can argue from the greater to the lesser.
iii) It is certainly possible to offer a more general definition of causality, one that is neutral on questions of temporal priority or simultaneity. Consider the definition of David Lewis: A is the cause of B just in case B would not exist unless A existed. (Cf. "Causation," Journal of Philosophy , 70:556-67.) The question, then, is whether temporal relations are an essential feature of causality, or merely a special case and incidental mode of causality.
iv) Le Poidevin never explores the possibility of a temporal effect of an atemporal cause. And if time had a beginning, then this conclusion is forced upon us.
Many cosmologists, in fact, concede as much. They identify time with the space-time continuum, and they believe that the universe had a point of origin. Now the science behind this may or may not be true, but the point is that this is considered to be a live option. And if cosmological arguments, such as the Kalaam version, prove a temporal origin of the universe, then they yield this additional conclusion as well.
iv) Le Poidevin is assuming, without benefit of argument, that our notion of causality is derived from observation. But this is viciously circular. You can only recognize what you already know. Empiricism cannot very well bootstrap its own criteria. If you bring some chips to the table, you can win more chips; but if you bring nothing to the table, you can't play the odds. Even a game of chance has some ground rules that are not derived by chance.
Why can it not be known a priori that every contingent thing has a cause? (13). And if it cannot be known either a priori or a posteriori, then science can know nothing at all. For science is the science of the contingent.
In chapter 2, Le Poidevin discusses the ontological argument in relation to modal realism. One of the problems facing the atheist is not just the general question of why there's something rather than nothing, but the specific question of why there's "this" rather than "that." For, on the face of it, it seems that many things could have been otherwise.
And this, in turn, raises a host of related questions. Where do these unexemplified possibilities exist? In the nature of the case, they don't exist in the actual world—inasmuch as they are unrealized alternatives to the actual world. Something merely possible doesn't exist at all except as a property of some other existent—of an agent or agency or property-bearer. What selects for the instantiation of one possibility out of the many?
This has less to do with the ontological argument than a version of the cosmological argument, based on the principle of sufficient reason—a la Leibniz. However, it does get mixed up with modal metaphysics, so there is some interconnectedness.
I'm afraid that Le Poidevin's treatment of modal realism strikes me a simply incompetent. It looks as though he's trying to embed Plantinga's modal version of the ontological argument in Lewis' ontology of possible worlds. But Plantinga has his own framework, which is not at all the same as Lewis' conceptual scheme.
Lewis erases the distinction between concrete and abstract objects. Lewis collapses abstracta into concreta. For Lewis, all possible worlds are equally concrete and real.
Plantinga, by contrast, maintains an abstract/concrete dualism. Possible is to abstract as actual is to concrete. Plantinga also draws another distinction. All worlds, whether possible or actual exist, and are, in that sense, equally real, but all worlds to not obtain. Only the actual world obtains. So Le Poidevin's analysis is incoherent.
There is also a danger of getting carried away with picture language, as when we talk about God existing "in" a possible world.
The theological distinction between the necessity, actuality, and possibility goes back to the Scholastic distinction between God's absolute and ordinate power, free and necessary knowledge. A possible world is a world which God could possibly make. So a possible world is simply a colorful way of objectifying God's omniscience and omnipotence. To reify a possible world as something that God inhabits is to be beguiled by a figure of speech.
Now, there is an underlying reality to this manner of speech. Possibilities inhere in the necessary existence and essential attributes of God. God's relation to the realm of possibility is more akin to a creative writer's relation to his fictional story ideas.
As to the status of the ontological argument, this is a very complicated question.
i) As a general matter, how you rate the ontological argument has less to do with the specifics of the argument than with your general theory of knowledge. If you're in the nominalist/empiricist camp, you're predisposed to reject it; if you're in the realist/rationalist camp, you're predisposed to accept it.
ii) Anselm's argument takes the form of a logical dilemma. An atheist cannot deny the existence of God unless he has a concept of God. This concept is the idea of a greatest conceivable being. What is more, even though the concept is mental, it is a concept of an extramental being. Although the thought of God subsists in the mind, the object of thought is the idea of God as an extramental existent. That much is necessary even for purposes of negation.
So what is the gold-standard that redeems the object of thought? What is it that makes either the affirmation or negation of a proposition to be true or false? And what if the proposition is about a being that is greater than the thinker who thinks him?
Gaunilo replied by introducing the counterexample of the blessed isles. But this fails on couple of counts: (i) it raises a similar dilemma, for what is there that grounds both the bare possibility and its practical denial? Of what is this possible? Of what is this false? To what does it refer? (ii) Unlike the idea of the blessed isles, the idea of God is the idea of a being that is more than an idea.
The ontological argument is deceptively simple, for it quickly draws us into deep waters regarding the truth-conditions of intentionality, possibility, necessity, and negation. This is what, in turn, invited a modal version of the ontological argument—a la Plantinga. But this can be pressed a step further, for what is there that lays the final foundation of modality, if not the greatest conceivable being?
This is why the ontological argument, like Dracula, is so often pronounced dead and buried, yet continues to rise from the grave. For there is just too much at stake. It cannot be denied without denying the possibility of denying anything else.
Chapter 3 begins by reiterating the confused exposition and analysis of modal realism in chapter 2. Hence, this part of the argument can be set aside.
Next, Le Poidevin attacks the prime mover version of the cosmological argument on the grounds that causal explanations cannot be extrapolated beyond the bounds of natural law and space-time.
Before addressing the specifics, one is struck by how often the appeal to natural law forms the bedrock of secularism. This seems especially ironic. Where does the notion of natural law come from, anyway?
Isn't this a personification? And does it not represent a hangover from faith in the Creator of the world? God is the Lawgiver who legislates the natural and moral order of world as the work of his hands.
How can you deny the existence of God by invoking natural law if natural law presupposes his existence? How do we cash out the metaphor once we take God out of the picture? The atheist is like a man who locks himself inside a prison cell and tosses the key out the window.
Back to the particulars of his case, he says that "as an explanation [the prime mover] is a dead loss: it tells us nothing about the first cause other than that it is a cause" (37). But this objection is wrong on two counts:
i) Certainly it advances our understanding to know "that" something had a cause, even if we didn't know "what" was the cause in question. Knowing "that" the murder had a cause is the presupposition for a homicide detective to investigate "what" caused the murder.
ii) A Christian apologetic has a very clear concept of what the prime mover is like. Although he may, in part, infer his existence from the origin of the world, his information is not limited to that inference alone.
Le Poidevin then says that we cannot demand a causal explanation for the universe as a whole inasmuch as causality is an intra-cosmological rather than extra-cosmological principle. Without the universe, there are no natural laws; without natural laws, there are no causal explanations. By way of reply:
i) This simply begs the question. Whether the universe is self-explanatory is the very point at issue. Rather than an argument for secularism, this is just another paraphrase of secularism. At the very least, then, it raises the question of the burden of proof.
ii) Cosmology doesn't suffer from any inhibitions when it comes to theorizing about the origin of the universe as a whole. And this holds true whether you subscribe to a cyclical (closed-time model) or linear theory of time.
iii) Le Poidevin is assuming, without defending, a particular brand of epistemology, according to which the principle of causality is an inference abstracted from a generality of inductive instances. But this is a roadmap to nowhere.
Le Poidevin then says that "two things cannot be related as cause and effect unless there are laws connecting one type of event with another." (37). By way of reply:
i) One problem with this criticism is that it fails to engage the opposing position. Le Poidevin is writing against theism. As such, he needs to ask himself what the theistic alternative would be, and how that stacks up against his own position.
Although theism is generally committed to some level or degree of second causes, yet—in principle—that is inessential to the idea of divine causation. To begin with, a Christian would say that God is the primary and immediate cause of all secondary agents and agencies. This is especially so if we regard God as timeless, such that he enacts the entire world in one indivisible fiat. In addition, many miracles are the effect of direct divine action. Finally, some Christians regard the category of miraculous events as a special class or limiting case of a more uniform and universal principle of divine agency (occasionalism).
Now, Le Poidevin would reject all this, but the problem is that he has failed to offer any sort of rebuttal to the opposing position, even though that is the stated aim of his argument. Why should a Christian accept his premise? A Christian doesn’t share his wholesale investment in natural law.
ii) He fails to distinguish between natural laws as descriptive and prescriptive or proscriptive. He seems to ascribe a causal efficacy to natural laws, as though a law of nature were the same thing as a force of nature. Here he seems to be bewitched by the power of a metaphor or shorthand expression. Natural laws don't do anything; they don't make anything happen—any more than a speed limit makes me drive at 60 mph. A traffic law is not an internal combustion engine.
In fielding the argument that the world was the work of a personal agent, Le Poidevin says two things. To begin with, he reiterates his claim that causal priority assumes temporal priority. But I've already dispatched that objection in chapter 1.
Second, he says that intentions are contingent, so that trying to explain a contingent world by recourse to divine intent only pushes the problem back a step and thereby generates an infinite regress.
But this is sophistry, for the adequacy of the answer depends on the scope of the question. When a homicide detective traces a murder back to the murderer, he has solved the crime. Sure, this is not a total explanation. We can still ask why the murderer killed his victim. But if we can't answer that question, does that imply that the crime scene is a discrete, self-contained event?
Le Poidevin's objection is a non sequitur. I don't need to explain everything to explain anything. Failure to explain the murderer's motives does not render the murder self-explanatory. The crime requires a criminal.
In chapter 4, Le Poidevin tries to debunk the teleological argument by first offering two different interpretations of the argument, either as an argument from analogy or an argument from probability, and then attacking both lines of argument.
Before touching on the specifics, I'd just register an initial and elementary caveat. For I don't regard either analogy or probability as an essential feature of design.
Let's take two illustrations of Paul Janet's Final Causes (Scriber 1893). I drop a crystal on the floor. It shatters along geometric lines. I conclude that this could not be by chance. The outward pattern bespeaks an inward pattern, the effect a preexistent structure. I don't need to repeat the experiment.
Suppose, though, someone would object that, for all we know, this could be a freak accident, a natural anomaly. So, to be on the safe side, we should repeat the experiment.
Very well, but this draws attention to the old quandary regarding the limits of induction. How many times does the same result have to turn up for us to detect design or dub a law of nature?
This brings us to Janet's second illustration. I'm awakened by the sound of a clock striking four o'clock in the morning. Strictly speaking, I "perceive" it striking four, but I only "hear" it striking one o'clock four times in a row.
The inner relation is not given in nature. Sensation only registers a temporal sequence of discrete particulars. All I hear is one, then one, then one, then one. I never hear "four."
This is the irony of induction. You can never learn induction by induction. Unless the mind has an innate ability to discern certain patterns and classify natural kinds, induction would be dead in the water.
You don't distinguish glass from crystal by repeatedly smashing both objects on the floor. All it takes is one instance to distinguish a random from a non-random distribution pattern of the shards. This is not an argument from either analogy or the law of averages.
Hence, the "design inference" does not depend on any background information, and that's a good thing, too—otherwise induction would generate a vicious regress.
Moving on to a few of the particulars of his objection, Le Poidevin regards the theory of evolution as having inflicted heavy damage on the analogical version of the design argument (45-56). Of course, this objection is only as good as evolution itself, and is vulnerable to its own set of objections. Certainly his own explanation raises many more questions than it answers. If biological organisms are cast in the die of nature, how can they be ill-adapted to their natural surroundings? What accounts for this in-built adaptibility to begin with? If some species are ill-adapted to their environment, how do they survive long enough to adapt?
It takes an intelligent observer to appreciate the survival value of an opportune adaptation. But natural selection is blind. So an atheist must assume a God's-eye view of nature in order to deny a God's-eye view of nature. He must resort to a top-down, goal-directed explanation to make sense of a bottom-up, non-directive process. The process has a purpose unknown to the process. Hence his ontology and epistemology tug in opposing directions. Naturalism needs a viewpoint above brute nature to explain naturalism. Naturalism denies supernaturalism any admittance through the front door, but allows it in through the back door, on an incognito basis.
Le Poidevin's discussion of statistical probability (50-51) suffers from the same crippling confusion over modal realism that disabled his attack in chapter 2. There is an interrelated quality to a number of Le Poidevin's arguments that is either a strength or a weakness. It is a virtue to see that many issues are interrelated. And it lends the treatment a certain elegance and efficiency to apply a few basic principles of analysis. But by that same token, a few strategic errors introduce a cumulative error rate into the analysis.
Towards the tail-end of this chapter and into the next, he discusses the anthropic principle, and actually concedes that the theist has the better of the argument here. I would, though, just touch on the common countermove, which says that there's nothing remarkable about the existence of intelligent life, for absent intelligent life, there'd be no one to mount the anthropic argument in the first place!
This strikes me as a disguised description rather than an explanation. To borrow an illustration from W. L. Craig, if I survive a firing squad, I have good reason to wonder how all those bullets managed to miss their target. To say that if I hadn't survived the ordeal I'd be in no position to pose the question is hardly an answer in itself.
Chapter 6 is rather odd because Le Poidevin answers his own objections, but doesn't seem to register that fact. He says that virtue might be dependent on God in a couple of respects: (i) "since he created things in such a way that doing certain things would be good for us" (78); (ii) "God has some properties that we would consider good in a human" (79). The first explanation would form a natural law ethic, the second an exemplarist ethic. And both these ways of grounding the good are complementary. The former is grounded in the will of God, the latter in his nature.
He then poses the question of how we become aware of moral values, if they exist independently of us (80,85). This fails to distinguish between the ontological and epistemic aspects of morality. They exist independently in the sense that exemplary virtues inhere essentially in the nature of God, and exist whether or not we exist. However, God has created us with a natural constitution and exemplary virtues. They inhere continently in human nature, as ontologically dependent on the nature and will of God.
As to how we come to know them, Le Poidevin answers this question, in part, with reference to divine command theory (83). This, however, is not an arbitrary fiat. God has given us a conscience, as well as a revealed code of conduct, and these are preadapted to the Creator/creature distinction. The command is not what constitutes our obligations; rather, it discloses our obligations. That is how moral values supervene on natural facts (85). Thus the recourse to divine command theory does not generate a vicious regress (83). It is only natural that a finite creature ought to obey an omniscient, omnipotent, and benevolent Creator.
Different conceptions of right and wrong exist in part because of different degrees of revelation (special and general), as well as the effect of sin, which seeks self-justification and thereby redefines the moral law to excuse its own criminality.
Le Poidevin says that these difficulties are not faced by atheism, which can resort to moral subjectivism and emotivism. But the problem with this contention is twofold:
i) These are not genuine difficulties for Christian theism.
ii) Subjectivism and emotivism are descriptive, whereas morality is prescriptive and proscriptive. Appeal to feelings doesn't adjudicate between good and bad feelings, or between conflicting feelings. It is devoid of moral discrimination. On this view there is no problem of evil because there is no gratuitous evil, no distinction between good and evil whatsoever. It just comes down to feelings—yours and mine.
In chapter 7, on the problem of evil, Le Poidevin says that the greater good defense is unable to account for the sheer amount of suffering, some of which seems to be purely gratuitous (89).
This objection is very vague. First of all, is the objection quantitative (degree of evil) or qualitative (kind of evil)? These are two very different objections. Is there some quantifiable threshold beyond which the amount of evil would count against the existence of God? It is hard to see what calculus could be applied to that question.
As to the quality of evil, this is high ambiguous. Evil can be purposeful in one respect, and purposeless in another. Adam's motive in precipitating the fall are very different from God's in ordaining the fall.
Again, something can be both a means and an end, or either one or the other, depending on the viewpoint and incentive of the agent or coagent.
The gratuity of evil exposes its true character. Its shock value has pedagogical value (Rom 7). Such is the paradox of gratuitous evil, which, at another level, ceases to be merely gratuitous. The wonton abuse of a first-order good makes possible the emergence of a second-order good.
In chapters 8-9, Le Poidevin defends religion without God. This is, to varying degrees, a widespread phenomenon in Christendom. You have churchgoers who grew up in the faith without growing into the faith. They have no faith, but faith has spoiled them for unbelief. They have nothing better to replace it with, and, indeed, the alternative is far too austere and severe to contemplate head-on, so they live and die like a dog that would rather starve to death upon his master's grave than linger on without his master's company.
Far from being an argument for atheism, this exposes the tragic outlook of atheism to an uncommon degree.
Le Poidevin distinguishes between scientific realism and antirealism (instrumentalism, positivism). But by way of reply, scientific theories could be true or false, but because "we have no privileged perspective on reality" (109), we could not tell which, if any, are true. Again, some scientific theories could "function" as useful fictions, even if they were not fictional, but positively false. They would be useful to the extent that they sustain some correlation with the real world, yet that need not be descriptive. A "code" language (109) sets up a closely controlled correspondence between the ciphertext and the plaintext, yet their point of contact is functional rather than formal.
There is a parallel between the moves in a chess game and the plotline of a Lewis Carroll novel (Through the Looking Glass), but the action of the former does not resemble, in a representational sense, the action of the latter.
Indeed, the very fact that there is, on the one hand, some sort of gap between appearance and reality, in spite of which we manage to make our way in the world, is—itself—a theistic argument. For it would take a common mind to coordinate their interactive relations, like a VR program.
Again, there are local as well as global varieties of realism and antirealism. One could be a global realist about abstract objects, and a local antirealist about concrete objects—admitting "that" there are extramental sensibles, but denying that we can know "what" they are like.
In addition, we could say that there is at least one agent who does enjoy a privileged perspective on the world, and that is the Maker of the world. Moreover, we could say that God, in some measure, shares his privileged perspective with us by means of divine revelation.
In chapter 9, one of the problem's with Carnap's deflationist theory is that it looks suspiciously like a position that imposes on reality rather than deriving from reality. It stipulates the questions and answers in advance by spreading an interpretive grid over the debate. But the exercise is highly artificial, for this is not what motivates scientific theorizing. People ask questions and seek answers about the real world. Their theories may be wrong or wrongheaded, but scientific investigation intends to be descriptive.
Likewise, God-talk is intended to be descriptive. Theological antirealism is parasitic on theological realism. It takes over traditional categories and formularies that were meant to be descriptive of the truth, whether in literal or analogical terms, and treats them as fictional. It substitutes a pretended meaning for an intended meaning. To treat traditional theology as make-believe is—itself—an exercise in make-believe.
In his tenth and final chapter, Le Poidevin deploys the B-theory of time to defend the enduring value of life without an afterlife. On the A-theory, every human achievement is a sandcastle, erected at the ebb-tide of life, and washed away at the high tide of death.
Le Poidevin tries to counter this dismal notion by contending that if the B-theory of time is true, then all times are equally and eternally real, such that "regardless of whether all traces of my life will be destroyed, it will always be the case that I once lived and did various deeds. These truths can never be obliterated, even if the evidence for them is" (144).
What are we to make of this? Atheists have often tried to salvage some consolation in the thought that their deeds live on in the hearts and thoughts and deeds of those they influence.
But the problem with this sort of solace is that it is a sentiment which, in the nature of the case, can only appeal to the living and not the dead. It assumes a viewpoint which death denies, for if death marks the extinction of individual awareness, then it cannot share in this satisfaction. It is a view of death poised on the living, and toppled from its perch by the cessation of consciousness. So this sentiment is a spectral projection.
In sum, I conclude that Arguing for Atheism is a failure from start to finish. This is no reflection on Le Poidevin, who is an astute and ingenious advocate for his cause, with a deft touch at many points. But he cannot surmount the overwhelming handicap of arguing for a false thesis. Logic can only follow the facts.