Atheism: Encyclopedia entry backup

I'm the author of the Internet Encyclopedia of Philosophy entry on Atheism. Here's an alternate copy of that article.

Atheism

What is atheism?

The term “atheist” describes a person who does not believe that God or a divine being exists. Worldwide there may be as many as a billion atheists, although social stigma, political pressure, and intolerance make accurate polling difficult.

Unless otherwise noted, we will use the term “God” to describe the divine entity that is a central tenet of the major monotheistic religious traditions--Christianity, Islam, and Judaism. At a minimum, this being is usually understood as having all power, all knowledge, and being infinitely good or morally perfect. We can identify the particular traits that the divine being in different religious traditions have in addition to these three omni- properties. The Christian God, it is thought, also sent Jesus to be sacrificed for the salvation of human kind; he rewards belief and piety with eternity in heaven, punishes sinners, and so on. When necessary, we will use the term “gods” to describe all other lesser or different characterizations of divine beings, that is, beings that lack some, one, or all of the omni- traits.

The God concept is the central focus of philosophical arguments for atheism for several reasons. First, that being is the essential, common thread in the western monotheistic traditions, as mentioned above. More than 3 billion people now can be said to subscribe to the view that God exists, and countless more in the past have agreed. Furthermore, the existence of such a being, more so than any other characterization, would have profound metaphysical, personal, moral, social, and historical implications. Without exaggeration, the existence of such a being would be the single most important fact ever acknowledged by human beings. Another motivation for focusing atheistic arguments on the notion of a divine being that embodies conceptual absolutes as the omni-God does, is that by implication many of the conclusions we can draw about that being will apply a fortiori to other, lesser beings. Furthermore, a being that is not the positive culmination of all possible power, knowledge, and goodness, it has been argued, would not be worthy of the title “God” and would not be worthy of our attention in the same way. Many atheists and theists have agreed on at least this much.

We can make two useful pairs of distinctions concerning the term “atheism.” It has come to be widely accepted that, at a minimum, to be an atheist is to lack a belief that a God or gods exist. We can follow Antony Flew (1984) and label this inclusive sense of the term as negative atheism. Parallels for this use of the term would be terms such as “amoral,” “atypical,” or “asymmetrical.” So negative atheism would includes someone who has never reflected on the question of whether or not God exists and has no opinion about the matter and someone who had thought about the matter a great deal and has concluded either that she has insufficient evidence to decide the question, or that the question cannot be resolved in principle. Both people lack a belief in God, but in importantly different ways. So the position traditionally characterized as agnostic—neither believing that God does exist nor believing that God does not exist, is included in negative atheism.

A positive atheist is someone who believes that God does not exist; she affirms that no such being exists. So positive atheists are negative atheists, but negative atheists need not be positive atheists. An analogy is useful. If a person believes that there is no such thing as unicorns, then she is a positive atheist concerning unicorns. Someone who has never heard of them would be a negative atheist with respect to unicorns. So would a person who has thought about the question, and is not sure whether they exist or not. (Martin 1990)

Atheism can be narrow or wide in scope. The narrow atheist does not believe in the existence of God (an omni- being). A wide atheist does not believe that any gods exist, including but not limited to the traditional omni-God. The wide positive atheist denies that God exists, and also deny that Zeus, Gefjun, Thor, Sobek, Bakunawa and others exist. The narrow atheist does not believe that God exists, but may not taking a stronger view about the existence or non-existence of other supernatural beings. One could be a narrow atheist about God, but still believe in the existence of some other supernatural entities. (This is one of the reasons that it is a mistake to identify atheism with materialism or naturalism.)

Separating these different senses of the term allows us to better understand the different sorts of justification that can be given for varieties of atheism with different scopes. An argument may serve to justify one form of atheism and not another—alleged contradictions within a Christian conception of god, by themselves, for instance, do not serve as evidence for wide atheism. But presumably, reasons that are adequate to show that there is no omni-God would be sufficient to show that there is no Islamic God.

Epistemological Approaches to Atheism

Justifications for atheism have taken forms that can be usefully divided into several categories. For the most part, atheists have taken an evidentialist approach to the question of God’s existence. That is, atheists have taken the view that whether or not a person is justified in having an attitude of belief towards the proposition “God exists,” is a function of that person’s evidence. “Evidence” here is understood broadly to include a priori arguments, arguments to the best explanation, inductive and empirical reasons, as well as deductive and conceptual premises. An asymmetry exists between theism and atheism in that atheists have not offered faith as a justification for non-belief. That is, atheists have not presented non-evidentialist defenses for believing that there is no God.

Not all theists appeal to faith, however. Many are evidentialist theists. The evidentialist theist and the evidentialist atheist may have a number of general epistemological principles concerning evidence, arguments, and implication in common, but then disagree about what the evidence is, how it should be understood, and what it implies. They may disagree, for instance, about whether the values of the physical constants and laws in nature constitute evidence for intentional fine tuning, but agree at least that whether God exists is a matter that can be explored empirically or with reason.

Many non-evidentialist theists may deny in one way or another that the acceptability of God belief depends upon evidence, reasons, or arguments as they have been classically understood. Faith or prudential based beliefs in God, for example, will fall into this category. The evidentialist atheist and the non-evidentialist theist, therefore, may have a number of more fundamental disagreements about the acceptability of believing despite inadequate or contrary evidence, the epistemological status of prudential grounds for believing, or the nature of God belief. Their disagreement may not be so much about the evidence, or even about God, but about the legitimate roles that evidence, reason, and faith should play in human belief structures.

It is not clear that arguments against atheism that appeal to faith have any prescriptive force the way appeals to evidence or arguments do. The general evidentialist view is that when a person grasps that an argument is sound, that imposes an epistemic obligation on her to accept the conclusion. Insofar as having faith amounts to believing contrary to or despite a lack of evidence, one person’s faith that God exists does not have this sort of inter-subjective, epistemological implication. Failing to believe what is clearly supported by the evidence is ordinarily irrational. Failure to have faith that some claim is true is not similarly culpable.

Justifying atheism, then, can entail several different projects. There are the evidential disputes over what information we have available to us, how it should be interpreted, and what it implies. There are also broader meta-epistemological concerns about the roles of argument, reasoning, belief, and religiousness in human life. The atheist can find herself not just arguing that the evidence indicates that there is no God, but defending science, the role of reason, and the necessity of basing beliefs on evidence more generally.

Friendly atheism. William Rowe has introduced an important distinction to modern discussions of atheism. If someone has arrived at what they take to be a reasonable and well-justified conclusion that there is no God, then what attitude should she take about another person’s persistence in believing in God, particularly when that other person appears to be thoughtful and at least prima facie reasonable? It seems that the atheist could take one of several views. The theist’s belief, as the atheist sees it, could be rational or irrational, justified or unjustified. Must the atheist who believes that the evidence indicates that there is no God conclude that the theist’s believing in God is irrational or unjustified? Rowe’s answer is no. (Rowe 1979, 2006)

Rowe and most modern epistemologists have said that whether a conclusion C is justified for S will be a function of the information (correct or incorrect) that S possesses and the principles of inference that S employs in arriving at C. But whether or not C is justified is not directly tied to its truth, or even to the truth of the evidence concerning C. That is, a person can have a justified, but false belief. She could arrive at a conclusion through an epistemically inculpable process and yet get it wrong. Ptolemy, for example, the greatest astronomer of his day, who had mastered all of the available information and conducted exhaustive research into the question, was justified in concluding that the Sun orbits the Earth. A medieval physician in the 1200s who guesses (correctly) that the bubonic plague was caused by the bacterium yersinia pestis would not have been reasonable or justified given his background information and given that the bacterium would not even be discovered for 600 years.

We can call the view that rational, justified beliefs can be false, as it applies to atheism, friendly or fallibilist atheism. The friendly atheist can grant that a theist may be justified or reasonable in believing in God, even though the atheist takes the theist’s conclusion to be false. What could explain their divergence to the atheist? The believer may not be in possession of all of the relevant information. The believer may be basing her conclusion on a false premise or premises. The believer may be implicitly or explicitly employing inference rules that themselves are not reliable or truth preserving, but the background information she has leads her, reasonably, to trust the inference rule. The same points can be made for the friendly theist and the view that he may take about the reasonableness of the atheist’s conclusion. It is also possible, of course, for both sides to be unfriendly and conclude that anyone who disagrees with what they take to be justified is being irrational. Given developments in modern epistemology and Rowe’s argument, however, the unfriendly view is neither correct, nor conducive to a constructive and informed analysis of the question of God.

Atheists have offered a wide range of arguments and justifications for non-belief. A notable modern view is Antony Flew’s presumption of atheism (1984). Flew argues that the default position for any rational believer should be neutral with regard to the existence of God, and to be neutral is to not have a belief regarding its existence. "The onus of proof lies on the man who affirms, not on the man who denies. . . on the proposition, not on the opposition,” Flew argues. Beyond that, coming to believe that such a thing does or does not exist will require justification, much as a jury presumes innocence concerning the accused and requires evidence in order to conclude that he is guilty. Flew’s negative atheist will presume nothing at the outset, not even the logical coherence of the notion of God, but her presumption will be defeasible, or revisable in the light of evidence. We shall call this view atheism by default.

The atheism by default position contrasts with a more permissive attitude that some people take regarding religious belief. The notions of religious tolerance and freedom are sometimes taken to indicate the epistemic permissibility of believing despite a lack of evidence in favor or even despite evidence to the contrary. One is in violation of no epistemic duty by believing, according to the common view, even if one lacks conclusive evidence in favor, or even if one has evidence that is on the whole against. In contrast to Flew’s jury model, we can think of this view as treating religious beliefs as permissible until proven incorrect. This sort of epistemic policy about God or any other matter has been controversial, and a major point of contention between atheists and theists. Atheists have argued that we typically do not take it to be epistemically inculpable or reasonable for a person to believe in Santa Claus, the Tooth Fairy, or some other supernatural being merely because they do not possess evidence to the contrary. Nor would we consider it reasonable for a person to begin believing that they have cancer because they do not have proof to the contrary. The atheist by default argues that it would be appropriate to not believe in such circumstances. The epistemic policy here takes its inspiration from an influential piece by W.K. Clifford (1999) in which he argues that it is wrong, always, everywhere, and for anyone, to believe anything for which there is insufficient reason.

There are several other approaches to the justification of atheism that we will consider below. There is a family of arguments, sometimes known as exercises in deductive atheology, for the conclusion that the existence of God is impossible. Another large group of important and influential arguments can be gathered under the heading inductive atheology. These probabilistic arguments invoke considerations about the natural world such as widespread suffering, nonbelief, or findings from biology or cosmology. Another approach, atheistic noncognitivism, denies that God talk is even meaningful or has any propositional content that can be evaluated in terms of truth or falsity. Rather, religious speech acts are better viewed as a complicated sort of emoting or expression of spiritual passion. Inductive and deductive approaches are cognitivistic in that they accept that claims about God have meaningful content and can be determined to be true or false.

Deductive Atheology

Many discussions about the nature and existence of God have either implicitly or explicitly accepted that the concept of God is logically coherent. That is, for many believers and non-believers the assumption has been that such a being as God could possibly exist, but they have disagreed about whether there actually is one. Atheists within the deductive atheology tradition, however, have not even granted that God, as he is typically described, is possible. The first question we should ask, argues the deductive atheist, is whether the description or the concept is logically consistent. If it is not, then no such being could possibly exist. The deductive atheist argues that some, one, or all of God’s essential properties are logically contradictory. Since logical impossibilities are not and cannot be real, God does not and cannot exist. Consider a putative description of an object as a four-sided triangle, a married bachelor, or prime number with more than 2 factors. We can be certain that no such thing fitting that description exists because what they describe is demonstrably impossible.

If deductive atheological proofs are successful, the results will be epistemically significant. Many people have doubts that the view that there is no God can be rationally justified. But if deductive disproofs show that there can exist no being with a certain property or properties and those properties figure essentially in the characterization of God, then we will have the strongest possible justification for concluding that there is no being fitting any of those characterizations. If God is impossible, then God does not exist.

It may be possible at this point to re-engineer the description so that it avoids the difficulties, but now the theist faces several challenges according to the deductive atheologist. First, if the traditional description of God is logically incoherent, then what is the relationship between a theist’s belief and some revised, more sophisticated account that allegedly does not suffer from those problems? Is that the being that she believed in all along? Now? What were the reasons that led her prior to considering the atheological arguments to believe in that conception of God? Secondly, if the classical characterizations of God are shown to be logically impossible, then there is a legitimate question as whether any new description that avoids those problems describes a being that is worthy of the label. It will not do, in the eyes of many theists and atheists, to retreat to the view that God is merely a somewhat powerful, partially-knowing, and partly-good being, for example. Thirdly, the atheist will still want to know on the basis of what evidence or arguments should we conclude that a being as described by this modified account exists? Fourthly, there is no question that there exist less than omni-beings in the world. We possess less than infinite power, knowledge and goodness, as do many other creatures and objects in our experience. What is the philosophical importance or metaphysical significance of arguing for the existence of those sorts of beings, and advocating belief in them? Fifthly, and most importantly, if it has been argued that God’s essential properties are impossible, then any move to another description seems to be a concession that positive atheism about God is justified.

Another possible response that the theist may take in response to deductive atheological arguments is to assert that God is something beyond proper description with any of the concepts or properties that we can or do employ. So complications from incompatibilities among properties of God indicate problems for our descriptions, not the impossibility of a divine being worthy of the label. Many atheists have not been satisfied with this response. The theist has now asserted the existence of and attempted to argue in favor of believing in a being that is, by their own admission, something that we cannot form a proper idea of, one that does not have properties that we can acknowledge; it is a being that defies comprehension. It is not clear how we could have reasons or justifications for believing in the existence of such a thing. It is not clear how it could be an existing thing in any familiar sense of the term in that it lacks comprehensible properties. Or put another way, as Patrick Grim notes, “If a believer’s notion of God remains so vague as to escape all impossibility arguments, it can be argued, it cannot be clear to even him what he believes—or whether what he takes for pious belief has any content at all.” (2007). It is not clear how it could be reasonable to believe in such a thing, and it is even more doubtful that it is epistemically unjustified or irresponsible to deny that such a thing is exists. It is clear, however, that the deductive atheologist must acknowledge the growth and development of our concepts and descriptions of reality over time, and she must take a reasonable view about the relationship of those attempts and revisions in our ideas about what may turns out to be real.

Single Property Disproofs

Deductive disproofs have typically focused on logical inconsistencies to be found either within a single property or between multiple properties. Philosophers have struggled to work out the details of what it would be to be omnipotent, for instance. It has come to be widely accepted that a being cannot be omnipotent where omnipotence simply means to power to do anything. This definition of the term suffers from the stone paradox. An omnipotent being would either be capable of creating a rock that he cannot lift, or he is incapable. If he is incapable, then there is something he cannot do, and therefore he does not have the power to do anything. If he can create such a rock, then again there is something that he cannot do, namely lift the rock he just created. So paradoxically, having the ability to do anything entails being unable to do some things. As a result, many theists and atheists have agreed that a being could not have that property. A number of attempts to work out an account of omnipotence have ensued. (Cowan 2003, Flint and Freddoso 1983, Hoffman and Rosenkrantz 1988 and 2006, Mavrodes 1977, Ramsey 1956, Sobel 2004, Savage 1967, and Wierenga 1989 for examples). See the entry on Omnipotence for details.

It has also been argued that omniscience is impossible, and that the most knowledge that can possibly be had is not enough to be fitting of God. One of the central problems has been that God cannot have knowledge of indexical claims such as, “I am here now.” It has also been argued that God can’t know future free choices, or God cannot know future contingent propositions, or that Cantor’s and Gödel proofs imply that the notion of a set of all truths cannot be made coherent. (Everitt 2004, Grim 1985, 1988, 1984, Pucetti 1963, and Sobel 2004). See the entry on Omniscience for more details.

The logical coherence of eternality, personhood, moral perfection, causal agency, and many others have been challenged in the deductive atheology literature. See bibliography.

Multiple Property Disproofs

Another form of deductive atheological argument attempts to show the logical incompatibility of two or more properties that God is thought to possess. A long list of properties have been the subject of multiple property disproofs: transcendence and personhood, justice and mercy, immutability and omniscience, immutability and infinitely love, omnipresence and agency, perfection and love, eternality and omniscience, eternality and creator of the universe, omnipresence and consciousness. (Blumenfeld 2003, Drange 1998b, Flew 1955, Grim 2007, Kretzmann 1966, and McCormick 2000 and 2003)

The combination of omnipotence and omniscience have received a great deal of attention. To possess all knowledge, for instance, would include knowing all of the particular ways in which one will exercise one’s power, or all of the decisions that one will make, or all of the decisions that one has made in the past. But knowing any of those entails that the known proposition is true. So can God have the power to act in some fashion that he has not foreseen, or differently than he already has without compromising his omniscience? It has also been argued that God cannot be both unsurpassably good and free. (Rowe 2004).

Failure of Proof Disproof

When attempts to provide evidence or arguments in favor of the existence of something fail, a legitimate and important question is whether anything except the failure of those arguments can be inferred. That is, does positive atheism follow from the failure of arguments for theism? A number of authors have concluded that it does. They taken the view that unless some case for the existence of God succeeds, we should believe that there is no God.

Many have taken an argument J.M. Findlay (1948) to be pivotal. Findlay, like many others, argues that in order to be worthy of the label “God,” and in order to be worthy of a worshipful attitude of reverence, emulation, and abandoned admiration, the being that is the object of that attitude must be inescapable, necessary, and unsurpassably supreme. (Martin 1990, Sobel 2004). God would not be the sort of being that we would merely infer inductively from evidence left behind in the universe. Such a being wouldn’t be contingent and the proof of its existence would not arise from contingency. That is to say that of all the approaches to God’s existence, the ontological argument is the strategy that we would expect to be successful were there a God, and if they do not succeed, then we can conclude that there is no God, Findlay argues.

Ontological arguments have attempted to show God’s existence on entirely a priori grounds from an analysis of the concept of the greatest logically possible being. But these attempts to prove God have not met with success, as most see it. Findlay says, “the general philosophical verdict is that none of these 'proofs' is truly compelling.” John Hick (1977), the well-known philosopher and theist, “we have in each case concluded, in agreement with the majority of contemporary philosophers, that these arguments fail to do what they profess to do.” And Alvin Plantinga says, “To show that these models are true, therefore, would also be to show that theism and Christianity are true; and I don’t know how to do something one could sensibly call ‘showing’ that either of these is true. I believe that there are a large number (at least a couple dozen) good arguments for the existence of God; none, however, can really be thought of as a showing or demonstration. As for classical Christianity, there is even less prospect for demonstrating its truth.” (Plantinga 2000) See the entry on the Ontological Argument.

The implications of Findlay’s argument for deductive atheology here is that Findlay makes a case that not just any being will be adequate to fill God’s shoes, as it were. God, in order to be God, must be a conceptually maximal being. And if the project to derive God’s existence through ontological arguments fail, then we can take that as indicator of more than merely the failure of those arguments. We should conclude that there is no such being. The Ontological Argument has all or nothing consequences for God’s existence.

Inductive Atheology

The view that there is no God or gods has been criticized on the grounds that it is not possible to prove a negative. No matter how exhaustive and careful our analysis, there could always be some proof, some piece of evidence, or some consideration that we have not considered. God could be something that we have not conceived, or God exists in some form or fashion that has escaped our investigation. Positive atheism draws a stronger conclusion than any of the problems with arguments for God’s existence alone could justify. Absence of evidence is not evidence of absence.

Findlay and the deductive atheological arguments attempt to address these concerns. But the question of inductive or probabilistic justifications for negative existential claims has figured centrally in modern atheism. The response to the “You cannot prove a negative” criticism has been that it invokes an artificially high epistemological standard of justification that creates a much broader set of problems not confined to atheism.

The general principle seems to be that one is not epistemically entitled to believe a proposition unless you have exhausted all of the possibilities and proven beyond any doubt that a claim is true. Or put negatively, one is not justified in disbelieving unless you have proven with absolute certainty that the thing in question does not exist. The problem is that we do not have a priori disproof that many things do not exist, yet it is reasonable and justified to believe that they do not: the Dodo bird is extinct, unicorns are not real, there is no teapot orbiting the Earth on the opposite side of the Sun, there is no Santa Claus, ghosts are not real, a defendant is not guilty, a patient does not have a particular disease, so on. There are a wide range of other circumstances under which we take it that believing that X does not exist is reasonable even though no logical impossibility is manifest. None of these achieve the level of deductive, a priori or conceptual proof. The objection to inductive atheism undermines itself in that it generates a broad, pernicious skepticism against far more than religious or irreligious beliefs. Mackie (1982) says, “It will not be sufficient to criticize each argument on its own by saying that it does not prove the intended conclusion, that is, does not put it beyond all doubt. That follows at once from the admission that the argument is non-deductive, and it is absurd to try to confine our knowledge and belief to matters which are conclusively established by sound deductive arguments. The demand for certainty will inevitably be disappointed, leaving skepticism in command of almost every issue.” (p. 7) If the atheist is unjustified for lacking deductive proof, then it is argued, then it would appear that so are the beliefs that planes fly, fish swim, or that there exists a mind-independent world.

The atheist can also wonder what the point of the objection is. When we lack deductive disproof that X exists, should we be agnostic about it? Is it permissible to believe that it does exist? Clearly, that would not be appropriate. Gravity may be the work of invisible, undetectable elves with sticky shoes. We don’t have any certain disproof of the elves—physicists are still struggling with an explanation of gravity. But surely someone who accepts the sticky-shoed elves view until they have deductive disproof is being unreasonable. It is also clear that if you are a positive atheist about the gravity elves, you would not be unreasonable. You would not be overstepping your epistemic entitlement by believing that no such things exist. On the contrary, believing that they exist or even being agnostic about their existence on the basis of their mere possibility would not be justified. So there appear to be a number of precedents and epistemic principles at work in our belief structures that provide room for inductive atheism. But these issues in the epistemology of atheism and recent work by Graham Oppy (2006) suggest that more attention must be paid to the principles that describe epistemic permissibility, culpability, reasonableness, and justification with regard to the theist, atheist, and agnostic categories.

Below we will consider several groups of influential inductive atheological arguments

The Santa Claus Argument for Atheism

Martin (1990) offers this general principle to describe the criteria that render the belief “X does not exist” justified:

A person is justified in believing that X does not exist if

(1) all the available evidence used to support the view that X exists is shown to be inadequate; and

(2) X is the sort of entity that, if X exists, then there is a presumption that would be evidence adequate to support the view that X exists; and

(3) this presumption has not been defeated although serious efforts have been made to do so; and

(4) the area where evidence would appear, if there were any, has been comprehensively examined; and

(5) there are no acceptable beneficial reasons to believe that X exists. (pg. 283)

Many of the major works in philosophical atheism that address the full range of recent arguments for God’s existence (Gale 1991, Mackie 1982, Martin 1990, Sobel 2004, Everitt 2004, and Weisberger 1999) can be seen as providing evidence to satisfy the first, fourth and fifth conditions. A substantial body of articles with narrower scope (see bibliography) can also be understood to play this role in justifying atheism. And a large group of discussions of Pascal’s Wager and related prudential justifications in the literature can also be seen as relevant to the satisfaction of the fifth condition.

One of the interesting and important questions in the epistemology of philosophy of religion has been whether the second and third conditions are satisfied concerning God. If there were a God, how and in what ways would we expect him to show in the world? Empirically? Conceptually? Would he be hidden? Martin argues, and many others have accepted implicitly or explicitly, that God is the sort of thing that would manifest in some discernible fashion to our inquiries. Martin concludes, therefore, that God satisfied all of the conditions, so, positive narrow atheism is justified.

Problem of Evil

At least since Epicurus said, “Is he willing to prevent evil, but not able? then is he impotent. Is he able, but not willing? then is he malevolent. Is he both able and willing? whence then is evil? Is he neither able nor willing? Then why call him God?" the problem of evil has been one of the most compelling arguments in favor of positive atheism. Both deductive and inductive forms of the argument are included here under inductive arguments for atheism for clarity.

It is useful to distinguish between different kinds of evil and different problems. Moral evil is apparently pointless suffering or death that is inflicted by humans on other sentient creatures. We say “apparently” because few will deny that there are many instances of suffering that appear to be pointless, hence the conflict with the existence of a being who would and could prevent them, but the issue under contention between many theists and atheists has been whether those instances of apparently pointless evil really are pointless. If in fact they are not pointless, as many theists argue is possible, the problem of evil may not be as big a problem for theism as alleged. Moral evil can be divided into suffering that is inflicted by humans on other humans (HH), or by humans on non-human animals (HA). Natural evil is apparently pointless suffering and death that is the result of natural forces such as hurricanes, plagues, and droughts. Nature inflicts suffering on humans (NH) and on non-human animals (NA)

The strongest form of the argument, if successful, is often characterized as the deductive or logical problem of evil. The existence of the suffering and the existence of God are logically incompatible. It is not possible for both claims: “God exists” where it is understood that God is all knowing, all powerful, and all good and “Evil exists” to both be true. But we know that there is evil, therefore it is not true that there is a being that is God. (Mackie 1955,1982, and Sobel 2004).

Inductive forms of the argument can allow that God could possibly coexist with evil in the world, but they argue that it would not be reasonable to think that he in fact does. The most famous and influential version of this argument is William Rowe’s:

1. There exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse.

2. An omniscient, wholly good being would prevent the occurrence of any intense suffering it could, unless it could not do so without thereby losing some greater good or permitting some evil equally bad or worse.

3. Therefore, there does not exist an omnipotent, omniscient, wholly good being. (Rowe 1979)

The deductive and inductive arguments and the kinds of evil can be represented this way:

Deductive Problem of Moral Evil: The existence of God is not logically compatible with the existence of all instances of HH and HA.

Deductive Problem of Natural Evil: The existence of God is not logically compatible with all instances of NH and NA.

Inductive Problem of Moral Evil: The existence of instances of HH and/or HA makes existence of God improbable.

Inductive Problem of Natural Evil: The existence of instances of NH and/or NA makes the existence of God improbable.

Theistic Rejoinders to the Deductive Argument.

Since it aims to make the strongest possible case against God, the deductive problem is most vulnerable to rebuttals. What would it take to refute the logical problem argument? If the charge is that it is not possible for God to coexist with evil, then the critic need only show that it is possible; responding to the logical problem of evil alone does not require arguing for the truth of any particular justification of evil that God has.

Freedom, it has been frequently argued, could be a sufficient good in itself to justify moral evil. And since God (whose power is confined at least to that which is logically possible) cannot bring it about that we are both free and that we choose good, our free evil choices produce HH and HA suffering, but that suffering is our responsibility, not God’s. The suffering serves a greater good and it is at least possible that there is no other way for God to achieve that good, so it is possible that he has a morally defensible reason for allowing it.

The soul-building defense alleges that by creating a world with free beings who can experience the good and evil results of their actions, it is possible for them to grow intellectually and morally and achieve virtue in ways that not even an omnipotent God could accomplish through direct creation. It is possible that creating a world where there are instances of HH, HA, NH and even NA is God’s best means of achieving moral virtue. And that moral virtue or its possibility is sufficiently valuable to justify their existence. Even the inflexibility of natural laws that sometimes produce disasters and profound suffering catalyzes the expansion of human knowledge, power, and moral responsibility.

Theodicies such these are thought by many to establish that it is at least possible that some forms of suffering are not pointless. That would indicate that the deductive problem of evil for positive atheism for that form of suffering fails.

The discussion about the logical incompatibility of God and evil is far from settled. A number of philosophers in recent years have persisted with arguments that the existence of evil precludes the possibility of God’s existence. (Gale 1991, Sobel 2004, Martin 1990, Smith 1997 for instance) See the entries on the Problem of Evil, the Freewill Defense, and Soul-Building for more.

Theistic Rejoinders to the Inductive Argument

Much recent work has been done on the inductive argument, and not surprisingly, attention has focused largely on Rowe’s first premise. Rowe’s example of a fawn that is horribly burned in a forest fire started by lightening away from any human contact has come to serve as a watershed case for theists and atheists. In the history of all sentient beings, human and animal, and among all of the cases where they have suffered, if there is a single case of suffering that serves no greater good, or that does not serve to avoid some equal or greater evil, then it would seem reasonable to conclude there is no God. Surely it is reasonable to think that there are many of those cases, argues Rowe (1979, 1984, 2006).

Rather than arguing that there are no such cases, an influential set of theistic responses under the heading skeptical theism have given arguments that we are not in a position to know whether or not there are or have been such cases. We have a limited capacity to grasp the total cosmic picture from God’s perspective. There could be long term effects that we do not see or appreciate, or the alternatives may be worse. So skeptical theists have suspended judgment about premise 1.

Atheists for the most part have found these rejoinders implausible and inconsistent with what they take to be obvious instances of evil that can and should be prevented, but aren’t. The debate has shifted to other epistemological issues surrounding the conditions under which appearances of pointless evil warrant the inference to the existence of pointless evil. (Rowe 1979, 1984, 2006, Wykstra 1984, Alston 1996, van Inwagen 1996) See the entry on the Problem of Evil.

God, Atheism and Cosmology

Questions about the origins of the universe and cosmology have been the focus for many inductive atheism arguments. We can distinguish four recent views about God and the cosmos. Naturalism: On naturalistic view, the Big Bang occurred approximately 13.7 billion years ago, the Earth formed out of cosmic matter about 4.6 billion years ago, and life forms on Earth, unaided by any supernatural forces about 4 billion years ago. Various physical (non-God) hypotheses are currently being explored about the cause or explanation of the Big Bang such as the Hartle-Hawking no-boundary condition model, brane cosmology models, string theoretic models, ekpyrotic models, cyclic models, chaotic inflation, and so on. Big Bang Theism: We can call the view that God caused about the Big Bang 13.7 billion years ago Big Bang Theism. Intelligent Design Theism: There are many variations, but most often the view is that God created the universe, perhaps with the Big Bang 13.7 billion years ago, and then beginning with the appearance of life 4 billion years ago, God supernaturally guided the formation and development of life into the forms we see today. Creationism: Finally, there is a group of people who for the most part denies the occurrence of the Big Bang and of evolution altogether; God created the universe, the Earth, and all of the life on Earth in its more or less present form 6,000-10,000 years ago.

Taking a broad view, many atheists have concluded that neither Big Bang theism, Intelligent Design Theism, nor Creationism is the most reasonable description of the history of the universe. Before the theory of evolution and recent developments in modern astronomy, a view wherein God did not play a large role in the creation and unfolding of the cosmos would have been hard to justify. But now, internal problems with those views and the evidence from cosmology and biology, indicate that naturalism is the best explanation.

Justifications for Big Bang Theism have focused on modern versions of the Cosmological and Kalaam arguments. Since everything that comes into being must have a cause, including the universe, then God was the cause of the Big Bang. (Craig 1995) See the entries on Cosmological

Arguments and the Kalaam argument.

The objections to these arguments have been numerous and vigorously argued. Critics have challenged the inference to a supernatural cause to fill gaps in the natural account, as well as the inferences that the first cause must be a single, personal, all-powerful, all-knowing, and all-good being. It is not clear that any of the properties of God as classically conceived in orthodox monotheism can be inferred from what we know about the Big Bang without first accepting a number of theistic assumptions. Infinite power and knowledge do not appear to be required to bring about a Big Bang—what if our Big Bang was the only act that a being could perform? There appears to be consensus that infinite goodness or moral perfection cannot be inferred as a necessary part of the cause of the Big Bang—theists have focused their efforts in the problem of evil discussions just attempting to prove that it is possible that God is infinitely good given the state of the world. Big Bang Theism would need to show that no other sort of cause besides a morally perfect one could explain the universe we find ourselves in. Critics have also doubted whether we can know that some supernatural force that caused the Big Bang is still in existence now or is the same entity as identified and worshipped in any particular religious tradition. Even if major concessions are granted in the cosmological argument, all that it would seem to suggest is that there was a first cause or causes, but widely accepted arguments from that first cause or causes to the fully articulated God of Christianity or Islam, for instance, have not been forthcoming.

In some cases, atheists have taken the argument a step further. They have offered cosmological arguments for the nonexistence of God on the basis of considerations from physics, astronomy, and subatomic theory. These arguments are quite technical, so these remarks will be cursory. God, if he exists, knowing all and having all power, would only employ those means to his ends that are rational, effective, efficient, and optimal. If God were the creator, then he was the cause of the Big Bang. But cosmological atheists have argued that the singularity that produced the Big Bang and events that unfold thereafter preclude a rational divine agent from achieving particular ends with the Big Bang as the means. The Big Bang would not have been the route God would have chosen to this world as a result. (Stenger 2007, Smith 1993, Everitt 2004.)

Teleological Arguments for and Against the Existence of God

We can divide modern teleological, or design, arguments for the existence of God into two groups: arguments from biological complexity that focus on organisms, and arguments from physical nature and complexity that emphasize the fundamental physical laws that apply to all matter.

Before Darwin, a design argument from biological complexity like William Paley’s famous Watch argument had some considerable force (2006). Paley’s argument proceeds by an analogy: were you to find a watch on the beach with intricate, well-adjusted parts that fit together and that all contribute to performing a larger function (telling time), you would reasonably conclude that the watch had been designed and built by a designer with a deliberate set of intentions in mind. Such a thing would not come about by the blind, mindless forces of nature. By analogy, we find similar complexity, adjustment, and form in biological organisms. The inner workings of the eye with the remarkable coordination of the retina, lens, optic nerve, cornea, iris, and other parts has been a favorite example of advocates.

Biological research into the phenomena of natural selection has shown that what Paley thought was impossible has occurred for billions of years on Earth. Adaptive variations in organisms, inherited through genetic transmission, accumulate over time and give the appearance of purposeful design in the relationship between the parts of the organism and its ability to function in its environment. Among the vast majority of those who understand the theory, there is a consensus that it is a better, non-teleological explanation of biological complexity than the Creationism hypothesis. (Dawkins 1976, 1986, 2006)

Intelligent Design Arguments: A few arguments that aspects of biological complexity are best explained by God have emerged from discussions of evolution. These arguments grant the general claim that some form of natural selection occurred in evolutionary history and that it is responsible for the most part for that appearance of biological design. But there are specific structures in some organisms or episodes in evolutionary history that cannot be completely explained by neo-Darwinian selection. Evolution occurred, but God must have periodically intervened to guide the process in order to produce life as we find it. (Behe 1996, Dembski 1998)

These arguments have been heavily criticized. In some cases, biologists have simply explained the process whereby natural selection sans God could have produced the examples of biological structures that were alleged to be impossible on evolution. Intermediate fossils have been found, and many of the gaps in our understanding of the process and history of evolution have closed. The probabilistic reasoning that sought to prove a supernatural source has been critiqued. (Sober 2002, 2003, 2007, Kitcher 1982). As the objections have mounted, many have become more convinced that there is little need for God in our account of the history of life.

Physical Design or Fine Tuning Arguments: Among philosophical theists, much attention has shifted to new teleological arguments that accept evolution, but argue for God on the basis of various parameters, laws, and constants of physics, chemistry and cosmological history that are conducive to the existence of life in our universe. According to these arguments, given all of the other configurations of a physical world that are possible, many of them non-biophilic, the hospitability of this universe to life suggests a designer with a purpose. If the cosmological constant had been even slightly greater or less, or if gravity had been slightly stronger, for instance, there would be no life and no humanity in the universe. (Leslie 1996, Swinburne 1979)

Problems with Arguments from Design: The general difficulty for design arguments is deriving the strong conclusion that God must have been responsible when much lesser beings or forces are sufficient to explain the evidence. Critics have argued that even if we grant the conclusion that there is evidence of purposeful design manifest in the world, powerful aliens could be responsible who are running a cosmic zoo for humans. Almost three centuries ago, Hume argued that the evidence may well point to a baby god, an idiot god, a committee of gods, a cruel or absent minded god, an absentee god, or some other possibility (Hume 1935). The evidence that design theists cite is not sufficient to motivate the conclusion that a being with infinite power, knowledge, and goodness must have been responsible. If it was designed at all, some inferior being seems to be better indicated.

Another way to understand the point is that any properties that we would find in the universe that might identify it as an artifact will be properties that we have observed in artifacts that we have created. But the properties that we find in our artifacts that indicate that they were created all directly or indirectly reveal our limitations, our inabilities, and other finite aspects of our creation. So the properties indicative of creation that we might find in the universe will also point to limitations, inabilities, and finite aspects of the creator of the universe. That is, our best examples of artifact creation are anthropomorphic—they reflect our limited natures. So if we find features in the universe that hearken to examples of known human creation, at best, they will suggest an anthropomorphic, non-divine creator. But if God really does exist, then the universe was not created by a limited, finite being, and the markers of intelligent design in it will not resemble our own. So pointing to anthropomorphic design features in the universe that we can recognize actually suggests that God, a divine being who would be capable of much more, was not responsible. And if God did not create the universe, that suggests that God does not exist

Furthermore, physicists and astronomers are actively considering a wide range of theories, some of them that can make testable predictions, about why there is something rather than nothing, why we find these lawlike regularities instead of none or some others, and why the laws we find appear to be conducive to life. Given the history of scientific inquiry, if a non-supernatural answer to a question seems to be pending, it would be premature and irresponsible to insist that only a theistic answer will suffice.

.

Bayesian Arguments

Many of the modern teleological arguments for the existence of God have employed Bayesian calculations of probability. Bayes’ Theorem provides us with a way to calculate the likelihood that a hypothesis is true given some observations. (See entry on Bayes Theorem). For example, with a patient with a specific set of symptoms, we know from past experience that those symptoms are more likely to occur as a result of the patient’s having disease X than disease Y. When we are not in possession of the necessary background data, Bayes’ theorem also allows us to make a subjective calculation for an individual involving what they believe to be the likelihoods of those observations and relevant explanations. In these instances, the theorem gives us a measure of epistemic probability or the degree of surprise that S would have to make an observation given one hypothesis or another. In the case of God, Bayesian arguments for theism have attempted to show that the occurrence of our biophilic universe, for instance, would be very surprising if the atheistic hypothesis were true. But the occurrence of a universe with life would be likely, or at least less unlikely, if there were a God. Therefore, it is likely that there is a God, or at least more likely than the alternative.

The atheist’s challenge to Bayesian arguments have focused on those prior assumptions that factor into S’s subjective degree of surprise. A subject’s degree of surprise that some observations will occur very often has little to do with their actual probability. I might attach a very low value to the probability of robberies occurring in my neighborhood that does not take into account the fact, unknown to me, that my neighborhood has the highest robbery rate in the country. I might be surprised to see their of them in one week, nevertheless, their occurrence would not therefore be improbable. So while the God explanation might be tempting to explain various observations we are making about the universe, its origins, and its limits, all of the answers and information that we do not have suggests that we should not attach too much weight to our subjective sense of surprise, or unfounded prior probabilities.

In this vein, many of the modern arguments from design are probabilistic; it is exceedingly improbable that just this set of natural laws would occur, or that all of the matter in the universe would behave according to the same natural laws. But the probability judgments here are not based upon some larger set of base rate data about the frequency of lawful and unlawful universes and their being conjoined with a God or no God. We can have no such data with only one universe. Natural laws in the universe are what make probability judgments possible, so these arguments seem to be claiming that probability itself is improbable. On what grounds would we base that claim? All that we seem to have is our subjective sense of surprise, which as we have seen, can be a highly unreliable indicator of what is actually likely or unlikely. These appeals to probabilities to prove God appear to be either ill-formed, subjective in a fashion that undermines their rational force, or they rest upon claims whose prior probabilities the atheist contests.

The Argument Against Design

In William Paley’s famous analysis, he argues by analogy that the presence of order in the universe, like the features we find in a watch, are indicative of the existence of a designer who is responsible for the artifact. Many authors—David Hume (1935), Wesley Salmon (1978), Michael Martin (1990)—have argued that a better case can be made for the nonexistence of God from the evidence.

Salmon, giving a modern Bayesian version of an argument that begins with Hume, argues that the likelihood that the ordered universe was created by intelligence is very low. In general, instances of biologically or mechanically caused generation without intelligence are far more common than instances of creation from intelligence. Furthermore, the probability that something that is generated by a biological or mechanical cause will exhibit order is quite high. Among those things that are designed, the probability that they exhibit order may be quite high. But that is not the same as asserting that among the things that exhibit order, the probability that they were designed is high. Among dogs, the incidence of fur may be high, but it is not true that among furred things, the incidence of dogs is high. Furthermore, intelligent design and careful planning very frequently produces disorder—war, industrial pollution, insecticides, and so on.

So we can conclude that the probability that an unspecified entity (like the universe), which came into being and exhibits order, was produced by intelligent design is very low.

Arguments from Nonbelief

Another recent group of inductive atheistic arguments has focused on widespread nonbelief itself as evidence that atheism is justified. The common thread in these arguments is that something as significant in the universe as God could hardly be overlooked. The ultimate creator of the universe and a being with infinite knowledge, power, and love would not escape our attention, particularly since humans have devoted such staggering amounts of energy to the question for so many centuries. Perhaps more importantly, a being such as God, if he chose, could certainly make his existence manifest to us. Creating a state of affairs where his existence would be obvious, justified, or reasonable to us, or at least more obvious to more of us than it is currently, would be a trivial matter for an all-powerful being. So since our efforts have not yielded what we would expect to find if there were a God, then the most plausible explanation is that there is no God.

One might complain that we should not assume that God’s existence would be evident to us. There may be reasons, some of which we can describe, others that we do not understand, that God could have for remaining out of sight: revealing himself is not something he desires, remaining hidden enables people to freely love, trust and obey him, remaining hidden prevents humans from reacting from improper motives like fear of punishment. Remaining hidden preserves human freewill

.

The non-belief atheist does not find these speculations convincing for several reasons. In religious history, God’s revealing himself to Moses, Muhammad, Jesus’ disciples, and even Satan himself did not compromise their cognitive freedom in any significant way. Furthermore, attempts to explain why a universe where God exists would look just as we would expect a universe with no God have seemed ad hoc. Some of the logical positivists’ and non-cognitivists’ concerns surface here. If the believer maintains that a universe inhabited by God will look exactly like one without, then we must wonder what sort of counter-evidence would be allowed, even in principle, against the theist’s claim. If no state of affairs could be construed as evidence against God’s existence, then what does the claim “God exists,” mean and what are its real implications? Alternately, how can it be unreasonable to not believe in the existence of something that defies all of our attempts to corroborate or discover?

Theodore Drange (2006) has developed an argument that if God were the sort of being that wanted humans to come to believe that he exists, then he could bring it about that far more of them would believe than currently do. God would be able, he would want humans to believe, there is nothing that he would want more, and God would not be irrational. So God would bring it about that people would believe. In general, he could have brought it about that the evidence that people have is far more convincing than what they have. He could have miraculously appeared to everyone in a fashion that was far more compelling than the miracles stories that we have. But it is not the case that all, nearly all, or even a majority of people believe, so there must not be a God of that sort.

J.L. Schellenberg (1993) has developed an argument based upon a number of considerations that lead us to think that if there were a loving God, then we would expect to find some manifestations of him in the world. If God is all powerful, then there would be nothing restraining him from making his presence known. And if he is omniscient, then surely he would know how to reveal himself. And perhaps most importantly if God is good and if God possesses an unsurpassable love for us, then God would consider each human’s requests as important and seek to respond quickly. He would wish to spare those that he loves needless trauma. He would not want to give those that he loves false or misleading thoughts about his relationship to them. He would want as much personal interaction with them as possible. But of course, these conditions are not satisfied. So it is strongly indicated that there is no such God.

Schellenberg gives this telling parable:

You’re still a small child, and an amnesiac, but this time you’re in the middle of a vast rain forest, dripping with dangers of various kinds. You’ve been stuck there for days, trying to figure out who you are and where you came from. You don’t remember having a mother who accompanied you into this jungle, but in your moments of deepest pain and misery you call for her anyway: “Mooooommmmmmm!” Over and over again. For days and days. . . the last time when a jaguar comes at you out of nowhere. . . but with no response. What should you think in this situation? In your dying moments, what should cross your mind? Would the thought that you have a mother who cares about you and hears your cry and could come to you but chooses not to even make it onto the list? (2006, pg. 31)

Like Drange, Schellenberg argues that there are many people who are epistemically inculpable in believing that there is no God. That is, many people have carefully considered the evidence available to them, and have actively sought out more in order to determine what is reasonable concerning God. They have fulfilled all relevant epistemic duties they might have in their inquiry into the question. And they have arrived at a justified belief that there is no God. If there were a God, however, evidence sufficient to form a reasonable belief in his existence would be available. So the occurrence of widespread epistemically inculpable nonbelief itself shows that there is no God.

Atheistic Naturalism

The final family of inductive arguments we will consider involves drawing a positive atheistic conclusion from broad, naturalized grounds.

Methodological naturalism can be understood as the view that the best or the only way to acquire knowledge within science is by adopting the assumption that all physical phenomena have physical causes. This presumption by itself does not commit one to the view that only physical entities and causes exist, or that all knowledge must be acquired through scientific methods. Methodological naturalism, therefore, is typically not seen as being in direct conflict with theism or having any particular implications for the existence or non-existence of God.

Ontological naturalism, however, is usually seen as taking a stronger view about the existence of God. Ontological naturalism is the additional view that all and only physical entities and causes exist.

Among its theistic critics, there has been a tendency to portray ontological naturalism as a dogmatic ideological commitment that is more the product of a recent intellectual fashion than science or reasoned argument. But two developments have contributed to a broad argument in favor of ontological naturalism as the correct description of what sorts of things exist and are causally efficacious. First, there is a substantial history of the exploration and rejection of a variety of non-physical causal hypotheses in the history of science. Over the centuries, the possibility that some class of physical events could be caused by a supernatural source, a spiritual source, psychic energy, mental forces, or vital causes have been entertained and found wanting. Second, evidence for the law of the conservation of energy has provided significant support to physical closure, or the view that the natural world is a complete closed system in which physical events have physical causes. At the very least, atheists have argued, the ruins of so many supernatural explanations that have been found wanting in the history of science has created an enormous burden of proof that must be met before any claim about the existence of another worldly spiritual being can have credence. Ontological naturalism should not be seen as a dogmatic commitment, its defenders have insisted, but rather as a defeasible hypothesis that is supported by centuries of inquiry into the supernatural.

As scientific explanations have expanded to include more details about the workings of natural objects and laws, there has been less and less room or need for invoking God as an explanation. It is not clear that expansion of scientific knowledge disproves the existence of God in any formal sense any more than it has disproven the existence of fairies, the atheistic naturalist argues. But physical explanations have increasingly rendered God explanations extraneous and anomalous. When Laplace, the famous 18th century French mathematician and astronomer, presented his work on celestial mechanics to Napoleon, the Emperor asked him about the role of a divine creator in his system, for instance, Laplace is reported to have said, “I have no need for that hypothesis.”

In many cases, science has shown that particular ancillary theses of traditional religious doctrine are mistaken. Blind, petitionary prayer has been investigated and found to have no effect on the health of its recipients, although praying itself may have some positive effects on the person who prayers. Geology, biology, and cosmology have discovered that the Earth formed approximately 3 billion years ago out of cosmic dust, and life evolved gradually over billions of years; the Earth, humans, and other life forms were not created in their present form some 6,000-10,000 years ago. Many alleged miraculous events have been investigated and debunked.

Wide, positive atheism, the view that there are no gods whatsoever, might appear to be the most difficult atheistic thesis to defend, but ontological naturalists have responded that the case for no gods is parallel to the case for no elves, pixies, dwarves, fairies, goblins, or other creates. A decisive proof against every possible supernatural being is not necessary for the conclusion that none of them are real to be justified. The ontological naturalist atheist believes that once we have devoted sufficient investigation into enough particular cases and the general considerations about natural laws, magic, and supernatural entities, it becomes reasonable to conclude that the whole enterprise is an explanatory dead end for figuring out what sort of things there are in the world. Consider that the U.S. and British patent offices have instituted policies refusing to consider any more patent applications for perpetual motion machines.

The disagreement between atheists and theists continues on two fronts. Within the arena of science and the natural world, some believers have persisted in arguing that material explanations are inadequate to explain all of the particular events and phenomena that we observe. Some philosophers and scientists have argued that for phenomena like consciousness, human morality, and some instances of biological complexity, explanations in terms of natural or evolutionary theses have not and will not be able to provide us with a complete picture. Therefore, the inference to some supernatural force is warranted. While some of these attempts have received social and political support, within the scientific community the arguments that causal closure is false and that God as a cause is a superior scientific hypothesis to naturalistic explanations have not received significant support. Science can cite a history of replacing spiritual, supernatural, or divine explanations of phenomena with natural ones from bad weather as the wrath of angry gods to disease as demon possession. The assumption for many is that there are no substantial reasons to doubt that those areas of the natural world that have not been adequately explained scientifically will be given enough time. ( Madden and Hare 1968, Papineau, Manson, Nielsen 2001, and Stenger)

Increasingly, with what they perceive as the failure of attempts to justify theism, atheists have moved towards naturalized accounts of religious belief that give causal and evolutionary explanations of the prevalence of belief. (See Atrans, Boyer, Dennett 2006)

Cognitivism and Non-Cognitivism

In 20th century moral theory, a view about the nature of moral value claims arose that has an analogue in discussions of atheism. Moral non-cognitivists have denied that moral utterances should be treated as ordinary propositions that are either true or false and subject to evidential analysis. On their view, when someone makes a moral claim like, “cheating is wrong,” what they are doing is more akin to saying something like, “I have negative feelings about cheating. I want you to share those negative feelings. Cheating. Bad.”

A non-cognitivist atheist denies that religious utterances are propositions. They are not the sort of speech act that have a truth value. They are more like emoting, singing, poetry, or cheering. They express personal desires, feelings of subjugation, admiration, humility, and love. As such, they cannot and should not be dealt with by denials or arguments any more than I can argue with you over whether or not a poem moves you. There is an appeal to this approach when we consider common religious utterances such as, “Jesus loves you.” “Jesus died for your sins.” “God be with you.” What these mean, according to the non-cognitivist, is something like, “I have sympathy for your plight, we are all in a similar situation and in need of paternalistic comforting, you can have it if you perform certain kinds of behaviors and adopt a certain kind of personal posture with regard to your place in the world. When I do these things I feel joyful, I want you to feel joyful too.”

So the non-cognitivist atheist does not claim that the sentence “God exists” is false, as such. Rather, when people makes these sorts of claims, their behavior is best understood as a complicated publicizing of a particular sort of subjective sensations. Strictly speaking, the claims do not mean anything in terms of assertions about what sorts of entities do or do not exist in the world independent of human cognitive and emotional states. The non-cognitivist characterization of many religious speech acts and behaviors has seemed to some to be the most accurate description. But for the most part, atheists appear to be cognitivist atheists. They assume that religious utterances do express propositions that are either true or false. Positive atheists will argue that there are compelling reasons or evidence for concluding that in fact those claims are false. (Drange 2006, Diamond and Lizenbury 1975, Nielsen 1985)

It would appear that the non-cognitivists are at least partially correct. Many religious utterances are non-cognitive. Religious ceremonies, rituals, and liturgies tend away from assertions that are true or false and more towards some kind of religious expressionism. Recognizing the extent to which religious behaviors are non-cognitive can clarify the disagreement between theists and atheists. If our speech acts or behaviors shift from cognitive to non-cognitive and back, we would do well to be aware of it. But many religious utterances are clearly treated as cognitive by their speakers—they are meant to be treated as true or false claims, they are treated as making a difference, and they clearly have an impact on people’s lives and beliefs beyond the mere expression of a special category of emotions. So non-cognitivism does not appear to completely address belief in God.

Future Prospects for Atheism

20th century developments in epistemology, philosophy of science, logic, and philosophy of language indicate that many of the presumptions that supported old fashioned natural theology and atheology are mistaken. It appears that even our most abstract, a priori, and deductively certain methods for determining truth are subject to revision in the light of empirical discoveries and theoretical analyses of the principles that underlie those methods. Certainty, reasoning, and theology, after Bayes, Wittgenstein, Quine, and Kripke are not what they used to be. The prospects for a simple, confined argument for atheism (or theism) that achieves widespread support or that settles the question are dim. That is because, in part, the prospects for any argument that decisively settles a philosophical question where a great deal seems to be at stake are dim. The existence or non-existence of any non-observable entity in the world is not settled by any single argument or consideration. Every premise will be based upon other concepts and principles that themselves must be justified. So ultimately, the adequacy of atheism as an explanatory hypothesis about what is real will depend upon the overall coherence, internal consistency, empirical confirmation, and explanatory success of a whole worldview within which atheism is only one small part. The question of whether or not there is a God sprawls onto related issues and positions about biology, physics, metaphysics, explanation, philosophy of science, ethics, philosophy of language, and epistemology. The reasonableness of atheism depends upon the overall adequacy of a whole conceptual and explanatory description of the world.