Criticism of EAAN

Plantinga's argument has been criticized in detail from the viewpoints of natural sciences and philosophy of science. Within the field of evolutionary science, critics admit that organisms can be deceived, although not to the degree that Plantinga's argument would suggest (Ruse ; Boudry ). Several authors have criticized Plantinga for attaching “low or inscrutable” probability values to evolution's capability of producing true beliefs without any empirical support. For example, Branden Fitelson and Elliott Sober () have argued that his probability arguments are vague and would require more specific probabilities. Others have argued that evolution actually does care about truth (Law ; Wilkins and Griffiths ; Boudry and Vlerick ), because every organism needs to be able to navigate its environment in a reasonably reliable manner: organisms need a representation of the world that is accurate enough for survival and reproduction. Systematically wrong beliefs would thus be mostly selected against according to evolutionary theory. Recently, researchers have investigated the possibility of “adaptive misbelief,” that is, beliefs that are biologically adaptive despite their falsity. For example, unrealistic perceptions of one's capabilities and talents, as well as unwarranted optimism about the future, could lead to higher performance and thus be fitness‐enhancing. Even if adaptive misbeliefs exist, however, they remain rare, and no single category of adaptive misbelief is undisputed. By and large, as Ryan McKay and Daniel Dennett () argue, true belief pays off, and adaptive misbeliefs may be the exceptions that prove the rule.

Critics of Plantinga's argument also note that although our cognitive faculties are indeed fallible and can sometimes give rise to false beliefs, Plantinga presents too strong a dichotomy, making either all of our beliefs or none of them reliable (Boudry and Vlerick ). In fact, it has long been known that our cognitive faculties are sometimes prone to biases (Nickerson ; Schacter, Guerin, and St. Jacques ; Wilson and French ), but these biases can only be detected against a background of accurate and reliable belief formation. If our biases were as pervasive as Plantinga suggests, scientists could not even investigate them, because they could never be escaped from.

William J. Talbott () has also examined the EAAN and concludes that its definition of rationality depends too heavily on our past errors. However, rationality encompasses the capability of entertaining new hypotheses in the light of new evidence. Similarly, Michael Levin () has noted that although we can be misled by our faculties, “we can find this out and correct for it by comparing the predictions of our theories to observation.” In addition to assessing our individual cognitive faculties, Plantinga also disregards the quality of collective cultural knowledge and the improvement this brings through experimentation and observation (Childers ). Hence, even the very “belief that one's perceptual faculties are unreliable can be rationally revised in the light of appropriate perceptual evidence” (Talbott ). William Ramsey () has noted that some cases of unreliable cognition are predicted by the theory of evolution, precisely because evolution is unguided and subject to several constraints. Evolution is a satisficing process that finds imperfect but workable solutions to adaptive problems. By contrast, this pattern would not be expected of intelligently designed organisms. Thus, the existence of cognitive biases could backfire against theism.

In response to the claim that true beliefs are adaptive, Plantinga (, 8–9) has discussed the possibility of evolutionarily adaptive or fitness‐enhancing but still false beliefs. For Plantinga, beliefs can be false but still useful, for example, a person could believe that hazardous phenomena are caused by witchcraft but their avoidance would still be beneficial regardless of the false belief. Critics have argued that Plantinga relies on contrived examples, and has failed to show how false beliefs can be systematically adaptive (Fales ; Boudry and Vlerick ). In addition, Plantinga (, 10) admits that it would be useless to calculate exact probabilities for truth values but simultaneously claims that vague estimates would be sufficient for the EAAN. He asserts that it is hard “given materialism, to envisage a way in which the content of a belief could get causally involved in behavior” and reinforces his claim that the evolutionarily produced physiological faculties producing beliefs would not be reliable.

New Arguments against EAAN

Although the naturalistic or philosophical criticism of Plantinga's claims has produced a large amount of philosophical debate, the issue of the theistic alternative is scarcely scrutinized. Fitelson and Sober () have noted that a theist is faced with a similar skeptical predicament and cannot just presuppose “the fact that the human mind was designed by a God who is no deceiver.” This would lead to the same type of global skepticism that Plantinga accuses naturalism of. While Fitelson and Sober () do not examine this possibility further, we shall concentrate on this issue in detail, starting with Plantinga's (, 3) presumptions about the qualities of the God in his argument. He assumes that his God is rational, omniscient, and omnipotent, and that the Bible is an infallible source of true information: “a proper source of knowledge and justified belief” (Plantinga , 26; 1991; 2008, 3–4). Accordingly, he holds Scripture in high esteem, stating that “no doubt science can correct our grasp of Scripture; but Scripture can also correct current science” (Plantinga ).

We aim to analyze Christian theism in the form associated with Plantinga's EAAN (holding the Christian belief system as a reliable source of knowledge) as well as a more general tenet of theism (supernatural intervention) to see if these worldviews would produce warranted beliefs. It is one thing to show that Plantinga's argument fails to undermine naturalism. It would be even more interesting to show that it backfires against his own worldview. We decided to investigate the Bible as such and to see if the divine attributes described therein provide any epistemic warrant for our beliefs. God's characteristics, as presented in Biblical texts, are relevant for Plantinga's argument, as he accepts that the Bible is the word of God (see also Plantinga ). The specific aim of this essay is to assess Plantinga's argument as follows: If we accept his claim that evolutionary naturalism is self‐defeating, does the theistic worldview fare any better? Does (Christian) theism offer us a way out of global skepticism or does it fall prey to the same problems? We assessed this by (1) examining the Biblical texts for instances that would jeopardize the warrants for proper beliefs and by (2) discussing the theistic/religious worldview regarding supernatural events and miracles and their effects on the possibility of producing reproducible scientific patterns or theories.

Potential Counterarguments

The assertion that theism produces reliable, “true beliefs” could, perhaps, be salvaged if some issues were resolved. (1) If we could spot deceit or predict situations where it occurs, maybe we could dismiss it as an insignificant possibility. Furthermore, (2) if the Biblical verses suggestive of divine deceit would not affect theistic scholars, such as Plantinga, or (3) if the instances where deceit is mentioned could be explained away by Biblical scholars in a rational manner, then the EAAN would potentially have more support.

While some verses mentioned above indicate that people could be victims of divine lies, it could be argued that deceit is only aimed at sinners/unbelievers/enemies of God (Clarke , 330; Greenberg , 368–369; Gorday , 113) and that a sufficient level of Christian–theistic faith would no longer make one a target of potential deceit. This, however, easily leads into the “no true Scotsman” fallacy, which Bradley Dowden () defines as an “ad hoc rescue of one's generalization in which the reasoner re‐characterizes the situation solely in order to escape refutation of the generalization.” For example, a member of a religion/belief system who has committed a violent act is easily dismissed as not being “a true follower” of that religion. In Plantinga's case, the definitions of “adequate” faith would be extremely difficult to construct: those who have enough faith are not deceived, and the ones who are deceived must not have had enough faith. Even if the scholar regarded him/herself as theistic and obedient to his/her beliefs, it would be very hard to assess objectively that he/she is good enough a Christian not to be affected by delusions. Is not everyone burdened with original sin, as Christian doctrine stipulates? Is it not presumptuous to think that one is among the few “pure” believers, untainted by sin and free from divine deception? The Christian worldview warns against such hubris and human vanity (e.g., Philippians 2:3–4). Problems also arise when trying to construct a method for recognizing deception. If no deception is apparent, does that mean that God has spared us? In fact, the impression of not being deceived may itself be a part of God's deceptive plans. The most thorough form of deception is one in which the victim is blissfully unaware of being deceived. Of course, it is hard to demonstrate that it is impossible for a theist and a proponent of the EAAN to escape from this skeptical predicament. In any case, we invite them to formulate reliable methods to determine who could fall victim to divine deception.

We acknowledge that the Biblical texts can and do offer many alternative interpretations that are sometimes mutually incompatible. For instance, the first chapter of the Bible (Genesis 1) has produced both the doctrine of young‐earth creationism, with a literal creation within six days approximately 6,000 years ago, old‐earth creationism that interprets Biblical “days” as long periods of time, and various metaphorical interpretations for the proponents of intelligent design and theistic evolution (reviewed in Nieminen and Mustonen ; Nieminen, Mustonen, and Ryökäs ; Nieminen, Ryökäs, and Mustonen ). The texts selected above concentrate on the possibility of the Biblical deity having qualities that would not be compatible with the development of true beliefs. Interpreting these passages as metaphorical is certainly a possibility, but this would not save Plantinga's () argument. Instead, it would make the theistic approach perhaps even more difficult, because accepting different interpretations as equally valid would give us another defeater for reliable beliefs. It would lead us into a predicament in which the possibility of arriving at reliable conclusions about sacred texts would be low, which means that there would be no reasons to favor belief in a trustworthy deity over a deceptive deity. This is the same skeptical situation that Plantinga attributes to naturalism cum evolution.

Dismissing or disregarding the Bible verses containing deception will not do. While they may not be considered the most significant ones regarding Christian faith, there are quite a few Biblical scholars who accept them as involving actual deception. This is exemplified by the Isaiah verses, where deceivers are defined as spirits sent by Yahweh causing delusion/confusion (Wildberger , 256; 2002, 403–404), and the Ezekiel examples referring to “the shocking idea that God misleads those who anger him into sin” (Greenberg , 369). Ralph W. Klein (, 265) suggests that in the 2 Chronicles verse “Yahweh himself has sent out that lying spirit,” and a commentary by James W. Clarke (, 330) on 2 Thessalonians also mentions deliberate delusion caused by God: “there has come upon them a strong delusion, to make them believe what is false. This delusion has come from God.” We acknowledge that there might be several interpretations among Biblical scholars regarding these verses. However, simply dismissing them or interpreting them as metaphorical would also undermine Biblical theism, as this would eventually lead to skepticism about the truth value of the Bible. Either the human scribe would have received erroneous information from God, he would have misunderstood it, or the message would have been distorted over time. Thus, warranted belief based on the Bible fails in any case and, if we apply Plantinga's model to theism, would eventually lead to Biblical theism being self‐defeating, once again similar to the situation described above. If Biblical scholars of various denominations cannot agree on the possibility of divine deceit, this alone makes the theistic alternative to naturalism suspicious and unreliable, instead of providing the warrant that Plantinga (, 11–12) is looking for. Furthermore, it is clear that Plantinga is expecting to find evidence in contrast to mere belief (Plantinga , 4), making religious belief something that requires extrinsic proof. This sets Plantinga apart from those theologians who wish to keep religious faith and reason/evidence in their respective and mutually exclusive domains (Dawes ). In this essay, we follow the example of Gregory W. Dawes (), as we are critiquing the purported warrant of (Christian) theism, not religious faith itself.

But suppose that the Bible would not contain any verses suggestive of divine deception. Suppose that all the verses we mentioned were absent or that theologians could conveniently interpret them to apply only to the category of unbelievers or heretics who have angered God. Even that scenario would be of little comfort to the theist. If God really was prone to deception, why would he inform us about this in his revelation? Indeed, to do so seems to undercut the effectiveness of deception. If God really wanted to deceive us, he would remain silent about this in his revelation, pretending that he is wholly benevolent and wants us to know nothing but the truth. In other words, the fact that an omnipotent being presents himself as trustworthy and reliable in some revelation—supposing this being was really its author—does not give us ample reason to conclude that he is. The theist may object that the theist God is by definition benevolent, and that large‐scale deception would not be compatible with this image of God. But to that we can retort along the lines of the classical theological response in the face of evil and suffering in the world: God may have reasons that we cannot fathom. For all we know, living in an illusion produced by divine delusion could really be for our own benefit. Perhaps, God knows that we cannot handle the truth, being only finite beings, and has therefore decided to subject us to a benign form of paternalistic deception? In short, once you posit the existence of invisible and omnipotent beings, which by definition have unlimited powers of deception, all bets are off.

Miracles, Rationality and Causality: General Theistic Arguments against EAAN

Another issue that is central to the theistic or religious worldview is the supernatural involvement of a deity, Biblical God or other, in history and in people's lives, or a “God who can interfere at will in his creation and violate the natural order” (Boudry ). There are alleged supernatural apparitions, communication of texts, and miracles conjured up by a deity or transmitted through prophets. These phenomena are in contradiction to a rational view of the world governed by causal laws. It would be difficult to predict how natural phenomena proceed, for instance, looking at some Biblical examples, when the dead do not remain dead (Luke 7:13–15), a sea is parted (Exodus 14:21) or a rock brings forth water (Exodus 17:6). In addition, devout believers put forward assertions of miraculous experiences, for example, supernatural healing and protection. These occurrences violate physical laws and break the usual course of events determined by natural sciences and empirical evidence. This undermines the notion of a deity that prizes regularity and constancy (Plantinga , 4) in His creation.

The possibility of supernatural intervention makes it impossible to rely on the laws of nature, as they can be broken or suspended by a deity at any given time. A similar notion has been expressed previously by Maarten Boudry (): “What reasons do we have… to assume that God will not act capriciously, that he will continue to sustain the natural order?” If we look at Plantinga's (, 11–12) conditions for warranted beliefs, this most effectively disproves clause II about a proper cognitive environment. Miracles that occur in an unpredictable fashion create an environment that is uncertain and beyond the rational rules of cause and effect. In addition, clause III is weakened, because supernatural intervention makes it difficult to form consistent, true beliefs. If scientific research were based on a worldview that includes the possibility of miracles, the laws described by science would not hold universally, as miracles could always overthrow the natural, lawful outcome. We would not be able to know

• if phenomena were natural based on the laws of physics or if they were supernatural deviations from the natural chain of causality;

• if a pattern of cause and effect can be deciphered, where similar conditions would sometimes lead to the expected outcome (e.g., a fatal disease not cured by prayer) and sometimes to a situation that defies natural laws due to supernatural intervention.

Potential counterarguments are quite similar to the ones introduced above in the context of supernatural deceit. If it was possible (1) to identify miracles, at least a significant proportion of the time, and (2) to predict situations where supernatural intervention would be expected, miracles could be assessed rationally and in a statistical manner in scientific studies.

However, there is a controversy even about the concept of a miracle within theistic and religious communities. Some authors define them as violations of the laws of nature, while others claim that miracles suspend the laws of nature; and still others feel that even natural phenomena (rainfall, and so on) can sometimes be regarded as results of supernatural causation (Ward ; Driscoll ). The actual presence of miracles in the theistic and religious worldviews is hardly questioned. John T. Driscoll takes them to be “so interwoven with our religion, so connected with its origin, its promulgation, its progress and whole history, that it is impossible to separate them from it” (, 346). To make these alleged phenomena subjects of scientific study, they would need to have reliably testable effects. Usually, however, the identification of miracles is based on personal accounts and testimonials of eyewitnesses. This inevitably leads to a multitude of interpretations that have not been assessed in any systematic manner that would provide a tool to recognize supernatural intervention reliably. In addition, events can be taken to be supernatural although they are subtle and practically indistinguishable from the natural course of events (Boudry and De Smedt ). Thus, complex stochastic and chaotic processes (weather, success in sports), poorly understood phenomena (being cured from a disease), and alleged conjunctions of natural and supernatural causes (yielding additional strength for a duel) can make the presence of supernatural causality impossible to discern consistently. Indeed, the notion that God “moves in mysterious ways” can be seen as a form of divine deception again, with a deity disguising his miracles behind a screen of natural causes and effects. The deity, especially the Christian God, does not offer himself up for observation but instead moves about in subtle and indiscernible ways. This makes it impossible for us to determine with any confidence whether or not a miracle has taken place. For all we know, deities could be working miracles all the time, without us being aware of it. Therefore, this potential counterargument fails.

Or is it possible to imagine a situation where miracles take place in a predictable manner? For example, a certain disease or accident could be reversed by supernatural intervention if the patient/victim was innocent from the Christian point of view. This is quite similar to the possibility of finding out who would and would not fall victim to deception, as discussed above. Without any objective criterion for assessing Christian “innocence,” however, this could lead to the “no true Scotsman” fallacy (Dowden ). Again, we invite the proponents of the EAAN to elaborate on their claim by assessing the difficulties that supernatural intervention poses for their assertion. For example, being a “pure” or a “good” believer is no warrant against being the target of supernatural harassment. Especially regarding Christian theism, the story of Job in the Old Testament describes a totally pious man experiencing supernaturally induced sickness and loss of family and fortune because of a capricious bet between God and Satan. In this case, even Biblically proven worthiness does not save Job from being the target of supernaturally induced tribulations admittedly inflicted by Satan but simultaneously allowed by the omnipotent God.