Christianity
16
Christianity
34
Islam
18
Hinduism
13
Buddhism
6
Other religions
13
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 515
We have some quantitative data about the situation in the Western world. For example, the “BBC’s ‘Soul of Britain’ survey in 2000 found that 76 per cent of the population . . . had some kind of spiritual experience” (Hick 2006, p. 17). The category of spiritual experience is considerably broader but other statistics show that the proportion of people having RE
(largely theistic) is not low:
In the United States a 1975 National Opinion Research Center inquiry in which people were asked ‘Have you ever felt as though you were close to a spiritual force that seemed to lift you out of yourself?’ found that 35 per cent of those asked said that they had, and a Princeton Research Center . . . survey in 1978 also recorded 35 per cent. In Britain at the same time a National Opinion Poll of a sample of 2000 reported 36 per cent. (Hick 2006, p. 35) The real proportion may be higher because of the taboo factor I have mentioned.
The said results were obtained with the more impersonal method of a poll. Hay and Morisy fi nd out that when they interview people and try to build up mutual trust and let them take time to recall, the positive response rate rises dramatically to 62–67 percent (Hay 1994, p. 11).
Quantitative data about TE in other countries are not available. But Christianity is nowadays truly a world religion, and there are a large number of Christians in every continent. For example, there are at least, say, 30–40 million Chinese Christians. I know from my own experience (treated as a kind of informal sampling) that a signifi cant proportion of them have TEs, some of them quite a lot and quite dramatic (see Cases 1 and 2). I cannot offer any exact fi gure, but no one can deny that many Christians living in very diverse places in the contemporary world, perhaps running up to hundreds of million, have had at least one TE. If we factor in the experiences of Muslims and people in other theistic traditions, and TEs even in cultures not particularly theistic (e.g. the Chinese culture), the fi gure will be even more impressive.7 In short, TE indeed has a broad base across cultures, eras, and even religious traditions.
Some critics charge that TE is a kind of odd experience (Vardy 1990, p. 103; Draper 1992, p. 159), and hence it should be subjected to initial skepticism. However, it seems wrongheaded to label a kind of experience shared by at least tens of million of people odd.
Again Hay has got it right:
post-Enlightenment, secular models of reality have come to dominate contemporary understanding,” and this leads to “a failure on the part of many scientists to attend seriously to the phenomenology of religious experience. This has produced a distorted understanding and dismissal of what appears to be a widespread and normal fi eld of human experience. (1994, p. 1; italics added)
7. For some people in a more specifi c theistic tradition, for example, Christianity, they may want to dismiss the theistic experiences in other traditions in toto. The CTA I adopt does not advise us to do that, and I do not think a Christian theist need to do that. He can appeal to the concept of general revelation or common grace in the Christian tradition. I submit indeed it is plausible to think that God can be experienced in different cultures (e.
g. the Chinese culture) at least as the Creator, Supreme Being in the universe, or just a very powerful divine being.
Moreover, we are only talking about the PFJ here. That we should not dismiss the TEs in other traditions or cultures from the very beginning does not entail that we will accept all of them as veridical in the end. Especially in this chapter, I mainly want to defend a less specifi c subtype of TE, such as the sense of presence of God, and the awareness of the transcendent being implicit in different forms of TE.
Typology of TE
Numinous experience
Otto’s description of the numinous is still a good one: mysterium tremendum et fascinans.
Yandell spells it out in this way: the subject “seems to experience an awesome . . . majestic and overpowering being and the subject . . . has a feeling of being a creature in the presence of the creator . . . the subject seems to experience a being who is unique in kind and intensely alive” (1984, pp. 9–10).
Theistic mysticism
Whereas in a numinous experience God is felt to be other than the subject, in theistic mysticism, God is experienced in a union of the soul with God. God is searched from within in this type of love mysticism (e.g. Teresa of Avila) (Pike).
Sense of presence of God
It is the more commonplace intuitive awareness of God. This awareness is often fused with a feeling of calm assurance and peace (Baillie, Beardsworth). Whereas extraordinary types of TE usually last for a relatively short duration, it is possible for some to maintain the sense of presence of God for a relatively long period (Brother Lawrence).
Divine–human encounter
In this type of TE, God is experienced as the Eternal Thou. Farmer has laid his fi ngers on the crucial features: “First, central in the living awareness of God as personal is something which happens . . . in the sphere of the will. The religious man is aware of a certain peculiar type of resistance . . . the resistance, namely, of absolute, sacred, unconditional values” (1935, p. 23). Second, there is also the awareness of God as the “fi nal succour. . . . The unconditional demands, the values of God, are apprehended as pointing the way to the highest self-realisation, the fi nal security of man” (1935, p. 25). These experiences of encounter are fused with characteristic emotions which accompany personal interaction (e.g. warmth, gratitude, joy of communication, sense of being personally addressed).
Experience of grace
Awareness of God’s unconditional demands may also occasion awareness of one’s moral failings as “sinning against God” and awareness of one’s moral impotence as “spiritual death.” This may also open up the possibility of experiencing divine forgiveness of sin and renewal of moral personality through reconciliation with God. Characteristic emotions are sense of guilt, peace of forgiveness, and joy of liberation.
Experience of personal growth in God
We should also note the personalizing character of some TEs (Cases 6 and 7). A person who experiences God may also experiences personal growth at the same time: heightened
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 517
sensitivity to people and moral values, a revitalized conscience, a greater concern for people and willingness to sacrifi ce, and so on. In a word, he fi nds himself becoming more like an (integrated) person and his telos fulfi lled in his life in God.
Baptism in the Spirit
It is “an overwhelming sense of being set free from sinful selfi shness . . . deep interior peace. . . . the gift of tongues” (Dorr 1978, p. 40). There are a lot of these nowadays, since the charismatics constitute one of the fastest growing groups in contemporary world.
Conversion
The content of a conversion experience actually overlaps with the aforementioned types of TEs and it is only marked out by the dramatic and sudden change that is caused by the experience. There may be a drastic change in personality, beliefs, and direction of life.
Corporate TE
Many TEs occur in the context of the “fl ight of the alone to the Alone.” However, in Christianity, the corporate nature of these experiences is always emphasized:
after Pentecost, the Spirit descends on the disciples. . . . then it is said that the disciples were all “of one heart and soul”. . . . So the descent of the Spirit forms the community and the community is such that in it everything is shared, even at the economic level. (Griffi ths 1989, p. 223)
So experience of God and experience of community are often inseparable, and I would call this kind of TE a corporate TE. To different degrees, I think a religious community in corporate worship and prayer (sometimes) is having similar experiences.
Mediated TE
Sometimes a person can have a TE that is mediated by other kinds of experience, for example, experiences of nature, art, conscience, fellowship with others, and saintliness of others. Sometimes the subject has focal awareness of God through subsidiary awareness of something else, for example, awareness of God through awareness of the beauty of nature.
At other times the awareness of God may just be present in the background while the person is focally aware of some other thing. We then have an oblique perception of God instead.
Sensory TE
A sensory TE is an experience of God that is mediated by sensory experiences (e.g. visions, auditions, dreams, stigmata).
Interpretive TE
It is a spontaneous interpretation of an event as God’s action or message but the event can be clearly described without using religious concepts. Experiences of God through answered prayers, guidance, miracles, healing, tongues, and so on are examples of interpretive TE.
Intuitive apprehension of God
All the aforementioned types are more or less mediated by some feelings, but some mystics also report a nonsensory awareness or intellectual vision of God in which all phenomenal content is absent. I call this an intuitive apprehension.
TEs within a common ontology
All the given kinds of TE are actually experienced by believers. It is foolish to elevate a certain kind of TE to the exclusion of other kinds. God’s relations to the world and to man are so multifaceted that it is unreasonable to think there is only one way to approach Him.
Consider an experience of drinking a cup of coffee, which is actually consisted of visual, olfactory, tactile, and gustatory experiences of the coffee. Considered in themselves, the smell of coffee and the visual sensations of a brown liquid are as diverse as any two things can be. Yet they are connected as modalities of the same SE of the same object. Similarly, God can be approached in many directions and manifested in diverse ways; yet the whole lot can be coherently explained by the categoreal nature of the same God:
(a) Since God is the transcendent and holy creator, numinous experience and experience of contingency are to be expected.
(b) If man is created in the image of God and their selves are ultimately grounded in God’s sustaining activity, then God can also be approached from within (theistic mysticism).
(c) Since God is Himself the Absolute Good, it is no surprise to have experiences of His unconditional imperative as well as love and succor, and also mediated TEs through conscience and morality.
(d) Since God is personal, we can have experiences of personal encounter, divine speech, emotional healing, and so on.
(e) God’s wisdom and power over nature makes it possible for Him to reveal His purposes through providential and miraculous acts (mediated and interpretive TE).
(f) God as the source of personality and community nicely explains the personalizing character of TEs and the occurrence of corporate TEs.
(g) God as Redeemer is correlated with our experiences of grace and conversion.
(h) God can act directly on the mind, and hence intuitive apprehension is possible.
So the experience of God is marked by a rich diversity in unity. Arguably, this pattern is already there in the Biblical portrayal of the divine interaction with man. For example, the Old Testament scholar Rowley says,
His voice is still heard in Nature and in history, in individual experience and in the personality of men and women who are attuned to his spirit . . . Of greater signifi cance than any of these separately, however, is that combination of factors dovetailing into one another. (1956, p. 47)
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 519
In the New Testament, Paul talks about a God who is above us, inside us, and among us (Ephesians). These descriptions can be nicely correlated with numinous experience, mystical or oblique experience of God, and corporate TE. Paul’s own experience of God refl ects the same rich diversity in unity: a conversion experience, experience of grace and self-transformation, experiences of providence and miracles, and probably some mystical experience (see 2 Corinthians 5).
The same intricate pattern of experience of God is also reported by persons in contemporary secular society. Consider Jackie Pullinger’s experience of God. She experienced God’s calling to serve Him and, fi nally, she settled down inside the Walled City of Hong Kong, trying to preach the gospel to the drug addicts, Triad gangsters, and the prostitutes.
Now it is still an expanding ministry. Her story consists of many “inner” and “private”
experiences of God, and experiences of guidance, providence, comfort, and illumination that result in practical actions of social concern. She herself undergoes self-transformation while many drug addicts experience conversion. There are also many experiences of tongues, healing, and miracles. Above all, all these experiences fi t nicely together in her lifelong experience of God that manifests beauty and an overarching purpose.
It is really important to keep sight of the whole of this evidential base. All these experiences are of God and the various types are to be expected once the nature of God is given.
So they are not like the links in a chain which is as weak as the weakest link. Instead, they are like the strands which are woven into a cable. The full evidential force of TE can only be seen when the rich diversity and coherence of TEs are simultaneously recognized.
Conceptual Coherence of TE
To be a well-established type of experience, the concept of TE needs to be conceptually coherent. Assuming the coherence of the concept of God, TE can still be incoherent because experience of God can be shown to be impossible. Several reasons have been adduced for this claim: Jonathan Harrison rejected TE “partly on the grounds that one cannot have immediate experience of dispositional characteristics, partly on the grounds that we can have immediate experience only of the contents of our own mind” (1999, pp. 257–8).
Forgie takes “theistic experience” to mean “an experience which is phenomenologically of God, i.e. an experience which, if veridical, would have to constitute an accurate perception of God and nothing else instead” (Forgie 1998, p. 317; italics added). But no experience can be phenomenologically of God. So TE is impossible. Some other critics put the emphasis on the alleged impossibility to recognize the infi nite attributes of God in experience (Davies).8
Initially, it sounds plausible to think it rather problematic to say the fi nite can experience the infi nite, or the transcendent can be wholly given in the immanent. Simply put, God is just too big for humans to experience. However, puzzles immediately arise when we realize similar questions can be raised about the possibility of seeing a physical object. We think we can see the Yellow River or the Great Wall of China but is it not equally obvious that the Yellow River or the Great Wall are also too big for humans to see? I can see a physical object, and a physical object is essentially a three-dimensional object, having a front and a 8. Everitt also thinks that God “is not even a possible object of sensory experience” (2004, p. 172). I think his objection basically hinges on a kind of favoritism toward SE, which is widely criticized in this chapter.
back, which cannot be given in our visual experience at the same time. So can I really see the object? Moreover, suppose Kripke is right that water is essentially H2O. Then we need to ask: when we see water, how can the property of being H2O be given in our experience?
The basic problem about these kinds of questions about experience of God or physical objects is that they have presupposed this requirement:
For S to have an experience of an object O having an attribute F, either the whole being of O (or at least O’s essential properties) has to be given in the experience, or that an infallible indicator of the state of affairs (O is F) is present (and perceived as such) in the experience.
Similar requirements such as that mentioned are in fact made by the traditional foundationalists. If experience is our only access to the external world, once we allow a gap between the experience and the world, how then can we guarantee the possibility of empirical knowledge? However, the lessons of epistemology in recent decades tell us that these requirements simply cannot be met even in our SE.
Consider our visual experience of water. What are given in the experience are two kinds of content: the propositional content of our epistemic seeming (“there is water over there”), and the nonconceptual content, which can be described in various ways: being appeared to waterly, this and that kind of qualia or sense data. The direct realists also contend that, when the experience is veridical, there is the direct awareness of water in virtue of water appearing to us. Water as a physical object cannot be, in any literal sense, be given in our experience, which is a mental event. Indeed this is a good ground for raising the question whether any experience of a physical object is possible. This worry is further aggravated by the realization that the concept of a physical object is the concept of something which can exist independent of all our experiences – how can this property be given in our experience? This ability to exist independently is also a dispositional property. How can we perceive that? What features of our experience tell us that physical objects can do that? More experiences only tell us those objects can continue to be perceived and experienced, and it does not entail that they can exist independently of our stream of experience.
Contemporary understanding of physical objects further complicates matters. For example, Russell thinks that “if physics is true there must be so little resemblance between our percepts and their external causes that it is diffi cult to see how, from percepts, we can acquire a knowledge of external objects” (1948, p. 213). Since the “table as a physical object, consisting of electrons, positrons, and neutrons, lies outside my experience” (1948, p. 236),
“I do not ‘see’ the furniture in my room except in a Pickwickian sense. When I am said to
‘see’ a table, what really happens is that I have a complex sensation which is, in certain respects, similar in structure to the physical table. The physical table, consisting of electrons, positrons, and neutrons, is inferred” (1948, pp. 237–8).
The problem then arises: if the physical table is inferred, then how do we justify this inference? The sensations certainly are not infallible indicators of the presence of the table.
Even the alleged direct awareness of the table can be faked. As Alston admits “the intrinsic indistinguishability of hallucinations and the real thing,” he has to concede that “even if hallucinations do not prove that perceptual experience is never a direct awareness of external objects, they certainly prove that it isn’t always that” (1999, p. 238). Of course, the propositional content of our experience can also be mistaken.
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 521
At this point, I think it is pertinent to ask what kind of approach we should adopt in relation to these questions. I suggest we cannot decide what capacities for experience and recognition are possible in an a priori manner. Instead it is more fruitful to look and see what are the actual experiences people claim to have and what sort of things they claim to recognize. If a concept of experience renders most of our actual experiences impossible, then what needs to be changed perhaps is the concept itself. In this spirit, I would suggest there is no need to insist on too strong a connection between experience and its object. As long as the appearance of a transparent colorless liquid reliably indicates the presence of water (H2O) in our context, and the water causes in an appropriate way our visual experience, then we can see water (H2O). Our experiences are fallible, and its reliability cannot be guaranteed. Suppose, unknowingly, we have been transported to Putnam’s Twin Earth. The appearance of water may then be misleading but this does not mean that here on Earth we cannot see water. The said analysis is used to explicate the concept of “experience of something.” I do not suppose it can be proved apart from our actual experiences.
It follows that when we have an experience of O, we do not need to require that O’s essential properties be given in the experience. Schoen argues that encountering maples does not require any knowledge, perception, or even detection of the essential nature of maples, which:
presumably . . . are tied somehow to matters of genetics . . . If perceived or detected objects are to be recognized as maples, reliable indicators must be processed more specifi cally. Such indicators . . . need bear no special relation to essences . . . Furthermore, it is not necessary that everyone use the same indicators. (1990, p. 9)
In this way, experiences of the Yellow River, the Great Wall, or a three-dimensional object are no longer puzzling.9
We can now come back to the question about the possibility of the experience of God.
Forgie’s concept of TE is certainly too strong. As he admits, his concept will render impossible SEs which are phenomenologically of individuals, or many ordinary objects such as a coffee maker. He also admits his concept of the phenomenological content of experience is not the only possible one, and that “on . . . an alternative account theistic experience would thus be possible” (Forgie 1998, p. 323). I have offered an alternative account earlier which is also similar to Layman’s account: “one can have a reliable experience of God without experiencing all of God’s features (though of course the experience must include reliable indicators of a divine presence, such as infi nite power and goodness)” (Layman 2007, p. 54).
Similarly, Harrison’s objection depends on a very narrow understanding of experience.
The concept of a physical object entails a lot of dispositional characteristics. Take the experience of color as example:
In reporting that I saw a red fl ower, I am normally reporting that I had a visual experience of a physical object having certain powers or dispositions, i.e. that under normal lighting, normal perceivers (in general) would also see something red if they looked at the fl ower . . . And of course, being very powerful . . . is a power or capacity. So, at least in principle, a direct 9. Even a critic of theism, Everitt, agrees here (2004, p. 173).
awareness of an entity as very powerful-or even almighty-seems to be possible. (Layman 2007, p. 54)
We can also use this model to understand how the content of a TE can convey the presence of God: “God might have created us in such a way that when He appears to us what appears to us phenomenologically is secondary qualities of God which emerge when God is so experienced” (Gellman 1994, p. 55). These secondary qualities may consist of some unique phenomenal properties of the appearance of God such as the “taste” of God that some mystics have reported. Our feelings and emotions can also be used as indicators of divine presence:
our own subjective, emotional reactions fi gure into our recognition of certain types of features.
Sometimes upon meeting a person we regard him as less than trustworthy primarily because he invokes in us a feeling of uneasiness. In this type of case, our subjective, emotional reactions serve (in combination with presentations made through sense experience) as indicators of certain types of qualities, and these indicators are very important in dealing with social realities . . . in the theistic mystical experiences . . . subjects were presented with personal presence, and they had an immediate impression of its infi nite power and goodness. Their subjective, emotional reactions were of profound awe, of feeling very “small” or fi nite, of being loved, of extraordinary peace, and so on. All in all, I think that’s what one might expect an experience of God to be like. (Layman 2007, p. 54)
For example, consider the experience of God’s infi nite love. According to Harrison, it is both a disposition as well as a mental state of other minds, and hence cannot be presented in immediate experience. However, making use of the given model, there can be an awareness of the love of God mediated by feelings of feelings of warmth and bliss – feelings which are analogous with what we feel in our interpersonal experience.10
Pace Harrison, when we attend to our concrete experiences, there is reason to believe we can feel the love, joy, sadness, or other mental states of other persons.11 Alternatively, we can be told by God about His love. Philosophers sometimes ignore the communication model because they are obsessed with the visual model. However, the former model is especially important in interpersonal experience: a lot of experiences of other minds are mediated by verbal and informal communication (e.g. symbolic acts) – say, God can give us a vision of the Cross of Jesus and impress on our mind or tells us by a “voice”
the meaning of this event. Something like this has in fact happened to many believers (Case 2).
Since it is an empirical fact that an awful lot of people seem to have experiences of God, in the absence of a good reason to think otherwise, one should conclude that TEs are indeed possible.
10. I think epistemologists are still bound by a narrow understanding of human knowing, and underestimate the cognitive role of emotions in human life. Wynn’s recent book (2005) defends the thesis that “emotional feelings can function as modes of value perception- in relation to God, the world, and individual human beings . . . they can also function as ‘paradigms’, and can therefore properly direct the development of our discursive understanding, in religious and other contexts . . .” (2005, p. xi). He explicitly applies this understanding to TE, and argues that even a TE’s phenomenal content is purely affective, this will not necessarily be epistemically dubious (2005, pp. 8ff).
11. Hallett has a good defense of this possibility, and use this as analogy for TE.
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 523
Intracoherence of TE
Of course, mere possibility is not suffi cient. We need to judge whether TEs deserve PFJ by the criteria of intracoherence formulated earlier. As for clauses (a), (b), and (c) in that criteria, we have noted that many people in almost every age, culture, and in many religions seem to have at least some plainer TE such as the sense of presence of God. Even nowadays, in some countries, the majority has some TEs. In other countries, TEs are at least shared and communicated to one another within theistic communities, which are signifi cant minorities in the society. So conditions (a) and (c) are certainly satisfi ed.
For many persons who have had TEs, they may have just a few of them. However, we have also mentioned the dramatic experiences of God, which quite a number of individuals (e.g. Christiana Tsai, Jackie Pullinger) have throughout their lives, and many more somewhat in the middle of these two groups. So condition (b) is reasonably satisfi ed for many, and eminently satisfi ed for a smaller number. I have also argued for the explanatory coherence between the TEs, and the conceptual coherence of TE. If my arguments are correct, conditions (d) and (e) are also satisfi ed. So barring the demonstration of conceptual incoherence of theism itself, we can conclude that TE is a relatively well-established type of experience.
The critics may argue that our criterion of intracoherence of a type of experience has too low a requirement. For example, TEs still fall short of being universal and are much less frequent to many people than other established types of experience such as SE or memory. However, the critics’ requirements are too stringent here. First, universality among human beings is certainly not a necessary condition for PFJ. Obviously, a kind of experience shared just by a minority of people can nevertheless still be veridical. Actually, not all people can see or hear. Conceivably, a nuclear war can happen which causes all but one to be blind. Furthermore, many are tone-deaf. It seems to be quite clear that even if much fewer people share the SEs we have, they are still at least prima facie justifi ed for us.
If we count all the people in the whole history, people having TE may outnumber those who have none. Moreover, in view of the fact that TEs occur to many diverse kinds of person, it is not implausible to say that the capacity for TE is at least potentially universal.
If it is true, the contingent fact that TEs are not actually universal is not that damaging.
The near universality of our current SEs is also contingent. Suppose there are aliens in many different places in the universe, and they all possess very distinct types of sense.
Humankind is only a particular community of perceivers. So our SE may not be really universal in the universe. Would this hypothetical fact show that our SE cannot even be regarded as having PFJ?
Furthermore, in most of our cognitive situations, when we have some initial presumption for the reliability of the experients, a suffi cient number short of the entire community are adequate to convince us their reports are at least worthy of initial support. Take science as example, the majority of people in fact do not know how to give reasons for the existence of black holes or quarks. However, as long as the science community agrees that they have good reasons to believe those exotic entities, I am reasonable to trust them even though the scientists involved are just the minority among human beings. It is also the case for testimony to events one has not witnessed. Suppose one person tells me that he has seen a rare bird species in Hong Kong. It has some weight for me but I may still doubt it (even not having strong reasons to doubt his integrity or reliability as a witness). However, when 30
people tell the same story to me, I think my initial worry should be suffi ciently dispelled.
Suppose I later learn of 300 people who also give the same testimony. The degree of PFJ of that testimony will be further enhanced, but it does not seem to be essential for its PFJ.
I suggest the critics often commit the superreliability fallacy, which strictly requires a BSJ to reach the level of superreliability before it can be regarded as having PFJ. Gellman rightly points out that we should not tie our notion of confi rming evidence to that associated with sensory perception. It is because:
Our ordinary physical-object beliefs are way overjustifi ed by confi rming evidence. We have extremely luxurious constellations of confi rming networks there. Hence it does not follow that were mystical claims justifi ed to a lesser degree than that, or not by similar procedure, that they would be un justifi ed. (Gellman 2001, p. 27)
The critics are right in pointing out the ways in which TE differs from sensory experience, and it indeed follows that TE has intracoherence to a lesser extent than sensory experiences. However, the hasty inference that TE is, therefore, not well established is a non sequitur, and is like arguing that, since a scientist is less brilliant than Einstein, he must be an incompetent scientist. The superreliability requirement will not only endanger the PFJ
of TEs but also that of moral experience, aesthetic experience, and so on. This consequence is also implausible.
We are not yet arguing for the veridicality of TEs: we are only asking that they deserve similar kinds of initial trust we give to many kinds of experience, and not to write them off. Is it reasonable to believe that all “God-experients” are either deceiving themselves or others? Gutting, for one, does not think so:
religion, throughout human history, has been an integral part of human life, attracting at all times the enthusiastic adherence of large numbers of good and intelligent people. To say that something that has such deep roots and that has been sustained for so long in such diverse contexts is nothing but credulity and hypocrisy is . . . extraordinary. (Gutting 1982, pp. 2–3) So TE is not extraordinary. It is part and parcel of the human situation. Instead it is the overly skeptical attitude toward TE which is extraordinary. Suppose we come to know the life story of a person who has dramatic experiences of God throughout his life. We fi nd that person honest, sane, and intelligent. We also fi nd his story corroborated by many others’
stories throughout history in many cultures. Is it not rash to say that all of them are entirely and chronically deluded? And for that person, he or she (e.g. Christiania Tsai, Jackie Pullinger) is not even entitled to trust his experiences at least initially? This does not seem to be reasonable.
The Structure of the CTA
The two basic components of CTA are the critical spirit and the attitude of trust. The CTA thinks that these two components are both essential elements in any viable epistemology, and they need to be kept in a kind of balance and healthy tension. It is similar to Thomas Kuhn’s emphasis on a kind of essential tension between tradition and innovation in science (this phrase is the name of one of Kuhn’s books).
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 525
Data gleaning – trust
The data for the CTA are all our experiences. The original incorporated epistemic seeming of a given experience is taken as the epistemological starting point and accorded PFJ. These presumptive data are defeasible. The basic principles are the weak Token PCT and moderate Type PCT discussed earlier.
Bonjour has provided a helpful classifi cation of different kinds of foundationalism.
While strong foundationalism requires basic beliefs to be infallible, moderate foundationalism allows basic beliefs to be fallible but insists that they are “suffi cient by itself to satisfy the adequate-justifi cation condition for knowledge” (if not defeated) (Bonjour 1985, p. 26). In contrast, weak foundationalism thinks that the “basic beliefs possess only a very low degree of epistemic justifi cation on their own” (Bonjour 1985, p. 28).
So my approach here is a kind of weak foundationalism. Gellman has moved from
moderate foundationalism to weak foundationalism because of the kind of criticisms offered by Gutting.12 In any case, the ARE can be mounted on the basis of either moderate foundationalism or weak foundationalism, but the latter is the one I adopt here (Kwan 2004).
Data sifting and epistemic ascent – critical trust
Ground-level sifting
Presumptive data can be defeated. Unfortunately, typically the defeater itself is also defeasible and there can be defeater-defeater as well as defeater-defeater-defeater, and so on.
Suppose we have a defeater D of an originally prima facie justifi ed belief B. Presumably, D
also has PFJ. Why can we not use the original belief to defeat the defeater instead? Bonjour has summarized Rescher’s suggestion in this way:
the basic idea is to fi rst segregate the total set of data or presumptions into maximal consistent subsets and then choose among these subsets. . . . one might then choose among the maximal consistent subsets on the basis of the plausibility of their members. (quoted in Bonjour 1985, pp. 224–5)
The idea is that we have to bring in coherence considerations to determine the weight of each presumptive datum. Once a presumptive datum coheres with many other presumptive data, its weight would be increased and it can serve to defeat another less weighty presumptive datum. A presumptive datum which confl icts with many other presumptive datum is then defeated. In general, we can formulate this methodological rule:
The rule of ground-level sifting
Always choose the consistent subset of our presumptive data which has maximal
weight.
12. I am not entirely sure that we have to do that. Perhaps moderate foundationalism is the better position (Van Cleve 2005, p. 173), but I do not need to settle the question here. I think that both are reasonable positions and the distinction between them may not be absolute.
Explanatory ascent and feedback sifting
We are not just happy to leave everything as it is. The data of experience (of various kinds) need to be ordered, explained, and made more intelligible. So the data of SE, when subjected to the search for order and intelligibility, yield the scientifi c framework. However, the framework itself can have feedback effect on the initial data. Some presumptive data of SE may be rejected due to theoretical reasons (e.g. the Principle of Simplicity). This can be captured by the following principle:
Principle of Epistemic Defeat
We should not believe that things are as they seem to be in cases when such a belief is in confl ict with the simplest theory compatible with a vast number of data obtained by supposing in a vast number of other cases that things are as they seem to be.
The CTA would regard this as the rational search for intellectual economy. In general, it is rational to believe in the theory which best explains our diverse presumptive data. It is the so-called inference to the best explanation (IBE). When the presumptive data confl ict and the best explanation cannot comprehend all the data, our best explanatory theory can serve as defeater of those “recalcitrant” data. That is also why a worldview can affect our feedback sifting because it is supposed to be an explanation of all our presumptive data.
Second-order critical principle
When we trust the majority of our presumptive data, those data may suggest to us that some types of presumptive data are not altogether reliable. For example, those presumptive data are found to be grossly inconsistent or they are contradicted by other well-established data. In such cases, we can formulate second-order critical principles. They are second-order principles because the justifi cation of these principles depend on our basic trust of our presumptive data, by and large, which is prescribed by the fi rst-order PCT. For example, our experiences may tell us that drunk people are prone to have hallucinations. So we can form this second-order critical principle: “Bizarre perceptual claims of drunk people are not to be trusted.” Of course, without the fi rst-order basic trust, we cannot establish the grounds for this principle, nor can we fi nd out who are the drunk people! These principles are in fact crystallized out of repeated applications of our Principle of Epistemic Defeat.
Consensus and testimony
Our data consist of personal experiences as well as testimonies. The Principle of Testimony, which will be assumed here, dictates that others’ testimonies are also presumptive data for one. We can formulate this:
Principle of Consensus
When an epistemic seeming is consensually corroborated, it is justifi ed to a much higher degree.
First, any consensus has to be discovered by each individual. Consensus is not, in any sense, given directly. Second, my principle does not entail that when an experience is not,
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 527
or even cannot be, consensually corroborated, it should then be doubted. The latter position would amount to a reverse of PCT.
The need for both foundation and coherence
The pattern of reasoning given earlier is similar to Susan Haack’s foundherentism, and her crossword puzzle metaphor is helpful (Haack 1994, p. 736). In short, an entry is much more likely to be correct if it not only corresponds to the clue given but also coheres with other entries which, in turn, correspond to their clues. We can formulate the following principle to capture this idea.
Principle of Epistemic Enhancement
When an epistemic seeming is coherent with other epistemic seemings, and its truth is coherent with or even suggested by the simplest theory which can explain many other epistemic seemings, then its degree of justifi cation will considerably be enhanced.
The Principle of Consensus is just a special case of the Principle of Epistemic Enhancement. So the CTA needs to appeal to both a foundation and coherence. Without some foundation (albeit weak), the process of justifi cation cannot take off the ground. Without coherence, we can hardly build a decent cognitive edifi ce on the foundation.
Comparison and confl ict
We can evaluate a type of experience by its coherence with other types of experience: intercoherence:
A type of experience E1 is coherent with another type E2 if
(a) the ontology of E1 is consistent with E2, and either
(b) one type helps to explain the nature, possibility, veridicality, and so on of the other type of experience, or
(c) the realization of the ontology of one type receives inferential support from the other type, or
(d) they are analogous in some respects (e.g. phenomenology, structure).
I would call a doxastic system a worldview if it is meant to incorporate all the phenomena into its scope and unify them by explaining the whole lot with reference to a few fundamental principles or ideas. The weight of a type of experience would also be affected by its worldview coherence: the degree of ease this type of experience can be accommodated within a coherent worldview with high explanatory power. This is just the application of the Principle of Epistemic Enhancement to a larger context. It is obvious that the possession of the above kinds of coherence to a higher degree would make a type of experience better established.
Principle of Comparison
A type of experience E1 is better established than E2 if either
(a) ceteris paribus, E1 has a higher degree of intracoherence than E2, or
(b) ceteris paribus, E1 has a higher degree of intercoherence than E2, or (c) ceteris paribus, E1 has a higher degree of worldview coherence than E2, or This Principle of Comparison seems to imply the following principle:
Principle of Confl ict Resolution
If type E1 confl icts with type E2 and that E1 is better established than E2, then E1 can serve as a defeater of E2.
Cognitive adjustment
When a prima facie justifi ed belief is defeated, what kind of cognitive adjustment should we make to our original cognitive structure? An experiential claim can be false in different ways and to different extent. Suppose I thought I seemed to see John on a street in London last Monday but this was defeated by many reliable witnesses. I ought to retract my claim.
Consider these possibilities: (a) I may be hallucinating, or (b) I actually saw someone who looks very like John. The second is at a higher level of epistemic seeming than the fi rst. All other things being equal, we should reinterpret the original experience so as to preserve as much truth in the original experience as possible. So (b) is rationally preferable. We can formulate the following:
Principle of Conservation
When an experience is defeated, it is rational to salvage as much noetic content as possible from that epistemic seeming, that is, to retain the highest undefeated level of epistemic seeming embedded in that experience.
The Impartiality Argument for the PCT
Different ways to defend the PCT
Defenders have offered several main lines of argument for the PCT. Some say that the PCT
is intuitively correct or even self-evident (Swinburne 1979, p. 254; Wall 1995, p. 19; Huemer 2001, p. 103). Others emphasize that the PCT is not a radically new invention. It is a principle which we do, in fact, employ in our actual epistemic practice (Mawson 2005, p. 166).
While the former focus more on our epistemic practice, John Hick argues that the PCT is implicit in our whole life: “We could not live on any other basis . . . Critical trust, then, is part of our working defi nition of sanity” (Hick 2006, p. 130). Still others argue that the PCT is justifi able on the basis of pragmatic utility (Rescher 1995, pp. 96–7). Rescher’s argument in fact has some similarity to Alston’s argument from practical rationality and impartiality. Alston argues that we cannot provide any non-circular proof for any of our socially established doxastic practices. How should we proceed then? To suspend our acceptance of the doxastic practices is either impossible or extremely costly. To replace the current doxastic practices with some others is no wiser because the new ones will not do better. So it is “eminently reasonable for us to form beliefs in the ways we standardly do”
(Alston 1993, p. 126).
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 529
Alston (1993) is aware of the possibility to “take our stand on one or more of these
[doxastic practices], and hold the others subject to judgment on that basis,” but this approach is “vulnerable to a charge of undue partiality in taking some of our fi rmly established doxastic practices for granted and requiring vindication of the others in the light of the former” (p. 126; italics added). Layman also presses this point about impartiality: Having seen that we cannot prove the reliability of sense experience-that we must give sense experience “the benefi t of the doubt,” how can we reasonably demand proofs for the reliability of other types of experience? To do so is to operate with an unjustifi ed double standard.
(Layman 2007, p. 45)
I will unpack this line of thinking in the following impartiality argument for PCT.
The impartiality argument
(T1) Impartiality Thesis
If we adopt a certain epistemological attitude toward a certain type of noetic experience, we should adopt the same attitude toward other types of noetic experience when we can fi nd no epistemologically relevant distinction.
(T2) Applicability Thesis
The Type PCT should be applied as a fundamental principle to at least some types of our noetic experience. (Note that “Type PCT” here actually means a principle identical to my generalized Type PCT except that its scope is unspecifi ed.)
(T3) Seamless Web Thesis
We can fi nd no clear-cut distinctions within the whole web of our noetic experience which are epistemologically relevant with respect to the applicability of the Type PCT.
Therefore,
(T4) The Type PCT should be applied as a fundamental principle to all types of noetic experience.
The argument seems valid. Let us look at the credibility of the premises.
Impartiality Thesis
(T1) is intuitively very plausible, and is also the implication of the general principle of impartiality: “treat similar cases similarly unless the relevant differences are shown.” If we let go this principle, then effectively we are abandoning any rationality in various kinds of discourse, for example, legal reasoning, personal judgment of scientists, literary criticism, social sciences, and philosophy. These consequences seem too costly. If we reject the Principle of Impartiality in epistemology, then we can just choose to treat different token experiences in arbitrary ways. It would also be diffi cult to justify how we can treat a type of experience, say sense perception, as an epistemological unit because certainly there are
differences between every two individual perceptions and it is hardly possible to show that each of this is not epistemologically relevant. If we have to discuss whether each perceptual claim is credible and which attitude we should adopt to this or that, epistemology would hardly be possible!
Applicability Thesis
One common response is that we should not accord PFJ to any kind of experience. The critics contend that we can justify claims of experience without appealing to the PCT. We will consider the cases of SE and memory.
Application of the Type PCT to SE
Alston (1991, 1993) has extensively and convincingly argued for the thesis that we cannot noncircularly show the reliability of SE. To avoid repeating his points all over again, I offer only brief discussions and put my focus more on memory. It may be argued that our belief in the general reliability of SE is justifi ed by inductive evidence. But it is not hard to see that in gathering this inductive evidence, we cannot avoid the use of memory: “an induction from past experiences to future experiences is only reliable if we correctly recall our past experiences. And what grounds have we got for supposing that we do? Clearly not inductive grounds- an inductive justifi cation of the reliability of memory-claims would obviously be circular. Here we must rely on the principle that things are the way they seem, as a basic principle not further justifi able” (Swinburne 1979, p. 256).
So in the end to “justify” ordinary perception inductively, we have to rely on the prima facie reliability of memory. But, of course, memory claims can be prima facie reliable as data concerning what the world is only if our past perceptions were prima facie reliable.
We are trapped in a circle if we deny PCT’s status as a fundamental principle.
Application of the Type PCT to memory
It is clear that our memory is fallible. How can we safeguard from mistakes and distinguish real memory from false ones or even imagination? Some philosophers have suggested that the reliable memory claims have intrinsic characteristics which are accessible to us, for example, vividness of the images associated or degree of conviction. However, after a long exploration in his book on memory, Von Leyden concludes, “since a false memory claim is as a rule qualitatively indistinguishable from a correct one, no memory experience alone can make it certain that what is alleged to be remembered really occurred” (1961, pp. 115–6).
Since reliance on memory is almost ubiquitous in formation of our beliefs, it is extremely diffi cult to see how there can be noncircular justifi cation of memory. Consider the suggestion that the general reliability of one’s memory is to be established by our perceptual experiences. However, unless we take the reliability of memory for granted to some extent, it is well-nigh impossible to gather suffi cient empirical evidence for memory. It is because our experiences almost all happened in the past. Without our basic trust in memory, what we can get from an even infallible SE is only and always a vanishing point.
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 531
Some philosophers suggested we can check our memories with documents and archaeological evidence (e.g. Hamlyn). However, if we do not already trust our memory of the meaning of the words, how can we be sure that we have interpreted the documents correctly? Moreover, what is the basis of my belief that the documents are reliable? Their reliability cannot be read off directly. Here we again need some inductive evidence from the past, which we can remember, to support it. So this kind of check is also in the end circular.
In the case of our memory beliefs, the appeal to the PCT as a fundamental prin-
ciple seems inescapable. I have surveyed quite a few authors, and found that almost all of them reach similar conclusions. For example, Russell claims that “the past occurrence is itself a premiss for my knowledge. It cannot be inferred from the present fact of my recollecting it except by assuming the general trustworthiness of memory” (1948, p. 205).
Other authors concur: “in the last analysis . . . we cannot justify or validate our memories further than by allowing a great number, possibly the majority, to be reliable and hence by assuming that the suffi cient conditions of remembering are in fact very often fulfi lled”
(Von Leyden 1961, p. 119). Ginet also thinks we “should trust the direct deliverance of one’s senses, or the direct perceptual deliverance of one’s memory, when they belong to the part that is coherent relative to all that their deliverance imply about one’s direct perceptions. Why should one do this? Why not?” (1975, p. 202). The last question is supposed to be a rhetorical one. To suspend our trust in them is simply “too intolerably frustrating to be borne and in practice impossible.”
Explanatory foundationalism
The CTA also emphasizes explanatory considerations, and indeed some scholars are forced by their explanationism to accept PCT (e.g. Lycan 1988, pp. 165–6). However, explanatory foundationalism wants to deny that the PCT is a fundamental principle by suggesting that the hypothesis of the external world can be justifi ed as the best explanation of our SEs (interpreted as appearance beliefs such as “it appears to me that a table is there”).
Alston (1993) has already criticized many versions of explanatory foundationalism (e.g.
Goldman, Slote). I want to discuss the more recent attempt by Jonathan Vogel (1998, 2005) here.
Vogel considers the isomorphic skeptical hypothesis (ISH), which is, in a way, parasitic upon the real-world hypothesis (RWH). “The relationships among causes and effects according to the ISH match those of the RWH. To that extent, it seems, the explanations provided by the one are no better or worse than the explanations provided by the other”
(2005, p. 75). For example, if our experiences are produced by the computer of a mad neuroscientist, then different portions of the computer disk can be supposed to occupy the explanatory roles we normally assign to familiar objects. This is called the computer skeptical hypothesis (CSH).
This kind of improved skeptical hypothesis apparently can do as well or as bad a job as the RWH does. However, he contends that necessary truths such as “two distinct objects connot be in the same place at the same time” in the RWH has some explanatory power, but ISH has to guarantee that by invoking “an extra empirical regularity,” say, one written down in the computer program of the superscientist. So the ISH is shown to be inferior by its “lack of simplicity” (2005, p. 77).
I doubt whether this really works. For argument’s sake, let us concede his point about extra empirical regularities in the CSH. However, we can equally argue that the ISH just needs a computer and a brain, whereas RWH needs myriad of separate objects. Hence, the former scheme is much more economical. Vogel is aware of this point but then he goes on to assert that “it is far from clear that, all by itself, positing fewer entities is a theoretical virtue” (1998, p. 355). Well, it seems to me quite clear that when we consider our practice of IBE, a theory which posits fewer entities is indeed simpler than another theory which posits more. A detective will not posit 10 murderers to explain a dead body when one can do. Astronomers will not posit 10 more planets to explain the deviation of Uranus’ orbit from Newton’s theory if just one (Neptune) can do, and so on.
Moreover:
we can compare our commonsense hypotheses about physical objects to, let us say, a Berkeleyean hypothesis about a very complex mind orchestrating the comings and goings of sensations. Which theory is simpler? Well Berkeley had just minds, mental states, and causation. The commonsense hypothesis has minds, mental states, causation, and
physical objects. On any criteria of simplicity, Berkeley seems to win. (Fumerton 1992, p. 165)
So in terms of simplicity, it is, by no means, clear that the RWH will win over skeptical competitors. To make a successful epistemic ascent possible, the Principle of Simplicity alone is not enough because its basic inclination it to trim things down. It needs to work together with the PCT.
Fumerton has further pointed out serious problems about IBE. Concerning the status of IBE, Vogel just says: “If . . . you are skeptical about IBE . . . , you will also be a skeptic about induction of all kinds . . . skepticism about IBE is exotic and, consequently, may be ignored” (2005, p. 79). It seems to me this kind of dismissal of exotic skepticism is quite dogmatic. I agree that the “skeptic has every right to insist . . . that one can and ultimately must ask questions concerning the legitimacy of the reasoning . . . One philosopher’s domestic species of skeptic . . . is another philosopher’s exotic skeptic” (Fumerton 2005, p. 94).
In any case, there are some general problems with explanatory foundationalism which even Vogel’s version cannot overcome. First, the basic trust in memory seems quite inescapable. The common data for the competing hypotheses such as RWH or CSH are the continued coherence of our experiences. But this datum we can only know through our memory.
Indeed the presupposition of memory in almost all attempts to vindicate SE’s reliability is not far from the surface. Or perhaps we can try to provide a coherentist justifi cation for memory, or even an IBE for the reliability of memory? Bonjour clearly points out the problems:
But then the issue arises of how, according to a coherence theory, the memory beliefs upon which any access to the fact of continued coherence must rely are themselves to be justifi ed.
Many philosophers have offered coherence theories of the justifi cation of memory beliefs, but such an account seems clearly to be involved in vicious circularity if the only reason for thinking that coherentist justifi cation is conducive to truth, and so that the memory beliefs in particular are true, relies on the existence of coherence over time and so on those very memory
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 533
beliefs themselves. The upshot is that there is no noncircular way for a coherentist to appeal to sustained or long-run coherence. (Bonjour 1999, p. 130)13
Second, memory is also vital in reasoning. Suppose we have a valid mathematical proof which needs 10 pages to write it out. When you come to the end and believe that the theorem is proved, what justifi es this? Not directly by the steps because we cannot hold all of them together in our minds and see all at once their connections; but by the memory that you have checked all the connections and found them convincing. So any reasoning with more than a few premises, including the explanatory argument for RWH, has to depend on the memory belief that the argument has proceeded correctly.
Third, how do we justify the appearance beliefs themselves? Sometimes we claim that they are infallible or incorrigible. But then how do we know that? Well, the answer can only be “it seems self-evident or obvious to me.” (The problem would be more serious if we believe that these beliefs are not incorrigible.) Some form of PCT(introspection) seems to be inescapable. The Applicability Thesis is also justifi ed then.14
Seamless Web Thesis
Suppose the critic wants to reject (T3). The basic strategy is to acknowledge SE and memory as the only BSJs. All other kinds of experience are then non-BSJs which must be justifi ed by the deliverances of the BSJ. However, how do we justify this kind of “discrimination,” which apparently violates the Principle of Impartiality? Perhaps we may claim that the BSJs alone are infallible, incorrigible, or uncontaminated by theory. But then it is doubtful that any type of experiential claim can aspire to this status, not even SE and memory.
We may claim the BSJs alone are public. But to know a type of experience is public, we need to fi rst know that it is shared by many people, and they agree quite a lot. But we cannot know that unless we presuppose PCT(SE and memory).
Some may claim that the BSJs alone are the most reliable, and SE and memory are the only candidates here. However, once it is admitted that the difference between the various kinds of experience is only a matter of degree, is it still plausible to cling to the qualitative distinction between the BSJs and the non-BSJs? Admittedly, SE and memory are more reliable but this judgment is only possibly arrived at when we give initial trust to our various kinds of experience. Furthermore, given two kinds of experiences E1 and E2, is it plausible to demand that E2 has to be validated by E1 before it has force if E1 is only more 13. Despite this criticism, Bonjour is ambivalent about the ultimate prospect of explanation foundationalism.
On the one hand, he thinks that “some explanation is needed for the combination of involuntariness and coherence, and that the conclusion advocated by the philosopher in question is thereby justifi ed as the best explanation of the facts in question. My own conviction is that such a inference, to Locke’s conclusion . . . , is ultimately cogent and can be justifi ed on a priori ground.” This sounds like he is hopeful but then it is puzzling that he immediately goes on to say that it is “a very familiar and serious problem for which no developed solution is yet available” (Bonjour 1999, p. 139). So at most he is only issuing a promissory note. He seems to be struggling here. I admire his heroic struggle with skepticism, and in the process he has not evaded the hard problems. In fact he also has doubts about whether the argument he promised can in the end succeed: “it still seems to me that some forms of skepticism are unavoidable and will simply have to be lived with” (Bonjour 1999, p. 129).
14. My argument does not need to start with SE or memory.
reliable than E2? Suppose a mutant species of human beings (X-men) start to emerge among us, and they possess a kind of super-perception which is much more reliable than our perception. They may insist only their super-perception is a BSJ, and that our perception needs to be validated by their superperception before it can be trusted. Would it be a reasonable requirement? No!
(T3) asserts that we can fi nd no relevant distinction between the various kinds of experience with respect to the applicability of the PCT. Although this is hard to prove conclusively, the earlier discussions suggest it may be well-nigh impossible to defeat (T3) once we accept (T2). If we grant (T2), then we are conceding that the Type PCT should be applied as a fundamental principle to, say, our SE as epistemic seeming, and on this ground alone. That means we accept SE for no other reason than that it seems to us to be true. Then when we consider experiences in other areas, the only relevant point seems to be whether they are also epistemic seemings. The other factors are just irrelevant. So (T3) follows. Anyway, the burden is equally on the critic who rejects the (T3) to show what is the relevant distinction.
Now I have defended the Type PCT and this is suffi cient for my ARE. However, the Type PCT also requires the weak Token PCT. If token experiences have no force at all, it is hard to understand why then tokens of a type can have PFJ.
Objections to the ARE
Disanalogy objection
Some critics argue that the serious disanalogy of RE or TE with SE undermines its veridicality. The logic is like this:
(A1) A kind of experience is cognitive only if it is analogous with SE, especially in the aspect of (e.g. having analogous tests).
(A2) RE (or TE) is not analogous with SE.
(A3) Hence, RE (or TE) is not cognitive.
It is instructive to consider a parallel argument against induction:
(B1) Induction is justifi able (or reliable) only if it is analogous with deduction.
(B2) Induction is not analogous with deduction.
(B3) Hence, induction is not justifi able (or reliable).
The second premise seems true. Deduction is truth-preserving while induction is not, and so on. The differences are quite obvious. However, there are problems for (B1). Why do we regard analogy with deduction as essential for a kind of inference to be reliable? Is it not implicitly assuming that there is and can only be one kind of reliable inference? Exactly because the two methods are concerned with different kinds of epistemological connection, is it not to be expected that the two methods should also show differences accordingly?
The fact that induction is disanalogous with deduction only suggests that the two cannot be reliable in the same way or to the same degree. It is still a live possibility that induction can be as reliable as it can be in its own realm. So this parallel argument against induction is not sound.
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 535
Similarly, the earlier argument against TE should be rejected. Why should we expect that there is only one kind of reliable experience? Is it not possible that we can have some access to different kinds of contingent truths about different aspects of the world? If it is the case, is it not to be expected that there will be several kinds of experience accordingly?
Each may be reliable in its own way and to different degrees. So (A1) seems to presuppose narrow empiricism whose employment of disanalogy objection is self-refuting. Typically, it is also claimed that the assumption that only SE and its like are reliable is also justifi ed by the facts of the matter (Daniels 1989, p. 489). But this justifi cation of SE is dependent upon the reliability of memory, which is also disanalogous with SE in many respects. If (A1) is upheld, would it not destroy justifi ability of memory as well? This is to undermine the earlier justifi cation of SE. The appeal to science to reinforce (A1) also suffers from similar problems. The actual success of science also depends on deductive reasoning and the capacity for making rational judgment, which are both very disanalogous from SE. So (A1) would damage justifi ability of science as well! So (A1) is quite implausible. On the contrary, if TE is, in general, veridical, we should positively expect TEs to be somewhat different from sensory experiences.
The bracket in (A1) can be fi lled in with different content, emphasizing a different aspect of disanalogy. Let us further examine the objection (from A1 to A3) which has been put forward by Richard Gale: REs “are cognitive only if they are subject to similar tests to those that SEs are” (1991, p. 316) (his prominent target is also TE). The reason for Gale’s negative assessment is that:
A veridical sense perception must have an object that is able to exist when not actually perceived and be the common object of different sense perceptions. For this to be possible, the object must be housed in a space and time that includes both the object and perceiver. . . . there is no RE analogue to this concept of objective existence . . . Because of this big disanalogy, God is categoreally unsuited to serve as the object of a veridical perception, whether sensory or nonsensory. (1991, pp. 326–7)
Gale’s discussions clearly display his epistemic chauvinism. Even if he is correct about the necessary conditions for a veridical sense perception, it is only one concept of an objective empirical particular. To argue that it is the only viable concept would have presupposed that physical object is the only possible kind of objective existent and that SE is the only viable kind of experience. It is just arguing in a circle. Gale admits that his argument would commit him to reject intellectual and moral intuition and the like as noncognitive. It also seems clear that introspection and some personal memories are not of spatiotemporal particulars. So on Gale’s criteria, they should also be rejected as noncognitive. These debilitating consequences of his argument in fact count against his argument. That Gale’s disanalogy argument has such sweeping consequences is a reason to suspect that his standard for cognitive experience is too stringent.15 Whether TE is called a cognitive experience does not really matter. The crucial question is whether TE is a BSJ. It is quite clear that Gale’s disanalogy objection offers an unreasonably restrictive criterion for BSJ in general. So his objection is not a suffi cient reason for discounting TE as a BSJ.
15. We should note that he himself appeals to necessary truth, for example, “no two empirical individuals of the same kind spatiotemporally coincide” (1991, p. 328). I would like to see how he is going to justify the said claim apart from an appeal to intuitions. He has not even hinted at the answer.
In any case, Gale’s objection does not undermine my argument for TE.16 The general idea is that the determination of analogy or disanalogy must be subsequent to our basic trust in experiences. The PCT is prior to knowledge about details of analogy or disanalogy.17
The confl icting claims objection
Many critics claim that, since REs are so various and mutually contradictory, we should regard all of them with suspicion. Even if we grant some force to REs, different REs cancel one another’s force in the end (Flew 1966, pp. 126–7; Martin 1986, pp. 87–8). So it seems that the challenge of RE to naturalism can be easily neutralized.
The application of the PCT to confl icting experiences
First, the existing contradictions between REs do not render the PCT inapplicable to them.
To apply the PCT to some experiences is to have initial trust in them and, if they are defeated, to salvage as much as possible from them. It does not entail that they are all or mostly reliable. There is no contradiction in saying that we should have initial trust in confl icting experiences. Indeed, almost all sorts of experience or doxastic practices produce confl icting beliefs sooner or later. Empirically speaking, no experience in which we trust is completely free from this problem. (Just think of the empiricists’ “argument from illusion.”) So why do we think that the presence of contradictions in RE should debar us from having initial trust, at least to some degree, in RE?
Now the critics seem to suggest a skeptical rule (SR):
(SR) When experiences or claims confl ict with one another, we should reject all of them.
Should we adopt the SR instead? I do not think so. Consider the confl ict of witnesses in the courts. It would be indeed stupid to reject all their accounts just because they confl ict!
It seems to be a rational strategy to try to reconcile their reports as much as possible. For example, a common core18 can be identifi ed. Another example: suppose a phenomenon occurred very briefl y, which led to confl icting reports – A reported seeing an airplane, B a spaceship, and C an air balloon. It is absurd to suggest that we should reject all their statements and think that nothing has ever happened! At the very least, we should accept the common content of their experiences: there is an unidentifi ed fl ying object. Moreover, 16. For a more detailed reply to Gale, see Gellman (2001, chap. 3).
17. Everitt’s critique of ARE also includes a form of the disanalogy objection. Initially, he directly attacks the tests of RE by applying the standards of SE (2004, p. 169). When he later considers the plausible suggestion that some kind of variability of tests across practices is to be expected, he just insists that these tests cannot be too different from the standards of SE (2004, p. 170). Besides being a mere assertion than an argument, it is unclear how different is “too different.”
18. Indeed it is not the case that a “common core” has to be shared by all the eyewitness accounts. Sometimes it is suffi cient that it is shared by the large majority of the accounts, provided that either the error of the deviant witness in that aspect can be explained or overwhelming explanatory power is attained by adopting the common core. Admittedly, there are borderline cases in which we have to rely on our judgments.
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 537
historical documents are also liable to massive contradictions. However, we do not deduce from this phenomenon that historical enquiry is entirely pointless. The job of the historian is to utilize all these materials to reconstruct the past by harmonizing them without producing too much strain in the overall interpretation. Many historical accounts of a momentous historical event (e.g. China’s Cultural Revolution) are contradictory. It is diffi cult to determine the exact course or nature of this event but it would be preposterous to deny that the Cultural Revolution has happened. All the given examples count against the skeptical policy and show that confl ict of presumptive data is not irremediable.
In line with the said suggestions, although there is a signifi cant degree of confl ict of the REs at the highest level of description, a certain common core can still be extracted from them at a lower level of description.19 For example, Davis carefully sifts through the data and suggests the following as the common core:
(i)
the mundane world of physical bodies, physical processes, and narrow centres of consciousness is not the whole or ultimate reality.
(ii) . . . there is a far deeper ‘true self ’ which in some way depends on and participates in the ultimate reality.
(iii) Whatever
is the ultimate reality is holy, eternal, and of supreme value; it can appear to be more truly real than all else, since everything else depends on it.
(iv) This holy power can be experienced as an awesome, loving, pardoning, guiding (etc.) presence with whom individuals can have a personal relationship . . .
(v)
. . . at least some mystical experiences are experiences of a very intimate union with the holy power . . .
(vi) Some kind of union or harmonious relation with the ultimate reality is the human being’s summum bonum, his fi nal liberation or salvation, and the means by which he discovers his ‘true self ’ or ‘true home’. (1989, p. 191)
All the REs point to the fact that there is another realm up there or beyond the naturalistic world. So even if REs have internal confl icts, arguably they can still lend support to ARE insofar as they tilt the balance away from naturalism. However, it is a kind of general support for the religious worldview instead of a specifi c support for theism at this level.
Reclassifi cation of RE
To evaluate the epistemic status of TE relative to other kinds of RE, the rule of ground-level sifting is helpful. What we need to do is to segregate the presumptive data into consistent subsets and then choose the one which has maximal weight. So let us reclassify REs in the following way:
(A) TE
(B) Theism-compatible non-TE
These REs, though not specifi cally theistic, are logically compatible with theism. It can be further divided into the following:
19. Davis (1989, chap. 7) provides detailed suggestions how the confl icts can be reconciled.
(1) Theism-friendly non-TEs: these experiences go quite well with theism and they can be readily interpreted theistically without distortion or implausible
reinterpretation.
(2) Theism-neutral non-TEs: these experiences are compatible with theism and they can be interpreted theistically; however, they do not strongly suggest a theistic interpretation, and they can also be interpreted in a way that goes against
theism.
(3) Theism-unfriendly non-TEs: these experiences are logically compatible with theism but seem to suggest an interpretation that is in tension with theism.
(C) Theism-incompatible non-TEs
These experiences, if veridical in their most ramifi ed description, are incompatible with theism because the Ultimate disclosed in these experiences is not personal.
I just explained my classifi cation scheme, and I will soon discuss what kinds of RE
can be placed within it. I have defended the claim that TE is a well-established type, and so the Type PCT is applicable to it. As far as other non-TEs are concerned, they are also covered by the Token PCT, which means that they at least have some force.
Now there seems to be a major confl ict between TE and theism-incompatible non-TE.
If we also take into consideration the theism-compatible non-TEs, then there are two competing groups of RE: the pro-theistic group, which consists of TE plus theism-friendly non-TEs, and the anti-theistic group, which consists of theism-incompatible plus theism-unfriendly non-TEs. (Since the theism-neutral non-TEs do not point in any specifi c direction, we can ignore them here.) The question is to decide which group has the greater weight. Before discussing this question, we need to clarify the nature of monistic mystical experience.
PCE and monistic mystical experience
How to accurately describe the monistic mystical experience is in fact a controversial question, and the mystics sometimes offer paradoxical descriptions. To clarify the problem, we need to distinguish these two experiences:
(1) Awareness
of undifferentiated unity, the One, the Self, and so on.
(2) A pure state of awareness which is undifferentiated unity.
The fi rst experience is more like an intuition, an insight into reality but the intentional structure of experience is intact. The second experience is the PCE and it does not have an intentional structure at all. They are logically incompatible: no person can have both experiences at the same time. I propose that the phrase “monistic mystical experience” should be used to refer to the fi rst intuitive experience rather than the PCE because the PCE in itself is not specifi cally monistic. Indeed it cannot yield support to monism by virtue of its epistemic seeming because there is no conceptual element in the PCE. The PCE is not without its conceptual diffi culties (see Katz) but we have empirical evidence for the PCE
(Forman 1990, pt. I).
Consider this report:
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 539
I would settle down, . . . and there would just be a sort of complete silence void of content.
The whole awareness would turn in, and there would be no thought, no activity, and no perception, yet it was somehow comforting. . . . I did not yet identify myself with this silent, content-free inner space. . . . Then . . . I began to recognize in it the essence of my own self as pure consciousness. Eventually, . . . “I” as a separate entity just started to have no meaning
. . . There is no thought, there is no activity, there is no experiencer. (Forman 1990, pp. 27–8).
The description is littered with paradoxes. Certainly, there was some experience akin to a PCE but it was also “somehow comforting.” Did this sense of comfort feature within the experience or not? If yes, the experience is no longer pure. If no, then the sense of comfort should be another experience subsequent to the PCE. This is probably the case. I also wonder how the person can report a PCE afterward. In the PCE, there is no thought and feeling and no awareness of any sort. How then can the subject later ascribe this experience to himself? This must be explained by the continuous operation of the memory, even if it is somehow hidden during the PCE. But whose memory and whose PCE? To give intelligible answers to this question, it is diffi cult to avoid positing a self who possesses this PCE
even though the sense of self is not present in the PCE. Indeed it is the interpretation adopted by the subject originally. But later the subject comes to another interpretation:
“there is no thought, activity and experiencer.” But this thought cannot be given in that experience. So this must be an interpretation imposed on the PCE due to subsequent refl ection. So the question as to why this interpretation is justifi ed has to be raised.
Perhaps we can conceive of the experiences as a unity. First, the subject has a PCE
without monistic mystical experiences. However, when he comes back from this contentless state, he immediately has some monistic illuminations. The two experiences seem to go together and the PCE in this context may then be legitimately interpreted as an experience of the Self. However, the same move is available to the theists. Pike argues that when the Christian mystics experience a “monistic” interval within the context of dualistic experiences of God, it is legitimate to think that this “moment of experience in which the soul detects no distinction between itself and God is a kind of experiential fi ction” (Pike 1992, p. 156). The spirit is simply “deluded by love into not noticing the difference between itself and God.” The upshot is that there is no simple way to connect PCEs with monistic mystical experiences, either conceptually or evidentially. So PCEs are theism-neutral.
Coherence and confl ict of TE with other REs
I have just argued that PCEs are theism-neutral. The main reason is that a PCE has no propositional content, and hence it cannot contradict theism. In fact in the earlier cases, some people can have both PCEs and TEs, and they do not see any contradiction between them.
I think nature mysticism is also theism-neutral. It can be interpreted in either a monistic or theistic way. For example, Jonathan Edwards’s experience of nature mysticism is explicitly integrated with a TE:
The appearance of everything was altered; there seemed to be, as it were, a calm, sweet cast, or appearance of divine glory, in almost everything. God’s excellency, his wisdom, his purity and love, seemed to appear in everything; in the sun, moon and stars; in the clouds and blue sky; in the grass, fl owers, and trees; in the water and all nature. (quoted in Hick 2006, p. 29)
The kind of unity sensed in nature mysticism can refl ect either the unity of all things rooted in their common source – their Creator God, or the kind of metaphysical unity envisaged by monism.
Second, let us assess the weight of the pro-theistic group. I have contended that TE is a well-established type of experience, and it is plausible to think it has the highest degree of intracoherence among REs. So TE is weighty in itself. Furthermore, many kinds of REs are compatible with theism and can readily be interpreted theistically, and hence theism-friendly:
(a) Ecstasy and peak experience
(b) Encounter of the Light Being in a near-death experience
(c) Experience of evil spirit, angels, or departed saints
(d) Experience of contingency – a spontaneous feeling that the world is somehow dependent on something beyond
(e) Experience of design – experience of being struck by the beauty and intricacy of the natural order and the feeling that this order is ultimately due to Intelligence (a) suggests a kind of spiritual depth in human beings which is more consonant with theism than with naturalism. If (b) and (c) are veridical, then there exist immaterial beings outside the bounds of nature. This is compatible with theism but incompatible with naturalism. (d) and (e) point somewhat in the direction of a transcendent Creator-Designer. If these experiences are veridical, they are certainly friendly with theism, and can even offer signifi cant support for theism. I think these experiences are not that uncommon but the empirical case cannot be made here. Anyway, as far as RE is concerned, the pro-theistic group has considerable weight.
Now let us assess the weight of the anti-theistic group. The prime examples of theism-incompatible non-TE are monistic mysticism and the experience of Nirvana. Since these experiences suggest the Ultimate is essentially impersonal, it confl icts with theism. I will examine this confl ict later. Let us now discuss theism-unfriendly non-TEs. Perhaps the experience of minor deities (e.g. visions of Kali or Buddha or Apollo) belong to this group.
Michael Martin argues that they are indirectly incompatible with theism because they are embedded in a worldview which is at odds with the theistic one. Martin has a point but I would like to emphasize that logically speaking, it is possible that both God and these minor deities exist at the same time. The tension only arises because we have a further assumption that if God exists, He would not have created these minor deities. Otherwise, we can just regard these beings as spiritual creatures in God’s world. However, perhaps given some plausible assumptions, these experiences may be theism-unfriendly. Since the confl ict is not directly given in the experiences, I think it is quite possible to remove the tension by appropriate adjustment of our underlying assumptions and reinterpretation of those experiences.
As I have argued, cognitive adjustment is frequently needed even in our SEs, memories, and scientifi c inquiry. Deliverances of SE and memory often directly and indirectly confl ict with one another. Some experimental data are frequently “indirectly incompatible” with other experimental data because the former are incompatible with the simplest law which are suggested by the latter. We might decide to stick to each and every experimental datum no matter what, but no cognitive systematization is then possible. If we are not willing to forgo at least some presumptive data and take theoretical gain into considerations, no
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 541
universal laws can be substantiated. For example, after performing an experiment, we have six data which fall roughly on a straight line but also an anomalous datum not even close to that line. I submit it is rational to discard the anomalous datum in favor of the six others.
This is what the CTA advises, and is also what the scientists in fact are doing.
Martin objects that acceptance of the PCT would result in a bloated ontology. This refl ects a lack of understanding of the CTA. He seems to think that the PCT automatically commits us to accept each and every experience of minor deity in toto. This kind of picture is not even true of our more mundane epistemic practices. So in accordance with the CTA, the following is the rational way to do:
• We search for the neatest ontology or worldview which can preserve the maximal weight of the presumptive data here.
• Some of these data can be rejected if they are more vulnerable to defeaters and it results in a suffi cient gain in simplicity or coherence. However, even in this situation, we should try to preserve their cognitive force as much as possible.
These considerations are just corollaries of the principles of the CTA. The theist does not need to give detailed answers about each and every RE in human history. He can be contented to point out that the argument for TE would be defeated only if the process of cognitive adjustment described results in the truncation of the whole lot of TE. I do not think it is the case. To the contrary, it seems to me that those lesser divine beings are not as widely experienced as God is. So it seems to be justifi ed, on all the evidence, to regard the existence of God as the defeater of the experiences of those lesser beings who are incompatible with existence of God, and not the other way round. Moreover, it is much simpler to accept theism, which is supported by the signifi cant weight of the pro-theistic group.
We also need to investigate how common are these experiences (e.g. experience of Apollo). We can raise these questions: does it survive the test of age and culture? Is it widespread? Does anybody claim to have an experience of Apollo which can be described as a “life in Apollo”? As far as I know, there is no evidence to show at least one-third of people in contemporary Western societies have had an experience of Apollo, and that signifi cant minorities in every culture have organized their life around the experiences of Apollo. I have never heard of a single Chinese having an experience of Apollo, but many of them have defi nitely experienced God.
A theist does not need to choose between blanket rejection or blanket acceptance. He can adopt a more nuanced, case-by-case approach toward these experiences of minor deities.
• For those minor deities who are not really experienced by a sizable number of people across eras and cultures, the related experiences can be treated as anomalous data- either rejected or shelved.
• There is no need to reduce all those REs to sheer self-deception or hallucination. For those minor deities who have a more substantial experiential basis, the related experiences can be explained as the work of spiritual beings, or treated as genuine phenomena in the spiritual world, of which we do not yet have a full understanding. In cases like an experience of Virgin Mary, it may readily be interpreted as a veridical vision that does not commit us to her literal existence in the manifested form.
• In any case, experiences of minor deities apparently point in a consistent direction beyond naturalism. If they can be readily defeated, say, by naturalistic explanations, then the problem of indirect confl ict with TE is removed. If the naturalistic explanations available are not really adequate,20 then the whole group of such experiences can be taken to support a less specifi c claim: there are some spiritual forces in the world. This would actually provide further support for the theistic interpretation of the world.
• These experiences may be reinterpreted by monism as part of the maya. For example, Kali can be regarded as just one aspect of the Ultimate (the One) rather than an independently existing deity. This is possible but it is hard to see how those experiences can provide positive support for monism.
The theistic interpretation is further reinforced by considerations of intercoherence and worldview coherence: we can also ask does anybody have an experience of Apollo or the like through experiences of nature, goodness, beauty, fi nitude, freedom, and so on? How does the veridicality of the experience of Apollo help to explain other experiences? Is there any worldview that revolves around the existence of Apollo which is coherent and having explanatory power, say, with regard to contemporary scientifi c fi ndings like fi ne-tuning? Is there ever a natural theology for the existence of Apollo? I think the theistic worldview will fare much better in these aspects. The entire content of this book is a support for this claim.
To sum up, experiences of minor deities taken at their most ramifi ed description may be somewhat theism-unfriendly. However, from the perspective of TE, there are plausible reinterpretations of experiences of minor deities (whereas the reverse does not seem to be true) that can salvage much of their epistemic force, which supports the claim that there are spiritual forces in this world. This is confi rming theism as well.
Is the ultimate reality personal or impersonal?
As for the contradiction between the personal versus impersonal understanding of the nature of the ultimate reality, I believe it is not as stark as it is commonly made out. The Personal Ultimate can manifest himself in a nonpersonal way. The manifestation can still be veridical and revelatory. Consider Yahweh’s epiphany to Elijah. God can be said to be manifested in the earthquake and the whirlwind but this is not yet a personal manifestation. However, the whole thing is transformed when the “still small voice” is added to the scene. The whole experience becomes an unambiguous personal manifestation. So nonpersonal manifestation need not mean that it is antipersonal. Indeed, the Old Testament scholar Rowley says, “we fi nd personal and impersonal factors woven together in what the Hebrews believed to be God’s manifestation of himself ” (1956, p. 45).
Wall suggests that:
experiences of God, some personal and some impersonal, may be analogous to viewing an object from different sides. Or perhaps it is like viewing an object from different distances: From a distance we do not see the needles of a pine, but up close we do; yet regardless of the 20. For example, those experiences of minor deities may be supported by paranormal experiences that are hard to discredit.
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 543
differences in the experiences, we do not begin to think that we have seen different trees.” So
“we might say that ‘from a distance’ we may see God as impersonal but ‘up close’ as personal.
(1995, pp. 319–20)
So in the end, there is a confl ict only if one claims that the Ultimate is essentially personal, and the other claims that the Ultimate is essentially impersonal. However, “we rarely, if ever, run across claims to the effect that someone was aware of experiencing God’s essential nature or some aspect of God as essential. In all cases I collected or examined I found neither claim made” (Wall 1995, p. 320).
So in the end, we do need to resolve the confl ict between the claim that the Ultimate is essentially personal and the contrary claim that the Ultimate is essentially impersonal. I now explore the confl ict of theism and monism of the Advaita Vedanta school.
Confl ict of theism and monism
The Advaita Vedanta school makes these claims:
(a) Brahman as the Impersonal Absolute is the Ultimate Reality in which there is no distinction and differentiation.
(b) Atman is Brahman, that is, every human self is metaphysically identical to Brahman; hence every self is metaphysically identical to another self.
(c) The world is ultimately maya (illusion).
Obviously, these claims confl ict with theism which asserts that the Ultimate is a Personal God who creates the world and distinct selves. Theism also claims that although all the latter are contingent upon the continuous creative act of God, they are nonetheless real and distinct from one another. In my CTA, we have resources to resolve this confl ict: Principle of Comparison and Principle of Confl ict Resolution. The crucial point is whether TE
is a better-established type than monistic experience (ME) or vice versa. Let us briefl y discuss this question.
Intracoherence
(a) It seems to me the people who have MEs are confi ned to a relatively small group of mystics in various religious traditions, and mainly Eastern ones, whereas I have argued that TE is present in various traditions and shared by a substantial portion of humankind.
(b) I have argued that tokens of TE do form intricate coherence relationships, whereas MEs do not exhibit many modalities nor do their tokens exhibit a complex pattern of coherence.
(c) It seems to me the articulation of the ME suffers from many conceptual diffi culties: how is an experience of Nirvana or the Absolute possible? A noetic experience has an intrinsic structure but the Absolute is structureless. If it is literally ineffable, how can it be communicated to form an intersubjective type of experience? Hick is also aware of this kind of problem:
to lose one’s individual identity completely, like a drop becoming part of the ocean –
a familiar simile in mystic literature – would be to lose the individual continuity of
consciousness and memory in virtue of which the mystic would later be able to report the experience. How could someone remember being in a state in which he or she no longer existed as a distinct individual? There must, surely, have been a continuing strand of consciousness to enable them later to speak about it, while still enjoying something of its bliss. (Hick 2006, p. 22)
I have defended here the conceptual coherence of TE. TE seems to have a higher degree of intracoherence than ME.
Intercoherence
The degree of intercoherence of TE is high. It does not confl ict with other basic types of human experience. It can even provide a framework for explaining them (e.g. moral experience and rational intuition). The following points concentrate on the intercoherence of ME with various kinds of noetic experience.
SE
It seems that SE discloses a world of multiplicity and change. Since SE is one of the best-established types of experience, then SE can serve as the defeater of almost all other types of experience. Now ME apparently comes in confl ict with SE, and this seems to be a serious problem. For example, Gale thinks that concerning MEs, their “descriptions are often contradictory or in confl ict with our best-established empirical beliefs, such as that there exists a multiplicity of distinct objects and events in space and time” (1991, p. 303). This criticism has been pressed by Madhva in the thirteenth century within the Hindu tradition itself (Copleston 1982, p. 80). The postulation of maya does not clearly help: is it not introducing multiplicity within Brahman itself? There is also the move of qualifi cation: there are degrees of reality, and SE discloses something real at its level but not ultimately real. I, however, have the problem of understanding its exact meaning. If it means that the physical world, although real, is only contingent, it does not confl ict with either SE or theism. If it means that the world is real only as an appearance, it again comes into confl ict with SE.
Moral experience
Monism also seems to come in confl ict with the basic assumption of moral experience, that is, there is an objective distinction between good and evil, right and wrong. It is questionable how this can be preserved in a system when All is One. Does it mean that in the end, we cannot distinguish Hitler from Mother Teresa, and Stalin from Gandhi? If so, it is a mockery of our ordinary consciousness, not to mention the morally dangerous doctrine that Brahman is beyond good and evil. Of course, many monistic mystics exhibit a high degree of moral consciousness and display virtues in their lives. What I am pressing is the logical problem: are the moral experiences of the monistic mystics compatible with their MEs? Furthermore, moral strife seems to presuppose a self distinct from Brahman:
the teaching of the Advaita school about the need for moral purifi cation, for detachment from selfi sh passions, and for the advance from ignorance to knowledge presupposes that it is in fact the human being who is being urged to turn to the One. It is the human being, not
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 545
Brahman, who is called upon to recognize his or her relationship with the One. If the Advaita really means that the human being must disappear altogether, he should say so in an unambiguous manner . . . Further, if the human being is appearance, to whom the appearance appear? To Brahman? If so, the Absolute presumably misleads itself, though why and how remains obscure. (Copleston 1982, p. 84)
TE, on the other hand, provides an ontological explanation or basis for the basic distinction in morality and the categorical imperative, and there is experiential fusion between moral experience and TE. Conscience sometimes mediates the voice of God and experience of grace is often experienced in and with the experience of moral improvement and so on (see Kwan 2006c).
Interpersonal experience
Our interpersonal experience reveals the otherness of the Thou. Again this seems to be in confl ict with the monistic doctrine that all these distinctions are just illusory. On the other hand, this type of experience is coherent with TE. For example, theism allows for the ecstatic union of souls, and the ecclesiological doctrine of the Body of Christ also suggests that although human persons are distinct, they are destined for intimate communion. So some insights of monism are preserved.
Experience of change
According to monism, change is an illusion. But “nothing seems plainer than that there are a multiplicity of things in the world around us and that these, and our experience of them, are constantly changing. To deny the reality of change, variety and multiplicity seems to be the most bizarre of all religious or metaphysical procedures” (Lewis 1969, p. 296).
So TE also seems to have a higher degree of intercoherence than ME.
Worldview coherence
I assume that the theistic worldview has substantial explanatory power (see other chapters in this book). If there is not a comparable rational case for monism, then TE will have an advantage over ME. This question cannot be fully settled here, of course.
As a summary, it seems possible to argue, with respect to clauses (a) and (b) (and possibly (c)) of the Principle of Comparison, that TE should be taken as the better established subtype of RE. Therefore, in accordance with the Principle of Confl ict Resolution, it is plausible to claim that TE would serve as the defeater of ME rather than the other way round.
Cognitive adjustment
Of course, it is not plausible just to dismiss the whole type of ME. For my case to be more convincing, some plausible postexperiential reinterpretation of ME is needed. First, let us consider the case of extrovertive mysticism which:
takes its start from the facts of the world around us, and . . . fi nds that all things run into one another in a unity in which all separate existence and variety is lost. There is nothing but the One, but the One is seen ‘in or through the multiplicity of objects’. (Lewis 1969, p. 302)
It seems that the core experience is a kind of nature mysticism which, as I have argued, is compatible with theism. On the other hand, there are also plausible interpretations of introvertive mysticism. Maritain suggests that:
The Hindu experience does appear therefore, to be a mystical experience in the natural order, a fruitive experience of the absolute, of that absolute which is the substantial esse of the soul and, in it and through it, of the divine absolute. (1966, pp. 97–8)
The PCE can be interpreted as an experience of the substance of one’s soul which is indeed grounded in God. Because the PCE is a purely negative experience, it is not surprising that a monistic intuition can occur after that. Although it is defeated, it can be seen to be a plausible error. Furthermore, theistic mysticism shows that the mystical path can be accommodated within theism, that is, it is not at all surprising that God will bring about the experiences of emptiness. According to Louis Dupre, negative theology may be the fi rst step to achieve spiritual poverty and humility. This enables us to see the emptiness of creaturely things and our inadequacies and liberates us from the attachment to things (Dupre 1981, p. 48).
Can it be done the other way round? Can the way of devotion be incorporated into the mystical path? Yes, but ultimately, it is more like a concession than an integration. The Advaiti n would allow that the bhakti is a legitimate way to Brahman, but ultimately, a loving devotion to Brahman, if he is consistent, must be regarded as based on ignorance. It is because devotion presupposes some sort of dualism and it confl icts with Advaita’s nondualism. So: theism can give us a profound and beautiful way of integrating the insights of prophets with those of contemplatives and combining the paths of devotional worship and mystical endeavour. But if you try to do it the other way round the main teachings of theism begin to disappear: devotion fades and the revelations wither. While theism can convincingly absorb and enrich the mystical path without detriment to the latter, the mystical path cannot absorb theistic belief without relegating it to second place. (Smart 1960, p. 72)
If this is the case, then the theistic way of cognitive adjustment preserves more PFJ than the monistic way. In accordance with the Principle of Conservation, the former is to be recommended.
Of course, the above is just the beginning of a critical dialogue between different religious traditions. I also have no time to discuss other theism-incompatible REs, for example, the experience of Nirvana (yet many things said earlier also count against these “mystical”
theism-incompatible REs). Nevertheless, I suggest tentatively that the CTA may have signifi cant resources to resolve the confl icts, and at least it is not obvious that TE is defeated by ME or other types.21 Yandell (1999, chap. 13) has also offered a strong critique of monistic non-TEs.
21. Perhaps some may object that my interpretation of ME is uncharitable. For example, Hick agrees that accounts of ME, when taken literally, are indeed incoherent. His solution is that we should interpret the mystical literature metaphorically: “the unitive language of Advaita Vedanta is . . . metaphorically . . . expressing a . . . vivid awareness of the limitless reality in which we are rooted” (Hick 2006, p. 23). I am not sure Hick’s revisionism is really true to the monistic mystics’ experience or traditions. In any case, a vivid awareness of the limitless reality in which we are rooted is compatible with theism because God is limitless, and we are indeed rooted in Him. So in any case, TE seems less problematic than ME. If taken literally, the ME is plagued by apparent incoherence. If taken metaphorically, it is not incompatible with theism, and it cannot be taken to offer support for monism over against theism.
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 547
Conclusion
The degree of confl icts among REs is usually grossly overestimated. Many REs are just different and far from being logically inconsistent. REs as a whole still support theism more than naturalism. Among the REs, TE is a relatively well-established type, and it is also widely corroborated by a diverse group of theism-friendly REs. It is, at least, compatible with many mystical experiences (e.g. nature mysticism). Moreover, in accordance with the principles of CTA, a plausible case can be made that the pro-theistic group has a greater weight than the anti-theistic group. So TEs are, by no means, defeated by confl icting REs.
I have a fuller treatment elsewhere (Kwan 2003).
Brief discussions of several objections to TE
I have treated quite a few major objections: logical gap objection, theory-ladenness objection, privacy objection, conceptual incoherence of TE objection, and disanalogy objection.
I have written at length responding to the confl icting claims objection because I think it is the most serious one. In my section on “Intracoherence of TE,” I have also implicitly dealt with the oddness of TE objection. There are still a few objections which I have not explicitly dealt with. I give some brief comments on them further in the discussion.
Impossibility of individuation objection
James Harris has built on Gale’s idea, and has made heavy weather of the problem of individuating God in TE. I think it is indeed impossible to prove that it is the same God all the time and not similar but different gods appearing in different occasions. However, I do not think the requirement of absolute or noncircular proof is realistic because even our SEs cannot satisfy this. Outside the critical trust framework, we cannot in fact prove that our experiences apparently of the same table are really caused by the same table, and not really the results of two qualitatively similar acts of deception by the Cartesian demon. I agree with Gellman’s (2001, chap. 3) idea that reidentifying physical objects is also a holistic practice, which is inescapably circular in the end. If we also accept the reidentifi cation of God as a holistic practice, then I think given the basic trust in TE (revelatory experiences of prophets and Jesus, the experience of maximal greatness-making properties in TE) and the Principle of Simplicity, the problem can be largely solved. Even if in the end we cannot do that, then at most we can say that the ARE does not uniquely support monotheism over against polytheism. But this would be small comfort for naturalists because polytheism is equally unacceptable to them.
The no criteria/uncheckability objection
Critics allege that there is no criterion to distinguish the veridical REs from the nonveridical ones. Of course, there are in fact criteria from within the religious framework, but the critics object that these are not objective, noncircular criteria. The fi rst line of response is to point out that it is also the case that SEs can only be checked by other SEs, and that it is also circular (see Kwan 2006b). This point should be obvious after discussions in this chapter. Moreover, my lengthy section on the “Structure of CTA” is, in fact, meant to show that there is a universal method which governs how second-order critical principles or
principles of epistemic enhancement arise from the experiences themselves. In fact, criteria of veridicality and unveridicality are just corollaries of these principles. I have also demonstrated how the rules of CTA can be used to settle the problem of confl icting REs. So the basic thrust of the no criteria objection has been neutralized.
The naturalistic explanation objection
The prior question that needs to be settled is when would a naturalistic explanation of TE
really constitute a defeater. I think it needs to fulfi ll several conditions:
(a) It can specify a set of causally suffi cient conditions for TE.
(b) We have reasons to believe that that set of conditions will render the veridicality of the TE unlikely.
(c) We have reasons to believe that set of conditions really obtain in the majority of cases of TE.
I doubt that any available naturalistic explanation can satisfy these requirements, but I have not said much about this here. Fortunately, many authors have a good response to this problem: Davis (1989, chap. 8), Yandell (1993, chaps. 6 and 7), Gellman (1997, chap. 5; 2001, chap. 5), and Griffi th-Dickson (2000, chap. 4). Wall’s book is entirely devoted to this issue and he utilizes concrete examples of REs to point out the inadequacy of various naturalistic explanations. I have also offered a detailed critique of projectionism (Kwan 2006a).
Gullibilism objection
Many critics like to produce alleged counterexamples to the PCT such as experiences of UFO, apparitions, and so on. They contend that the PCT leads to the acceptance of all these strange claims, and hence the approach is too gullible. I have taken care of this objection in the very fi rst stage of my argument. I only argue for a weak Token PCT, emphasize the need for critical sifting, and explain how the CTA can produce an orderly and coherent noetic structure.
The ARE in the Twenty-First Century
Although the ARE is hotly contended, I think Swinburne’s route of taking RE as prima facie evidence for the transcendent realm is a promising one. Following this line, I have defended a version of ARE in this chapter.
I have argued that the Type PCT is a fundamental principle. Even if my impartiality argument is not conclusive, I hope I have made, or at least suggested, a plausible case for the CTA, which is clearly delineated here. I contend that, since TE does have impressive empirical grounding and intracoherence, it deserves to be accorded PFJ. I have also argued that many popular alleged defeaters of TE do not really succeed. Of course, I cannot deal with all the proposed defeaters here. However, the recurrent pattern of discussions shows that the objections often commit fallacies of double standard, epistemic chauvinism, or requirement of superreliability. So those alleged defeaters are more implicitly the rejection of the PCT and CTA. Otherwise, it seems rather diffi cult to show the entire corpus of TE
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 549
to be unreliable. So it may not be unreasonable to believe that the critics are yet to produce a convincing defeater. It follows that belief in God can at least be tentatively justifi ed.
I do not expect that the ARE will gain consensus, but in any case, philosophical arguments rarely achieve this. The ARE is worthy of further exploration together with the deep epistemological questions in the twenty-fi rst century. It may well lead to fruitful developments in both philosophy of religion and epistemology.
References
Alston, W. (1991) Perceiving God: The Epistemology of Religious Experience. Ithaca, NY and London: Cornell University Press.
Alston, W. (1993) The reliability of Sense Perception. Ithaca, NY and London: Cornell University Press.
Alston, W. (1999) Perceptual knowledge. In J. Greco and E. Sosa (eds.), The Blackwell Guide to Epistemology, 223–42. Oxford: Blackwell.
Baergen, R. (1993) The infl uence of cognition upon perception: the empirical story. Australian Journal of Philosophy 71: 1, 13–23.
Bagger, M. C. (1999) Religious Experience, Justifi cation, and History. Cambridge: Cambridge University Press.
Baillie, J. (1962) The Sense of the Presence of God. London: Oxford University Press.
Beardsworth, T. (1977) Sense of Presence. Oxford: The Religious Experience Research Unit, Manchester College.
Bonjour, L. (1985) The Structure of Empirical Knowledge. Cambridge, ΜΑ: Harvard University Press.
Bonjour, L. (1999) The dialectic of foundationalism and coherentism. In J. Greco and E. Sosa (eds.), The Blackwell Guide to Epistemology, 117–42. Oxford, Blackwell.
Broad, C. D. (1953) Religion, Philosophy and Psychical Research. London: Routledge and Kegan Paul.
Cartwright, N. (1994) How we relate theory to observation. In P. Horwich (ed.), World Changes: Thomas Kuhn and the Nature of Science, 259. Cambridge, MA: MIT Press.
Chisholm, R. M. (1973) The Problem of the Criterion. Milwaukee, WI: Marquette University Press.
Copleston, F. (1982) Religion and the One. London: Search Press.
Dancy, J. (ed.) (1988) Perceptual Knowledge. Oxford: Oxford University Press.
Daniels, C. (1989) Experiencing God. Philosophy and Phenomenological Research XLIX, 487–99.
Davies, B. (1985) Thinking about God. London: Geoffrey Chapman.
Davis, C. (1989) The Evidential Force of Religious Experience. Oxford: Clarendon Press.
Donovan, P. (1979) Interpreting Religious Experience. London: Sheldon Press.
Dorr, D. (1978) Remove the Heart of Stone – Charismatic Renewal and the Experience of Grace. Dublin: Gill and Macmillan.
Draper, P. (1992) God and perceptual evidence. International Journal for Philosophy of Religion 32, 149–65.
Dupre, L. (1981) The Deeper Life. New York: Crossroad.
Edwards, R. (1972) Reason and Religion: An Introduction to the Philosophy of Religion. New York: Harcourt Brace Jovanovich.
Elgin, C. Z. (1996) Considered Judgment. Princeton, NJ: Princeton University Press.
Elgin, C. Z. (2005) Non-foundationalist epistemology: holism, coherence, and tenability. In M. Steup and E. Sosa (eds.), Contemporary Debates in Epistemology, 156–67. Oxford: Blackwell.
Everitt, N. (2004) The Non-Existence of God. London: Routledge.
Ewing, A. C. (1973) Value and Reality. London: George Allen and Unwin Ltd.
Farmer, H. H. (1935) The World and God. London: Nisbet and Co. Ltd.
Flew, A. (1966) God and Philosophy. London: Hutchinson and Co. Ltd.
Flew, A. and MacIntyre, A. (eds.) New Essays in Philosophical Theology. London: SCM.
Flew, A. and Varghese, R. A. (2007) There is a God: How the World’s Most Notorious Atheist Changed His Mind. New York: HarperCollins.
Forgie, J. W. (1998) The possibility of theistic experience. Religious Studies 34, 317–23.
Forman, R. K. C. (ed.) (1990) The Problem of Pure Consciousness – Mysticism and Philosophy. New York: Oxford University Press.
Frank, S. L. (1946) God with Us. London: Jonathan Cape Ltd.
Fumerton, R. (1992) Skepticism and reasoning to the best explanation. In E. Villanueva (ed.), Rationality in Epistemology, 149–70. Atascadero, CA: Ridgeview Publishing Co.
Fumerton, R. (2005) The challenge of refuting skepticism. In M. Steup and E. Sosa (eds.), Contemporary Debates in Epistemology, 85–97. Oxford: Blackwell.
Gale, R. (1991) On the Nature and Existence of God. Cambridge: Cambridge University Press.
Geivett, R. D. (2003) The evidential value of religious experience. In P. Copan and P. Moser (eds.),
The Rationality of Theism, 175–203. London: Routledge.
Gellman, J. (1994) Experiencing God’s infi nity. American Philosophical Quarterly 31: 1, 53–61.
Gellman, J. (1997) Experience of God and the Rationality of Theistic Belief. Ithaca, NY: Cornell University Press.
Gellman, J. (2001) Mystical Experience of God: A Philosophical Inquiry. Aldershot: Ashgate.
Ginet, C. (1975) Knowledge, Perception, and Memory. Dordrecht: Reidel.
Grice, H. P. (1988). The causal theory of perception. In J. Dancy (ed.), Perceptual Knowledge, 66–78.
Oxford: Oxford University Press.
Griffi th-Dickson, G. (2000) Human and Divine: An Introduction to the Philosophy of Religious Experience. London: Duckworth.
Griffi th-Dickson, G. (2005) The Philosophy of Religion. London: SCM Press.
Griffi ths, B. (1989) A New Vision of Reality – Western Science, Eastern Mysticismand Christian Faith.
London: Collins.
Gutting, G. (1982) Religious Belief and Religious Skepticism. Notre Dame, IN: University of Notre Dame Press.
Haack, S. (1994) Double-aspect foundherentism: a new theory of empirical justifi cation. In E. Sosa (ed.), Knowledge and Justifi cation, 729–44. Aldershot: Dartmouth.
Hallett, G. (2000) A Middle Way to God. Oxford: Oxford University Press.
Hamlyn, D. W. (1970) The Theory of Knowledge. London: Macmillan.
Harris, J. F. (2002) Analytic Philosophy of Religion. Dordrecht: Kluwer.
Harrison, J. (1999) God, Freedom and Immortality. Aldershot: Ashgate.
Hay, D. (1990) Religious Experience Today. London: Mowbray.
Hay, D. (1994) ‘The biology of God’: what is the current status of Hardy’s hypothesis? International Journal for the Psychology of Religion 4: 1, 1–23.
He, Guanghu. (1995) The root and fl ower of Chinese culture. Primordial Dao, 2: 29–56. (ԩܝⓀ,LJЁ
᭛࣪ⱘḍ㟛㢅Lj,ljॳ䘧NJѠ䔃, 1995ᑈ5᳜, ䷕29–56DŽ).
Hesse, M. (1981) Retrospect. In A. R. Peacocke (ed.), The Sciences and Theology in the Twentieth Century 281–96. Stocksfi eld: Oriel Press.
Hick, J. (2006) The New Frontier of Religion and Science: Religious Experience, Neuroscience and the Transcendent. Basingstoke, England and New York: Palgrave Macmillan.
Huemer, M. (2001) Skepticism and the Veil of Perception. Lanham, MD: Rowan and Littlefi eld.
Katz, S. T. (ed.) (1978) Mysticism and Philosophical Analysis. New York: Oxford University Press.
Kwan, K.-M. (2001) A critical appraisal of a non-realist philosophy of religion: an Asian perspective.
Philosophia Christi 3: 1, 225–35.
Kwan, K.-M. (2003) Is the critical trust approach to religious experience incompatible with religious particularism? A reply to Michael Martin and John Hick. Faith and Philosophy 20: 2, 152–69.
T H E A RG U M E N T F RO M R E L I G I O U S E X PE R I E N C E 551
Kwan, K.-M. (2004) Review of Jerome Gellman, Mystical Experience of God: A philosophical Inquiry.
Faith and Philosophy 21: 4, 553–60.
Kwan, K.-M. (2006a) Are religious beliefs human projections? In R. Pelly and P. Stuart (eds.), A Religious Atheist? Critical Essays on the Work of Lloyd Geering, 41–66. Dunedin, New Zealand: Otago University Press.
Kwan, K.-M. (2006b) Can religious experience provide justifi cation for belief in God? The debate in contemporary analytic philosophy. Philosophy Compass November.
Kwan, K.-M. (2006c) Moral arguments for the existence of God [addendum]. In D. Borchert (ed.), Encyclopedia of Philosophy, 2nd edn. Detroit, MI: Macmillan Reference USA.
Kwan, K.-M. and Han, S. (2007) The Quest for God in Chinese Culture and Contemporary China.
Paper presented in the Trinity Newman Lecture, July 16, 2007, Victoria University, Wellington, New Zealand.
Layman, C. S. (2007) Letters to Doubting Thomas: A Case for the Existence of God. New York: Oxford University Press.
Lewis, H. D. (1969) The Elusive Mind. London: George Allen and Unwin Ltd.
Lycan, W. (1988) Judgment and Justifi cation. Cambridge: Cambridge University Press.
Mackie, J. (1982) The Miracle of Theism. Oxford: Clarendon Press.
Maritain, J. (1966) Challenges and Renewals. Cleveland and New York: Meridian Books, The World Publishing Co.
Martin, C. B. (1959) Religious Belief. Ithaca: Cornell University Press.
Martin, M. (1986) The principle of credulity and religious experience. Religious Studies 22, 79–93.
Martin, M. (1990) Atheism: A Philosophical Justifi cation. Philadelphia: Temple University Press.
Mawson, T. J. (2005) Belief in God: An Introduction to the Philosophy of Religion. Oxford: Clarendon.
Miles, G. (2007) Science and Religious Experience: Are They Similar Forms of Knowledge? Sussex: Academic Press.
Moody, R. (1975) Life after life. New York: Bantam Books.
Newberg, A, D’Aquili, E., and Rause, V. (2001) Why God Won’t Go Away: Brain Science and the Biology of Belief. New York: Ballantine Books.
Oppy, G. (2006) Arguing about Gods. Cambridge: Cambridge University Press.
Otto, R. (1924) The Idea of the Holy. Oxford: Oxford University Press.
Owen, H. P. (1969) The Christian Knowledge of God. London: The Athleone Press.
Papineau, D. (1979) Theory and Meaning. Oxford: Clarendon Press.
Pike, N. (1992) Mystic Union – An Essay in the Phenomenology of Mysticism. Ithaca, NY: Cornell University Press.
Pullinger, J. (1980) Chasing the Dragon. London: Hodder and Stoughton.
Rescher, N. (1995) Satisfying Reason: Studies in the Theory of Knowledge. Dordrecht: Kluwer.
Rowe, W. (1982) Religious experience and the principle of credulity. International Journal for Philosophy of Religion 13, 85–92.
Rowley, H. H. (1956) The Faith of Israel. London: SCM Press.
Russell, B. (1948) Human Knowledge: Its Scope and Limits. London: George Allen and Unwin Ltd.
Schaefer, R. T. and Lamm, R. P. (1995) Sociology, 5th edn. New Yord: McGraw-Hill.
Schoen, E. L. (1990) The sensory presentation of divine infi nity. Faith and Philosophy 7: 1, 3–18.
Slote, M. (1970) Reason and Scepticism. London: George Allen and Unwin Ltd.
Smart, N. (1960) A Dialogue of Religions. London: SCM Press.
Swinburne, R. (1979) The Existence of God. Oxford: Clarendon Press.
Swinburne, R. (1981) The evidential value of religious experience, and Comments. In A. R. Peacocke (ed.), The Sciences and Theology in the Twentieth Century, 182–96, 303–4. Stocksfi eld: Oriel Press.
Swinburne, R. (1986) The Evolution of the Soul. Oxford: Clarendon Press.
Trueblood, E. (1957) Philosophy of Religion. Grand Rapids, MI: Baker.
Tsai, C. (1953) Queen of the Dark Chamber. Chicago: Moody Press.
Tsai, C. (2000) Queen of the Dark Chamber. Hong Kong: The Bellman House. (This is a Chinese book: 㫵㯛,ljᱫᅸПৢNJ, 佭␃: ᰼᯳ߎ⠜⼒,2000DŽ).
Van Cleve, J. (2005) Why coherence is not enough. In M. Steup and E. Sosa (eds.), Contemporary Debates in Epistemology, 168–80. Oxford: Blackwell.
Vardy, P. (1990) The Puzzle of God. London: Collins, Flame.
Vogel, J. (1998) Cartesian skepticism and the inference to the best explanation. In L. Martin Alcoff (ed.), Epistemology: The Big Questions, 352–9. Oxford: Blackwell.
Vogel, J. (2005) The refutation of skepticism. In M. Steup and E. Sosa (eds.), Contemporary Debates in Epistemology, 73–84. Oxford: Blackwell.
Von Leyden, W. (1961) Remembering: A Philosophical Problem. New York: Philosophical Library.
Wainwright, W. (1981) Mysticism. Brighton: The Harvester Press.
Wall, G. (1995) Religious Experience and Religious Belief. Lanham, MD: University Press of America.
Waterhouse, E. S. (1923) The Philosophy of Religious Experience. London: The Epworth Press.
Wynn, M. R. (2005) Emotional Experience and Religious Understanding: Integrating Perception, Conception and Feeling. Cambridge: Cambridge University Press.
Xi, Y. (1990) Song of the Valley. Petaluma, CA: Chinese Christian Mission. (This is a Chinese book: Ꮰ䳆,ljᑑ䈋П℠NJ, ЁֵᕦԜ䘧᳗, 1990DŽ).
Yandell, K. (1984) Christianity and Philosophy. Grand Rapids, MI: Eerdmans.
Yandell, K. (1993) The Epistemology of Religious Experience. Cambridge: Cambridge University Press.
Yandell, K. (1999) Philosophy of Religion: A Contemporary Introduction. Cambridge: Cambridge University Press; London: Routledge.
The Ontological Argument
ROBERT E. MAYDOLE
Ontological arguments are deductive arguments for the existence of God from general metaphysical principles and other assumptions about the nature or essence of God. There have been three very signifi cant developments in the history of ontological arguments. The fi rst is the ontological argument developed by St Anselm of Canterbury in the eleventh century. The second is the argument sketched by Descartes in the late seventeenth century and completed by Leibniz in the early eighteenth century. And the third development consists of the numerous ontological arguments of the twentieth century that explicitly utilize modal logic, particularly those of Malcolm, Hartshorne, Plantinga, and Gödel. My chief aim in this chapter is to logically evaluate logical reconstructions of each of these six arguments. I shall also present and logically discuss two of my own explicitly modal ontological arguments.1
The logical evaluation of a logical reconstruction of an argument often requires that we explicitly identify assumptions that are only implicit in the author’s original presentation of the argument. And in some cases, it might involve the inclusion of “plausible” philosophical principles that are consistent with the author’s worldview, principles that strengthen the argument if we include them among the premises of the reconstruction. My modus operandi will be to make each of the arguments as strong as possible before critically evaluating them. Even though I shall try to remain reasonably faithful to the intent of the original author of each argument, my main objective will be logical instead of historical.
A good deductive argument should be valid and have true premises. And if it is to be convincing it should not beg the question. Ontological arguments are frequently the target of parodies, perhaps more so than any other argument in philosophy. So in addition to checking the arguments we discuss for validity, truth, and question begging, I shall also test their vulnerability to being refuted by some of the well-known parodies in the philosophical literature. Like many other philosophical arguments of note, ontological arguments stand or fall on the acceptability of some very high-level and well-entrenched principles of metaphysics and logic. And, as we shall see, some ontological arguments are logically much stronger than what fi rst meets the eye.
1. It would not be possible for me to discuss all or even most of the ontological arguments in the history of philosophy in this chapter. One of the most comprehensive and fairest discussions of many of these is by Graham Oppy (1995).
The Blackwell Companion to Natural Theology Edited William Lane Craig and J. P. Moreland
© 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17657-6
1a. The Validity of Anselm’s Ontological Argument
Anselm expresses his ontological argument of Proslogium, Chapter ll as follows: Hence, even the fool is convinced that something exists in the understanding, at least, than which nothing greater can be conceived. For, when he hears of this, he understands it. And whatever is understood, exists in the understanding. And assuredly that, than which nothing greater can be conceived, cannot exist in the understanding alone. For, suppose it exists in the understanding alone: then it can be conceived to exist in reality; which is greater.
Therefore, if that, than which nothing greater can be conceived, exists in the understanding alone, the very being, than which nothing greater can be conceived, is one, than which a greater can be conceived. But obviously this is impossible. Hence, there is no doubt that there exists a being than which nothing greater can be conceived, and it exists both in the understanding and in reality. (1962 , p. 8)
There are a few key ideas in this passage that require our attention before we present the main argument. First, Anselm understands the predicate “is greater than” to mean the same as “objectively better or more worthy than.” It was commonplace for philosophers of the Middle Ages to order things into a great chain of being according to the degree to which they possess great-making properties, such as wisdom, power, goodness, and completeness, and existence-in-reality. What Anselm argues here is that God is an upper bound to the great chain of being. He makes this point very explicitly in his Monologion: Furthermore, if one considers the nature of things, one cannot help realizing that they are not all of equal value, but differ by degrees. For the nature of a horse is better than that of a tree, and that of a human more excellent than that of a horse . . . It is undeniable that some natures can be better than others. None the less reason argues that there is some nature that so overtops the others that it is inferior to none. (1998, p. 14)
Second, the conclusion of Anselm’s argument is that something than which nothing greater can be conceived has the property of existence-in-reality. Now if something has the property of existence-in-reality, then it exists, period. But the property of existence-in-reality is only one kind of existence. The property of existence-in-the-understanding is another. Notions, concepts, ideas, thoughts, beliefs, and so on are the kinds of things that do or might have existence-in-the-understanding. And tables, persons, angels, numbers, forces, and so on, and God, are the kinds of things that do or might have existence-in-reality. We can identify the realm of things that have existence-in-the-understanding with the totality of mental things that actually exist in minds; and we can identify the realm of things that have existence-in-reality with the totality of nonmental things that actually exist in the world.2 It is crucial, however, not to confl ate existence-in-reality with existence 2. Deane translates the phrase “esse in intellectu” as “exists in the understanding” (Anselm 1962), and Charlesworth translates it as “exists in the mind” (Anselm 1998). But McGill translates it as “stands in relation to the understanding” (Anselm 1967) because, he contends, “exists in the understanding” and “exists in the mind” both have a Cartesian connotation of existing in a substantial place of some sort with in which certain mental phenomena occur, which is not, he says, what Anselm intended. “For him the intellect is the intentional phase of human being [ sic]. It is man’s active openness towards reality, toward real entities through “understanding” and towards possible entities through “conceiving”. “It is never a self-enclosed place within which certain phenomena
O N TO LO G I C A L
A RG U M E N T 555
generally. For Anselm, mental things that have existence-in-the-understanding exist just as much as nonmental things that have existence-in-reality, but in a different way.
Third, it is inconceivable that one and the same thing could have both existence-in-reality and existence-in-the-understanding. Things that have existence-in-reality are very different kinds of things from things that have existence-in-the-understanding. A table, for example, is different from the concept or idea of a table. Likewise, it impossible to think of God, even qua pure spirit, as having existence-in-the-understanding, even if God fails to have existence-in-reality or even if “the fool is convinced that something exists in the understanding, at least, than which nothing greater can be conceived.” Either Anselm has been mistranslated or he misspoke and should have said that even the fool is convinced that the concept of something than which nothing greater can be conceived has existence-in-the-understanding; and instead of saying “whatever is understood, exists in the understanding,” he should have said “the concept of whatever is understood has existence-in-the-understanding” and so on. That said, a neo-Platonist such as Anselm might hold that things that have existence-in-reality and things that have existence-in-the-understanding might share many or most of the same properties, including the property of it being inconceivable for something to be greater.
Fourth, it is clear that for Anselm the phrase “that than which nothing greater can be conceived” should be understood as a defi nite description which refers to the one and only one thing than which nothing greater can be conceived, God, even if there is no such being.
Similarly, the proposition “the concept of whatever is understood has existence-in-the-understanding” should be understood in the context of his argument as saying that the concept of whatever a defi nite description that is understood refers to has existence-in-the-understanding. This is consistent with Anselm’s belief that even though we cannot fully and adequately understand God, we do at least have a partial concept of God as a being than which nothing greater can be conceived, and thereby we can understand our referential talk about God. The presupposition is that some referring singular terms and defi nite descriptions could be free of existential import, and quantifi ers should be allowed to range over possibilia (Girle 2003, chap. 4). Otherwise, some referential terms that refer to nonmental things, such as “God” and “the being than which nothing greater can be conceived,” would have to refer to mental things that have existence-in-the-understanding, which makes no sense; or those referential terms would have to have to refer to things that have existence-in-reality, which would make the Anselmian ontological argument beg the question.
Finally, in order to be able to test our Anselmian argument for validity, it will be useful to present it in standard form and to express it in the language of quantifi cation theory.
We shall use the following lexicon:
Ux =df x is understood
Sy =df the concept of y exists-in-the-understanding
Ex =df x exists-in-reality
Gxy =df x is greater than y
Fxy =df x refers to y
occur, such as ideas and inferences . . .” (McGill 1967, p. 82). The fi rst two translations are consistent with the assumptions and presuppositions of this chapter. McGill’s is not. Anselm was both a Christian and a neo-Platonist, and he had to be committed as such to the existence of minds or souls qua substantial entities within which mental phenomena occur.
Dx =df x is a defi nite description
d =df the defi nite description “(ɿx) ~©(∃y)Gyx”
g =df (ɿx)~©(∃y)Gyx
P(Y) =df Y is a great-making property
©. . . =df it is conceivable that . . .3
Here then is our logical reconstruction of Anselm’s ontological argument:
A1 The
defi nite description “that than which it is not conceivable for something to be greater” is understood. ( Premise)
(Dd & Ud)
A2 “That than which it is not conceivable for something to be greater” refers to that than which it is not conceivable for something to be greater. ( Premise)
Fdg
A3 The
concept of whatever a defi nite description that is understood refers to has existence-in-the-understanding. ( Premise)
(x)(y)((Dx & Fxy & Ux) ⊃ Sy)
A4 It is conceivable that something is greater than anything that lacks a great-making property that it conceivably has. ( Premise)
(x )( )[( ( )& ∼ & © ) ⊃ ©(∃ )
]
1
Y
P Y
Yx1
Yx1
x2 Gx2x1
A5 Existence-in-reality is a great making property. ( Premise)
P(E)
A6 Anything the concept of which has existence-in-the-understanding conceivably has existence-in-reality. ( Premise)
(x)(Sx ⊃ ©Ex)
A7 It is not conceivable that something is greater than that than which it is not conceivable for something to be greater. ( Premise)
∼©( ∃y)Gyg
Therefore,
A8 That than which it is not conceivable for something to be greater exists-in-reality.
Eg
The following deduction proves that this argument is valid:
Deduction4
1. Dd & Ud
pr
2. Fdg
pr
3. The conceivability operator need not be made explicit for this argument, since the deduction shows that the argument is valid in nonmodal fi rst-order quantifi cation theory. However, I include it because it will be needed later, and it also improves readability.
4. See Appendix 1 for the rules of inference and so on of the logic used in this chapter.
O N TO LO G I C A L
A RG U M E N T 557
3. (x)(y)((Dx & Fxy & Ux) ⊃ Sy)
pr
4. (x1)(Y)[(P(Y) & ~Yx1 & ©Yx1) ⊃ ©(∃x2)Gx2x1] pr
5. P(E)
pr
6. (x)(Sx ⊃ ©Ex)
pr
7. ~©(∃y)Gyg pr
8. Fdg & ~©(∃y)Gyg
2, 7 Conj
9.
(∃x)[~©(∃y)Gyx & (z)(~©(∃y)Gyx ⊃ z=x)
8, theory of descriptions
& (Fdx & ~©(∃y)Gyx)]
10.
~©(∃y)Gyν & (z)(~©(∃y)Gyz ⊃ z=ν)
9, EI
&
(Fdν & ~©(∃y)Gyν)
11. ~©(∃y)Gyν 10,
Simp
12. Fdν 10,
Simp
13. (P(E) & ~Eν & ©Eν) ⊃ ©(∃x
ν
2)Gx2
4 UI
14. (Dd & Fdν & Ud) ⊃ Sν 3
UI
15. (Dd & Fdν & Ud)
1, 12, Simp, Conj
16. Sν
14, 15 MP
17. Sν ⊃ ©Eν 6,
UI
18. ©Eν
16, 17 MP
19. ~(P(E) & ~Eν & ©Eν)
13, 11 MT
20. ~((P(E) & ©Eν) & ~Eν) 19
Com,
Assoc
21. ~(P(E) & ©Eν) ∨ ~~Eν) 20,
DeM
22. P(E) & ©Eν
5, 18 Conj
23. Eν
21, 22, DS, DN
24. ~©(∃y)Gyν & (z)(~©(∃y)Gyx) ⊃ z=ν) 10
Simp
25. ~©(∃y)Gyν & (z)(~©(∃y)Gyx) ⊃ z=ν) & Eν
23, 24 Conj
26. (∃x)[~©(∃y)Gyx & (z)(~©(∃y)Gyx) ⊃ z=x) & Ex]
25 EG
27. Eg
26, theory of descriptions
1b. The Truth of the Anselmian Premises
The fi rst conjunct of A1 is true by defi nition. The second is introspectively true. For it does seem as though we understand the phrase “that than which it is not possible for something to be greater.” As Anselm would say, “many people appear to understand it when they hear it – even the fool.” This assumes, of course, that the relational predicate “is greater than” is meaningful and understood by those who claim that they understand it when they use it and hear it. Such an assumption could be challenged, although not lightly and not without good reason. Perhaps the absence of a plausible theory of great-making properties would constitute a challenge to the meaningfulness of “is greater than.” However, the burden of proof of a claim that a word or phrase is meaningless must always fall on the challenger, especially when the word appears to be used with understanding by a great many people.
And the predicate “is greater than” is just such a term. Indeed, many philosophers from Plato to the present, including most neo-Platonists, scholastics, and rationalists, believe that the things of the world can be ordered in terms of both ontological and/or normative greatness, the absence of a nearly complete and coherent theory of great-making properties notwithstanding.
A2 also is analytically true, given the presupposition that in order for the defi nite description “that than which it is not conceivable for something to be greater” to genuinely refer, that than which it is not conceivable for something to be greater must be a member of the domain of some possible world.
Anselm constructs an insightful, but somewhat muddled, argument for the bogus proposition “whatever is understood exists in the understanding” that we might be able to use in support of the more plausible A3.
. . . When the fool hears mentioned a being than which a greater is inconceivable, he understands what he hears . . . Moreover . . . if this being is understood, it is in the understanding
. . . For as what is conceived, is conceived by conception, and what is conceived by conception, as it is conceived, so is in conception; so what is understood, is understood by understanding, and what is understood by understanding, as it is understood, so is in the understanding. What can be clearer than this? (1962, p. 157)
This passage suggests either a deductive or an analogical argument for A3.
The deductive argument is:
1 Concepts are in that which conceives of concepts.
2 Whatever is in that which conceives of concepts has existence-in-the-understanding.
3 Therefore,
the
concept of whatever a defi nite description that is understood refers to has existence-in-the-understanding.
And the analogical argument is identical to the deductive argument, save for 2* in place of 2.
2* Having the property of existence-the-understanding is like having the property of being in that which conceives of concepts.
The deductive argument is clearly valid, and the analogical argument certainly appears to be inductively strong, with a degree of inductive strength that is proportional to the degree of likeness between conceiving and understanding. The common fi rst premise is analytic.
Premise 2 is true if the degree of likeness between conceiving and understanding is 100
percent; and 2* is true if that degree of likeness is high, which it surely is, even if less than 100 percent.
A4 is only implicit in Anselm’s Proslogium. Yet it is so intuitively obvious that I know of nothing more intuitive and general from which we might infer it. We might, of course, classify it as analytic and say that it is built into the very meaning of being a great-making property that it potentially increases greatness when instantiated. But such a move would presuppose a better idea of greatness than I have been able to give. This is not to say that we do not know what greatness is, or that we do not know that some principles of greatness are true. Rather, it is that a good theory of greatness has yet to be constructed, as far as I know. And were such a theory to be developed, it would be tempting to view A4 as an axiom or fi rst principle.
Let us now turn to A5, arguably the centerpiece of Anselm’s ontological argument, but a proposition for which he never appears to argue. Note fi rst that A5 does not say that existence per se is a great-making property. It says rather that existence-in-reality is
O N TO LO G I C A L
A RG U M E N T 559
great-making. There is a big difference. If existence is a property, it applies, for Anselm, to both the contents of minds (existence-in-the-understanding) and the contents of worlds.
Existence-in-reality is narrower than existence, for there are things that exist (in-the-understanding) that do not exist (in-reality).
One interesting characteristic seemingly possessed by things that have existence-in-reality but not by things that only have existence-in-the-understanding is ontological completeness.5 Something is ontologically complete if and only if every property or its negation is a member of the set of all its properties. This computer on which I am writing is ontologically complete because it possesses every possible property or its negation, including the property or its negation of containing a hydrogen atom that was formed a split second after the Big Bang. The set of properties possessed by this computer is also infi nite.
But my rather limited idea of this computer, howsoever accurate and robust, is surely fi nite.
Nor does that idea include either the property or its negation of containing a hydrogen atom that was formed a split second after the Big Bang. And even if I now were to amend my idea of this computer to include such a property or its negation, there would always some other property and its negation, neither of which I attribute to my idea of the computer.
We can now formulate a plausible argument sketch for A5 that might appeal to an Anselmian:
1 Things that have existence-in-reality are ontologically complete. ( Premise) 2 The property of being ontologically complete has the property of being great-making.
( Premise)
3 For every property X and Y, X has Y if and only if everything that has X has a property that has Y. ( Premise)
4 The property of being ontologically complete has the property of being great-making if and only if everything that has the property of being ontologically complete has a property that has the property being great-making. ( 3, UI)
5 Everything that has the property of being ontologically complete has a property that has the property being great-making. ( 2,4, Equiv, Simp, MP)
6 Hence, everything that has the property of existence-in-reality has a property that has the property of being great-making. ( 1, 5, UI, HS, UG)
7 The property of existence-in-reality has the property of being great-making if and only if everything that has the property of existence-in-reality has a property that has the property of being great-making. ( 3, UI)
8 Hence, the property of existence-in-reality has the property of being great-making. ( 6, 7, Equiv, Simp, MP)
A6 is the Anselmian cognate of the proposition that it is possible for God to exist, a premise common to many ontological arguments. One way of showing that A6 is true would be via ultrarealism. As a neo-Platonist, Anselm would have thought of the ontology of the world as partitioned in three ways: particulars, Forms, and minds. The Forms are instantiated either as the properties of the particulars or as universal ideas in minds. While Plato himself believed at times that the Forms had a greater degree of reality than the 5. This is not true of things that exist in the mind of God, all of which are ontologically complete.
particulars, which he relegates in The Republic to the shadows, the Platonist must still think of particulars together with the Forms as real existents. Anselm could therefore comfortably think of the world as divided into things that have existence-in-reality and things that have existence-in-minds. Forms and particulars exist-in-reality, and universal ideas existin-minds. We can therefore say that for an Anselmian, if the concept of something has existence-in-the-understanding, then a universal idea of it exists-in-a-mind. Or we can say that if the concept of something has existence-in-the-understanding, then a it is possible that a universal idea of it exists-in-a-mind.
Some medieval neo-Platonists, the so-called ultrarealists, believed that the order of thought (the universals) and the order of extramental particulars correspond exactly.6 To each particular real chair, for example, there corresponds a universal idea of that chair, a so-called mental chair. The real chair and the mental chair instantiate exactly the same Forms, but in different mediums. Real chairs instantiate the Forms in matter. Mental chairs instantiate the Forms in-mind. Otherwise, real chairs and mental chairs have exactly the same nonexistential properties.7
It strikes me as a bit far-fetched to think that particulars and their corresponding mental replicas share exactly the same nonexistential properties. There is more to reality than what meets the mind’s eye, and vice versa. But because conceivability generally outruns actuality, it is less far-fetched to think that for each conceivable mental replica of something, it is conceivable that there exists another real particular that has each and every nonexistential property of the replica. Call this “weak ultrarealism.”
We can formulate two similar neo-Platonic and ultrarealistic arguments for A6. Let
“Mx” =df “a mental replica of x exists-in-a-mind.”
Argument 1
1 (x)(Mx
⊃ ©Ex)
2 (x)(Sx
⊃ Mx)
(x)(Sx
⊃ ©Ex)
Argument 2
1 (x)(Mx
⊃ ©Ex)
2 (x)(Sx
⊃ ©Mx)
(x)(Sx
⊃ ©Ex)
Argument 1 is valid in fi rst-order quantifi cation theory. Argument 2 is valid in an S4
modal-like extension of fi rst-order quantifi cation theory that licenses inferences from conceivable conceivability to conceivability. The same fi rst premise of both arguments is quite weak and diffi cult to challenge. The second premise of Argument 1 could easily be challenged by a non-Platonist and a nondualist. Yet it would be harder to challenge the second premise of Argument 2, since it too makes a very weak claim.
A7 appears to be self-evident. Yet we can show that it is true in a couple of ways. First, the following argument is valid and both premises are logical truths:
6. Some historians of philosophy believe that Anselm was an ultrarealist (Copleston 1961, p. 35).
7. A nonexistential property is a property other than existence-in-reality and other than existence-in-the-mind.
O N TO LO G I C A L
A RG U M E N T 561
1 (Y)(z)[z=(ɿx)Yx ⊃ Yz]
2 g=(ɿx) ~©(∃y)Gyx
~©(∃y)Gyg
Second, if we use Russell’s Theory of Descriptions to eliminate the defi nite description
“(ɿx) ~©(∃y)Gyx” from A7, we get:
A7a.
∼©(∃x)[∼©(∃y)Gyx & (z)(∼©(∃y)Gyz ⊃ z= x)& ( y
∃ )Gyx].8
If we then modestly assume that conceivability is equivalent to possibility, A7b becomes: A7b.
∼◊( ∃x)[∼◊(∃y)Gyx & (z)(∼◊(∃y)Gyz ⊃ z= x)& ( y
∃ )Gyx].9
But A7b is logically true in even the weakest of modal logics. Therefore, A7 is true.
1c. On Whether Anselm’s Ontological Argument
Begs the Question
An argument begs the question just in case belief in the truth of the conclusion is included among the reasons for asserting the truth of the premises.10 Some sound arguments sometimes beg the question. Consider the following valid argument:
Either 1 + 1 = 3 or God exists.
Not 1 + 1 = 3.
Hence, God exists.
While it is true that theists will believe that it is sound and nontheists believe that it is not, neither belief makes it so. Assume that no one believes that 1 + 1 = 3. If some theists believe that the argument is sound because they believe qua theists that the second disjunct of the fi rst premise is true, then they beg the question. Likewise, if some nontheists believe that the argument is not sound because they believe qua nontheists that the second disjunct is false, then they too beg the question. Yet the argument would not beg the question if its proponent believed that the fi rst premise is true for reasons that do not include the proposition that God exists. It would only be pointless.
William L. Rowe argues that Anselm’s ontological argument begs the question by
granting what it tries to prove (2001, pp. 39–41). According to Rowe, Anselm’s ontological argument boils down to one that defi nes God as a greatest possible being, and also counts 8. The description “(ɿx)~©(∃y)Gyx” occurs within the scope of “~©” and, clearly, has a secondary occurrence in A7. If we were to construe it as having a primary occurrence, then A7 would expand to “(∃x)[~©(∃y)Gyx & (z)(~©(∃y)Gyz ⊃ z=x) & ~©(∃y)Gyx],” and Anselm’s argument would thereby beg the question.
9. I realize that equating conceivability with possibility is controversial. But it is an issue that is beyond the scope of this chapter.
10. Strictly speaking, arguments do not beg the questions. Arguers do. Thus, an argument might beg the question for one person and not for another.
the property of existence-in-reality as great-making.11 These two things, he correctly argues, imply that nothing that fails to exist-in-reality can be a greatest possible being; but they do not alone imply that a greatest possible being actually exists-in-reality. However, if we also assume that a greatest possible being possibly exists-in-reality, we can then infer that a greatest possible being actually exists-in-reality because no such possible being could fail to exist-in-reality and still be a greatest possible being, given that the property of existence-in-reality is great-making. This means, Rowe then concludes, that the assumption that a greatest possible being possibly exists-in-reality is “virtually equivalent” to the concluding proposition that it actually exists-in-reality. “In granting that Anselm’s God is a possible thing, we are in fact granting that it actually exists . . .
the argument begs the question: it assumes the point it is trying to prove” (Rowe 2001, p. 41).
Rowe is wrong on two counts. First, he equivocates on the word “grant.” It can mean either “assume” or “implies.” We assume (grant) the premises of an argument. And in granting these premises, we are in fact granting (implying) its conclusion if the argument is valid. Surely, the fact that the premises of an argument imply its conclusion does not mean that the argument begs the question. Second, the proposition that a greatest possible being possibly exists-in-reality is not at all equivalent to the proposition that it actually exists-in-reality, Rowe’s use of the hedge word “virtually” notwithstanding. Indeed, the former does not even imply the latter, unless we assume that the other premises of Anselm’s argument are logical truths. But one of the other premises is that the property of existence-in-reality is great-making, which is not a logical truth. Moreover, even if the proposition that a greatest possible being possibly exists-in-reality did imply the proposition that it actually exists-in-reality, the argument would beg the question only if the latter were given as a reason for believing the former.
The upshot is that Rowe’s analysis does not show that his distilled version of Anselm’s ontological argument begs the question. Nor is there any reason to think that the reasons I have given for believing the premises of the expanded version of the argument I have presented in sections 1a and 1b of this chapter include the proposition that the greatest conceivable being has the property of existence-in-reality. Therefore, we can confi dently believe that it does not beg the question.
1d. On Parodies
A parody of an argument is a structurally similar argument with an absurd conclusion.
There are two ways parodies can refute what they parody. First, if the parody has true premises and the same logical form as the argument parodied, then the argument parodied 11. Rowe’s distilled version of Anselm’s argument can be expressed schematically, thus: 1 Some possible object exemplifi es the concept of God.
2 No object that fails to exist-in-reality could exemplify the concept of God. (Because God is defi ned as a being than which none greater is possible, and it assumed that the property of existence-in-reality is great-making.)
3 Every possible object either exists-in-reality or does not.
4 Therefore, God exemplifi es the property of existence-in-reality.
O N TO LO G I C A L
A RG U M E N T 563
must be invalid. Second, if both the parody and the argument parodied are valid, and the premises of the parody are at least as justifi able as the premises of the argument parodied, then the parody refutes the argument parodied because the argument parodied will also have to have at least one unjustifi able premise, and it will thereby fail to support its conclusion. In other words, if the premises of a valid parody are as justifi able as premises of the argument parodied, then it cannot be rational to believe that the premises of the argument parodied are true, for then it would be rational also to believe that the absurd conclusion of the parody is true. Conversely, if either the parody is invalid or some of its premises are arguably less justifi able than the respective corresponding premises of the argument parodied, then the parody per se fails to refute the argument parodied, and the argument parodied might well be sound.
Perhaps the most famous parody in the history of philosophy is the one formulated against Anselm’s ontological argument by Gaunilo, a contemporary of Anselm, who reasoned that one could infer the absurd conclusion that a greatest conceivable island exists-in-reality from premises structurally similar to those of Anselm’s argument. Let us add the following to our lexicon:
Ix =df x is an island
i =df (ɿx) ~©(∃y)(Iy & Gyx)
h =df the defi nite description “(ɿx)(Ix & ~◊(∃y)(Ix & Gyx))”
Then Gaunilo’s parody is this:
G1 Dh & Uh
pr
G2 Fhi
pr
G3 (x)(y)((Dx & Fxy & Ux) ⊃ Sy)
pr
G4 (x1)(Y)[(P(Y) & ~Yx1 & ©Yx1) ⊃ ©(∃x2)Gx2x1] pr
G5 P(E)
pr
G6 (x)(Sx
⊃ ©Ex)
pr
G7 ~©(∃y)Gyi pr
Therefore,
G8 Ei
This parody is valid but not sound because G7 is false. So it fails to refute Anselm’s argument. Replace G7 with G7a.
G7a.
∼©( x
∃ 2)(lx2 & Gx2i)
Then the premises of the new parody are true if Anselm’s premises are true; but the parody is not valid. So it fails to refute Anselm’s argument. Replace G7 with G7a, and replace G4
with G4a.
G4a. (x )( )[( ( ) & ∼
1 & ©
1 ) ⊃ ©(∃x 2 )(
2 &
2
1 )]
1
Y
P Y
Yx
Yx
lx
Gx x
This parody is valid but not sound because G4a is false.
Let “Lx2x1” be short for “x2 is exactly like x1 except for having Y in place of the negation of Y.” Replace G4 with G4b, and replace G7 with G7a.
G4b. (x )( )[( ( ) & ∼
1 & ©
1 ) ⊃ ©(∃x 2 )(
2
1 &
2
1 )]
1
Y
P Y
Yx
Yx
Lx x
Gx x
.
Since “©(∃x2)(Lx2i & Gx2i)” entails “©(∃x2)(Ix2 & Gx2i),” the resulting parody is valid.
Or let “Kx2x1” be short for “x2 is the same kind of thing as x1.” Replace G4 with G4c, and replace G7 with G7a.
G4c. (x )( )[( ( ) & ∼
1 & ©
1 ) ⊃ ©(∃x 2 )(
2
1 &
2
1 )]
1
Y
P Y
Yx
Yx
Kx x
Gx x
.
Since “©(∃x2)(Kx2i & Gx2i)” entails “©(∃x2)(Ix2 & Gx2i),” this resulting parody is also valid.
But do the last two parodies refute Anselm’s ontological argument? While it is certainly true that the addition of a great-making property to anything that conceivably has that property would result in something which would conceivably be greater than the fi rst thing would be without that property, there is no guarantee I can think of for believing that the thing that would result from the addition of a great-making property would be exactly like the fi rst thing sans that property, or even the same kind of thing as the fi rst thing is without that property. In other words, the addition of a great-making property might change the nature of the thing it is added to. Take the property of existence-in-reality as an example.
Things that exist-in-reality are very different from corresponding things that exist-in-the-understanding. A real table is different than the concept of a table, and a real tree is from the concept of a tree, and so on. So it is far from obvious that either G4b or G4c is true.
But A4 is true intuitively. Consequently, neither of these parodies refutes Anselm’s argument.
Oppy suggests that Anselm’s ontological argument can be successfully refuted by parodies that purport to establish the existence of different kinds of devils:
. . . Consider the formula ‘a being than which no worse can be conceived’. It seems that it would be worse if a very bad being existed both in the understanding and in reality than if it merely existed in the understanding. Consequently, it seems that if the Anselmian formula is understood as ‘a being than which no better can be conceived’ – then the Anselmian argument can be successfully parodied using this formula. (1995 , p. 181)
Let us add the following notations to our growing lexicon:
Vyx =df y is more evil than x
e =df (ɿx) ~©(∃y)Vyx
j =df the defi nite description “(ɿx)~◊(∃y)Vyx”
Now replace “G” by “V,” “g” by “e,” and “d” by “j” in our reconstruction of Anselm’s ontological argument. Assuming that proposition “Ee” is absurd, Oppy’s putative devils parody is the result. Does it refute Anselm? Answer: only if premises O4 and O7 are arguably just as justifi able as A4 and A7, respectively.
O4. (x )( )[( ( ) & ∼
1 & ©
1 ) ⊃ ©(∃x 2 )
2
1 ]
1
Y
P Y
Yx
Yx
Vx x
O7.
∼©( y
∃ )Vye
O N TO LO G I C A L
A RG U M E N T 565
Oppy suggests that O4 is true because an evil that exists-in-reality is more evil than it would be if it only existed-in-the-understanding. If the property of existence-in-reality is great-making, as it is assumed to be in the antecedent of O4, then an evil that exists-in-reality cannot be worse than any evil that does not. The addition of great-making properties make things better, not worse. So it is questionable that O4 is true, at least for the reason given by Oppy. Suppose, however, that an evil that exists-in-reality could be worse than one that does not. Then proposition “©(∃y)Vye” would presumably be true, and O7 false. The conclusion is that this parody also fails to refute Anselm’s ontological argument.
2a. The Validity of the Ontological Argument of Descartes
and Leibniz
Descartes expresses his ontological argument in “Meditation V,” thus: It is certain that I no less fi nd the idea of God, that is to say, the idea of a supremely perfect Being, in me, than that of any fi gure or number whatever it is; and I do not know any less clearly and distinctly that an [actual and] eternal existence pertains to this nature . . .
I clearly see that existence can no more be separated from the essence of God than can the idea of its three angles equal to two right angles be separated from the idea of a [rectilinear]
triangle.
. . . from the fact that I cannot conceive of God without existence, it follows that existence is inseparable from Him, and hence that He really exists; not that my thought can bring this to pass, or impose any necessity on things, but on the contrary, because the necessity which lies in the thing itself, i.e., the necessity of the existence of God determines me to think this way . . . (1952, pp. 93–4)
He expresses it more clearly and succinctly in his “Arguments . . . in Geometrical Fashion”: To say that something is contained in the nature of a concept of anything is the same as to say that it is true of that thing. But necessary existence is contained in the concept of God. Hence it is true to affi rm of God that necessary existence exists in Him, or that God Himself exists.
(1952, p. 132)
There are two key ideas in these passages. First, while Descartes says in “Meditation V” that existence is contained in the concept or essence of a supremely perfect being, it is clear from the context, and from the “Arguments . . . in Geometrical Fashion” that he really meant to say (or should have said) that God is a necessary being, where a necessary being is one that exists only if it exists necessarily.12 Moreover, it would (should) have been obvious to him that the categorical property of existence (or nonexistence) cannot be included in the concept of anything without ultimately begging the question of its very existence (or nonexistence).13 Thus, for Descartes, it is really the conditional property of existing necessarily 12. The property of existence for Descartes is not split into existence-in-reality and existence-in-the understanding, as it is for Anselm.
13. We might call this the Principle of Existential Noninclusion. This principle does not preclude asserting that a supremely perfect being exists, only that existence must not be included in its concept or essence.
if existing at all, not the categorical property of existence per se (and not even the categorical property of having the property of existence necessarily), that is included in the concept or essence of a supremely perfect being. In other words, whether God exists or does not exist, He is not the kind of being who can exist contingently. Relatedly, a supremely perfect being possesses supremity necessarily.
Second, we need to clearly understand what is implied by the proposition that a certain property is contained in the concept or essence of something.14 When we say, to use Descartes’ own example, that the property of having its three angles equal to two right angles is contained in the concept or essence of a triangle, we imply that every triangle has the property of having its three angles equal to two right angles. In general, if property Y is contained in the concept or essence of something of kind X, then everything that is an X
is a Y.
We are now in a position to logically reconstruct Descartes’ ontological argument.
Suppose
Rx =df x is supremely perfect
Nx =df (Ex ⊃ Ex)
C(Y, X) =df Y is included in the concept or essence of an X.
It would appear from the said quotations that Descartes’ argument might be as follows: D1 For every X and Y, if the property of being a Y is contained in the concept or essence of being an X, then necessarily everything that is an X is a Y.
(X)(Y)(C(yˆ[Yy], yˆ[Xy]) ⊃ (
z)(Xz ⊃ Yz)
D2 The property of necessarily existing if existing at all is contained in the concept or essence of a supremely perfect being.
C(yˆ[Ny], yˆ[Ry])
Therefore,
DC A supremely perfect being exists.
(∃x)(Rx & Ex)
It should at once be obvious, however, that this is an invalid argument. The only thing that relevantly follows from D1 and D2 is that everything that is supremely perfect necessarily exists if exists. But, if we add D3 and D4 as premises, then the argument is valid in S5
quantifi cational modal logic.15
D3 It is possible that a supremely perfect being exists.
◊(∃x)(Rx & Ex)
14. Descartes vacillates between talking about the concept of things and talking about their essence. While concepts are generally thought of as subjective, and essences as objective, both will work in his ontological argument.
15. Although Descartes must have had sharp modal insights in order to be able to see that his enthymeme was valid, he could not have been aware of the exact modal principles used, since modal logic was not developed formally until the twentieth century.
O N TO LO G I C A L
A RG U M E N T 567
D4 Necessarily, supremely perfect beings are necessarily supremely perfect.
(x)(Rx ⊃ Rx)
Then the following deduction proves that the argument D1, D2, D3, D4/DC is valid: 1 (X)(Y)(C(ˆy[Yy], ˆy[Xy]) ⊃ (z)(Xz ⊃ Yz)
pr
2 C(ˆy[Ny], ˆy[Ry]) pr
3 ◊(∃x)(Rx & Ex)
pr
4 (x)(Rx ⊃ Rx) pr
5 C(ˆy[Ny], ˆy[Ry]) ⊃ (x)(Rx ⊃ Nx)
1 UI
6 (x)(Rx ⊃ Nx)
2, 5 MP
7 (x)(Rx ⊃ (Ex ⊃ Ex)) 6,
df
“N”
8
((x)(Rx ⊃ (Ex ⊃ Ex)) & (x)(Rx ⊃ Rx)) ⊃
theorem16
(x)((Rx & Ex) ⊃ (Rx & Ex))
9 (x)(Rx ⊃ (Ex ⊃ Ex)) & (x)(Rx ⊃ Rx)
4, 7 Conj
10 (x)((Rx & Ex) ⊃ (Rx & Ex))
8, 9 MP
11 (x)((Rx & Ex) ⊃ (Rx & Ex)) ⊃ (◊(∃x)(Rx & Ex) ⊃ ◊(∃x)(Rx & Ex)) theorem 12 ◊(∃x)(Rx & Ex) ⊃ ◊(∃x)(Rx & Ex)
10, 11 MP
13 ◊(∃x)(Rx & Ex)