Religious Conversion and Loss of Faith: Cases of Personal Paradigm Shift?

Is Thomas Kuhn’s model of scientific revolutions in terms of paradigm shifts appropriately applied to cases of radical changes in religious outlook, and in particular conversion to faith, or loss of faith? Since this question cannot be addressed in purely a priori terms, three case studies of philosophers who have described significant changes in their own perspectives are examined. Part of the justification for such an approach is to see how changes in view seem from the first-person perspective. Although what is offered here is a very limited group of studies, each of them perforce brief, a key theme emerges: that the incommensurability between paradigms which is part of the Kuhnian model is mirrored by the shift away from atheist or humanistic perspectives, though not necessarily by loss of faith.


Introduction
In his celebrated account of revolutionary scientific theory change in terms of paradigm shifts, The Structure of Scientific Revolutions, the first edition of which appeared in 1962, Thomas Kuhn suggests that the decision to embrace a new paradigm in place of the old, on the assumption that the new paradigm can succeed in solving the problems which confront it, 'can only be made on faith' (Kuhn (2012: 157)). In the 1969 Postscript to the original edition of the book, he describes the state of mind involved in fully embracing the new paradigm as a 'conversion experience ' (ibid.: 203). The idea appears again in a later essay: 'an individual's transfer of allegiance from theory to theory is often better described as conversion than as choice' (Kuhn, 1973: 338). The words 'faith' and 'conversion', of course, have strongly religious overtones to our ears, though Kuhn does not make explicit a specifically religious parallel. Alan Chalmers, however, in his introduction to Kuhn's views, does make this explicit: the change in outlook occasioned by a shift from one paradigm to another (such as the move from pre-relativistic to relativistic physics) is analogous to religious conversion (Chalmers (1982: 96)). Analogous, note, not identical: there is no suggestion that scientific paradigm change is itself a religious change, but rather one which parallels religious conversion in certain respects. The comparison is intended to underline the radical nature of scientific theory change. A paradigm shift is not simply a matter of solving a theoretical problem, but rather coming to see the world in a completely different light. In another parallel, Kuhn talks of a paradigm change as a 'gestalt switch' (Kuhn (2012: 118)). To model scientific change on religious change, however, is to suppose that we have enough of a grasp of religious conversion for the analogy to be illuminating. And in 1962, this would no doubt have been a sensible strategy. But now the concept of paradigm shifts is no longer novel, but instead relatively familiar in the history and philosophy of science, we might consider running the analogy the other way. Rather, that is, than employing an unanalysed notion of religious conversion to explain the change from one way of doing science to another, we might try employing the latter ─ perhaps, even, an extreme and controversial construal of the latter ─ to try to illuminate the undoubtedly puzzling nature of religious shifts, from unbelief to faith, or from faith to loss of faith, or from agnosticism to atheism. This paper is intended as a small contribution to that project.
The extreme and controversial construal of paradigm shifts I have in mind is one which employs the notion of incommensurability. Can different paradigms be compared, to the extent that the shift from one to the other can objectively be counted as scientific progress, or is the appeal to criteria of evaluation only an available move within a paradigm? Relatedly, can we even establish common meanings of terms which would allow us to see the paradigms as disagreeing about the nature of x (for some relevant x: a type of entity, property, relation, or set of laws)? The incommensurability of paradigms is one of the most controversial themes of Kuhn's work, though it finds perhaps a more extreme expression in Paul Feyerabend's Against Method (1975). I should say at the outset, however, that I am not primarily concerned with exegetical issues, but rather with the potential for a concept in philosophy of science to cast light on a religious phenomenon.
I claim no originality in proposing that we run the analogy in the science to religion direction. Tomas Sundnes Drønen, for example, has offered a careful analysis of the extent to which Kuhn's model can be applied to the case of religious conversion (including the shift from one religion to another), making use of sociological studies of religious shifts, and reaching a conditional conclusion: that although there are clear parallels, we should not uncritically carry over the notion of incommensurability to the religious case, particularly in view of the criticism of this very feature in Kuhn's analysis of scientific theory change (Drønen, 2006).
One potential disanalogy between scientific and theory change we can usefully pinpoint before embarking on the project. In scientific theorising and practice, the language is impersonal. The question is what the world is like, not how you or I experience or relate to the world. Theories are shared, and attempts are made to replicate other scientists' experimental results. But in religious contexts, though there may be an impersonal language in which religious doctrines and community practices are articulated, nevertheless the first-personal perspective cannot be eliminated. The religious viewpoint offers to transform my life. Equally, the loss of faith may occasion a personal crisis. Both the search for faith, and the abandonment of it, is a personal journey. Scientific paradigm shifts happen to scientific communities, even if it is also true to say that this may sometimes be driven by the scientific work and theorising of individuals. And when we ask whether different scientific paradigms are commensurable, we will most likely think of these in abstract terms, rather than as the belief states of individuals. In contrast, religious conversion or loss of faith is primarily something that happens to individuals, even if this may sometimes spill over into the communities of which these individuals are part (think of the charismatic preacher, or the influential sceptic). So when we ask whether different religious viewpoints are commensurable or not, we can pose this in impersonal terms, as a question about religious propositions (theism versus atheism, for instance). Or we can pose it in personal terms, as a question about the individual thinker: can the person who has gained, or lost, a faith, make an objective comparison between the state they are in now and the state they were in before? We might find we want to answer the impersonal question one way, and the personal question a different way. At least, that is a possibility worth considering.
In this paper, I will focus almost exclusively on the first-personal perspective, the perspective of the individual who gains or loses faith. I hope this approach will complement Drønen's study of religious change in communities. A full investigation of religious change would require attention both to the individual and to the social level, and, moreover, to the interaction between them. So what is offered here is only the beginnings of an investigation into part of the picture. But one of the key points I want to make is the value of the anecdotal. It is difficult, and perhaps not appropriate, to legislate on the proper way to negotiate religious transitions in the case of individuals. An empirical approach seems to be called for. So we will look at three figures who have described their own shift of outlook: Anthony Kenny, representing loss of faith; Janet Martin Soskice, representing conversion; and David Cooper, who describes his shift from a humanisticexistential outlook to one which makes room for mystery. This third case is not necessarily to be identified with a religious conversion. Cooper himself does not present his own case as such, but he does acknowledge that the sense of mystery he identifies may be part of an explicitly religious outlook. My reason for choosing these individuals is that they are philosophers discussing their own change in perspective and so, naturally, are particularly sensitive to the philosophical dimension of that change. The question I shall ask in each case is whether we can model the change on Kuhnian paradigm shifts, a change which so transforms the subject's outlook as to make internal dialogue between old and new viewpoints almost beyond reach.
To provide more detail to the comparison I am trying to draw, let us begin by rehearsing Kuhn's account of scientific revolutions.

The Nature of Scientific Paradigm Shifts
In brief, Kuhn envisages three kinds of phase in the development of science. The first phase is one of pre-science, or pre-paradigm period, where there are no overarching or guiding principles, or criteria by which theoretical adequacy is to be assessed, but rather a haphazard assemblage of competing ideas (Kuhn (2012: 48, 61)). This is replaced by the emergence of a paradigm, which, at least in one of the senses in which Kuhn uses the term, can be characterised as a combination of theory, a set of explanatory aims, a set of problems to be solved, and a methodology which defines proper investigative practices, including, perhaps, specific pieces of apparatus. The period in which a given paradigm is dominant is described as 'normal science' (ibid.: 23f). Perhaps one of the more surprising features of Kuhn's analysis (and one which distinguishes it from other models) is that the testing of the dominant theory is not part of normal science (ibid.: 144). Experiments are not designed to verify or falsify the theory so much as to apply it. The justification for one theory's being adopted in place of another will be defined within the paradigm, and the new theory will typically share explanatory aims with the old theory: they may both be trying to explain the same phenomenon, for example. This second phase in which a single paradigm dominates continues until it faces anomalies: phenomena it cannot explain, or internal tensions. The pressure of these anomalies builds up and leads to a (sometimes extended) period of crisis until a new paradigm emerges. The transition from one paradigm to another constitutes the third kind of phase, a phase of 'extraordinary science' involving the further exploration of anomaly, new kinds of experiment, and new speculative theories (ibid.: 87). Scientific communities may be divided at this point, some clinging to the old paradigm, some abandoning it in favour of the new. With the new paradigm come not only new theories, but also a new agenda, a new set of problems, a new way of assessing theoretical advance. There may be a change in some of the fundamental concepts with which scientists operate, or a new piece of apparatus whose function would have been unintelligible under the old paradigm.
Let us illustrate this abstract description with three chemical examples. Consider medieval alchemy (not, in fact, one of Kuhn's examples). It is natural, perhaps, to dub this a period of pre-science. The most casual survey reveals an eclectic mix of images, symbols, concepts, and descriptions of phenomena. Yet there is some method here. There are specific aims: the construction, or discovery, of the philosopher's stone, or the elixir of life, the route to the perfection of the soul. And there is, perhaps as a corollary, the transformation of base metals such as lead into gold. The route to these goals is a series of operations on substances, including which we would recognise as distillation, condensation, and sublimation, bringing about changes in physical state, as well as those involving chemical changes, such as calcination. All of these are described, and pictured, in metaphorical language and imagery, linking substances to the heavenly bodies (the seven alchemical metals sharing signs with the seven planets of medieval cosmology), and to the progress of the soul. In medieval alchemy, it seems, the animate and inanimate worlds are profoundly connected at every level of their activities. We can identify, then, a set of goals, a set of procedures for attaining these goals, and a language to describe them (see, e.g. Read (1995)). The very fact that we can give these the single label of 'alchemy' strongly suggests that we have here a paradigm. It may not be scientific in our sense, but it is a science in the most general sense. It is an attempt, that is, to understand and manipulate the natural world.
This paradigm (or period of pre-science, if you prefer) failed to produce the concrete goals it set itself, though in other ways, aesthetic, literary, and technical ─ consider the development of alchemical apparatus ─ it might be judged a success. It was succeeded by early modern chemistry, ushered in by such texts as Robert Boyle's The Sceptical Chymist. Whereas the language and symbolism of alchemy were thoroughly metaphorical, or analogical (such as the links between metals and the heavenly bodies, for example: gold: Sun; silver: Moon; iron: Mars; and lead: Saturn), that of early modern chemistry aimed at literal description. For the alchemists, base metals were in principle transmutable. For the early modern chemists, in contrast, the elements are immutable, except insofar as they participate in chemical reactions. But such reactions never involve the transformation of one element into another. The aims of the alchemists are replaced by a different set of aims, to do with an understanding of how the elements combine (in what proportions) in chemical reactions. There is a focus on quantity and measurement. And in due course, the list of agents of chemical change came to include a power of which the alchemists could not have dreamed: electricity.
In the early twentieth century, there is another paradigm shift in chemistry, a more focused one, with the advent of the electronic theory. This makes conceptually possible a thorough reduction of chemical phenomena to changes in electronic configuration. It also provides the theoretical basis for a phenomenon discovered in the late nineteenth century: radioactivity. The transmutation of the elements is back on the table, but this time it is linked to the radioactive decay of the transuranic elements and to the production of artificial elements by nuclear bombardment. Now take a case much discussed by Kuhn: the transformation in the way combustion and calcination were understood (Kuhn (2012: 53f)). For Joseph Priestley, a proponent of the phlogiston theory, the action of heat on red calx of mercury in an enclosed container is the absorption by the calx of phlogiston in the air, leaving 'dephlogisticated air' and metallic mercury. But from the point of view of the oxygen theory of combustion, championed by Lavoisier, this gets things exactly the wrong way around. Rather than combining with a component of the air, the heated calx (that is, mercuric oxide) releases the oxygen with which it was combined. Phlogiston is a myth. For Kuhn, the transition from the phlogiston theory to the oxygen theory is a clear example of a paradigm change. Yet the case is very different from the transition from alchemy to chemistry. The phlogiston and oxygen theories are both trying to do exactly the same thing, which is to explain what is going on in the processes of combustion and calcination. 'Dephlogisticated air' is functionally equivalent (at least in this context) to oxygen. The two theories are, in a sense, mirror images of each other. And they are answerable to the same set of data: the fact that the calx loses rather than gains weight during combustion, for example (a point in favour of the oxygen theory). Still, the transition is a central part of what became known as the Chemical Revolution.
Kuhn notes four features of scientific revolutions: (i) They involve choices between competing paradigms. A paradigm is not abandoned without being replaced by a new one (ibid.: 78). (ii) Competing paradigms are, at least to an extent, incommensurable. To put it dramatically, but perhaps metaphorically, adherents to different paradigms live in different worlds (ibid.: 117-118, 134). (iii) Scientific progress is not a matter of accretion, with older views somehow incorporated in new ones (ibid.: 98-103). (iv) Scientific progress through revolutions need not, and perhaps should not, be seen as driven by the need to converge on the truth (ibid.: 169-72).
Since (iii) and (iv) follow from (ii), let us explore (ii) further. (i) We will come back to in section IV; (iv) we shall leave until the end.

First-Personal and Impersonal Incommensurability
A period of normal science, suggests Kuhn, is characterised by puzzle-solving rather than testing of the paradigm (ibid.: 36-42). There is substantial agreement over the data to which different solutions are answerable. But where competing paradigms are concerned, comparison is more difficult. Each paradigm may define its own criteria of success. To take again a non-Kuhnian example, we unhesitatingly judge modern chemistry to be 'better' than medieval alchemy -but this judgement is made from the perspective of modern chemistry. From that perspective, a relatively transparent language is preferable to a mystical or metaphorical one. Similarly, we value the power of a theory to provide precise predictions of phenomena; experiments must be precisely repeatable, and so on. Moreover, the data to which the theories of modern chemistry are answerable are themselves described in the language of modern chemistry (the notorious 'theory-dependence of observation': see Hanson (1958)). But what if we expect our chemical outlook to make room for the divine, for the ensoulment of matter, to connect the properties of substances with cosmology? What if we look to see in the stages of chemical transformation a pattern akin to a pilgrim's progress of the soul? We will find little comfort or enlightenment in the prosaic, if complex, landscape of modern chemistry. Far more emotionally satisfying will be the mystery and richness of alchemy.
Then there are issues of translation. During a period of normal science, competing solutions will substantially share a vocabulary. There may be disagreement (as, at one time, there was) over whether chlorine is a compound or an element, or whether ammonia is a hydroxide of an element ('ammonium') or not, but the concept of 'element' and 'compound' will be the same in both cases. But when we are dealing with different paradigms, even what appear to be the same terms may have completely different meanings. The words 'sulphur' and 'mercury' occur in both alchemy and in modern chemistry, but do they mean the same thing in both contexts? Does 'sublimation'? Does 'putrefaction'? And some ideas in one paradigm may have no counterpart in the other, like 'conjunction', or 'stereoisomer'.
Before we rush to apply this schema to changes in religious outlook, we should note that the danger (though not everyone will see it as such) of appealing to incommensurability is that it may naturally lead to truth-relativism: we cannot judge a given proposition to be absolutely true but only true relative to a paradigm. But we can temper this by pointing out that the incommensurability in question is, in part, an epistemological one. It is not that there is no one truth at which scientific theories aim, but rather that we cannot adopt a sufficiently neutral perspective from which to judge what that one truth is. There is also a semantic aspect: apparently, the same terms are given different meanings in each paradigm, and some terms defined within a paradigm may have no application outside that paradigm. But this semantic incommensurability only takes as far as what William Newton-Smith calls 'trivial semantic relativism': the same form of words may express a truth relative to one context and a falsehood in another, but only because they express different propositions in those contexts (Newton-Smith (1980)).
With that cautionary note sounded, let us now consider whether the model of paradigm shifts in science can appropriately be applied to radical changes in religious outlook: conversion to a faith, or loss of faith.
As noted above, although a shift in perspective can happen to individuals as well as communities, the distinction between impersonal, community-level perspective change, and individual, first-personal perspective change is more significant in the religious case precisely because of the personal nature of religion, its demands on the individual. In the case of science, incommensurability, where there is no neutral standpoint by which to judge competing paradigms, is something which either occurs at both the personal and community levels, or neither. There does not seem to be a reason for thinking that incommensurability of competing scientific paradigms could arise for the individual but not for the community, or vice versa. Is the same true of religion, given the significance of the first-person perspective? It is worth at least asking whether problems of incommensurability in religion might arise in certain cases just at the personal level. Analytic philosophy of religion is largely composed of attempts to compare competing theories: theism versus atheism, divine command versus natural law theory, timeless versus temporal conceptions of deity, accounts of omnipotence, or realist versus non-realist conceptions of religious language, theistic versus non-theistic conceptions of the spiritual life, and so on. For debates of this kind to be possible, the competing theories must be regarded as commensurable. We can consider the rivals from a neutral standpoint and hope to be able to judge which better meets our (it is to be hoped) neutral criteria of success (e.g. internal consistency, explanatory power, ontological or theoretical parsimony, and compatibility with other independently justified beliefs). Granting all this, we may nevertheless want to make room for the idea that, when it comes to a shift in outlook within the first-person perspective, such comparisons are difficult. The individual undergoing such a transformation may, indeed, feel that they are in a better place, intellectually, morally, spiritually, or emotionally but also admit that they cannot articulate reasons for the shift which they would expect to have any force to anyone with a contrary viewpoint. Any attempt at such justification would appear circular. Moreover, they may feel that they cannot communicate their new outlook to anyone who does not share it. They may even have difficulty attempting to recover the content of their earlier beliefs. The obstacles to commensurability may arise just at the first-person level.
If this distinction between impersonal and first-personal incommensurability seems rather dubious, consider an analogy. The following state of affairs is one we can readily suppose is instantiated: p is true and x does not believe that p. 'x' stands for any impersonal way of picking out an individual. But now replace x by the first-person pronoun: p is true and I do not believe that p. This is not a thought that we can entertain without contradicting ourselves. Or consider a Cartesian insight: I can think 'x does not exist' but cannot sustain the thought that I do not exist. So it appears that there some states of affairs which can be contemplated, and accepted, from an impersonal or third-personal perspective, but not from the first-person perspective. In the case of incommensurability, the relevant difference is between 'x's belief that p is objectively more rational than y's belief that not-p' and 'My current belief that p is objectively more rational than my former belief that not-p'.
It is now time to consider some first-personal accounts of shifts of religious belief state and ask whether these are appropriately described as a shift in paradigm.

Anthony Kenny
In his autobiographical A Path From Rome, Anthony Kenny provides a detailed, highly reflective, and engrossing account of his intense religious education, his training for the Catholic priesthood, and the thinking which led to his gradual realisation that he should return to the lay state, which he did in 1963. What prompted this realisation was a combination of considerations, some concerning the emphasis accorded to church hierarchy, but in very large part to do with the intellectual basis for Catholic doctrine, and indeed for theistic belief in general. This could certainly be described as loss of faith, as in the final interview with his archbishop he confessed that he no longer had faith in God, and could only pray as an agnostic might, in the hope that someone might be listening. These doubts were in turn prompted by careful philosophical reflection on the nature of religious language and thought, a topic he explored both in his licentiate dissertation at the Gregorian University of Rome, and in his doctoral thesis, for which he worked in Oxford, coming into contact with contemporary methods of linguistic analysis, and some of their most prominent exponents.
The dissertation for the licentiate was particularly concerned with the threat that logical positivism appeared to pose for theological language. According to the principle of verifiability, meaningful assertoric sentences are either analytic truths (that is, true by virtue of their logical form, or the meanings of their components) or empirically verifiable. But theological statements are not analytic (unless one is prepared, as Kenny was not, to endorse the soundness of the ontological argument). Nor, on the other hand, do they seem to be empirically confirmable or discomfirmable, as theological statements appear to be compatible with any observation statement. One conclusion, drawn in Richard Braithwaite's essay 'An Empiricist's View of the Nature of Religious Belief' (1955), is that theological statements are not propositionally assertoric, but rather expressions of commitment to a certain way of life. What differentiates religions which largely share their moral outlook are the stories by which the moral messages are conveyed. Though sympathetic to Braithwaite's treatment, Kenny suggested that genuinely assertoric theological statements could be given content, and justification, through mystical experience. By the time he came to write his doctoral thesis, however, this struck him as inadequate. As he explains: ...my initiation into the philosophy of Wittgenstein at Oxford, while it removed the bogey of positivism, equally took away my confidence in religious experience as providing justification for the assertions of natural theology. I began to think -as I do to this day -that belief in God can be rationally justified only if the traditional proofs of the existence of God are valid. Faith will not do instead of proof, for faith is believing something on the word of God, and one cannot take God's word for it that He exists. Belief in God's existence must be logically prior to belief in revelation. (Kenny, 1985: 147) The path from faith to agnosticism which Kenny describes does not seem to fit the model of paradigm shift. There is, to be sure, the awareness of anomaly. From the later perspective, the earlier attempt to defend the meaningfulness and rationality of theological beliefs is found wanting. Speaking of his dissertation, he says: I find it painful to re-read after twenty-five years: not only am I embarrassed at the philosophical naivety of much of its philosophical content, but I am astonished at the thinness of the intellectual underpinning of a lifetime commitment to the service of the Church. Emotionally...I was at this period totally committed to the priestly life; as devoted to the Church as the most adoring lover to his mistress. But the emotional investment provided no more adequate a basis for a life of fidelity to the priestly ideal than infatuation does for a lasting marriage. The intellectual rationalization which the dissertation provides for the religious emotion looks, from this distance, merely a licence to wallow in make-believe. (Ibid.: 111) This may be a harsh judgement, but it indicates that Kenny finds no conceptual difficulty in recovering that earlier perspective. There is, then, no hint of incommensurability. And, while the specific reasons for finding it inadequate may not have clearly presented themselves at that earlier stage in his life, they do at least appeal to the kind of considerations of rationality which exercised him then.
Interestingly, the issue of incommensurability does appear in the licentiate dissertation, when he suggests that faith may actually provide content to religious propositions, so that what the theist come to accept is not necessarily what the atheist denies. Kenny quotes a paragraph which begins 'There must be a sense in which dogmatic propositions "mean something different" to the believer and unbeliever.' (Ibid.: 104) Again, however, the later judgement is unforgiving: This paragraph is full of muddles. What a proposition means must be independent of its truth-value; one cannot take it to mean one thing if one believes it and another thing if one disbelieves it. Belief and disbelief are contrasting attitudes towards one and the same proposition, and two sentences cannot express the same proposition if they have different meanings. (Ibid.: 105) It is undoubtedly true that the content of a proposition does not vary according to whether the proposition is true or false. It is also true that, in cases of genuine disagreement, one and the same proposition is the object of assent or denial. But there may be cases of apparent disagreement -assent or dissent to a certain form of words -where the truth-value of a single proposition is not at issue. And we cannot rule out a priori individual cases where a change in religious perspective cannot be reduced to a change of attitude to a single proposition. But in any event, Kenny's own change of perspective, as reported by him, does not invoke, or suggest, any issue of commensurability between earlier and later perspectives.
Recall Kuhn's thesis that scientific paradigms are not abandoned without there being an alternative paradigm to move to: 'The decision to reject one paradigm is always simultaneously the decision to accept another' (Kuhn (2012: 78)). The word 'simultaneously' here does not rule out an extended period of crisis where normal science is replaced by a period of extraordinary science, although that period does not go on indefinitely. But does Kenny's case provide a religious counterpart to this? It would seem not. The move from faith to agnosticism signals, not so much the abandonment of a paradigm, as a shift in epistemological attitude to it. But there is more than one model for loss of faith. One might imagine a religious paradigm being abandoned without a succeeding paradigm being adopted, a different way of making sense of morality and human purpose. Seeing loss of faith in these terms provides a ready explanation of its potentially shattering nature. Alternatively, there may emerge from this crisis a different religious paradigm, not so much a different religion, but a different interpretation of religious discourse and practice, such as is instanced by the non-realist approach (or approaches) of Don Cupitt (1980Cupitt ( , 1984Cupitt ( , 2012.

Janet Martin Soskice
Our second case study represents a religious transformation in the reverse direction: towards faith, though this is not a precise mirror-image of Kenny's transition. Janet Martin Soskice began adult life as an atheist (in Antony Flew's wide sense of one who is not a theist: Flew (1972)): I can remember being an atheist, or perhaps an agnostic, for in those days I did not think much about God one way or another. I knew at that time there were Christians -and I suppose adherents to other faiths whose adherence I judged 'ethnic' and thus not applicable in my own instance. As for Christians, I assumed they came in two sorts -those who went along for a sense of belonging or a sense of nostalgia, rather like attending a bridge club or square dancing; and evangelicals, who were clearly under-educated and over-excited. I can remember, from my lofty 21-year-old height of wisdom, thinking that it must be soothing to be one of the latter, all of whose problems in life could be seen to be answered. In my assumed clarity about religion I, in fact, knew nothing at all, in my own case it was only dramatic conversion which turned me round and put my feet on a steadier, more modest path whose depths are fathomless. (Soskice, 2009: 77) Immediately, we see a clear contrast with the causes of religious change in the case we discussed in the previous section. For Kenny, it was the accumulating weight of reasons, acquired over years, which eventually tipped the scales. For Soskice, it was sudden, datable, and essentially experiential: In my own case...faith came from a dramatic religious experience. It was not theatrically dramatic -I was not rescued from shipwreck by passing dolphins, or saved from falling to my death off a cliff-face by a gracefully placed liana; nevertheless it was dramatic to me. I was in the shower, on an ordinary day, and found myself to be surrounded by the presence of love, a love so real and so personal that I could not doubt it. (Ibid.: 77) This account is immediately preceded by a suggestion, perhaps more than a suggestion, of a paradigm shift between old and new viewpoints: Can I even say to those who, it seems to me, stand where I once stood (the cultured despisers of religion, as Schleiermacher might have said) what I now feel I know, and don't know, about God? It would be hard. Because it is not just that faith gives new answers to old questions -it gives new questions, a new world where even the most educated come as babes, born again. (Ibid.: 77) As described, the transition seems to satisfy a key criterion for thinking of a perspective change as a paradigm shift: the old and new viewpoints are not wholly answerable to the same set of desiderata, demands for explanation, or data. Rather, with the shift of viewpoint comes a new set of desiderata -and certainly a new set of data ─ which might not even be recognised by the old viewpoint as legitimate. Indeed, the old viewpoint would dismiss the 'data' the new viewpoint attempts to capture as illusion. And we can discern another obstacle to communication between the viewpoints. For the new viewpoint apparently makes available something which was not available before, a way of thinking about God. The shift is not just, as Kenny might see it, from denial to acceptance of a being, but also a shift from the thirdpersonal to the second-personal in relation to that being: Above all, I felt myself to have been addressed, not with any words or for any particular reason and certainly not from any merit -it was in that sense gratuitous -but by One to whom I could speak. (Ibid.: 78) Second-person thinking is, arguably, of a different kind, not only from impersonal thinking, but also from first-and third-person thinking, and cannot be reduced to the latter (See Salje (2017) for a defence of this thesis). In contemplating the possibility of God's existence, as an agnostic might, one views the object of thought third-personally. To view it second-personally requires the complete absence of doubt concerning its existence. This way of thinking is not available outside of the theistic paradigm, though that it is not to say that the idea of it is conceptually unavailable outside that paradigm.
Is it appropriate, however, to talk of the period before the conversion as one which is governed by a paradigm? I think it is, in this case. Soskice reports her earlier attitude towards religion in a way which makes it clear that she had an interpretation of religious attitudes, even if that interpretation is later dismissed as jejune. And we may plausibly suppose that, being highly reflective, she would have had views on the nature of morality, at the least that it had no theistic basis. So it seems quite appropriate to class this change in orientation, at least, as a genuine paradigm shift.

David Cooper
We now turn to our third, and perhaps most complex case. In 'Mystery, World and Religion', David Cooper describes the process by which he came to the view that reality, as it is independently of any human perspective, is mysterious, in the sense of being incapable of articulation or even conceptualisation (not 'discursable', is how Cooper puts it). Unlike Soskice, the transition was not primarily occasioned by an extraordinary experience. As he explains: Many people ─ those we call 'mystics', but not just those ─ have come to a sense of the mystery of reality through special experiences which they have had or claim to have had. My own life has not, I hope, been entirely devoid of Tintern Abbey moments ─ ones of a kind which, if people are forced to speak of them, have them calling on poetic vocabulary of 'mysterious presence', of 'something 'deep' that 'rolls through all things', and so on. Though, in my own case, it has tended to be littler and more humbler things than great ruined abbeys which have had me groping for the appropriate terms ─ things like the frog-plopping, bird-cheeping and bamboo-rustling that have been the occasions for Zen priests to communicate their sense of the mystery of things. But it is not these experience or moments which have fashioned my path to mystery. This has been a more intellectual, a more philosophical path. Though ─ who knows? ─ maybe this philosophizing would have struck me as too dry, too abstract, and would have failed to go deep with me, except in conjunction with experiences that I am tempted to construe as intimations of mystery. (Cooper (2009: 51-52)) Cooper outlines three contrasting views concerning the nature of reality: (1) Absolutism: there is a discursable way the world is in itself, independently of a human perspective.
(2) Humanism (of a rather extreme kind): there is just the human world, the world as it is from our perspective. (3) Mystery: there is a way the world is in itself, independently of a human perspective, but it is not discursable.
Influenced by writers such as Sartre, Husserl, Heidegger, and Merleau-Ponty, Cooper was initially drawn to humanism, and away from absolutism. Science appears to warrant absolutism, but the claims of science to provide the perspective-independent view of reality as it is are, he explains, 'bogus': The scientific account of the world owes too much to an all-too-human 'ground-plan', and to the privileging (driven by discernibly practical, technological interests) of a certain kind of explanation (causal, roughly), to be viewed as offering an objective, absolute account of the world and as having succeeded, therefore, in weeding out the human contribution. (Ibid.: 53) Humanism, of the 'raw' kind presented here, he nevertheless came to see as 'hubristic', in the sense that it attributed to us a capacity to live with the idea that nothing we do has any value beyond ourselves. Our beliefs and values are answerable to no higher standard, and so, in the end, are baseless. To confront this would, in fact, be intolerable (we might compare this state to that of someone who has lost their faith without yet finding anything to replace it). And so, Cooper found himself embracing the third option: there is something beyond the world as we construct it, but not something we can articulate. This is not the place to assess such a view, but it is worth noting how Cooper responds to the objection that, if we cannot articulate, or conceive, the nature of mind-independent reality, it cannot guide us. The champion of Mystery (the capital 'M' here signalling the name of an outlook) is in a position little better than that of the humanist. That, at any rate, is the inevitable conclusion if Mystery is no more than the bald statement that the world as it is in itself is not discursable. What more is needed, suggests Cooper, is a means of somehow attuning oneself to the mysterious, in such a way that one's life can be informed by it. This is the natural opening for religion, for are religions not precisely ways of attuning oneself to mystery? Cooper lists a number of religions that are explicit about this, including Daoism, Mahayana Buddhism, and medieval Christian and Islamic mysticism. And he has no objection to the sense of mystery, combined with a confidence in great thinkers who are receptive to mystery and define the pattern of a life in response to it, as 'religious'. What he resists, however, is the suggestion that the mysterious simply be identified with God, for God appears in theistic religions as an eminently discursable being, even in the writings of negative theologians whose professed principle is that such discourse is impossible.
Can we describe the transition Cooper describes from humanism to Mystery, a transition which is, at least potentially, a kind of religious conversion, as a paradigm shift? It certainly seems to chime with Kuhn's description of the proponents of competing paradigms as living in different worlds, or ─ less metaphorically and more prosaically ─ inhabiting radically different representations. Whether or not there is a reality which is not discursable is a significant issue intellectually (and emotionally), and on it depends the choice between significantly different kinds of life. The project of attunement to a mysterious reality is not one that makes any sense on the humanist picture (nor indeed on the absolutist picture). It is not immediately obvious, however, that there is an irresolvable incommensurability here. Cooper's own reasons for Mystery appear to be that the alternative positions are internally unstable, not that they fail to match a standard which makes sense only once one has acknowledged mystery.
It nevertheless remains the case that what gets represented by these three very different outlooks does not seem to be one and the same world, or the same mind which is doing the representing. We can approach this point through Cooper's warning against a dualistic, or disjunctive, picture of mysterious reality on the one hand, and on the other, the phenomenal world, the world as it appears, which is somehow causally produced by the world as it is in itself. He thinks this dualism is ultimately incoherent (Cooper (2009: 55-56). Presumably, he would not dispute that there is some distinction here, for it boils down to the distinction between the discursable and the non-discursable, without which Mystery cannot even be articulated. The problem rather seems to be that any attempt to give an account of the way in which the phenomenal world is produced, through interaction between mind and world, is doomed to failure. For 'mind', as part of reality as it is in itself, is thereby part of the mysterious. So we cannot say where mind ends and the rest of reality begins, nor what kind of interaction could generate the world as it appears. Contrast absolutism, for whom both mind and the rest of the world, and the interaction between them, are discursable. Here clear limits are set to what 'mind' could refer to. There are no similar limits to what 'mind' could refer to within Mystery. It makes no sense, then, to say that Mystery and absolutism are different representations of mind and world and their relations, despite the fact that in our initial presentation of these positions, that is exactly what those theories looked like. A similar problem emerges when we compare humanism and Mystery. In humanism, mind, world, and mind-world interactions are just so many representations (by what or whom, one wonders). Here too, then, the spectre of semantic incommensurability raises its head. In consequence, conversion from humanism to Mystery, absolutism to Mystery, absolutism to humanism, or any other permutation, looks very much like a paradigm shift.

Conclusion
Our central question was whether Thomas Kuhn's conception of scientific revolutions as paradigm shifts could be applied to cases of religious conversion or loss of faith (or a shift from agnosticism to atheism). Rather than examine such changes at the level of communities, however, we have tried to look at things through the perspective of the individual. This seems appropriate, as conversion and loss of faith is something that happens primarily at the individual level, and sometimes quite independently of events in society at large. In pursuing this question, we have looked, though rather briefly, at three first-person accounts of a significant shift in outlook: respectively, from faith to agnosticism, from atheism to faith, and from humanism to Mystery. The first two are explicitly religious, the third at least potentially so. In two of these cases (Soskice and Cooper), we found significant parallels with the Kuhnian model. In particular, the incommensurability of competing scientific paradigms has echoes in the extent to which, in these two cases, one outlook robs competitors of intelligibility for the person concerned. And we can describe in terms of paradigms the gulf between, on the one hand, the reflective and untroubled absence of faith preceding religious conversion, and, on the other, the void following the loss of faith, both of which might merit the name of 'atheism'. One is governed by a paradigm, the other not. That is one indication of the fact that 'atheism' is not a single category of outlook, and of the appropriateness of the increasingly used plural label 'atheisms'.
Three case studies hardly constitute a systematic study. But using case studies in the first place simply follows Kuhn's own method, although his interpretation of those case studies amounts to a retelling of the history of science. In contrast, the preceding narrative is, if revisionary, only modestly so. And, unlike Kuhn for most if not all the cases he discusses, I have had the luxury of the protagonists' first-hand accounts.
As Kuhn himself notes, only at the very end of The Structure of Scientific Revolutions does the discussion explicitly concern itself with the notion of truth. But when he finally turns his attention to this, the results are explosive. He compares the progress of science to biological progress, as viewed by Darwin's theory of natural selection. Pre-Darwin, biological progress was thought to be goal-driven, to be teleological in nature. Pre-Kuhn, scientific progress tended to be viewed as similarly goal-driven, the goal being truth. But, just as Darwin replaced that teleological account with one in terms of the survival of the fittest, so Kuhn proposes that we see scientific progress, not as goal-driven (where the goal is conceived as truth), but rather the result of selection of theories by their fitness to preserve scientific activity (Kuhn (2012: 169-172).
Can we draw similar conclusions concerning religious change? Here, the distinction between communities and individuals, impersonal versus first-personal religious change, is particularly pertinent. Viewed from the outside, shifts in religious perspectives in communities may seem not so much a matter of convergence on the truth, but rather driven by social forces, prompting a move from one experiment in living to a more viable one. But for the individual ─ at least if these case studies are at all representative ─ whether the shift is a result of a religious experience, philosophical reflection, or a shattering loss of faith, the goal is nothing less than the truth.