I am grateful to Josh Reeves for his engagement with the “After Science and Religion” project and his nuanced defense of the dialogue position in the science and religion field. I take his commentary seriously, not least because he recently has offered what in my view is one of the most perceptive analyses of the state of play in science‐religion discourse (Reeves 2018). So this is a conversation worth having. While there are a number of possible topics of discussion, in what follows I focus attention primarily on the two chief concerns that Reeves sets out: “that Harrison's project is too skeptical towards the categories ‘science’ and ‘religion’ and places too much emphasis on naturalism being incompatible with Christian theology.” On the first point, I wonder if my position has been misunderstood and welcome the opportunity to clarify it. On the second, I think there is both a degree of misunderstanding but also some genuine disagreement. In this latter case, it is important to acknowledge our differences and identify what is at stake in our respective positions.

Before addressing Reeves’ two chief concerns it is worth saying something briefly about the project, its shared commitments, and its relation to both the theological emphases of Rational Orthodoxy and my own historical work. Much of the credit for initiating the “After Science and Religion” project must go to Paul Tyson, and in addition to me, there were two other chief investigators involved—John Milbank and Tom McLeish. All up, there were more than 20 other participants who contributed to the main outputs of the project, After Science and Religion (Harrison and Milbank 2022) and New Directions in Theology and Science (Harrison and Tyson 2022). Needless to say, perhaps, we did not reach agreement on every issue (see esp. McLeish 2022a, 277–81). But it was a signal achievement of the endeavor to have so many able and articulate thinkers—many of whom had not hitherto been part of the conversation—offering some fresh thoughts on the topic of science and religion. All this is by way of making a subtle distinction between “my project” and a project in which I was an enthusiastic participant.

What were the major points of convergence? Foremost, was a determination to take history seriously. History is often given the role of simply providing diverting anecdotes about the science and religion of yesteryear; or it is regarded as a worthy, if largely unnecessary, prelude to the substantive discussion. In this project, historical developments were understood as providing key insights into the present relations between science and religion and to some degree as offering potential grounds for a critique of them. To take a specific example, a historical or genealogical approach can be deployed to expose the hidden indebtedness to theology of such secular modern institutions as science, the social sciences, and politics. There is a connection, then, between my own work on the theological origins of aspects of modern science (e.g., Harrison 1998, 2007), and John Milbank's well‐known genealogy of the modern social sciences (Milbank 2006). Related to this, because history conducted in this genealogical mode involves a careful investigation of the contingent aspects of the past that have brought us to where we are presently, it also enables us to envisage different modes of engagement between “science” and “religion” to those that presently obtain. “Natural philosophy” offers one such example (McLeish 2022a), as do the historical alternatives of “enchanted immanentism” and “enchanted transcendence” identified by Milbank (2022).

A second point of convergence was a determination to take theology seriously. Theologians in the project were interested in thinking more along the lines of a theology of science, rather than theology and science (see McLeish 2022b; Tyson 2022). Such an approach was intended partly to redress a situation that sees a lot of one‐way traffic between science and theology—largely the result of the present prestige of the natural sciences on the one hand and the waning influence of theology on the other. The recommendation to think in terms other than a one‐sided “dialogue” follows from this, since the preliminary step is to understand science from a historical and theological perspective. This differs from approaches that simply take the deliverances of the sciences at face value, as if they were not only unquestionably authoritative but also completely innocent of cultural and ideological baggage. To put it another way, this approach enables us to question the apparent theological neutrality of current scientific knowledge. This again takes us back to history and genealogy, since a theologically informed history of science identifies ways in which modern science is already grounded in theology and is in some sense conducted within an implicit theological framework (in spite of its professed naturalism). Attending closely to the forgotten theological foundations of the sciences, made evident in the study of their history, thus offers one starting point for a renewed discussion between science and theology. To put it another way, the sciences already have a tacit relationship with theology. This needs to be made explicit in order to facilitate a new kind of conversation.

Dispensing with “Science” and “Religion”?

It is also important to remind ourselves that the historical circumstances that led to the formation of the categories “science” and “religion” were different in each instance. This means that they perform different kinds of work. Crucially, “science” (along with “scientist” and “scientific method”) was first deployed in the nineteenth century in its modern sense as part of a boundary‐establishing exercise. Religion, along with ethics, theology, and metaphysics, was firmly placed on the other side of that boundary. (This constitutes some of the difference between science and natural philosophy.) Part of the intention of the conscious adoption of this new terminology was also to exclude clerics and amateurs from the practice of “real” science. We see a similar motivation in the partial displacement of “natural history” by the new term “biology,” the latter being regarded as properly scientific. What this history suggests is that “science,” especially in the conjunction “science and religion,” is not innocent and neutral, but is already performing hidden work that shapes the nature of the relation and how we talk about it. This is apparent not only in the consideration of the history of the terminology, but even in the present when we compare the English “science” to terms such as the German Wissenschaft or, going further afield, the Arabic ‘Ilm. While these latter expressions represent our best attempts to render the English “science,” their different histories means that they do not quite mean the same thing. The science‐religion discussion will often have a different resonance when expressed in other languages, just as our modern Anglophone discussions differ from preceding Western and medieval ones.

“Science,” then, is to some extent a way of delineating an area of enquiry, uniting it, and legitimating it. For these reasons, it is typically embraced by those who see themselves as practicing it. Scientists, in other words, are generally happy with the designation (although typically they will self‐identify with a particular scientific discipline rather than just “science” (Soskice 2022, 145). With “religion” the situation is different since it evolved as an outsider's term. It makes no appearance in the canonical documents of any of the faith traditions and is not a particularly important theological concept. In fact it has been the subject of internal theological critique, most famously by Swiss theologian Karl Barth who declared “religion” to be unbelief, maintaining that revelation was the abolition (Aufhebung) of religion (Barth 1956, 280–325). Adherents of other “religions” have expressed similar reservations. Christians, Buddhists, Hindus all have reasons not to think of themselves as subscribing to a religion (Smith 1978, 125f). So while both categories are potentially problematic, there is an asymmetry between them in terms of who might wish to query them, and why. It is important to understand in whose interest it is that we speak in these terms, and as a matter of observation (rather than normative recommendation), I note that the those within what we refer to as “the religions” have been less happy with their designation than scientists have been with theirs.

Related to the point about the usage of the categories, another reason I do not believe that we can simply abandon “science” and “religion” is to do with the fact that these are the familiar terms in which many, if not most, individuals express particular concerns: Does science conflict with religion? Are religious commitments compatible with being a scientist? And so on. Our conversations need to begin with the specific form in which doubts and concerns are expressed. Think here of the story of the tourist asking for directions to Dublin and receiving the unhelpful response “If I were you, I wouldn't be starting from here.” Discussions need to begin where people are, however far from the ultimate destination they might be. Nonetheless, there is a certain wisdom in the response given to the Irish tourist, because obviously the ease of reaching a destination is determined by the starting point. In the same way, in relation to the science and religion question there is a certain validity to the response: “I reject the premise of the question.” Sometimes acute conundrums can be ameliorated by analyzing the terms in which they are expressed. The general insight here, to borrow from Wittgenstein, is that some philosophical puzzles are not so much solved, as dissolved. This is because they arise out of unhelpful conceptual commitments or the limits of language rather than representing some ultimate puzzle about the world.

Recent empirical work conducted by the Theos Foundation on attitudes to science and religion in Britain offers a good example of how this works in practice. When survey respondents were asked about “science and religion” they tended to see more conflict between them than when were asked about science and specific religions—say science and Buddhism, or science and Christianity. The same was true when they were asked about specific sciences and religion: physics and religion, or microbiology and religion (Spencer and Waite 2022). This shows that the simple strategy of breaking down generic terms such as “science” and “religion” already moves people in the direction of thinking differently about a situation that they naturally imagined to be best characterized in a particular way. It is possible to begin with the categories in question, but then move on to a more nuanced and critical approach.

The projects associated with “science‐engaged theology” offer a further illustration of the creative possibilities once we leave these broad categories behind. As leaders of one such project, John Perry and Joanna Leidenhag (2021) explain, what their version of the project seeks to do is move beyond “the trans‐historical categories” of “science” and “religion” and focus instead on how scientific findings within sub‐disciplines of particular sciences might have a bearing on quite specific theological problems. They themselves regard this as a “second way” of responding to the issues raised by The Territories of Science and Religion (see also Harrison 2021, Davison 2022). ¹ Again, the idea is that by looking beyond global categories to more tightly focused puzzles as they arise within the practice of theology, at least some of the difficulties that attend more generic treatments might be avoided. This approach also precludes situations in which physics or quantum mechanics or certain versions of evolutionary theory get to stand in for science as a whole with the relations to “religion” being cast in a particular light as a consequence.

Finally, another reason that it would be impractical to dispense with “science” and “religion” is that the problems generated by their uncritical usage apply to a whole range of expressions within the conceptual vocabulary of the modern West. If we start cancelling particular concepts, there is no telling where it will end. Two prime candidates for further conceptual analysis in the science‐religion space are “supernatural” and “belief” (which will receive some attention in my next book). Examination of the history of “supernatural” (already essayed in Henri de Lubac's brilliant Surnaturel, 1946) suggests that the emergence in the West of a two‐tiered ontology—signaled by the gradual uptake of the word “supernatural” and eventually “supernaturalism”—was a precondition for modern naturalism and secularization (de Lubac 1946, 1996; cf. Milbank 2014). “Faith/Belief” has a similarly interesting history, gradually losing traditional connotations of “trust” and being transformed into an epistemic vice. It follows that while we might think that the meaning of the proposition “belief in the supernatural” is relatively straightforward, that is far from true. However, these are hardly categories that we simply abandon either, despite the fact that their use can generate pseudo‐problems. How a supernatural being might act in the realm of causally closed nature is one such conundrum that arises partly out of the question‐begging deployment of loaded categories “supernatural,” “Being,” “causally closed,” and so on. This points to the fact that “naturalism” is not an innocent category either but a notion with a particular history that generates novel difficulties that did not arise when the cosmos was conceptualized in rather different ways. This brings us to Reeves's second concern, to do with the compatibility of theology and naturalism.

Naturalism and Christian Theology

On the topic of naturalism I want to offer three observations. First, at a surface level it just seems obvious that certain versions of naturalism are not consistent with a commitment to theism. Second, when we probe deeper and consider the historical sources of contemporary naturalism we find that to a surprising degree monotheistic traditions contributed to its emergence. Third, and more specifically, one aspect of that contribution relates to a natural‐supernatural distinction that initially arose out of theological developments in the medieval and early‐modern West.

Reeves mostly offers an accurate analysis of my characterization of naturalism and its incompatibility with theism. However, I think we part company on exactly what this means, and what follows from it. My basic position on naturalism is relatively straightforward and, to my mind at least, common‐sense. While, notoriously, the term “naturalism” has a number of different senses, I am using it here in a standard way, in keeping with the definition offered by philosopher David Papineau: “The great majority of contemporary philosophers would happily accept naturalism as just characterized—that is, they would both reject ‘supernatural’ entities, and allow that science is a possible route (if not necessarily the only one) to important truths about the ‘human spirit’.” (Naturalism and Christian Theology) Papineau goes on to say that naturalism entails the view that “there is nothing more to the mental, biological and social realms than arrangements of physical entities.” On the face of it, this stance is not compatible with theology, theism, or with core Christian doctrines such as creation and providence, or the miracle of the resurrection. Certainly, I think this is how advocates of naturalism such as Papineau would see things. So I take it as obvious that naturalism, thus understood, is incompatible with traditional theism and with specific theological doctrines.

To get beyond this, in science‐religion conversations, the standard move is to specify that the strong view outlined above—ontological or metaphysical naturalism—is neither required for the practice of science nor implied by it. What is called for is the adoption of a methodological stance. For the purposes of the exercise—that is, for the conduct of science—we act as if there were no supernatural entities or spiritual substances operating in nature. This solution rests on a well‐established distinction between metaphysical and methodological naturalism. But again, it seems obvious to me that this methodological naturalism is not prima facie compatible with the approach of theology either, since the latter traditionally takes God, a “supernatural” being, to be its primary object. Theology can hardly adopt this methodological principle. (There is a question of whether natural theology might do so. My own sense is “no,” but that warrants more attention than can be given here.)

I agree that methodological naturalism helps obviate difficulties for the practicing theistic scientist, since the issue of the reality of the divine or the spiritual realm is simply bracketed out in the practice of science. This means that both theists and non‐theists can be engaged in a common enterprise. This is no doubt why the formal idea of methodological naturalism was promoted by Christian philosophers and scientists (de Vries 1986; Harrison and Roberts 2019, 2). However, it follows that the deliverances of science are not “neutral” with respect to the starting assumptions of theology. To expand on this further: what often seems to be assumed is that science and religion offer different perspectives on the same basic reality—nature—and that accordingly, as truth‐seeking enterprises focused on the same object, in principle there can ultimately be no conflict between them. This is often expressed in terms of a presumed complementarity between “the book of nature” and “the book of scripture.” But in fact Christian theology understands the natural world as God's creation, whereas the object of scientific investigation is “nature,” understood from the perspective of methodological naturalism as devoid of divine influence and uninhabited by any spiritual entities. It follows that we cannot simply put science and theology into direct conversation with each other as if these differences of orientation were not there. On this point, I have some sympathy with those who maintain that when combined with a strongly realist view of science, methodological naturalism implies metaphysical naturalism (see, e.g., Forrest 2000; Boudry, Blancke and Braeckman 2010).

The recommendation is not that we need to adopt some non‐naturalistic version of science. Certainly, existing contemporary efforts at an explicitly non‐naturalistic science seem unpromising. ² Instead, we need simply to recognize the limitations that science (as currently practiced) places on itself and understand what follows for attempts to put science and theology into direct contact with other.

To move to my second point, a paradoxical complication to all this arises out of the theological genealogy of naturalistic science. The idea of the immutable uniformity of nature, which naturalists typically take to be constitutive of their stance, was originally grounded in the theological notion of a divine lawgiver and laws of nature (Henry 2004; Harrison 2019). Ironically, then, naturalism turns out to be a secularized version of a theological interpretation of the natural order. Once the relevant history is taken into consideration, it is possible to see that at least some elements of modern naturalism are just theological ideas under another description. (Harrison and Roberts 2019; Harrison 2020; Jordan 2022). This is related to the sociological thesis that certain religious traditions can act as agents of desacralization (Gauchet 1999; Weber [1904] 2003; Harrison 2018 ). Examples would be the way in which Patristic thinkers denied the divinity of the heavens and the reality of pagan oracles, or the Protestant contraction of the sacramental channels through which divine grace could operate along with its skepticism about contemporary miracle‐working.

Obviously, naturalists tend not to see things this way, with naturalism typically regarded as inimical to religious interpretations of the world. Relevant to this point, as Reeves has noted, I wonder about the cumulative effects of the practice of methodological naturalism, and whether it might have a habituating or formative effect. I do not know if this is the case—it is an empirical question that warrants investigation. But if it does, it would influence the dispositions of scientists outside the specific context of scientific investigation. Theistic commitment, left at the door of the laboratory, might eventually be still left there on the way out. My guess is that this kind of generic naturalistic outlook, as the result of disciplinary formation, is actually more common in the social sciences than the natural sciences. Training in the social sciences often involves the inculcation of historical myths of science‐religion conflict (Aechtner 2015, 2019). All that said, while philosophers and science‐and‐religion scholars talk a lot about methodological naturalism, I am not sure that it is a fundamental preoccupation of practicing scientists—as if the invocation of supernatural causes was an ever‐present occupational hazard that had to be constantly resisted.

Thirdly, there is a deeper problem with naturalism, as hinted above. Just as “science” and “religion” have their own histories that determine the shape and outcomes of the present conversation, so, too, “natural” and “supernatural.” The latter term appears in the West from about the twelfth century onwards, but only in the early modern period does it begin to be strongly contrasted with “natural.” The relevant “isms” come later, in the nineteenth century. Prior to this, “natural” causation required divine agency and it would have been nonsensical to think in terms of natural and supernatural causation as somehow in competition with each other. It was not the emergence of modern science that changed this understanding of nature. On the contrary, and oversimplifying a little, modern science was predicated on changed theological understandings of the world and of the natural‐supernatural relation. This is not the occasion for a full account of these historical transitions. (For intimations of this argument see de Lubac 1996). Suffice it to say for now that the history of the natural‐supernatural distinction turns out to be as relevant for contemporary science‐religion discussions as the history of the concepts “science” and “religion.” This means that while naturalism as typically defined is prima facie incompatible with theism, its history suggests that aggressive assertions that naturalistic science has demonstrated the irrationality of religious view of the world turn out to be conceptually confused or even self‐defeating. Rhetorical deployments of naturalism, understood as an overarching metaphysical truth that is somehow justified by science is mostly bluster and bluff. Modern science is not as naturalistic as it imagines itself to be, since the long‐standing assumption of the uniformity of nature along with the notion of laws of nature was a contribution from theology.

Reeves concludes his commentary with a reflection on “our changed social environment.” Here, he expresses a concern that the skeptical agenda that he associates with Radical Orthodoxy “seems little different to the skeptical dismissal of science from creationists in the United States, who think they can ignore the empirical evidence of the sciences because scientists do not begin with their same foundational premises” (my emphasis). I disagree. I would say rather that the challenge of Radical Orthodoxy is not so much to science, but to scientism—science moving out of its lane (Schindler 2022). Be that as it may, I share Reeves’ general concern about the challenges to the authority of science. (For a nuanced account of this problem see Reeves 2021). But the conditional “in the United States” is revealing. In the United States, we certainly do encounter religiously motivated antievolutionism, refusal to acknowledge the reality of climate change that is often aligned with religious commitment, and an antiscientific resistance to mask‐wearing mandates and other measures to curb the spread and mortality rate of COVID‐19. The presidency of Donald Trump no doubt provided additional reasons to be concerned about the authority of science. But the United States is not the whole world. Many countries, including my own, are not blighted by alliances of right‐wing politics, Christian nationalism, and science skepticism. In short, from a wider perspective, the present and parochial problems of the United States are not likely to be universal ones. Accordingly, solutions to those problems will not be universal. I fully understand, given the social context in which Reeves is writing, why he would be particularly sensitive to theological approaches that might appear to give succor to the “enemy.” And I fully agree that local circumstances might demand local strategies. But these practical matters of strategy operate at a different level to more general and universal questions that theology seeks to grapple with. Perhaps these local circumstances, and the different audiences we are seeking to address, accounts for at least some of the differences in our respective approaches.