In my article on the blurred relationship between theistic evolution (TE) and intelligent design (ID), I made a collection of arguments toward the conclusion that an increasingly popular naturalistic formulation of TE fails to demarcate itself from ID (Leidenhag 2019). It has become popular for theistic evolutionists to propose that God only works through and within natural processes, and to contrast this account of God's action in evolution with the interventionist deity of ID. While the idea that God only works through and within natural processes comes in many forms, and is formulated against the background of a variety of metaphysical frameworks, I gave this theory of divine action the summarizing title, Natural Divine Causation (NDC). On NDC, God acts in a noninterventionist manner, in harmony with natural laws, and in congruency with the natural processes of the cosmos. For those who support the notion of NDC, which includes a number of prominent figures in the science‐religion dialogue, this view of God's providence provides a way beyond the interventionist picture of ID, whereby God occasionally breaks or suspends the laws of nature. As opposed to the “gappy” God of ID, NDC strives to uphold a compatibilist picture of the relationship between divine activity and natural processes. What is most important to the argument of my previous article is that those who support NDC suggest that such a model of divine action provides a line of demarcation between TE and ID. I located several causal problems facing NDC, particularly regarding causal overdetermination, and suggested that it consequently fails as a line of demarcation between TE and ID. When drawn with the pen of NDC, the line between TE and ID becomes blurred. And, when the line of demarcation between these theories is blurred, then the idea of TE and ID as competing alternatives dissolves into a falsely polemical image.

I would like to thank Christoffer Skogholt for his engaging and thought‐provoking response. In his response, Skogholt argues that my critique of an NDC accommodation of TE fails to hit its target. More specifically, Skogholt opposes my main argument, namely, that NDC seemingly makes God‐talk ontologically redundant. Before I go on to defend my overall argument, it is worth clearing up a major confusion that has emerged in Skogholt's article. I am not claiming that TE as such makes divine action redundant. That is, I am not opposed to TE. Rather, I take issue with those scholars who construe TE, and divine action more broadly, in a naturalistic manner, as exemplified by NDC. Given this confusion, Skogholt's response fails, unfortunately, to engage properly with my article.

I will, first, defend my previous argument that NDC makes God‐talk redundant. Second, I will highlight the promise of partial causation for framing an adequate version of TE, thus clearing up a major confusion that has emerged in Skogholt's response. Finally, I will briefly consider Skogholt's discussion of the problem of evil. It will be seen that NDC faces its own share of theological and philosophical challenges in light of the reality of natural evil in evolution.

DEFENDING THE REDUNDANCY ARGUMENT

My main argument is that the model of NDC is a theologically and philosophically costly strategy for framing God's actions in the world. As Skogholt seems to think that no scholar adopts NDC (if only this were true!), it is worth rehearsing what NDC means for our understanding of divine action, and how it impacts the science‐religion discussion more broadly. It is, in fact, rather surprising that Skogholt engaged so little with the model of NDC, and even more surprising that he opens his article with a brief outline of Aquinas's distinction between primary and secondary causation, which, while theologically and philosophically interesting, bears little resemblance to NDC. It would be a serious mistake to equate Aquinas's notion of divine providence with the theistic naturalism and panentheistic models of Clayton, Peacocke, Griffin, and Johnston. In fact, Peacocke and Griffin have both critiqued a Thomistic distinction between primary and secondary causality, as well as the model of double‐agency, for employing a dualistic ontology and for ultimately entailing a supernaturalistic interruption of natural laws (see Peacocke 1993, 148–49; Griffin 2000, 38–40; Griffin 2004, 8).

Once again, my argument is not aimed at theistic evolution as such, and so I have no immediate objection to the sort of distinction between primary and secondary causation that we find in many versions of theistic evolution. Aquinas's theology and NDC paint very different pictures of God's ongoing activity in the world, and Skogholt's response fails to engage with much of my argument due to this conflation of Thomism and NDC. Why start off with a Thomistic account of divine action when this account is so different from the naturalistic ambitions of Clayton, Peacocke, Griffin, and Johnston? ¹

Those who adopt NDC—which includes Philip Clayton, Arthur Peacocke, David Ray Griffin, Mark Johnston, and to some extent Deborah Haarsma—suggest that the divine action debate and our views of God's interaction with evolutionary forces need to proceed on a “presumption of naturalism” (Clayton 1997, 171; Leidenhag 2019, 926). This presumption states that “for any event in the natural order … its cause is a natural one as opposed to a supernatural one.” In a similar vein, Arthur Peacocke argues quite strongly that the “only dualism which such a [naturalistic] stance accepts is indeed that between God and all‐that‐is, the ‘world’; it rejects any dualisms within the natural order itself, including humanity” (Peacocke 2007, 9). In Griffin's terminology, this amounts to a “minimal naturalism,” which means that “the world's most fundamental causal principles are never interrupted” (Griffin 2000, 44). For Griffin, this constitutes the “most fundamental assumption of the contemporary scientific worldview” (Griffin 2000, 12). Griffin grounds this claim in the process theological rejection of creatio ex nihilo. This, according to Griffin, post‐biblical doctrine has not merely caused a dualistic and antagonistic relationship between God and creation, but renders, furthermore, the problem of evil acute (I will return to this claim toward the end of the article). According to Griffin, creatio ex nihilo distorts the Christian tradition and puts key Christian claims regarding divine agency in opposition to a scientific worldview. Rather than grounding our view of creation in creatio ex nihilo, we need to frame it in accordance with a minimal naturalism that respects the integrity of the natural world.

Although Mark Johnston launches his contribution from a somewhat different metaphysical angle, he suggests similarly that a main task for today's theologians is to formulate a model for divine action that “is in no way at odds with the form of the natural realm disclosed by science: that is, a causal realm closed under natural law” (Johnston 2009, 119–20). This is, contrary to a materialistic and scientistic version of naturalism, a “legitimate naturalism” that “arises out of the proper respect for the methods and achievements of science” (Johnston 2009, 43). Deborah Haarsma does not explicitly endorse a naturalistic starting point for theological considerations of divine providence. Nevertheless, her account raises issues regarding causal compatibilism between divine action and natural processes in the way that she stresses that we can “perceive design in nature even when scientists have a complete natural explanation” (Haarsma 2017, 222). ²

Hence, these scholars are united in seeking to explicate a model of divine action that can bring together two key claims:

1. God acts in the natural domain.

2. All events within the natural domain are (a) consistent with natural law and (b) causally reconstructible in terms of natural causation (Clayton 2004, 163).

The second claim leads, of course, to significant methodological and ontological restrictions for how we can imagine God's active involvement in the world. What they propose is, therefore, a form of theological compatibilism, which suggests that one can maintain a nondualist and noninterventionist view of divine action. God acts, but in a natural mode. My overall claim is that this model is fallacious and renders divine action ontologically meaningless. Skogholt, however, seems to think that these scholars are proposing a slightly weaker thesis, claiming that they are only naturalistic in the sense that God acts in a noninterventionist manner, and so they do not deny God's influence altogether (Skogholt 2020, 690). Proponents of NDC may like to argue for something closer to what Skogholt attributes to them. However, my main point is that they are unable to do so because they cannot uphold a sufficiently robust notion of divine influence.

Unfortunately, their commitment to a nondualistic relationship between divine causation and natural causation makes God's actions causally redundant. The naturalistic assumptions that underpin NDC means that when natural causes have been fully accounted for there is no causal work for God to do. This is why I take the main challenge for NDC to be a causal one; namely, the causal problem of overdetermination.

Skogholt seemingly thinks that my concern with NDC is that it makes divine action scientifically obsolete, in that God cannot enter into scientific practice and explanations. As he writes: “since God is scientifically redundant, God is, Leidenhag apparently thinks, made redundant tout court” (Skogholt 2020, 687). This is an unfortunate mischaracterization of my argument, but it helps to highlight the difference between my core argument and the concerns of ID theorists who argue for the inclusion of their theory in science curriculums. ID theorists fault theistic evolutionists for making divine action redundant in the sense that God is pushed out of scientific explanations. Since I do not think that God's actions can be verified by methods of natural science, I do not share this concern. Although my previous article focused on a critique of a particular account of divine action popular among some recent proponents of theistic evolution, it should not be taken as a positive argument for ID, which has significant problems and shortcomings of its own.

By drawing on Jaegwon Kim's influential work on issues of causal exclusion (Kim 1993, 1998, 2005, 2006), I made the argument that NDC, by negating any ontological difference between divine action and natural processes, is vulnerable to the criticism that it makes divine action causally redundant.

Let us look at Skogholt's discussion of Kim's argument as it pertains to divine action, to see if my argument still stands. His main point is that my critique does not hold water as “no theist could possibly claim that God's actions are ontologically secondary to natural processes, as mental phenomena are to physical phenomena in non‐reductive physicalism” (Skogholt 2020, 693). Skogholt thinks that no theist could hold to such a view, and I certainly agree that no theist should hold such a view. Skogholt is spot on with his concern that the prioritization of natural processes over divine activity is an unsustainable and undesirable position for any theist to affirm. This is exactly what the NDC position amounts to by framing divine activity in accordance with the stated “presumption of naturalism,” namely, that all divine acts within nature have to be reframed in naturalistic and scientific language.

For these naturalistically inclined thinkers, God's actions in the world need to be mediated by natural processes, as no dualism is allowed within physical reality. This is what Clayton maintains when he makes the broader assertion that there is “no qualitative or ontological difference between the regularity of natural law as expressing the regular or repetitive operation of divine agency and the intentionality of special divine actions” (Clayton 2004, 84). This means, following Johnston's legitimate naturalism, that God's “activity is manifest in and through natural processes alone” (Johnston 2009, 119). There is, as Peacocke says, “no need to look for God as some kind of additional factor supplementing the processes of the world” (Peacocke 2001, 474).

Hence, what these thinkers suggest is that there is a degree of supervenience, that is, ontological dependency, between divine causation and natural processes, such that divine acts need to be mediated by the natural regularities of the world. Although Skogholt thinks that no theist adheres to such an articulation of supervenience, this is in effect what a naturalistic construal of divine agency comes down to, which follows from the presumption of naturalism. The question, then, is not if these theists espouse supervenience, but whether such a construal is theologically and philosophically successful. Given that Skogholt simply claims that no theist adopts such a view, he never engages with my argument on this issue.

Moreover, while Skogholt addresses the issue of supervenience, he avoids or fails to address the main point in my engagement with Kim's work on causality and explanation, namely, that the NDC account of theistic evolution makes divine action causally redundant. The compatibilist picture of NDC entails that for any event in the natural order, this event can be given an entirely natural explanation in terms of the categories of the sciences. There are no gaps in nature, as the theistic naturalists often want to remind us of. However, such a view makes it exceedingly difficult to give any causal relevance to divine action in nature. God's action cannot be construed as an “additional influence” (Peacocke) in the world since the world's “causal principles are never interrupted” (Griffin), and all events have natural causes as opposed to supernatural causes (Clayton). Without having to repeat my main argument, I drew on Kim's discussion of causal overdetermination to show that this naturalistic articulation of God's relationship to the workings of the world renders the very notion of divine action causally superfluous. Divine action becomes an unnecessary metaphysical add‐on to an already complete natural view of the world. In what way does God add anything to event e if the cause of e is entirely natural? This dilemma is the Achilles’ heel of the NDC approach to theistic evolution. However, Skogholt's main counterargument against this critique, if I interpret him correctly, is that since God created and sustains the natural order, God is always causally involved in its development, and hence far from redundant. He references, in relation to this point, Peacocke's affirmation that “God continuously gives them [natural processes] existence” (Skogholt 2020, 689), and that God as “the creative source of all that is” brought “into existence a cosmos with fecund possibilities” (Skogholt 2020, 689). Although this clearly speaks to God's continuous sustaining action of the cosmos, it does not address how God's “special actions,” or special providence, are possible under these naturalistic constraints. It is special divine actions that are under consideration in the debate between TE and ID, because both camps claim a superior account of God's guidance, influence, and providence in conjunction with evolution. The affirmation that God created a world in the first place, or that God sustains creation, is not the issue at hand.

One might get the impression that these theistic naturalists, when reading Skogholt's article, are content with God's general sustaining activity, but this is far from true. ³ These NDC proponents suggest that God can act in particular and decisive ways in the world. Hence, Peacocke, for example, emphasizes frequently the theological necessity of special divine action, that “God could bring about the occurrence of particular events and patterns of events—those which express God's intentions. These would then be the result of ‘special, divine action’, as distinct from the divine holding in existence of all‐that‐is, and so would not otherwise have happened had God not so intended” (Peacocke 2006, 274). Likewise, Clayton's counter factual principle (Leidenhag 2019, 921–22) suggests that “had God not acted in helping to produce some effect, the effect would not have been identical to the state of affairs we in fact observe” (Clayton 2008, 104). Indeed, Clayton stresses that God acts in the minds of believers in a revelatory mode, and that God communicates directly with people in ways that are not fully capturable in scientific discourse. This is, according to Clayton, the level on which the “divine lure” operates (Clayton and Knapp 2011, 58). On the process theological perspective of Griffin, the divine reality acts in specific manners by “persuading” natural entities. This is “God's providential activity in us” which “is also the ultimate source of our moral and aesthetic ideals” (Griffin 2000, 97). These scholars stress God's more direct actions in order to pre‐empt the accusation that their proposals collapse into sheer deism (see regarding the problem of deism in their accounts of divine action: Clayton 1997, 192, 208; Griffin 2000, 28, 64, 261; Griffin 2004, 30; Johnston 2009, 50).

My main argument is that these scholars cannot successfully square their adherence to the presumption of naturalism with the view that God acts in specific ways in the natural order, without this producing Kim's causal overdetermination problem. If these scholars suggest that events within the natural order can be sufficiently explained through natural processes alone, then this leaves no causal room for God's special providence. This is, as I explain in my article, the problem of positing multiple, sufficient explanandum for one single event, e (Leidenhag 2019, 916–18). NDC is positing more causes than necessary for event, e, hence, this model encounters what Kim calls causal overdetermination. If one holds, as suggested by NDC, that e can be sufficiently explained in terms of some natural process, then God's providence is not adding anything causal to the picture. Skogholt never mentioned the problem of overdetermination and has done nothing to make me question this central argument. Indeed, his incredulity that any theist could propose something as bizarre as NDC and his desire to downplay the naturalistic dimension of proposals by Peacocke, Clayton, Griffin, and Johnston is encouraging to me. This misunderstanding arises largely from Skogholt's apparent equation between Aquinas's theology of divine action, which is not the focus of my critique, and NDC. As Skogholt never engages with the problem of causal overdetermination for NDC, my argument still stands.

THEISTIC EVOLUTION, ID, AND PARTIAL CAUSATION

Skogholt's response indicates that I take all forms of theistic evolution to be problematic. This is not true, and in my article I offer a constructive route toward a more viable form of theistic evolution through the idea of partial causation (Leidenhag 2019, 918–20). Partial causation allows for God and natural processes to jointly produce certain events in the natural order. My discussion of this is, however, strangely absent in Skogholt's response.

Let me briefly recap this view of causality and its implications for theistic evolution. Partial causation starts off from the intuitive idea that we need to appeal to a variety of causes in order to explain some event, occurrence, property, object, and so on. I used Kim's example of an automobile accident, and noted that we might need to appeal to more than one factor—perhaps an icy road, congested traffic, alcohol intoxication, and so on—to explain why there was an accident (Leidenhag 2019, 918–19). The upshot of this view of causality is that it avoids the problem of overdetermination as it does not posit several complete explanations for one single event. Applied on the issue of God's providence in nature, this view suggests that divine causality and natural causation might be “relevant to different aspects, properties, or features of the explanandum; hence, there is no epistemic competition,” and no issue of overdetermination (Leidenhag 2019, 919). I, therefore, suggested in my article that “partial causation is the best way to frame the theological and philosophical claims of” theistic evolution (Leidenhag 2019, 920). Once again, it surprises me that Skogholt seemingly ignored this discussion. Through the idea of partial causation, I seek to indicate an alternative route for framing the relationship between God and evolution. I would have appreciated Skogholt's response to this discussion.

Perhaps Skogholt assumes that framing divine action in this way entails ID, and that partial causation, therefore, is intrinsically incompatible with any meaningful formulation of theistic evolution? Indeed, I suggested that such a view of theistic evolution would pull it “closer to the antinaturalistic assertions of ID” (Leidenhag 2019, 919). Let me add some assuring words: partial causation does not necessarily entail ID as such, and adopting this view of causation does not mean that one needs to adopt the central claims of the ID community. First of all, a core claim within ID is that the divine presence in creation, in natural processes, and in organisms, is scientifically detectible. This is a claim that I take issue with.

When viewed from the perspective of faith or from within theology, various features of the natural world as described by science corroborate theological claims and indicate God's existence. However, I do not think (as ID proponents do) that we can infer such theological claims directly from scientific findings or observations. Divine design, or guidance, is neither scientifically verifiable nor falsifiable. Moreover, ID aims to propose a research programme on the basis of design. I doubt the success of this proposal, given the general agnostic attitude that many ID theorists take with regard to the identity of the Designer (see Leidenhag 2019, 928, footnote 1).

It should be noted that an NDC formulation of theistic evolution subscribes, ironically, to a similar methodological assumption of ID, namely, that the issue of divine design is something that needs corroboration through science. For the NDC proponent, this means that divine action needs to be articulated in a naturalistically acceptable manner and through scientific categories and theories. For ID proponents, it means that divine action is scientifically verifiable. These groups proceed on seemingly shared assumptions regarding how theology can and should be related to the sciences.

The task of articulating divine providence is a theological one, and not a scientific one. It is primarily a theological matter for which science can be relevant and helpful, but which needs no scientific validation. This sets my view apart from both NDC and ID. In fact, I agree with those who have argued critically that such tendencies to subjugate theology to the natural sciences reveal a tacit commitment to scientism (see, e.g., Taede Smedes's 2004 critical treatment of the divine action projects of John Polkinghorne and Arthur Peacocke).

To conclude this section, I suggest that a successful account of theistic evolution needs to be articulated along the lines of partial causation, which safeguards the irreducibility of divine causation. I would have appreciated Skogholt's input on this issue, and so I am disappointed to see that his response did not engage with my proposal.

IS NDC BETTER SUITED TO RESPOND TO THE PROBLEM OF EVIL?

Another rather perplexing feature of Skogholt's response is when he states that a second core argument of my article is that “if theistic evolution escapes the redundancy‐objection it is so vulnerable to the problem of natural evil” (Skogholt 2020, 686). Worth noting again is that my article does not critique theistic evolution per se, merely a naturalistic construal of it as exemplified by Peacocke, Clayton, Griffin, Johnston, and to some extent Haarsma. Because of this misunderstanding, a large part of Skogholt's discussion is irrelevant to my overall argument. He mentions several sophisticated responses—Christopher Southgate's, Bethany Sollereder's, John Polkinghorne's—to the problem of evil and suffering in an evolutionary world, yet none of these scholars adopt an NDC account of theistic evolution. Indeed, I find myself largely agreeing with such responses. The problem, here, once again is that Skogholt mistakenly thinks that my argument against a naturalistic construal of theistic evolution constitutes an argument for ID. This becomes clear when he suggests that the problem of evil is a problem for all theists, including those theists who affirm ID. He writes, “One can argue that the philosophical‐theological framework of ID is actually less well suited for handling the existence of natural evil” (Skogholt 2020, 692). I agree fully with Skogholt's remark on this point, and it should be noted here that I stated in my article that I find “the claims of ID theologically thorny, philosophically dubious, and scientifically problematic” (Leidenhag 2019, 910). It could very well be the case that ID is particularly ill‐equipped for dealing with the problem of evil.

Putting these misinterpretations aside, I am glad that Skogholt emphasizes the issue of natural evil as I believe that such a feature of the natural world poses significant challenges to a naturalistic accommodation of theistic evolution and divine action in general. In fact, the NDC proponents under consideration present the problem of evil as a case for naturalizing God's active presence in nature. Clayton, Griffin, and Peacocke link the problem of evil to an interventionist depiction of God. Clayton writes, “If one offers a ‘rich’ account of divine action, according to which God is frequently altering states of affairs, then one must also be prepared to provide an account of why he does not act at other times” (Clayton 1997, 11). Griffin offers a similar concern when he suggests that the interventionism of a supernaturalist depiction of God is “contradicted by the world's evil,” such as the Lisbon earthquake of 1755 (Griffin 2000, 27). The problem of supernaturalism, according to Griffin, is that “any evil in the world could have been prevented by God” (Griffin 2000, 28). Peacocke affirms a similar conclusion about the theological deficit of interventionism: “In the classical perception of God as transcendent and as existing in a space distinct from that of the world, there is an implied detachment from the world in its suffering. This renders the problem of evil particularly acute. For God can only do anything about evil by an intervention from the outside, which provokes the classical dilemma of either God can and will not, or he would but cannot” (Peacocke 2004, 105).

A traditional view of God, whereby God can perform miracles above and beyond natural regularities and laws, faces certainly a number of theological and philosophical challenges in light of the reality of evil. Why God permits evil remains a mystery and a problem. The severity of this problem should not be underestimated or lightly dismissed. However, the relevant question here is whether NDC offers any advantages to the traditional (“interventionist”) picture. If it did, then despite the problems of overdetermination outlined above, this would be a substantial reason to seriously consider NDC. However, NDC offers no advantages over traditional, nonnaturalistic accounts of divine action with regard to the problem of evil. Indeed, NDC has its own set of challenges in accounting for the existence of natural evil. While traditional models struggle to account for God's inaction with regard to the permission of evil, NDC struggles to avoid the more severe conclusion that God's action participates in natural evil and that God may ultimately be unable to defeat the evils of this world. I will list some concerns about this below.

CONCLUSIONS

This article has defended my previous argument regarding the causal problems of NDC in light of Skogholt's response, and how it leads to a blurred line between theistic evolution and ID. As Skogholt, unfortunately, underestimates the naturalistic ambitions of Philip Clayton, Arthur Peacocke, David Ray Griffin, and Mark Johnston, his response to my critique fails in many ways to hit its target. This is clear in the way that he uses Aquinas's theology as an interpretive lens for all accounts of theistic evolution. However, Aquinas proposes a significantly different metaphysical picture of divine causation in the world compared to those who affirm NDC. Given this confusion, I devoted the first part of my article to defending the very idea that these scholars seek to make divine action naturalistically acceptable.

Given that Skogholt indicates that I take all forms of theistic evolution to be intrinsically problematic, I recapitulated and defended the claim in my previous article that partial causation delivers a promising route for theistic evolution. As I remarked above, I find it surprising that Skogholt never engaged with my discussion on this, given that it could better frame God's activity in evolution. Toward the end of this article, I briefly highlighted the ways in which the problem of evil comes to the surface within an NDC account of theistic evolution, and that NDC does not necessarily fare better than traditional theism in light of natural evils in an evolutionary world. To conclude, I do not think that Skogholt has successfully undermined my argument that a naturalistic accommodation of theistic evolution is vulnerable to the problem of causal overdetermination.