One of the means that secularists have used to achieve dominance in the culture over the last 250 years has been the manipulation of language. Key terms have been modified, and new terms coined, which slant the ‘rules of engagement’ between Christianity and secularism against Christianity. Three terms in particular: ‘natural law’, ‘miracle’, and ‘methodological naturalism’ have been affected. If we do not expose and correct this sophistry, an honest debate is not possible. At root, these issues reflect the clash between worldviews that must ultimately be accepted for reasons outside of science.
One of the most common ways to talk about well-established scientific ideas is to refer to them as ‘the laws of nature’, ‘natural law/s’, or ‘scientific law/s’. These phrases conjure up notions of an orderly universe, and convey the idea that science can teach us much about the universe. The notion of ‘law’ can be a helpful metaphor for understanding the regularity we see in how the universe works, but it is just a metaphor.
One thing true of all metaphors is that they are only helpful to a point before they break down. They do so because, from the point of view of formal logic, all metaphors break the law of non-contradiction. When they are pushed too far, different things end up equated in ways that amount to false statements. For example, saying “Mark is a teddy bear” with reference to Mark’s gentleness equates the physical softness of the teddy bear with the gentle character of Mark, which amounts to saying that Mark has a gentle character. However, saying that “Mark is a teddy bear” with reference to Mark’s physical makeup does not work because Mark, a human, is clearly not made of stuffing! Not all the attributes of teddy bears can be attributed to Mark, and vice versa. This doesn’t invalidate the use of metaphors because everyone intuitively knows this, and we don’t use a metaphor to say that two different things are logically identical—only that they are identical in some significant and relevant property in a particular sense.
However, as people jettisoned the Bible and its doctrine of providence and came to accept a ‘fuzzy positivism’ in its place from the 19th century onwards, many people seemed to forget that the notion of ‘natural law’ is just a metaphor.2 Because of this, and that the metaphor is so fitting, we have forgotten to pay attention to those areas where the regularities of nature do not correspond to human laws. Failure to recognize the metaphorical nature of ‘natural law’ has brought undue confusion over how the world actually works. The problem is that people often conceive of ‘the laws of nature’ as materially equivalent to human laws. If someone does what human law forbids, it is described as ‘breaking’ or ‘violating’ the law. People then apply this same notion of law to ‘the laws of nature’, and consider a miracle in the same way they consider an illegal act—as if some precept has been ‘broken’ or ‘violated’.3,4 The biggest mistake people make when they equate ‘natural law’ and human law is that they make natural law prescriptive. However, there are two major errors with understanding the concept of ‘natural law’ like this.
First, when people think of ‘natural laws’ as prescriptive, there is in fact a subtle but significant difference in what ‘prescriptive’ connotes in comparison to human laws. Human laws are prescriptive messages—they are coded information. Human laws are not inextricably bound to any given physical form—one can communicate the law “you shall not murder” in any human language, and in numerous different physical formats (e.g. writing on a piece of paper, Morse code, speech, etc.). Moreover, a human law cannot render the act of murder physically impossible. Herein lies the difference—when ‘natural law’ is seen as prescriptive, it is not conceived of as merely coded information, but is something materially able to make certain states of affairs physically impossible. However, ascribing ‘natural law’ a tangible existence like this commits the fallacy of reification, which treats abstract concepts as concrete objects. ‘Natural law’ is an abstract concept—it is coded information, like human laws, which formally describes regular patterns observed in nature. As such, ‘natural laws’ have no more power to cause the regularities of nature than a redraw of a map of New Zealand can change the physical coastline.5
Second, although human laws and ‘natural laws’ are both coded information, they are different types of messages. Human laws prescribe what people should and should not do, e.g. the law forbidding murder. The law itself says nothing about how people actually act—it is a command, not a statement of fact. ‘The laws of nature’ are different: they are formal descriptions of regular patterns observed in nature—i.e. they are statements of fact. They either correspond with reality or they do not. However, it is meaningless to apply the notion of correspondence with reality to a human law such as forbidding murder, because it is an imperative command, not a statement of fact. Conversely, the statement: “murder is ethically wrong” is a statement of (ethical) fact, and is generally implied to form the basis for the law forbidding murder. This means ‘the laws of nature’ are not formally prescriptive, and are more accurately seen as descriptive concepts.
Consider the First Law of Thermodynamics. It is one of the most well-established natural regularities that we have discovered. It states that the total amount of energy in an isolated system remains constant over time. It is often restated in this form: “Energy cannot be created or destroyed; it can only change form”. However, as Roger Larmer points out, the two formulations are not equal.6 The first formulation is deducible from the second, but not vice versa.
The first formulation details the conditions in which energy remains constant; it says nothing about whether there are any agents outside the isolated system that can create or destroy energy. It is a simple statement of inductive inference that has ontological limits, reflecting the inherent limits of experimental methodology. Both naturalism and theism can accommodate this formulation of the First Law of Thermodynamics. The First Law is consistent with an assumption that matter/energy has an eternal and necessary existence, which is foundational to many forms of naturalism. However, miracles in which energy is created or destroyed do not invalidate this formulation of the law because they are performed by beings that are not bound by the ‘isolated system’. Therefore, the theist who believes that the universe is a created, contingent reality in which secondary physical causes operate can also affirm this formulation of the First Law. It just means that the universe is not an isolated system with respect to God (and possibly other non-physical agents). In this instance it is God as the First Cause that grounds the First Law of Thermodynamics as a reality of secondary causality. In other words, the First Law of Thermodynamics is an accurate description of the way God sustains the universe.
The second formulation is a claim that energy is completely indestructible. It follows necessarily from this that energy has both an eternal and necessary existence, which a priori rules out theism and creation ex nihilo. This formulation amounts to a statement of metaphysics, not experimental physics. It also necessarily entails at least a de jure naturalism.7
The first formulation is clearly a superior scientific formulation because it doesn’t go beyond the bounds of experimental inductive inference—it maintains that there are limits to what the experiments can validate. It doesn’t tell us what is as if science can actually do that by itself; it merely sums up what we regularly observe. And that is all ‘natural laws’ based on science can ever say.
Miracles (typical candidates include Jesus’ Resurrection, the creation of life, and the creation of the universe ex nihilo) are commonly seen as ‘violations of the laws of nature’, adopted from David Hume (figure 2).8 This conjures up the notion that miracles are somehow illegitimate and unnatural, as if they destroy the harmony of the natural world.9,10 This can be seen in sources that are sceptical about the possibility of miracles and in sources that wish to affirm miracles in some way. Buzzwords such as ‘rupture’,9 ‘suspension’,11 and ‘intervention’,12 are used to describe their relation to ‘natural law’ with a clear pejorative connotation. Sometimes, it’s used to flat out deny the possibility of miracles,9 sometimes it’s used to deny that miracles are ever provable,13 and sometimes it’s used to describe an aversion to a particular way of understanding God’s action.14
All these ways of conceiving of miracles are beholden to Humean categories, which either exclude the possibility of miracles a priori or consider miracles a ‘problem’ of divine action to be solved. As a result, they make miracles sound like a problematic and/or incoherent concept that inevitably dissuade people from considering miracle claims. However, the only thing miracles are practically unanimously believed to demonstrate is that if one has ever occurred, then naturalism is false.15 Therefore, the lack of clarity and coherence is the fault of the definitions, not the events ‘miracle’ describes. Therefore, it is invalid and unhelpful to work with Humean categories for the definition of ‘miracle’. Defining miracles as ‘violations of the laws of nature’ tends to silence any possible criticism of a naturalistic worldview.
There are plenty who think that miracles are impossible by definition, and for this they appeal to a particular understanding of miracles as ‘violations of natural law’.16 However, this necessarily reifies the concept of ‘natural law’ (we’ve already looked at the problems with this), and it blatantly assumes metaphysical naturalism (which is a worldview with no logical basis17). Such question begging is palpably ironic, as Chesterton incisively pointed out:
Moreover, this ‘violation of the laws of nature’ definition of miracles renders the concept of ‘miracle’ incoherent, which defeats the purpose of a definition. With the naturalistic (i.e. ‘miracles can’t happen’) assumption behind the modern connotations of ‘natural law’, the definition of ‘miracle’ becomes ‘a violation of what is inherently inviolable’, which is self-contradictory. It’s no wonder that so many people think that ‘miracles’ so defined are impossible! However, events such as Jesus’ Resurrection simply don’t fit this definition, so another definition needs to be found.
Agnostic New Testament scholar Bart Ehrman follows Hume very closely in defining that a miracle is the most implausible explanation for any event by definition.13 Ehrman gives two reasons for this. First, he argues on the basis of frequency of an event—since billions of people have, for example, never walked on water or risen from the dead, it is also inherently improbable that Jesus ever did such things. Second, he insists that historians, as a matter of method, cannot speak about God qua historiography. However, there are a number of flaws in his reasoning.
First, it is highly disputable whether the constancy of ‘natural law is in fact a universal testimony of humanity. There have been many reports of miracles both ancient and modern that are based on solid evidence.19 Second, it assesses the intrinsic probability of a miracle occurring from the relative frequency of miracles—miracles are so rare that we can assume that whenever a miracle claim is made, it is false.20 However, the determining factor for the intrinsic probability of a miracle is not the relative frequency of miracles, but the existence of an agent able and willing to perform a miracle. So, if we assign a high prior probability to atheism, then miracles are indeed intrinsically improbable. However, if we assign a high prior probability to biblical theism, then specific miracles, such as the Incarnation and the Resurrection, become highly probable. So the prior probability of miracles is worldview dependent, and not intrinsic to the event. Whatever one believes about the prior probability of miracles, it is clear that they should not be defined by notions of prior probability because all that does is smuggle naturalistic assumptions into the concept before we get a chance to investigate the claim.
Second, why is it impossible to speak about God as an actor in history? That God is unobservable is irrelevant because neutrinos are also unobservable, and yet they are postulated to explain the effects observed in physics experiments. In the same way, when God causes a miracle, there are physical effects. That God is non-physical is also irrelevant because in terms of our ability to observe, there is no difference between a neutrino and God—physicality makes no difference. What matters is whether the unobservable explanatory entity we invoke provides the best explanation for the event and its context. Therefore, if miracles are real-world events, they are historical events with theological implications, not theological ‘events’ devoid of history.
Finally, the reasons given for his definition contradict each other.21 The problem is that the first reason assumes they can be assessed as historical events, and the second reason assumes they cannot. To talk of miracles as events with respect to history, even inherently implausible ones, is to talk of miracles as historical events. However, if historians cannot speak about miracles as history, then that amounts to saying that miracles are not historical events. Therefore, Ehrman ends up saying that we cannot speak of historical events as history—a self-contradictory statement. Putting ‘God’ into the mix doesn’t make the reasoning any less self-contradictory.
In essence, Ehrman wishes to have his Humean cake and eat it too, and as a result ties himself up in knots. However, this is what occurs when one follows Hume’s intent in his definition of miracles.
Humean notions of miracles and ‘natural law’ affect even those who profess some form of theism. The Divine Action Project is a consortium of scholars from different backgrounds that have produced a series of articles and books since 1988. They have set about understanding divine action as non-interventionist—i.e. God does not intervene in the cosmos as an outsider, but rather interacts (even miraculously) in the world in a way that doesn’t involve ‘suspending’ ‘natural law’.12 There is a particular aversion to conceiving of God’s action in the world as ‘intervention’,14 ‘breaking natural law’,21 ‘suspending natural law’ 11—all connotations attributable to Hume. The idea is that God would not set up a world to run in such a way and then interfere with the way that it runs, as this would make God inconsistent.
The sticking point seems to be the idea that God is ‘interventionist’. But what exactly does it mean for God to be ‘interventionist’? Plantinga has demonstrated that this prevailing notion of ‘interventionism’ is simply too vague to provide anything useful to interact with in dialogue about divine action and miracles, regardless of whether we perceive the world in a Newtonian or quantum mechanical way.22 Besides, how can God ‘interfere’ in a world that he owns and runs? On the scriptural understanding of divine action (i.e. creation and providence) notions of ‘intervention’ are superfluous. As one example of many, Robert Russell’s ‘Non-Interventionist Objective Divine Action’23 fares better if we just drop the ‘non-interventionist’ from the account entirely. Why should we be beholden to a concept not only foreign to Scripture, but also so vague as to be a superfluous add-on?
Plantinga identified the three problems alleged by the ‘Divine Action Project’ as problems with an ‘interventionist’ account of divine action: (1) ‘intervention’ happens so seldom that it does few people any good, (2) it destroys the integrity of creaturely free will, and (3) it means that God is arbitrary.24 Plantinga’s answers all three brilliantly: (1) God is not beholden to our conceptions of his action—remember Job, (2) the integrity of free will is not predicated by the absence of ‘intervention’, but on the presence of regularity and predictability, which are grounded in God’s unchanging character, and (3) a combination of answers (1) and (2).25 The irony is that these look like typical atheistic complaints voiced against theism! These problems are the same regardless of what words we use to describe God’s action with, and sticking ‘non-interventionist’ at the head of a fairly typical theistic account of divine action doesn’t solve the ‘problems’. The fact is that the ‘Divine Action Project’ researchers go far beyond what we can actually know about God’s action, and they do so in a way that ironically is beholden to the post-Enlightenment ways of thinking about divine action they so vehemently criticize.
Fundamentally, all these understandings of ‘miracles’ are anachronistic. They owe their popularity to the Enlightenment, which sought to establish a clean break between revelation and the ‘real’ world of reason and experience.25 ‘Natural law’ was recast in a deistic sense alien to its original usage, where the deity wound up the cosmic ‘clock’, and it has been running by itself ever since.26,27 However, the Bible doesn’t conceive of miracles in this way, and neither did most of the ancient world.28,29 As N.T. Wright points out with regard to the words used for what we would call a ‘miracle’ in the Gospels:
This creates a picture of miracles as marvels and/ or signs within the real world which had behind them an extraordinary power as their cause—whether from the person themselves or from the agent they claimed to represent.31 This is very different from the popular connotations that miracles as ‘violations of the laws of nature’ tends to attract. It means that, regardless of whether we can observe the one who caused the miracle, if miracles occur they produce detectable effects in the real, empirical world by definition. In other words, if miracles occur, they occur in this world, period. Miracles are therefore matters of history and experience. As such, miracle claims cannot simply be dismissed by assuming naturalism a priori because, if miracles occur, it necessarily follows that naturalism is false.
This tends to sit rather uncomfortably with many today because we intuitively recognize that there is a big difference between miraculous events and everyday experience. And there is a big difference; miracles are identified as historical events that create a set of circumstances for which natural regularity fails as an explanation.32 However, this distinction between ‘natural law’ and ‘miracle’ is not meant in any absolute Humean sense, which is really what most of the objections to miracles tend to assume. Rather, it provides a valid and useful heuristic distinction that helps us recognize that there is a qualitative experiential difference between ‘natural law’ and ‘miracle’ from a human perspective in history. In a similar way, we can make a valid and useful distinction between, for example, chemistry and biology as different scientific disciplines, while acknowledging that the distinction between the two is often fuzzy.
While human experience forms the basis for the distinction between ‘natural law’ and ‘miracle’, it is God’s action with respect to his creation that unites the concepts. ‘Miracle’ and ‘natural law’ draw a distinction between events that both come under the umbrella of God’s providence. God upholds the laws of nature, by regular sustaining action, and He can specially act—and allow other immaterial beings to act in the physical world (i.e. cause miracles)—without the world going haywire. Therefore, the laws of nature are not the fundamental reality upholding the cosmos; but God’s will is.
In the end, all the definitional and methodological problems with miracles can be explained by an implicit assumption of Humean categories. Under such assumptions ‘miracle’ is defined in unclear and incoherent ways, so they prejudice the investigator against the possibility of any event to which such a definition is improperly applied, such as Jesus’ Resurrection, the creation of life, or creation ex nihilo. And since none of these events are ‘miracles’ in any Humean sense, it is an unfair prejudice. Definitions matter—and so do words. It is better to go with a definition of miracles (and the Bible has presented a way of understanding miracles all along) that is both clear and consistent,29,33 so provides a far better way to understand miracles than do post-Enlightenment categories.
Another common term used when discussing science/ religion issues today is methodological naturalism (MN), which maintains that science can only explain what happens in the universe in terms of observed or testable natural mechanisms. According to theistic evolutionist John Haught, MN “maintains that as far as scientific knowing is concerned, nature is all there is”.34 It is an assumption of method, not an ontological claim, and is held up as equivalent to the scientific method.35 It is typically justified by appealing to the utility of science. But that begs the question that MN is the hallmark assumption of science. And the underlying question remains: why does the world run in such a regular manner? So how do proponents of MN address these issues?
MN has been used as an excuse to justify metaphysical naturalism, the worldview that says that nature is all there really is. The argument goes: since we have apparently never observed anything that contradicts MN, and anything that isn’t science is irrational superstition, so we can justly reify ‘natural law’ and proclaim that nature is all there is.36 This is the constant refrain of the so-called ‘New Atheists’— Richard Dawkins, Daniel Dennett, Christopher Hitchens, and co. Theists who embrace MN as equivalent to science clearly refuse to move from MN to metaphysical naturalism, because they rightly refuse to buy into the pure scientism of the metaphysical naturalists—there’s more to this world than science can ‘tell’ us.36,37 Nevertheless, the movement in the other direction is obvious and straightforward: MN is an obvious deduction from metaphysical naturalism (figure 3). Assuming that nature is all there is as a matter of method is clearly not a problem for someone who already assumes that nature is all there really is.38 The issue has to do with word association—it’s natural to associate ‘methodological naturalism’ with the worldview of metaphysical naturalism. On the other hand, can we associate a method called ‘methodological naturalism’ with a non-naturalistic worldview so easily? Attempts to do so abound, no doubt.36,38,39 The problem is that the popular media (and wider culture) is now pervaded by a naïve positivism40 that promotes the close association of MN and metaphysical naturalism. So this move is an easy one to make if we use the language of MN to describe science; both terms have ‘naturalism’ in them, so people understandably assume that MN is simply the methodological handmaiden of metaphysical naturalism.
So, in spite of this rather clear tendency to move from MN to metaphysical naturalism, why do so many theists continue to support MN? Many (both secularists and theistic supporters of MN) consider MN as a necessary demarcation criterion between ‘science’ and ‘non-science’: a line is drawn between ‘science’ (equated with MN) and ‘non-science’ (e.g. miracles) to separate them from each other. Some think that this tactic is a valid way to distinguish science as a discipline from other disciplines, such as theology and history.36 However, such a view is demonstrably false. This tactic was used explicitly in the McLean vs Arkansas Board of Education41 and Kitzmiller vs Dover42 cases to stop any questioning of evolution or introduction of Intelligent Design into American public school science classrooms.43 Thanks to a positivistic mindset that dominates mainstream academia and the wider culture,41 anything branded as ‘science’ is automatically judged reliable, while anything that is branded as ‘non-scientific’ receives the opposite treatment—it becomes branded as illogical, sloppy, or backwards. And since MN is the criterion by which this is judged, anything that has even the slightest hint of supernaturalism is automatically deemed irrational.
Moreover, this tactic is ultimately invalid because the term ‘science’ has proven impossible to define. There is no discipline that can be ruled out as science without ruling out at least some disciplines and theories that are intuitively understood as science.44 What’s more, this has been is widely known, agreed upon and even considered uncontroversial among philosophers for nearly 30 years!44
It is important to note that this tactic is not just employed by secularists; many theistic evolutionists45,46 (and even those sympathetic to their cause47) also employ this same tactic. However, in doing this they seem to display a destructively ambivalent attitude. Many of them are happy to adopt this demarcation tactic, but only with regard to evolution, long-age geology, abiogenesis, and the age of the cosmos, unlike the secularists who use MN more consistently. The irony is that many of them believe in miracles, such as Jesus’ miracles, and especially his Resurrection, on exactly the same grounds as creationists—the same grounds creationists use to defend the miracle of special creation. And yet, many of them have practically the same dismissive and derisive attitude towards ID advocates and creationists as secularists do (sometimes more so!), when in fact these theistic evolutionists have practically the same epistemology of miracles as creationists! If it were merely a disagreement over the necessity of a miracle for the origin of life, then the issue would be much less acrimonious. However, what is simply a miracle claim for a creationist is commonly construed as ‘non-science’, and even ‘anti-science’, by theistic evolutionists. In a society where labelling something ‘science’ gives it intrinsic authority and reliability, and denying the label marks something as irrational, such language is not merely the language of disagreement—it is the language of derision. Such language is used to silence a dissenting position, not engage it in fruitful debate. The irony is that in using such derision they are sawing off the branch they are seated on.
Why even use the term ‘methodological naturalism’ at all?48 Some feel it accurately describes how scientists have always worked. However, this can only be adduced if one radically redefines ‘naturalism’.49 As Hooykaas demonstrates,50 science blossomed with the deconstruction of aprioristic scholastic speculations with an emphasis on God as the orderly and free sovereign sustainer of nature. This meant that they still viewed nature as rational and subject to laws because the God who sustains it is himself rational and orderly, but man’s reason couldn’t figure out that order a priori—he had to drop his idealistic preconceptions and actually look at nature to figure out how God orders it. Thus science blossomed not under the adoption of MN as commonly understood, but under the assumption of biblical theism. The other thing is that Christianity comes with terms already geared to help us understand the role of science.51 So why bother adopting terms that will inevitably be counterproductive to describe how we do science when Christian theology already has adequate terminology? Moreover, even in a proximate sense, the validity of the scientific method rests on the assumption that the cosmos is ordered, not that natural cause and effect is the norm.52 We can see the difference when we look at metaphysical naturalism. There is no reason inherent in naturalism that nature should be ordered because there is no reason natural cause and effect must be ordered.53 Therefore, order in the cosmos is an ad hoc assumption for naturalism.
Both the orderliness of the cosmos and the impulse for a posteriori investigation of the natural world are simple deductions from the attributes of the biblical God.53 So, perhaps a better term for the scientific method might be something like ‘methodological Yahwism’—i.e. assume Yahweh exists, and science can proceed. The obvious objection is that Yahweh’s existence cannot be assumed because such language would prejudice science in favour of Christianity, and science is supposed to be ‘worldview-neutral’.54 But this confuses the usability of the method with its epistemological grounding because it invalidly supposes that science is an autonomous discipline. Just because the method can be used by people of all different worldviews, it doesn’t mean that the scientific method is deducible from all those worldviews. There is a reason that science only blossomed in a Christian context—it provided the reason not just to expect an ordered world, but gave motive to go out and investigate it as well.28 Moreover, if one objects to ‘methodological Yahwism’ on these grounds, then it necessarily follows that one must also object to ‘methodological naturalism’ on the same grounds. As using ‘methodological Yahwism’ would imply that the scientific method inherently favours Christianity, so using ‘methodological naturalism’ to describe the scientific method implies that science inherently favours a naturalistic worldview as opposed to any other. It conveys the idea that all scientists (including self-conscious theistic scientists) are naturalists, at least implicitly. Therefore, methodological naturalism is not a worldview-neutral term.55
What’s worse is that, if anything, naturalism is one of the most anti-scientific-method worldviews around.52 Naturalism has nothing to do with an ordered cosmos (the true methodological assumption of the scientific method). At least ‘methodological Yahwism’ would be a historically accurate term for the reason science blossomed, and would be philosophically justifiable! More worldview-neutral terms would be ‘methodological orderliness’ or ‘methodological regularity’, which also make plain the true assumption of the scientific method which ‘methodological naturalism’ obscures. It seems that ‘methodological naturalism’ is yet another term used (intentionally by some, unintentionally by others) for the advancement of metaphysical naturalism.
Terms such as ‘natural laws’, ‘miracle’, and ‘methodological naturalism’ have been used in the public square very effectively by secularists to shut out any notion that God interacts with the real world. Some of these, such as ‘miracle’ and ‘natural laws’, were older terms given new meanings that ruled out theism. Others, such as ‘methodological naturalism’, are terms favourable to anti-Christian ideals coined to express the fundamental tenets of what scientists do in such a way that it makes science sound intrinsically anti-Christian and anti-Bible.
What is the solution? A lot of hard work. We need to define our terms clearly, properly identify the worldviews at play, and clearly identify where those worldviews conflict. A number of creationists have been conducting such a program already,56 but much more needs to be done. We need not fear—Christianity is ultimately the only defensible worldview because it’s the only true one.
I wish to thank the two anonymous reviewers for their comments and suggestions, which were extremely helpful in revising an earlier draft of this essay.