An Examination of Van Til’s "Argument from the Uniformity of Nature"

A typical strategy of presuppositionalism is to focus on some area of philosophy which has historically been surrounded by controversy – a “problem” of philosophy in which consensus has historically been elusive and debate continuously ever-raging – and proclaim that the controversy is neutralized by adopting a specific brand of theism. It does not seem to matter to apologists that such a move does nothing to increase our understanding of the problem in question or that it invites yet new problems which apologists cannot resolve. This is because solving the problem was never their actual goal in the first place. On the contrary, their goal is to convince themselves of the alleged truth of their self-imposed delusion and to bamboozle as many unwitting sideliners as possible. This assessment is only confirmed by the fact that, even when the defects of their theistic “solution” to such philosophical quandaries are pointed out, apologists will continue on as though their defenses were entirely tenable. 

The appeal to ignorance underlying such a strategy should not be difficult to detect. Instead of pointing to empirical evidence demonstrating the existence of supernatural beings (e.g., prayer fulfillment, curing diseases by “laying on of hands,” restoration of amputated limbs, resuscitation of decedents, in-person meetings with angels – or the risen Jesus for that matter, etc.), apologists seek to put non-believers on the spot to “account for” some fundamental recognition about reality and articulate full-blown philosophical explanations solving some centuries-old debate found only in the hallowed chambers of academia. Wouldn’t it be most ironic if the ignorance which such apologetic strategies are purportedly aimed at exposing and exploiting actually haunts the proponents of those strategies in the first place? 

For apologists, a confession of ignorance on any point, especially one involving how the human mind operates, is like blood in water attracting sharks. It is into this gap in one’s knowledge that the apologist smells an opportunity to insert his god-belief as the answer which eludes the smartest thinkers. And the reason why otherwise intelligent individuals haven’t discovered the solution which is possessed by even five-year-old pew-sitters is that their minds are clouded with “sin” – moral enmity with the creator of the universe which prevents them from “seeing” the answer that’s right there before them.

If all philosophical problems were so easy! Just claim to believe in a supernatural being and voilà! Problem solved!

A stock example of this can be found in James Anderson’s paper If Knowledge Then God (IKTG), where he presents a formalized case for theism which exploits the so-called “problem of induction” as a conundrum for which resolution has purportedly evaded secular thinkers allegedly because of their “unbelief.” Anderson describes this problem as follows:

The infamous “problem of induction,” brought forcefully to our attention by Hume, refers to the deceptively difficult task of accounting for the rationality (construed in terms of truth-directedness) of inductive inferences. Why should it be thought eminently reasonable to make generalizations about future events on the basis of past events or to posit “causal laws” on the basis of finite observations of coincidental occurrences? Such inferences are grounded on the assumption of uniformity and order in nature, but the task of justifying that assumption without reasoning in a vicious circle has proven all but intractable. (IKTG, p. 21)

Oddly, I must have overlooked the part in the biblical narrative where the topic of the uniformity of nature, let alone the problem of induction, was instructively addressed. In fact, the only tie to Christianity as such that Anderson makes for this debating strategy does not come from words attributed to Jesus or an epistle attributed to the Apostle Paul, but rather from a quotation of a 20th theologian, namely Cornelius Van Til:

Our argument… would be that the existence of the God of Christian theism and the conception of his counsel as controlling all things in the universe is the only presupposition which can account for the uniformity of nature… (The Defense of the Faith, p. 103; quoted in IKTG, p. 22, emphasis Anderson's.)

Thinking back to the time when I sat regularly in church (multiple times per week, in fact), I don’t ever recall any discussion of Hume, the problem of induction, or the assumption that nature is uniform. I was already familiar with these topics since I had several years earlier taken introductory philosophy courses in college and had done a fair amount of side reading. So I would have recognized these concepts if they had ever been raised. On the contrary, what I recall was the overt suspicion that was encouraged from the pulpit for anything that smacked of “men’s wisdom” (cf. I Cor. 2:5 et al.); if it wasn’t discussed in “Scripture,” it did not warrant serious attention. Indeed, given the hostility for all things secular which pervaded my fellow churchmen at the time, I can confidently say that they would have been just as suspicious of Van Til as they would have been of Hume.


The Argument

 Anderson’s rendering of Van Til’s “argument from the uniformity of nature” runs as follows:

[P1] If theism is not the case, then one cannot account for the uniformity of nature presupposed by inductive reasoning. 

[P2] If one cannot account for the uniformity of nature presupposed by inductive reasoning, then beliefs based on inductive reasoning are not warranted. 

[P3] Beliefs based on inductive reasoning are warranted. 

[C] Therefore, theism is the case.

First off, it’s curious that the argument does not explicitly seek to conclude that a god exists. Rather, believers must be satisfied with the lukewarm “theism is the case,” which can be interpreted in a variety of ways. What exactly does it mean to say that “theism is the case”? Presumably it means “theism is true,” which would in turn need to be interpreted as “the belief that a god exists is true.” Why not disentangle beliefs as such from the thrust of the argument and focus on what is supposed to be the case regardless of anyone’s beliefs? Also, there’s nothing in the guts of this argument which indicates that only Christian theism could satisfy its call for a theistic “account for” the uniformity of nature. Unless this is just the start of a game full of “but wait, there’s more,” I see nothing here that calls for the threesome of Christianity’s trinity, a dying and rising Jesus, a formula for atonement, etc. But we shall not trifle with these matters. Instead, let’s consider the premises of this argument in order of their presentation.

The initial premise, P1, apparently rests on the unstated assumption that the uniformity which we observe in nature must be the product of conscious activity, specifically supernatural conscious activity. To the extent that this in fact is what P1 tacitly assumes (and given the argument’s goal of validating theism, it must in fact assume this), the fact that this assumption is hidden from view and not stated and defended outright raises the suspicion that apologists like Anderson recognize deep down that the presuppositionalist agenda would have no chance of lifting off the platform if it were openly affirmed.

But why suppose that the uniformity of nature is a product of conscious activity? We certainly do not observe any conscious entity installing uniformity on an otherwise disuniform nature. Can the apologist point to any evidence that consciousness has such an ability to begin with? Or are we supposed to be satisfied with merely imagining this to be possible, and leave it at that? Even more to the point: can the apologist point to any evidence of an existent that does not by its nature, apart from what any consciousness does, exhibit uniformity such that it would need some conscious activity to make it uniform? The view that the uniformity of nature is a product of conscious activity reduces the uniformity of nature to a merely subjective projection, which can only mean that the theistic “account for” the uniformity of nature has no objective basis.

Concerns of this sort are systemically ignored, not – I suspect – because they’ve been probed and pondered and apologists are still formulating their theories on the matter (I’ve not found any discussion of these concerns anywhere in the presuppositionalist literature that I have perused), but because the presuppositionalist argument is not in fact aimed at increasing our understanding on the matter in question. On the contrary, its goal is to put the non-Christian on the defensive by miring him in an intricate tangle of pointless bickering, which suffices to divert attention away from what Christianity actually affirms and keep it safe from critical scrutiny. Meanwhile, no matter what answers the non-Christian provides in response to the apologist who is purportedly so interested in the fundamentals of induction, his answers will be rebuffed for committing one fallacy or another, even if no such fallacy can legitimately be found.

Such practices are just one of several reasons why I often suspect that apologists themselves do not actually believe the worldview they champion and defend, but are instead most likely using their apologetic tactics for some ulterior social purpose to shore up some psychological deficiency – e.g., propping up their own vanity, enlarging their influence over others, seeking for approval of others, hoping to impress others, fluffing their plumage, etc. After all, they have a confessional investment to protect, and admitting that their religious affirmations are faulty is a bigger cross than the one they posture as shouldering in a public display of feigned virtue. It may be that they do not actually believe it, but fear disbelieving it, so they go through the motions “just in case.”

If, contrary to what theism would have us imagine, the uniformity which we observe in nature is something consciousness discovers - as opposed to something which consciousness creates or wishes into place, then we are on track towards an objective understanding of the uniformity we observe in nature. The objective view is that the uniformity of nature is essentially the applicability of the axiom of identity to all of existence. The axiom of identity is the recognition of the fact that to exist is to be something specific, to have an identity, to have a nature. Because to exist is to have identity, everything which exists has identity. It would be a self-contradiction to say something exists and yet deny that it has any identity whatsoever. Similarly, causality is the axiom of identity applied to action.

Contrary to many philosophers’ denials, actions do in fact have identity, and this is demonstrable by the fact that every human language has verbs which distinction one category of action from all other actions. Writing is distinct from skiing just as sleeping is distinct from skydiving. When a company CEO is giving a speech, we do not say that he is grooming his dog or stocking a grocery shelf. All these actions are distinguishable because they have identity. And the law of causality is the affirmation that the action of an entity depends on the entity’s nature. Since identity is concurrent with existence, uniformity is inescapable; there is no actual alternative to the uniformity we observe in nature just as it is not an option for an entity to be itself and not something other than itself.

But it’s a mistake to suppose that merely providing an “account for” the uniformity of nature explains induction. As I pointed out in an entry on the topic which I posted over a year ago, if nature is uniform, it’s uniform for all creatures great and small. It’s uniform for me just as it is for my cat. But my cat will never be able to draw inductive generalizations like I can. Inductive reasoning is a type of reasoning, an activity of the mind, an activity which human minds can perform. The objective view is that induction is essentially a conceptual process of cognition, which explains why my cat cannot draw inductive inferences even though nature is just as uniform for him as it is for me. My cat does not have a consciousness capable of identifying and integrating its objects in conceptual form, but I do. And already just by forming our first concepts, we are in effect forming primitive generalizations. The concept ‘tree’ for example, which we form on the basis of perceptual awareness of a few specific trees in our firsthand experience, includes all trees – all trees that exist now, that have existed, and that will exist. Concept-formation, then, provides the cognitive blueprint for inductive reasoning by expanding man’s awareness beyond the reach of his senses to include existents which he will never perceive. The uniformity of nature, qua the concurrence of identity with all of existence, is a necessary precondition for inductive reasoning, but by itself it is not sufficient: we need to be able to form concepts. Thus a full answer to “the problem of induction” requires a robust theory of concepts. But Christianity has no theory of concepts – not even a bad one. So presuppositionalism is pointed in the wrong direction if the goal is to expand our understanding of induction.

P2 is also suspect. Why is it automatically the case that “if one cannot account for” the uniformity we observe in nature that “beliefs based on inductive reasoning are not warranted”? I do not accept this premise at all. Consider children: even children draw generalizations based on their relatively limited experience, and quite often they are in fact warranted. A child can tell you that if you tip over a glass of water, it will spill its contents out and get everything beneath it wet. And he would be right! But ask that same child to “account for the uniformity of nature presupposed by [his] inductive reasoning,” and he’ll likely look at you with a dumbfounded and yet unphased blank. He might even reply sassily that it’s obvious. If so, he’d be closer to the truth than if he claimed an imaginary being wished uniformity onto the realm of existents. But while such a reply would not earn him very high marks in a 600-level philosophy course, his generalization about tipping over a glass full of water is nevertheless wholly warranted. One can even base predictions on it. That’s a testament to how powerful the human mind is.

Applying P2 to what we can certify about theism, given that it leads believers down the completely wrong track of assuming that the uniformity we observe in nature must be the product of the conscious activity of a being which is only available to the human mind by means of imagination, we can know that, given the falsehood of such an assumption inherent in the apologist’s worldview, presuppositionalism itself cannot account for the uniformity of nature. Proposing a false account of something does not count as presenting a cogent account for the phenomenon in question. So Anderson’s argument implicates itself, and I’d say fatally so.

Now Anderson does include a footnote to P2 which apparently anticipates such objections as I have raised here. There he writes:

The charge of epistemic level confusion can be avoided if premises [P1] and [P2] are understood to refer to an inability in principle rather than one indexed to a particular reasoner. If Susan cannot account for the uniformity of nature, it does not follow that her inductively-inferred conclusions are unwarranted; but if the uniformity of nature cannot be accounted for in principle, by any human reasoner, then the warrant of all inductively-inferred conclusions is cast into doubt. (IKTG, p. 22)

Given this disclaimer, I’d say Anderson’s argument needs revision, and not one reserved for footnotes. For on a clear reading of P2, the objection that philosophically uninformed individuals draw inductive conclusions that are fully warranted all the time is obvious. Even if one wants to say that “in principle” a child is unable to “account for the uniformity of nature presupposed by inductive reasoning” (and given his lack of cognitive tools, that would indeed be the case), it is still a fact that many inductive assessments reached by children are nevertheless wholly warranted. I suspect even many Christians who are sympathetic to enlisting the problem of induction as a debating device would acknowledge this.

Suppose P2 were revised to read as follows:

If one cannot account for the uniformity of nature presupposed by inductive reasoning on non-Christian grounds, then beliefs based on inductive reasoning are not warranted.

This, I think, is closer to what apologists want essentially to argue, for this version pins the root of the deficiency at issue on non-Christian thought, which is what apologists are seeking at the end of the day to implicate. This makes explicit what Anderson likely implies with his insertion of “in principle.” But even here I suspect that some Christian thinkers would find this version dissatisfactory on the grounds that, even if one cannot account for the uniformity of nature on non-Christian grounds, their beliefs based on inductive reasoning are still warranted in spite of their non-belief since their god nevertheless still ensures that nature is uniform. I doubt very much that believers are going to find it acceptable to suggest that their god’s activity in “the created order” is suspended because someone fails to accept their god-beliefs.

Unfortunately, had Anderson qualified P2 of the Argument from the Uniformity of Nature in such a manner, it’d be clear that he was just arguing in a circle at that point. I can only suppose that Anderson struggled with several versions of this premise before settling on the one he eventually published, still finding it necessary to include his qualifying footnote.

At any rate, it is not the case that one needs to look to any form of supernaturalism to “account for” the uniformity of nature given the truth of the Objectivist axioms: nature is uniform in the sense that identity is wholly concurrent with existence. Since there is no such thing as an exception to the fact that to exist is to have identity, there’s no such thing as a disuniform nature. And this state of affairs is not a product of conscious activity – it obtains and persists regardless of what anyone thinks, beliefs, wishes, desires, emotes, doubts, prefers or imagines. It obtains and persists independent of conscious activity.

The third premise – “beliefs based on inductive reasoning are warranted” – can only spell the defeat of Christian theism, indeed of any theism which endorses a doctrine of “miracles.” For belief that miracles are possible would at the bare minimum need to allow for the possibility that (some) consciousness has the ability to negate the uniformity we observe in nature. The biblical narrative demonstrates what believers must accept as reality in regard to water raining from the sky (when the a supernatural consciousness decides in anger to send it), water in pots (it can become wine when the right consciousness wishes it so), water in lakes (one can walk on it if the right conscious actions are taken). If nature were indeed uniform, angered decision-making will never cause rain, wishing will never turn water into wine, and believing with all your might will never make it possible for you to walk on unfrozen water. By contrast, if consciousness possessed the ability to defy the uniformity of nature, no reliable inferences about the nature of anything, whether individually or collectively, could be advanced with any confidence whatsoever, for those inferences would always be subject to being overturned if the right conscious actions were performed. But if in fact nature is uniform and “beliefs based on inductive reasoning are warranted,” one would be fully justified in rejecting the miracle tales found in the biblical record for their blatant inconsistency with the recognition that nature is uniform. Thus linking the uniformity of nature with theism is an exercise in self-contradiction.

So much for the argument from the uniformity of nature as it is presented here. Each premise is riddled with defects which can be easily detected, so it seems to me that anyone defending it seriously must either not really know what they’re talking about or have some goal other than advancing knowledge of the nature of induction in mind.

Of course, it could be a mixture of both. For Christian apologists under the Vantillian spell, the problem of induction is an opportunity, an occasion to stump for theism, not a real problem to be solved by increasing one’s understanding of this vital feature of human cognition. And this is clear from the apologist’s focus on the uniformity of nature as though explaining this were all that is necessary to validate induction as a form of inference. As I pointed out above, if nature is uniform, it’s uniform for all organisms which possess consciousness. But while my cat will never be able to draw the general conclusion that eggs will crack open if they’re dropped on a hard surface, human beings can. So there’s more to induction than merely grasping that nature is uniform and explaining why it is uniform. Induction is a cognitive process, and one will never find a passage anywhere in the Christian bible which explains this process.


Generalizations about Future Events

In setting up his formulation of the Argument from the Uniformity of Nature, Anderson poses the following question, which we saw above, as an encapsulation of the philosophical quandary which gives rise to the problem of induction:

Why should it be thought eminently reasonable to make generalizations about future events on the basis of past events or to posit “causal laws” on the basis of finite observations of coincidental occurrences? (IKTG, p. 21)

If we dial things back a few steps, let us ask: how is it that we are able to form the concept ‘table’ – a mental integration which includes every table which exists now, which has existed in the past, and which will exist in the future, on the basis of “finite observations”? No one who has formed the concept ‘table’ will ever observe every table which the concept subsumes, and yet this inability does not prevent us from forming a mental unit which encompasses all tables, even hypothetical ones. Could this ability to form open-ended units subsuming things we will never observe on the basis of incidental observations play any role in this? I argue that it in fact does.

To address Anderson’s question, let’s start by analyzing it. So I ask, then: What is an “event”? However it may be defined, it’s clear from common use that an “event” involves some kind of action. But then we should ask: What is action? We can use other words to denote the same phenomenon – e.g., motion, movement, activity, process, behavior, operation, etc. – but ultimately this concept cannot be analyzed in terms of more fundamental concepts. Like the axiomatic concepts of existence, identity and consciousness, the referents of the concept ‘action’ “can be specified only ostensively, by pointing to instances” (Leonard Peikoff, Objectivism: The Philosophy of Ayn Rand, p. 7). “’Action’ is the name for what entities do” (Ibid., p. 13).

But like “events,” actions are not phenomena which exist independently of the entities which perform them. An action is an action performed by some entity just as an event is an occurrence involving one or more entities engaged in some kind of activity. To speak of actions and events is to include the entities which perform those actions and participate in said events. As David Kelley puts it, “you can’t have a dance without a dancer.” So to speak of events, we are in fact speaking about entities which exist, which means we’re speaking about existence.

When we form the concept ‘action’, and by extension related concepts like ‘event’, we have formed a mental unit which includes all actions which are taking place now, which have taken place, and which will take place. This is the task of concept-formation, and it requires no appeal to supernatural beings to understand or perform. And the product of this process expands our awareness beyond what we perceive firsthand, in fact beyond what we will ever perceive firsthand. So just by forming the concept ‘action’, we have in place one of the rudimentary concepts necessary for induction, for already our awareness has extended beyond the reach of what we can directly observe.

The next question I would ask is: To what does “future” refer? And similarly, to what does “past” refer? These of course are temporal concepts, distinguishing the actuality of actions which have already taken place (“the past”) as against those which are taking place now (“the present”) and those which may happen at a later time (“the future”). It’s important to keep in mind that temporal concepts do not denote concretes as such: “the past” is not an existent available to our perceptual faculties, and neither is “the future.” Rather, they denote entities and their actions collectively with respect to when they performed their actions. But having formed the concept ‘action’, we now have awareness of actions which have taken place in the past as well as those which may take place in the future.

Notice that when we speak of events in the context of Anderson’s description of the problem of induction, we need to modify “events” with temporal descriptors (“past” and “future”) in order to draw an internal distinction. Why would we need to do this? The answer is because ‘event’ is a concept which in its unmodified state includes all events, whether occurring now, having occurred in the past, or going to occur in the future. Notice also that location is not restricted by use of a modifier – e.g., events which happen in North America as opposed to events which happen in Asia. This is because “past events” denotes events which have taken place regardless of where they may have taken place, just as “future” events includes events which may happen in the future regardless of where they might take place. Thus we see that two aspects – time and place – are omitted from the scope of the meaning of the concept ‘event’ and thus would need modification applied in order to distinguish one subcategory from another, namely temporal subcategories in the present case. The very need to use modifiers to make such distinctions confirms the validity of the Objectivist analysis of concept-formation as the means by which we acquire awareness of things beyond the reach of our senses in open-ended mental units.

We do this with other concepts as well, whether they are concepts denoting organisms (dogs, cats, ants, human beings, etc.), inanimate naturally occurring objects (rocks, rivers, clouds, hills, snowflakes, etc.) or man-made objects (tables, pens, tractors, mailboxes, etc.). As Peikoff puts it, “the concept ‘table’ integrates all tables, past, present, and future, regardless of these variations among them” (Objectivism: The Philosophy of Ayn Rand, p. 84). The same is the case for other concepts as well.

Thus, when we speak of events as such, unless otherwise noted by some means (e.g., by use of modifiers like ‘past’ or ‘future’), we’re talking about all events regardless of when and where they may take place. Time and place are irrelevant to the applicability of causality. We did not need to distinguish between past and future events in order to form the concept ‘event’, and yet the concept ‘event’ gives us a cognitive grasp of all events regardless of time and place. So why would we need to distinguish between past and future events in order to formulate generalizations about the causal structure of events? The answer is: We don’t need to do that. And we don’t need to do that because when we talk about events, whether past, present or future, we are not talking about different realities or different universes with opposing norms of physics, but rather the one and the same reality in which causality is identity applied to action regardless of time or place. This goes right back to the axiom of existence: Existence exists, and only existence exists, and this axiom delimits the field of rational inquiry. (In fact, we make inductive inferences about past "events" all the time; try investigating a crime scene or performing an audit of last year's financials some time.)

All this is to underscore the need to understand induction as a conceptual process whose operation is already active in forming the very terms used to call the reliability of induction into question in the first place. The “problem of induction” as it is argued by skeptical philosophers and enlisted by presuppositionalists as an apologetic device, ignores the conceptual basis which informs inductive reasoning. “Conceptual awareness is the only type of awareness capable of integrating past, present and future” (Ayn Rand, Introduction to Objectivist Epistemology, p. 57).


Conclusion

To summarize, the Argument from the Uniformity of Nature is hereby refuted. The uniformity of nature is not a product of conscious activity – either human or supernatural, but rather the concurrence of identity with all of existence. This is something we discover by looking outward at reality, whereas the Christian position requires us to look inward and treat the contents of our imagination as though it were real and thereby pretend that something imaginary is responsible for ensuring by means of conscious activity that nature is uniform. Since existence exists independent of conscious activity and since the imaginary is not real, the theistic “account for” the uniformity is not only untrue, it is also a completely unnecessary hypothesis.

Moreover, induction is a conceptual faculty, one that is not explained fully by simply “accounting for” the uniformity of nature, but rather must be understood as an extension of the process of forming concepts. Induction is essentially the application of the principle of causality (action has identity and is dependent on the nature of the entity performing it) to entity classes. We form concepts of entities (entity classes) on the basis of objective input which omits time and place as essential characteristics – thus expanding our awareness of objects beyond the range of our senses. Our concept of ‘egg’, for instance, includes all eggs, past, present and future, and applying the principle of causality to this entity category enables us to draw the general inference that dropping an uncooked egg on a hard surface will result in its shell breaking and its loose and gooey contents spilling out.

There is no mystery here, no need to postulate a supernatural consciousness beyond the universe, no justification to give up on the human mind as though surrendering to mystical notions were warranted. Indeed, the very reason why apologists think it’s necessary to point to a god in order to answer the problem of induction (presumably that the human mind cannot know anything beyond the range of one’s senses) can only cancel out the theist’s “solution” to that problem, for it involves the claim to know something beyond the range of one’s senses (for we certainly do not perceive the believer’s god).

For more on the matter, see the following entries on my blog:

Presuppositionalism and Induction 

Presuppositionalism and Induction: Apologists Courting Hume 

Presuppositionalism and Induction: Exhuming Hume 

Presuppositionalism and Induction: The Humean Condition 

Presuppositionalism and Induction: Thoughts on the Uniformity of Nature

by Dawson Bethrick