IBRI Research Report #22 (1984)

David P. Hoover

Copyright ©1984 by David P. Hoover. All rights reserved.


Gordon Clark is an uncommon presuppositionalist. Dating roughly from the publication of his “Wheaton Lectures” in 1968, he has increasingly stressed the complete impossibility of empirically acquired knowledge. According to Clark, the Bible, and only the Bible, can be known. All observation based truth-claims and all inductive arguments are logically worthless for apologetic purposes. This essay attempts to follow the progression in Clark’s thought which leads to such a thoroughgoing skepticism (with regard to a knowledge of the contemporary world). The essay looks, in turn, at Clark’s reasons for rejecting the reliability of perception, Clark’s adoption of the Bible as axiomatic, and Clark’s propositionalism. Although these subjects are treated critically, the aim of the author is to achieve clarity on what Clark holds and why.


Although the author is in agreement with the doctrinal statement of IBRI, it does not follow that all of the viewpoints espoused in this paper represent official positions of IBRI. Since one of the purposes of the IBRI report series is to serve as a preprint forum, it is possible that the author has revised some aspects of this work since it was first written. 

ISBN 0-944788-22-X

Gordon Clark's Extraordinary View of Man and Things


The writings of Gordon H. Clark are remarkable in a number of ways. The style is exceptionally clear, witty, and fast-paced—engaging. More importantly, Clark’s philosophical position, while claiming few pure disciples, has commanded the respect of virtually all fellow Christian apologetes—including even archrival John Warwick Montgomery who esteems Clark to be one of the two greatest living Christian philosophers (the other Montgomery considers to be Alvin Plantinga).1  In 1968 a Festschrift appeared honoring the philosophical career of Gordon Clark (The Philosophy of Gordon H. Clark). The vo1ume contains Clark’s “Wheaton Lectures” and a medley of critical responses written by several Christian philosophers and thinkers along with Clark’s own replies to his critics. In putting the essays of the Festschrift together, Ronald Nash has performed a very valuable service for all who want to become acquainted with Professor Clark’s considerable contribution to the apologetic literature of the past fifty years.

But most remarkable is the fact that Clark is an uncompromising Idealist and Rationalist within an intellectual milieu which puts a premium on an empirical approach to knowledge and to knowledge acquisition. To a nonphilosophically literate evangelical community, the labels “Idealist” and “Rationalist” convey, perhaps, little information. On the other hand, to those with some background in philosophy (or apologetics) they may even function as code words, marking off battle lines. My own intent for introducing them into a discussion of Clark’s thought is that they behave descriptively, not evaluatively. My own bias is quite opposite Clark’s regarding the overarching epistemological and metaphysical issues, but it is surely possible and profitable to conduct a friendly “in-house” discussion of the issues which divide us. My general aim in this essay is to provide an introduction to Clark’s presuppositionalism, a presuppositionalism, as will be seen, that stands in stark contrast to the presuppositionalism of Cornelius Van Til. More narrowly, I want to offer clarification to several of the more radical consequences of Clark’s general philosophical position. Although these consequences are not hard to identify in Clark’s writings, it seems to me that Clark is not widely understood concerning these within the evangelical community. As might be gathered from the essay title, one fundamental question is, What sort of world of men and things does Christianity stand to make sense of if Clark is correct in his main line of thought? To suppose we get an inkling into the answer to this question merely by noting that Clark is a presuppositionalist is a serious mistake. As with any presuppositionalist, Clark begins with what he cannot prove, but there the similarity stops. Clark differs with other presuppositionalists with regard to the nature of what is presupposed and with regard to the prospect of additional knowledge content based on what has been presupposed.


Whether in A Christian View of Men &Things(‘52), Three Types of Religious Philosophy (‘77), or in his more recent Language & Theology (‘80), Gordon Clark delivers, in a highly readable style, a relentless opposition to empiricism. But unlike any other presuppositionalist (at least that I’m aware of) Clark’s opposition to empiricism is based on what Clark regards as the logical impossibility of inductively acquired knowledge.

Clark’s opposition to empiricism can be appreciated along two related but distinct lines: (i) No particular inductive truth-claim can be validly argued since inductive arguments are, one and all, formally invalid. But (ii), and even more basically, the defender of inductively acquired knowledge must be committed to a theory of images (or sense-datum theory) and hence cannot escape skepticism. For convenience I will call the first problem “the ‘illogic’ of inductive reasoning” and the second problem “the non-cognitivity of sentience”.

(i) The “illogic” of inductive reasoning
There is perhaps no emphasis for which Gordon Clark is better known than his stress on rigorous logical demonstration. Throughout his writings he places great importance on beginning with a first principle or axiom and then validating each successive step by the canons of deductive logic. Arthur Holmes has drawn attention to this fundamental methodological standpoint as Clark’s commitment to “the geometric ideal, a certain kind of system, one modeled on ‘geometry with its theorems and axioms, its implications and presuppositions.’”2  In philosophy or apologetics, Clark insists, “one must begin at the beginning, not half-way along the road; and from the beginning every step must be validated.”3  It is not surprising, then, that Clark regards any reasoning chain that contains formal “logical gaps” as simply fallacious. Valid reasoning, for Clark, is the only profitable reasoning one can engage in, and it is definitionally deductively valid reasoning. Inductive arguments whose premises merely support their conclusions are necessarily bad arguments—in fact, worthless.

Clark’s uncompromising position on induction is easily illustrated, but needs to be taken seriously in its radical consequence for apologetics as well as for human knowledge about the creation. To illustrate we go to the final chapter of Three Types of Religious Philosophy. In that work Clark cites an exchange that took place between some Westminster Seminary students (along with Westminster professor Vern S. Poyithress) and John Warwick Montgomery (then professor at Trinity Evangelical Divinity School). The students’ response was to an article written by Montgomery and published in Christianity Today.4  The article, entitled “Having a Fuddled Easter?”, made the point that presuppositionalist reasoning is fuddled since its apriorism cannot allow evidential proof or confirmation for the resurrection of Jesus. Montgomery’s point is that the aprioristic reasoning principle which requires one to simply start with the Bible and all its content on faith would be extremely disabling if applied generally. How, for example, could one “distinguish the Bible [one] claims to start with apriori, from Playboy magazine?”5  Is not a presuppositional beginning a beginning from within an evidential vacuum and therefore entirely arbitrary?

The Westminster responses to Montgomery’s challenge are evaluated by Clark as “quite good” but not delivering the most telling blow to Montgomery’s position. The Westminster men, according to Clark, imply that Montgomery’s inductivism is correct in principle but inappropriate due to man’s fallen nature and sin. To this Clark responds that:

the fact of sin is really irrelevant. Those who stress human depravity in this argument give the impression that Adam before the fall was and the elect in heaven will be able to construct valid arguments, based on sensations, to prove the veracity of God. But the crushing and basic reply to the Montgomery position is that all inductive arguments are formal fallacies. Historical arguments to prove the resurrection and all sensory arguments to prove its significance are as bad as and indeed worse than trying to prove that a triangle contains two right angles with a protractor. The method is impossible, and it is impossible for an innocent Adam as for guilty Barabbas. Sin cannot make a valid argument fallacious, nor can perfect righteousness make a fallacy valid.6
To make explicit what is perhaps already obvious, Clark’s position on reasoning (arguing) can tolerate no non-formalizable elements at all. Reasoning either exhibits formal completeness, and so validity, or is formally incomplete, and so fallacious. But notice the thunderous consequences, consequences that Clark himself seems quite cheerful about. (i) There are, for Clark, no non-problematic “data” in the creation around us because no sentiently discriminated “datum”—the presence of my typewriter before me, for example—can be deductively argued to certainty. Otherwise put, the best argument that I can produce in behalf of the truth-claim that I am seated before my typewriter is an inductive argument—hence a fallacious argument and logically worthless. But if so, then (ii) the very method that would deny to Montgomery an evidential support for the truth-claim that Jesus rose from the dead would also deny all scientific access to the creation, whether in pursuit of archaeological artifacts or, say, in the development of a medical treatment for acquired immune deficiency syndrome (AIDS). Clark is rigorously consistent with these consequences in taking up apologetic method. For the Christian apologete there can be no archaeological truths. The apologete can take up archaeology only in its potential for logically embarrassing the opponent to Christianity. That is, “scientific discourse” can be taken up only ad hominem against the individual who puts credence in science.7
(ii) The non-cognitivity of Sentience
So far, Clark’s position has been seen to exclude inductive argument as a viable instrument for apologetic discourse. But it seems to me that Clark makes a distinguishable and even more basic and strenuous point about induction. The first point concerned logical formalism alone. Inductive arguments are not deductively valid arguments. But what of the content of the premises of an inductive argument? Can content taken up in propositions be gleaned in some non-argumentative way? Is simple observation possible? Consider the Center for Disease Control in its energies to find a way to fight AIDS. The testing of blood samples, the interviewing of AIDS victims, ostensibly requires the sentient examination of data. The doctors and scientists make observations and systematically keep track of their findings by propositionally focusing them. But what of the cognitive access that issues in the all-important observations? Are the human senses of sight, hearing, and touch sufficiently reliable to conduct CDC’s most important medical projects? These are certainly odd questions. The very legitimacy of questioning the reliability of sentience (the capacity to discriminate the environment by the senses) seems to imply that there are perhaps other (non-sentient) modes of cognitive access, other methods for finding out, that might be used instead. And if sentiently gleaning content goes part and parcel with inductive method, then the question becomes whether there is a better method than observation and testing to find a cure for AIDS. But what might an alternative be like? Divination? Inspecting tea leaves? Saying the first (second, third) thing that pops into one’s head? (Observing tea leaves is, of course, ruled out.) Or perhaps there is no such disease as AIDS since its occurrence is alleged on the basis of observation! And by parity of reasoning there are no diseases afflicting anyone! But if we find that we cannot help but admit that we do have the capacity for making observations in God’s world and that sentience is inextricably tied up in the making of observations, we will find Clark’s position even startling. But it is important to state that Clark’s position cannot be put away by a string of rhetorical questions. Clark’s position is that sensation is unintelligible, that one cannot coherently specify how sensation results in cognition. Why does he hold this view?

All too briefly, let us survey the arguments. Empiricism, says Clark, is wrong because it can provide no account of how the “raw reports” of sight, sound, touch, taste and smell graduate to rational cognitions—interpretations. According to Clark, any would-be theory about the acquisition of knowledge by means of the senses must depend on a theory of images (with the exceptions of Berkeleyan and subjective idealism and Pyrinhonian skepticism).8  This difficulty is developed as follows.

The problem of constructing knowledge from images is at least four-fold: (a) By definition, an image is not the real thing and is at best merely a representation of the real thing. This representational function, however, cannot be established without some way of getting back of the images, at least on some occasions, to certify that the images really do “picture” a mind-independent reality. And on empiricist principles we can never do this. (b) The having of images needs to be universal for mankind if empiricist epistemology is to be correct theory, but some persons claim not to be able to evoke the sort of images that are requisite, say, for the empiricisms of Hume, Russell, or the early Wittgenstein. An example of such an image might be the imaging of the color red in the dark (or in the absence of any red object), running a tune through one’s mind, or calling up a “picture” of one’s mother. Gordon Clark, for one, along with a small but impressive percentage of his former students, finds these feats impossible to perform.9  But if empiricism requires them, then empiricism is defective on a very fundamental point. (c) From sensations, sense-data, images, pictures, sensabilia, or whatever one wishes to call them, it is impossible to get to an ordered and interpreted perception.10  For how might one construct the concepts of the highly interpreted world of perceptual common sense out of mosaics of sense-data? Clark lauds Brand Blanshard’s attempt to exhibit perception as inference from sensory givens but refuses Blanshard’s effort because the latter “nowhere shows how to distinguish a true perception from a fallacious inference.11  And (d), there is a family of arguments designed to show the unreliability of sense perception. The arguments are the stock in trade of empirical skeptics and go by a variety of names: the argument from illusion, the argument from differential certainty, the argument from hallucination, and the argument from the physiology of perception.12  Although Clark may not advance these skeptical arguments under these names, he is quite concerned to make their main points. I take these points to be two: first, what one is directly or immediately aware of in sense perception are nothing but immaterial entities called sensations (or sense data); and second, that since the sense data of misperception or of distorted perception are phenomenally indistinguishable from so-called veridical perception, it is impossible to empirically tell veridical and non-veridical perception apart. But if the two cannot be told apart, then the distinction between the two sorts of perception is without a basis.

This is much too quick but my main intent is to array in a very cursory fashion a number of arguments that Clark would consider as at least collectively decisive against the cognitivity of sentience. And the radical consequence for apologetics is that the very idea of perceptual evidence from the creation becomes otiose. Also, although my intent is not primarily a refutation of these arguments, it is worth noting that within the philosophical literature of the past forty years there are hard-hitting replies to all of these arguments (or argument types). Two brief comments on the above arguments will have to suffice here: (i) For the most part, present day empirically oriented philosophers are not image theorists in Clark’s sense. They do not see their position(s) as requiring, in the manner of Brand Blanshard, a successful logical reconstruction of interpreted perceptual wholes out of a mosaic of static sensory atoms.13 And (ii) it is questionable whether the battery of arguments (d) establishes that the objects of sensory awareness are always immaterial sensory data rather than, say, physical objects. Am I, for example, touching (and feeling) typewriter keys or immaterial tactile sensations? It is at least linguistically odd to speak of seeing, hearing, touching, tasting or smelling sensations. Why not admit sensation in a cognitive account of sentient discrimination by saying that in the having of certain sensory stimuli one sees the chair, hears the telephone ring, touches the desk, etc.?14  As for the argument from phenomenal indistinguishability, it can be contested whether, say, dream phenomena and waking phenomena really are phenomenally indistinguishable. But even conceding that they are indistinguishable, it surely doesn’t follow that the obiects of the two sorts of perception have identical ontic statuses. That is, the object of my waking state awareness can be ontically dissimilar from the object of my dream state awareness without my having perceptual criteria for marking the crucial sensory object versus physical object distinction. A holograph tree and a physical tree, for example, might have identical phenomenologies to a given percipient without being the same sort of object. So even if a dream tree is phenomenally indistinguishable from a waking tree, it does not follow that if the dream tree is nothing but collected sensations the waking tree is also nothing but sensations.

To summarize, it is the sort of opposition to empirically acquired knowledge that has concerned us. Clark resists empiricism because he believes that sentiently acquired knowledge content is logically impossible, because empirically acquired knowledge is inconsistent with (Clark’s) rationalism. Because no observational truth-claim can be argued to deductive certainty (from an axiomatic beginning), and because perception itself cannot be rationalistically constructed from sensation, Clark categorically rejects empirical knowledge—whether of one’s own self or of the creation. The question then becomes how Clark can acquire knowledge content about himself, others, or the world if God has not given man the capacity for sentient discrimination of objects in the world. To answer, we need to refer to Clark’s “geometric ideal” in reasoning—a doctrine which is tantamount to maintaining that knowing is co-extensive with what is deductively provable. There is never knowledge without proof and there cannot be proof without an axiomatic (and hence indemonstrable) starting point. Thus there can be a Clarkian knowledge of God, the world, self, and others if, but only if, these candidates for knowledge can be rigorously deduced from a suitable axiom. And this brings us to Clark’s view of the Bible—i.e., its nature as truth-—and his “geometric” utilization of it.


Gordon Clark and Cornelius Van Til offer a very strong contrast in their respective views of “men and things.” Both men are presuppositionalists in the important sense that they begin their understanding of the world through the “eyes” of God’s revelation; they begin with what they cannot prove—namely, the Bible. The Bible functions for them as a necessary condition of all possible proving. But it seems to me that the nature of this presupposed Bible, and so the method for utilizing it with respect to understanding one’s world of men and things, are strikingly different between the two thinkers. I hope to state this difference as descriptively as can be done, but my remarks may at least appear evaluative if the reader s intuitions are more with Van Til than Clark. Also, it is worth remarking that Van Til’s ultimate concern for presupposing is that one’s presuppositions have complete explanatory power for the world one experiences. Clark’s ultimate concern for presupposing, it seems to me, is more nearly the philosophical one of overcoming skepticism (by exhibiting at least one logically irrefutable truth).15

To return to the nature of the presupposed Bible (the presupposed truth of all Biblical content),16  Van Til, from square one, needs a Bible which will give him confidence about who he, Van Til, is, about how the world is which he, Van Til, must make his way in. The Bible, that is, must be a lamp for one’s feet and a light for one’s path in the sense that the Bible stands to illuminate and explain one’s contemporary situation. In contrast to this, given Clark’s categoric denial that man is endowed to have sentient access to the world around him, the Bible’s significance is bound up in its status of being true regardless of the prospect of its certifiable application to one’s contemporary world. If the Bible is true then skepticism is defeated—period.17  We’ll now draw out some of the implications of Clark’s view of the Bible—i.e., the Bible taken in the strict logical sense as an axiom.

(i) The Word of God as Axiom
There are, perhaps, two senses of the term axiom in common use—one sense employed in everyday parlance and the other by logicians and those concerned with formal systems. In the loose, or everyday, sense we might say, for example, that God’s existence is axiomatic. In saying this we might mean to imply that God’s existence is fundamental to all else that exists, that God’s existence is so basic that it cannot be meaningfully challenged, or that God’s existence is in some way self-evident. But if we intend the proposition “God exists” to be an axiom in the logically strict sense—in the sense that an axiom (or axioms) head(s) a system— then we imply, in addition, that the axiom’s usefulness in the production of further insight is by way of extracting formal entailments with certain logical rules of inference. If a rule of inference cannot deliver it, the “insight” is not validly implied by the axiom—or, simply, the axiom does not imply it. Now imagine that all knowledge is a single system headed by an axiom (or axioms) in the logically strict sense. If a certain proposition—for example, “I am seated before my typewriter”—is neither mentioned in nor deducible from the axiom(s), then that proposition is not a part of knowledge .

It is Gordon Clark’s conviction that any item of knowledge must be part of an axiom system whose exclusive access is by deductive technique. Knowledge content for Clark, is deduced from an axiom, not (also) sentiently gleaned from an outer environment. In his “Wheaton Lectures” Clark identifies his axiom formally as “The Bible is the Word of God. “18  From this axiom one can easily deduce such propositions as “In the beginning God created the heavens and the earth” and “Jesus wept.” With a little greater logical sophistication, perhaps, it is possible to deduce “The Apostle Paul once left his cloak at Troas.” The first two statements are easy because they occur explicitly in the Axiom (Gen. 1:1 and Jn. 11:35) and can be retrieved as axiomatically true, taken singly, by invoking the rule of simplification. This rule states that whenever one has a conjunction of true statements (in our case, the conjoined statements of the Axiom), one can infer the truth of each statement taken separately. In notation, given that P and Q are both statements and that (P & Q) is itself a true statement, then either conjunct (whether P or Q) taken by itself will be a true statement as well. “The Apostle Paul once left his cloak at Troas” is more difficult because it must be derived from a number of propositions and cannot be inferred by simplification alone. The difficulty is compounded further because the Biblical sentence which serves as its only logical support is an imperative and not an indicative:

The cloak that I left at Troas with Carpus, when you come, bring, and the books, especially the parchments. II Timothy 4:13
Although I should want to affirm the truth of “The Apostle Paul once left his cloak at Troas” on the basis of this imperative, it is not clear to me that the text from II Timothy formally entails it. It would be preferable to say, I think, that “The Apostle Paul once left his cloak at Troas” is conversationally implied than to say that it is formally implied. If formal implication is the case, one must be able to give a formally complete chain of propositions from the Axiom to the entailed statement. Consider the following set:
1. The Bible is the Word of God. (Axiom)
2. The Bible is true. (Character of God, 1)
3. II Timothy 4:13 is a sentence in the Bible.
4. II Timothy 4:13, however, is an imperative and does not take a truth-value .
5. But if the Bible is true, then an imperative in the Bible implies the factuality of what is conversationally implied. (Assumption)
6. It is conversationally implied that the Apostle Paul left his cloak at Troas in Paul’s request to Timothy to bring it to him. (Assumption)
7. “Paul left his cloak at Troas” is factual. (1-6)
8. Therefore, the Apostle Paul once left his cloak at Troas. (1-7)
But I have obviously relaxed the requirement of formal implicitness in (5). (5) is an assumption, not a logical entailment of (2). There are additional implications of II Timothy 4:13, though not formal ones. It seems to me that implied in Paul’s request is that Paul remembers leaving his cloak at Troas and that Timothy has the capacity for searching for and finding the cloak. Moreover, Timothy has the capacity for making the discriminations needed to make his way to Rome from Troas—certainly no mean feat on Clarkian assumptions. It is far from obvious that Clark would concede these additional implications. For example, he seriously questions the trustworthiness of memory in a reply he makes to Ronald Nash:
Whether one wishes to prove that Brutus killed Caesar, or whether the law of gravitation explains or describes planetary motion, or whether memory is trustworthy, or whether sensation is the basis of perception, or whether sensation itself is only a word without meaning—all these entail problems so difficult that none of us can be sure of any answer.19
And what Clark has to say about induction rules out what I find implicit about Timothy’s capacity to respond to Paul. Although it is surely possible that God revealed to Paul that he left his cloak at Troas (never mind how), I do not see how, on any ordinary regarding of the letter in question, Timothy can be allowed a capacity for finding the cloak and for travel to Rome if sentient cognitivity is denied to him. Perhaps Clark can supply the answer here, I don’t know; my main point is that formal entailment gives out surprisingly early in much of the Bible and informally discerned implication consonant with the reader’s own experience must be relied on if the Bible is to be understood.
(ii) Defined & Undefined Terms in the Axiom
We have seen so far that Clark requires the Bible to be axiomatic in the strict sense and that the system that this axiom heads is the extent of knowledge. In the Festschrift, Clark states this position concisely:
For one last time therefore we must summarize and emphasize the whole argument. Consider the philosophy of science outlined in the preceding lecture. There it was claimed and argued that experimental science produces no knowledge whatever of the processes of nature. The laboratory can devise no method for determining whether the earth moves while the sun stands still or whether the sun moves while the earth stands still. Nor can the greatest amount of experimentation explain why two smooth pieces of marble adhere so stubbornly to each other. Neither can physics observe anything moving in a straight line. It is incorrect therefore to complain that the axiom of revelation deprives us of knowledge otherwise obtainable. There is no knowledge otherwise obtainable.20
The total extent of (human) knowledge, then, is the Bible. But we need to press this claim to see its more radical consequences. We have already seen the tenuousness of relying exclusively on formal entailment in order to understand, for example, what is being communicated by Paul to Timothy in II Timothy 4:13. Conversational implication is far richer than formal entailment, and yet our awareness of this net of implications determines our understanding of the text. Clark implies, however, that the contemplator of Scripture can bring no knowledge content to Scripture if that content cannot be deductively arrived at from the Axiom. This means, for example, that if “the heavens” and “the earth” are terms that can be understood and their referents known, both the definitions and the referents (designata) must be given in Scripture.

As Clark will readily acknowledge, there are both defined and undefined terms in the Axiom. The defined terms are the more prime objects entering into cognitive judgments because they, at least, can be understood as well as articulated by man—or so it would seem. Prime candidates for defined terms (for Clark) include the great doctrines of the faith: the atonement, total depravity, the sovereignty of God, etc. An example of an undefined term may be taken from Genesis 3:6. There it is affirmed that Eve could see the forbidden fruit, that the forbidden fruit “was pleasant to the eyes.” Obviously the Bible’s use of the verb to see in this instance cannot be intellectualized (although “see” is used this way several times in the same passage (vss. 5-7)). Eve is said to have engaged in an activity which engaged something called an eye. That is, Eve “saw” with, or by means of, her “eyes.” But “see” in this sense, and “eye” in this capacity, are nowhere defined or further analyzed in the Axiom. When asked at a debate at Covenant College what this distinctive sense of “see” means, Clark said that he did not know, but that physics couldn’t tell either. So although Eve did it, we do not know what she did when she “saw” the fruit because the term is nowhere defined. Nor is there any hope of knowing what “ground” or “tree” or “fruit” mean since none of these terms can be assigned a content. The terms themselves are essential to the Axiom (from the human perspective) only as logical ciphers, logical placeholders to make the system work formally.

But if this is the situation relative to the Axiom, what of the prospect of contemporary knowledge? Suppose, for example, that I want to determine whether “Hoover is using the Axiom” (1) makes sense, and (2) whether it is true. Since on Clark’s assumptions knowledge is the system headed by the Axiom (as identified), the individual terms if they can be assigned content, must be defined in the Axiom, and the proposition as a whole is certifiably true only if it is either mentioned in the Axiom or deducible from the Axiom. Taking the truth of “Hoover is using the Axiom” first, we see at once that it is neither mentioned in the Bible nor is it deducible from the Bible. It follows that I, Hoover, can never know whether I am using the Axiom—whether reading, applying, or reflecting on it. More drastically, perhaps, the very proposition “Hoover is using the Axiom” contains one prominent term for which no Clarkian meaning can be assigned. On the criteria already given, no acceptable definition of “Hoover” can be acquired. Hence, it seems to me, the formula “Hoover is using the Axiom” fails to constitute an intelligible proposition.

(iii) The Axiom as Revelation
Clark, as can be seen from the above quoted reply to Nash, calls his axiom the “axiom of revelation.” But it is of some importance to test this out. Can the Axiom, as discussed so far, have a revelatory function for modern man if it is the exclusive source of knowledge? Suppose you the reader were to get a letter from a friend disclosing the fact that he had left a valuable book at your house on his last visit and now requests that you bring it to him. Now imagine that the content of that letter (and so its disclosure) is the exclusive source of all knowledge. To expand your present insight you must formally deduce the letter’s logical entailments—but when you are out of entailments, you are out of knowledge. Suppose also that you yourself are not assigned a meaning in the letter and that such terms as “house,” “book,” “bed,” etc. occur undefined in the letter. What exactly is supposed to be revealed in this letter? And to whom? Since you yourself go undefined in the letter, knowledgeable self-reference is impossible—that is, you begin with the irony that you do not know whether you are even getting a letter! And even if we waived this, if “you” had supreme confidence in the truth of the letter because of your confidence in your friend’s trustworthiness (his letter is axiomatic for you), how could the letter make its disclosure to you if its key terms can function only as logical symbols? You might, it is true, celebrate the letter’s truth as proof positive that skepticism is defeated, but what of the revelatory function, what of the letter’s claim to your ongoing circumstances? The letter, surely, must lose its character as a personal disclosure to you if it is transmuted into an epistemological trophy. Remember that King Midas kept his daughter, but as a chunk of gold. In our little thought experiment about the letter, the point of the letter was not to refute skepticism (although I do not want to deny that it might have that function) but to move you to action by addressing your prior understanding and intellectual competence with new information.

Although the Bible is much larger than the letter imagined above, I think our thought experiment points up the weakness in Clark’s “epistemological victory.” Adam, by Clark’s own admission, was created as an intelligent agent who could receive and understand what God had to say to him. Human beings outside the scope of Biblical mention or entailment, however, seem not to be on this kind of intellectual par with Adam. Adam had a capacity to understand and apply what God said to him. Adam could see with his eyes. Adam could till a property he was told to have dominion over. Clark, however, cannot admit that he can see because neither physics nor geometric analysis of the Axiom can deliver the meaning of see. But if the capacity to see with one’s eyes, hear with one’s ears, feel with one’s sense of touch, etc. are necessary conditions for the exercise of dominion, then the exercise of dominion by modern men and women is impossible on Clark’s reckoning. Such capacities can be acknowledged only if their analysis is somehow implicit in the Axiom.


Although Gordon Clark is primarily concerned with theory of knowledge, he has also addressed in his writings the nature of “men and things.” Our concluding concern with Clark’s philosophical contribution to apologetics is primarily his view of the nature of persons, but also the more general implications of his epistemology for metaphysics. What sort or sorts of thing must one acknowledge as absolutely basic as one speaks about the creation? What sort of thing is man? What sort of thing is the mind of man? What difference does a philosophical commitment in these areas make for defending the faith?

Gordon Clark constructs a quite extraordinary view of men and things with the categoric refusal of empirically acquired knowledge—and he begins with epistemology. It is an epistemology that, in his own words, “entails a different metaphysics, a different view of what reality consists of, a different view of how one reality is related to another.”21  And it is an epistemology that is perhaps best understood as beginning with a commitment to the nature of the object(s) of knowledge. What sort of thing is a knowable? The progression from this point of departure may be put in four installments. First, a known object must be immediately apprehended by the mind. This is a basis for rejecting any view of knowables which requires that the candidate for knowledge be essentially unlike, or independent of, the mind.22  Empiricism accepts a fundamental dualism (i.e., most empiricisms) which would bar immediate access to many knowables (physical objects, for example) and interpose an intermediary between subject and object—a representation. Clark’s point is that in that case the object of knowledge would then not be:

the reality itself. Since the mind contains only the picture and never the ‘thing,’ there is no possibility of knowing whether the representation is similar to the object or not. To recognize a similarity between two things, they must be compared, and hence both must be in the mind. But if the reality is in the mind, the picture with its similarity is useless. If the reality is not in the mind, the picture, so far as we know, is a picture of nothing.23
It is crucial to take hold of the importance Clark attaches to the known object being in the mind which knows it. Suppose I claim to know that there are trees overarching my trailer. Clark would insist that if I did know this I could not know it representationally for the reason just quoted. But how then? I would have to know the reality itself-—viz., the reality of the very trees overarching my trailer. This reality itself would have to be in my mind and not any mere representation of it. Of course it is doubtful, to say the least, that I should know anything of this sort given Clark’s severe restriction on the scope of human knowledge. The trees in question are neither mentioned in, nor entailed by, Clark’s Axiom. But if this or any other piece of knowledge were to be in my mind it would follow that the nature of an object of knowledge would have to be mental.

Now note the above progression of thought. Without breaking philosophical stride, we’ve moved from epistemology to a fundamental metaphysical pronouncement. Clark’s epistemology entails Idealism. Thus from theory of knowledge Clark has begun to make good on his “different metaphysics,” his “different view of what reality consists of,” and his “different view of how one reality is related to another.”

Second, since all reality has been created by God, reality is thoroughly known by Him. But as we have just seen, to be known an object must be in the mind and hence mental in nature. But if all reality is known by God, reality is exhaustively mental in its nature; there is no room for any mind-independent objects at all and so Clark’s Idealism is thoroughgoing. There exists nothing that is not mental in its being.

Third, the character of the mentality of all things follows rigidly from the character of the items that are known. Knowledge is always knowledge of the truth and truth, in Clark’s view, is a quality of only propositions—that is, only propositions are the sort of thing that can be true (or false). Hence only propositions can be known. But since the range of the real and the range of the knowable coincide—or, alternatively put, since the set of all real objects and the set of all knowables is the same set—then given the doctrine of immediate apprehension, the character of reality itself is propositional. Even God is a proposition because He is thoroughly known to Himself! Hence only propositions exist and Clark’s Idealism is a thoroughgoing rationalistic Idealism. No mental entity can be accommodated that is not a proposition.

And fourth, if all reality is propositional, we come to understand Clark’s view of how one reality relates to another. Propositions, it would seem, relate only by logic. Propositions are not spatio-temporal objects. They are not facts or events. Unlike spatic-temporal objects, propositions do not occupy space or take up time. Unlike facts, they may be false. And unlike events, propositions do not occur and befall objects. Thus propositions do not interact causally: they do not affect one another by gravity or electro-magnetism, they cannot bump into one another, fall off shelves, or shatter. Clark’s propositions, rather, relate by logical implication, and they form, presumably, the one coherent system of truth. As such, Clark’s world of “men and things” is held together (sustained) by logic.

Perhaps we have made Clark’s position sufficiently clear but it can, perhaps, be put with a greater appearance of plausibility. Another way to end up with the propositionality of persons (and reality) is simply to hold to the coherence theory of truth without compromise. This was implied in the earlier account of the progression in Clark’s thought23 but was not developed. Briefly, the coherence theory of truth holds that there is only one ultimately coherent system of truth. Truth, that is, is a system of geometrically organized propositions. If a certain proposition is true, it is logically entailed by the system—or less stringently perhaps, a true proposition is one which is simply not inconsistent with any of the other propositions in the system. On the other hand, a proposition is false if and only if it fails to be entailed by the system—or, if and only if it is inconsistent with the system. The system contains all truth and only a truth can properly be said to be a knowable. But, according to Clark, “truth always comes in propositions,”24 so persons, if knowable, must be propositions. Let us take a closer look at this.

Clark is a strict advocate of the reductio ad absurdum refutation technique—indeed, it is the only refutation technique consistent with his principles. Perhaps in the space we have left we can sketch a reductio ad absurdum argument against Clark’s insistence that persons are propositions. This might take two forms (and perhaps there are others). (i) We might refute the contention that coherence is definitional of truth, or (ii) we might try and show that in Clark’s sense of proposition, the implicit meta-physic involved is insufficiently rich to accommodate even what Clark wishes to accommodate within his vision of men and things.

(i) Coherentism & Truth
Concerning the coherence theory of truth, and very briefly, I think the coherentist wrongly models human knowledge on the mathematical ideal of certitude. As a matter of historical record, none of the great Rationalists (e.g., Descartes, Leibniz, Spinoza) have been able to make such a “mathematicizing” of reality work. This, of course, is not a very telling criticism but it is important to appreciate that the Rationalists are strongly attracted to only one sort of certitude—the certitude generated by logical deduction. And the quest for a certitude of “men and things” so conceived is not given up even when the “mathematicizing” results in what certainly seems to be the patent distortion and even destruction of human experience. The telling criticism, however, is that, contrary to Clark’s belief, the coherence theory of truth does not provide a theory of the meaning of truth, but only a criterion for its recognition within a formal system. Thus, if presented with a proposition P, and I know that P is a member of a set of propositions which is consistent, then if that set is sanctioned by a true axiom, P must be true as well. The mere coherence of P is a sufficient criterion of its truth. But note that P’s content, by the coherence criterion, is irrelevant. My point is that a sure—fire criterion for the identification of a true proposition as a true proposition is not necessarily an insight into the meaning of truth.

But for the moment, let us suppose that the coherence criterion is also the meaning of ‘true’ as predicated of propositions. Suppose that Clark is right. Take the proposition, “Jesus Christ has come in the flesh.” (I John 4:2). The recipients of John’s letter, let us assume, rejoiced in this truth as over against the docetic teaching about the incarnate Christ. But in what, exactly, could they have rejoiced given that the coherence theory of truth is true? If we are indeed given the meaning of truth by the coherence theory, then John’s proposition’s truth consists in its formal consistency with the larger set of propositions of which it is a member. They rejoiced, then, that “Jesus Christ has come in the flesh” is logically consistent with all the other propositions they held for true. Conceivably, there may have been in that early church some rather eccentric academics and logicians who had the necessary expertise to delight in the logical fittingness of John’s statement, but did not the early church rejoice rather in this statement’s content? In what this statement is about? The Apostle John’s faithful witness is that we do not serve a docetic Christ, but a Christ who became flesh. The proposition in question, if we are to make sense of being encouraged and gladdened by it, is about a non-proposition—viz., the Incarnation of Christ. Its coherence may be a criterion or sign of its truth but its truth is more plausibly construed as its faithful recounting of what had taken place in the case of Jesus of Nazareth. So unless a mere mark of truth is the meaning of truth, coherence cannot be definitional of truth.

(ii) Persons & Propositions
It seems to me that Clark must account for the sorts of men and things found in the Bible. Men are the sort of thing that can fight and be killed in war, write letters, marry and have children, engage in teaching and learning, till the ground and go fishing, and a host of other activities. Clark’s philosophical anthropology must at the very least square with these data; his thought must at least acknowledge that human beings possess the God-given capacities to act in these ways. But puzzles arise, on Clark’s propositional monism. The properties (capacities?) of propositions seem quite unlikely when put with persons.

We might adjust to Clark’s proposal by listing ways we are prepared to speak of both propositions and persons and in this way make a preliminary effort to see whether it is meaningful to identify persons with propositions. That is, we may begin by conducting a thought experiment to see whether a proposition is the sort of entity or particular that could be a person. The principle which governs our thought experiment is sometimes called Leibniz‘s Law: if X and Y are the same thing, then whatever is truly predicable of X is also truly predicable of Y. So if persons are propositions, then generically (sortally) at least whatever is predicable of a proposition (as a basic sort of thing) ought, in logic, to be predicable of a person (as a basic sort of thing). But I find that a whole run of predicates join rather badly when an attempt of this kind is made. Consider the following combinations: One person in the Bible is named David. Is there a proposition named David? Persons in the Bible may be Philistine, Moabite, and Israelite. Can a proposition be Philistine, Moabite, or Israelite? Biblical oersons are said to be wise glad, or angry. Can a proposition be said to be wise, glad, or angry? And put the other way around, a proposition is either true or false Is a person the sort of thing that can be true or false? A proposition, on Clark’s reckoning, is a judgment. Is it meaningful to regard a person as a judgment? Propositions can imply or entail other propositions. Is it meaningful to speak of persons as implying or entailing propositions? other persons?

To my ear at least, the apparent identity Clark is after is a curious one indeed. The thought experiment, though not formally decisive, suggests that propositions and persons are of quite distinct and incommensurable categories. One way of seeing this clearly is to ask whether a Clarkian personproposition could be the sort of individual that could exercise moral agency. The concept ‘person', however else it might serve in our language, is the inherently moral designation of the human being. It is persons that have rights before the law. It is persons who stand to be violated or de-personalized by other persons—e.g., as in women being treated by men as mere sex objects or as in workmen being treated as mere labor power by management. And it is persons, in the spirit of Kant, that are possessed of freedom and inviolable dignity. If this is one correct way of employing the concept ‘person,’ it must be made room for in Clark’s philosophy. What does Clark say?

According to Clark, persons are propositions (i.e., “complex definitions”).25  Drawing from Leibniz, Clark affirms that a given person is precisely the set of propositions that make up his life history. And a person is all his propositions. From this it follows that personal identity, self-knowledge, is utterly problematic if not impossible. This follows for two equally sufficient reasons—the first posing a strictly epistemological problem and the second, a logical problem of individuation: (i) Epistemologically, at any point in one’s “autobiography” one has only a finite sample of a possibly infinite set of propositions that will go to constitute his identity. Recall that propositions based on sentient awareness do not count. Now if one has but a finite sample of propositions relevant for his/her identity, the probability that one has enough propositions for a correct identity judgment is either exceedingly small or zero (a finite number over infinity is zero). (ii) Logically, Clark’s theory identifies the person as a “complex definition”—or, to say the same thing in a way that will help us make the point about individuation more clearly, a person is a set of propositions. Persons, then, are individuated by propositional sets, by propositional totalities. This has the exceedingly odd consequence that persons can be told apart, so to speak, only by virtue of their entire life histories. Because no person in this life has the requisite omniscience to know total life histories, it follows that persons in this life are not the sort of thing that can be told apart. But perhaps more seriously—and this brings us around to the test question concerning the exercise of moral agency—if persons are thus individuated (by logically complete propositional sets), how is it imaginable that persons could be moral agents within their life histories? For in Clark’s sense of person, the person seems to be his life history and not a participant within it. And it does no good to invoke the omniscience of God. The fact that God can individuate the propositional sets in question still does not allow persons to be agents within history . Propositional sets are simply not the sort of entities that could commit adultery, care for the sick, or resist temptation. Thus it would appear that a propositional monism is insufficiently rich in basic categories to allow even the Biblical David to sin and repent.


Certainly a good deal more could be said about Clark’s constructive philosophy. It is enough, however, to indicate that Clark’s apologetics is rigorously tied to the philosophical orientation sketched above. Clark commits Christian apologetics to a defense of Christianity as a system of truth, a system of propositions which constitute mankind’s only subset of the one coherent system of propositions. A proper defense of this system may never invoke inductively acquired data in its support. Sentience is non-cognitive and inductive arguments are all fallacious. Instead the apologete shows up the inconsistencies of the non-Christian and does the best he can at exhibiting the coherence of the Christian system. He does this with the prayer that the non-Christian may be convinced of Christian theism’s coherence, and hence its truth. Of course Clark would admit that the non-Christian might come to believe in the Christian teachings for less than systematic reasons. One is a Christian because one believes the truth, not because one can also appreciate what truth one knows in all its logical and systematic elegance.

I conclude by indicating the most crucial difference all evidential apologetes have with Clark. This is Clark’s totally formalistic approach to defining knowledge and to theorizing about its acquisition. Clark simply refuses to acknowledge informal noetic capacities in man—viz., the capacities to see hear, smell, touch and taste objects in the environment. Clark restricts the human noetic endowment to sheer intellection with no epistemic access to the creation in which man lives. If, as I think is the case, Clark’s epistemology requires a serious distortion of human nature, to that extent it is ill-suited to serve as a basis for defending the Christian faith. Obedience to the Great Commission requires faith that we inhabit a real and discriminable world of real and discriminable sinners. Even fallen man is epistemically suited to the sort of world God has created. With our purely formal capacities—with our intellection—we may pursue mathematics, geometry, and logic, but human endeavor that involves trafficking in the world of God’ s creation is fundamentally dependent on informal cognitive capacities. An apologetic approach that despises empirical fit can never reach sinners who stand in need of being identified and interpreted by God’s Word. It is the saddest of ironies that rather than the Scriptures helping Clark to make sense of men and things, the Scriptures are construed in such a way so as to take away such a prospect. Construed as an axiom that can yield insight only by deductive logical technique, the Bible does not establish the contemporary Christian in a confidence of God’s guidance for him but remains stone silent about a reality extending beyond its own pages. With a Midas-like touch Clark has preserved God’s Word at the price of its revelatory function. Stark and logically frozen it sits, an epistemological trophy on the mantel in the Rationalist’s den.


1 Montgomery made this comment during an address at Covenant College (‘82).

2 “The Philosophical Methodology of Gordon Clark,” published in The Philosophy of Gordon H. Clark, ed. Ronald Nash, (Philadelphia: Presbyterian &Reformed Publishing Co., 1968), p. 219. Holmes here draws attention to Clark’s own characterization of his method in A Christian View of Men &Things, pp. 24-5.

3 Gordon H. Clark, Language & Theology, (Phillipsburg, N.J.: Presbyterian & Reformed Publishing Co., 1980), p. 132.

4 “Having A Fuddled Easter?” by John Warwick Montgomery, Christianity Today. (March 31, 1972).

5 Gordon H. Clark, Three Types of Religious Philosophy, (Nutley, N.J.: The Craig Press, 1973), p. 116.

6 Ibid.

7 See the final chapter of TTRP. The concluding paragraph in the book captures Clark’s position nicely:

. . .the dogmatic mode of argument is clear. It consists of an ad hominem attempt to convict the liberal of contradicting himself. The dogmatist does not attempt to prove the reliability of pottery dating, nor the contemporary principles of historiography. He is not really interested in them. In fact, he has, to his own satisfaction at least, shown that they are indefensible and untenable. (underlining my own)
8 These philosophies do not advocate or require an external world to be represented .

9 Language & Theology, p. 31.

10 Implicit in Clark’s assessment, of course, is that the relation between sensory units and perceptual wholes is that of logical implication. Clark’s question (as well as Blanshard’s) is how might sensation be propositionally construed so as to imply a resultant perception? However, if the relationship is not implication, Clark’s question is inappropriate; his response begs the question against the affirmer of empirically acquired knowledge. Note also that Clark’s rejection of Blanshard’s attempt to logically reconstruct percepts from sensory atoms counts against any strategy of that kind. The trouble is that Blanshard assumes the basic tenets of rationalism in order to reconstruct empirical knowledge. The empiricist, on the other hand, can maintain that there is an intimate and necessary relation between sensory stimuli and the perception facilitated thereby, but deny that stimuli and perception relate by logic. A stimulus, after all, is a stimulus a causal sort of thing.

11 Language & Theology, p. 132.

12 For a very readable discussion of these arguments, see George Pitcher’s A Theory of Perception (Princeton Princeton University Press, 1971), chapter 1, “Sense-Data and How to Avoid Them.”

13 If one begins with only logic and a congeries of marginally described atomic sense—data, it is surely impossible to deduce a dynamic world of perceptual common sense. From an aggregate of black dots which ostensibly occupy an area on a newspaper page, one cannot deduce the likeness of President Reagan. The President’s likeness is gestaltic in character, not discursive. The rationalist stacks the deck against the advocate of empirically acquired knowledge by the way he (the rationalist) poses the problem.

14 Cf. J. L. Austin’s Sense & Sensibilia, (London: 1962).

15 The Philosophy of Gordon H. Clark, p. 413.

16 What it means to presuppose the truth of all this content is itself far from clear. And the question is important beyond a mere philosophical quibble. For one thing, there is the character of truth itself according to the different presuppositionalists. Does truth attach to locutions within the various literary genre in a univocal way—e.g., in poetry, didactic passages, historical narrative? How does truth qualify, and provide implication for, imperatives, interrogatives, counter-factual subjunctives? When one avows that one “presupposes Biblical truth” one does not, ipso facto, answer this sort of question. And for another thing, when someone of Clark’s persuasion presupposes Biblical truth, he takes the Bible’s content as axiomatic. Van Til does not. Van Til, if I understand him, does not qualify the Biblical content by a pervasive structuring of formal logic. (It seems to me, however, that he still comes precariously close to rationalism in the way that he describes the Scripture’s explanatory function with respect to the creation.) Still others presuppose without rationalistic overtones. One way to specify this is by using the language of confession. “Confessing to be true” does not carry the logical overtones of “presupposing the truth of.” The reason for this, of course, is purely a matter of usage. Within the apologetics literature, “presuppose” is typically linked with Biblical truth as a system—a geometrically elegant system.

17 The Philosophy of Gordon H. Clark, ibid.

18 Ibid., p. 88.

19 Ibid., p. 414.

20 Ibid., p. 91.

21 Language & Theology, p. 32.

22 Gordon H. Clark, A Christian View of Men & Things, (GrandRapids: Eerdmans Publishing Co., 1952), pp. 318—323.

23 Language & Theology, p. 29.

24 The Philosophy of Gordon H. Clark, p. 411 (cf. also p. 413). Clark has always maintained that his critics fail to offer an alternative view. But part of the problem, I believe is not that Clark’s opponents fail to identify clear alternatives but that Clark’s assessment of competing views assumes the validity of his own view. If, for example, it is admitted, as I think it must be, that there is an inherent informality in the very meaning of truth and in the human being’s mode of cognizing truth, then the purely formalistic coherence theory is seen to be fundamentally inadequate. The point is, that the clarity of an alternative to Clark’s coherence theory may require a noetic capacity which Clark denies—viz., sentience. If there is, so to speak, ostensive clarity as well as discursive clarity, then Clark’s mere resolve to acknowledge only the latter must be regarded as a rationalist prejudice against sentience, not a refutation of sentience. That is, one cannot refute the position that men have a cognitive sentient capacity merely by showing that sentience is not reducible to discursive rationality. It is not incoherent to hold that there are two (not one) cognitive capacities. The coherence theory of truth is not proven by showing that ostension and discursion are irreducible to one another. In the final analysis, that is all that Clark’s arguments do. Clark has an a priori preference for discursive demonstrability over against ostensive (or inductive) demonstrability. What is important to see is that one cannot discredit the latter by showing that it cannot be analyzed in terms of the former. Moreover, the fact that informal cognition (knowledge by sentient acquaintance) is subject to well-known limitations and pitfalls does not argue that there can’t be sentient acquaintance with the world. To have a cognitive capacity does not entail the having of it to a superlative degree or the having of it without limitation with regard to the sort of capacity it is.

25 The Philosophy of Gordon H. Clark, p. 412.

Produced for IBRI
PO Box 423
Hatfield, PA 19440

You can contact IBRI by e-mail at: rnewman@biblical.edu
Return to the IBRI Home Page

Last updated: January 19, 2002