THE conclusion at which we had arrived at the end of last lecture was that the absolute opposition of faith and knowledge is one that cannot from any point of view be legitimately maintained; that, on the contrary, that which in faith is present in an unreflective form must be clearly grasped by thought and shown to be in its permanent elements capable of consistent and systematic statement in a theology or philosophy of religion. We cannot be satisfied with an appeal to immediate conviction, or avoid the toil and difficulty of investigation, by falling back upon “the faith once delivered to the saints”; nor is it possible to rest satisfied with the creation of mythological fictions, however comforting they may be; but we must be prepared to show that the truths of religion admit of rational defence and systematic statement. The possibility of constructing a philosophy of religion presupposes these two principles: firstly, that the universe is rational; and, secondly, that it is capable of being comprehended in its essential nature by us; and unless we are convinced of their truth, we cannot advance a single step. It will therefore prepare the way for a more positive treatment, if we ask what conception of the universe must be held on the assumption that it is in itself completely rational and can be known by us to be completely rational.

To take the last point first, it is manifest that, admitting the universe to be completely rational, it can only be known by us to be so, provided that our intelligence is not infected with an absolute limit, which for ever prevents it from transcending appearance and grasping reality as it actually is. For, as has often been pointed out, an absolute limit in intelligence is incompatible with the consciousness by intelligence that it is absolutely limited. If I know that there are absolute limits to my intelligence, it must be because, per impossibile, I have for once somehow escaped from those limits. For, the assertion that my intelligence is absolutely limited must be absolute; i.e., it must be an assertion which no increase in my power of comprehension would in any way affect; an assertion, therefore, which would be endorsed by an intelligence that knows all things as they are; in short, an omniscient intelligence. But, if the judgment that my intelligence is absolutely limited is true without any reservation, it must be only in reference to all other judgments that the limitation applies, not to this judgment itself. If there is any doubt of the truth of this judgment, it becomes doubtful whether other judgments may not be absolutely true; and if this judgment is false, its contradictory must be true, and therefore it must be false that all other judgments are not absolutely true. It thus seems obvious that, granting the truth of the hypothesis that the universe is rational, we cannot without contradiction defend the thesis, that there is an absolute limit in the human intelligence, which prevents it from knowing that the universe is rational.

It may, however, be admitted that the assertion of an absolute limit in the human intelligence refutes itself, and yet it may be contended that under the conditions of our knowledge we can never prove that the universe is rational. Our judgments, it may be said, are never more than a formulation of that partial and inadequate comprehension of reality which alone is ours. It is not necessary, it may be contended, to maintain that our judgments are absolutely false; all that need be held is that they are our human way of representing a reality which never actually comes within the circle of our intelligence. All our judgments in regard to ultimate reality, it may be urged, are of this character they are not false, and yet they are not absolutely true. For example, it is ordinarily held that the Unity, which we are compelled to postulate as the principle of all things, is a “person” or a “self-conscious intelligence.” And it must be admitted that “personality” is the highest category within our reach, and therefore may be legitimately enough predicated of the Absolute or God, provided we do not suppose it to be an adequate characterization of the ultimate Unity. That it is not adequate may at once be seen if we reflect that, in the strict sense of the term, “personality” has no meaning except when predicated of finite beings. We speak of a “person” when we are referring to a particular individual, who is and knows himself to be distinct from all other individuals; and, therefore, we cannot speak of God as a “person” without denying that he is the principle which all existence implies. But, though it cannot be an attribute of God, “personality” may fairly be said to represent or symbolize the real nature of God, and indeed to represent it in the highest way possible for our limited intelligences. We may therefore make use of the term “personality” when we are speaking of God, so long as we recognize that we are employing a metaphor, which only suggests or indicates that which it is not in our power to define. Certainly it is a more adequate way of describing the nature of God to say that he is a “person,” than to speak of him as an abstract “Power” or a “Substance”; and, if anyone is tempted to think of God in either of these inadequate ways, he may usefully correct that tendency by conceiving of him as a “person,” provided always he recognizes that God is something infinitely higher than personality, though what this something is no human mind can possibly tell.

This modified doctrine is not so obviously self-contradictory as the unqualified assertion, that the human intelligence is absolutely limited. It recognizes “degrees of truth,” refusing to say outright that by its very nature the human intelligence is shut out from all knowledge of reality as it absolutely is. But it can hardly be said to be a self-consistent doctrine. It goes on the principle that it is possible for the human intelligence to discern the limits within which it is confined, without being able in any way or in any sense to see beyond them. We can, it is assumed, tell that our judgments in regard to the nature of ultimate reality are not absolute, without being able to say positively what the content of an absolute judgment would be. Thus we are entitled to say, it may be urged, that the Absolute is not a person, much less a blind force or substance, because we can see that personality is applicable only to one individual subject as contrasted with another, while we are certain that God is not such an individual subject. But, surely, this doctrine must either be pushed further or entirely abandoned. If we know that God is not personal because he is not finite, it must be because we know him to be infinite, and that not in the purely negative sense that he does not belong to the class of finite beings—for in that case he might be nothing at all—but in the positive sense that he contains all reality within himself or is all-comprehensive. If we say that he is “super-personal,” we must have a positive ground for making the assertion. That being so, it must be possible to characterize God by a higher category than that of personality. If it is replied that we possess no higher category, then we must answer that in that case we have no ground for asserting that there is any higher category. But the contention can hardly be made good. For, to say that we have no higher category than personality is the same as saying that we have no conception of anything higher than abstract individuality—unless indeed we abandon the ordinary definition of personality, and employ it in a deeper sense, and then it may not follow that it is inadequate as a definition of God. The truth is, that in the very conception of God as infinite or all-comprehensive, we have, at least implicitly, a higher category than that of abstract individuality, since God is then conceived to be the prius of all abstract individuality. And when we go on to consider what is meant by personality, at least as understood by those who deny its applicability to God, we find that it is by no means “the highest category within our reach.” For, as we have said, personality is assumed to imply abstract individuality, and abstract individuality is demonstrably inadequate even as a characterization of man, not to speak of God. If it is said that every man is an abstract individual, i.e. an individual whose nature it is to be independent of all other things, it may easily be shown that such an abstract individual is a mere fiction. If any such individual actually existed, he would be independent, in the sense that he would be what he is even if all other beings were annihilated, or had never existed. But such a self-centred individual has no more existence than a centre without a circumference. Hence, personality must be interpreted in a higher way; or, what is the same thing, a higher category is required to define the nature of man. For, man is what he is, not in his isolation, but in his relations to other beings, the purely individual man being a mere abstraction. This higher category may be called self-conscious reason, and nothing less will adequately characterize the nature of man. Hence we must at least define God as self-conscious reason; and if this category also is found to be inadequate, we must replace it by others, until we have found one that is adequate. Whether or not the category of self-conscious reason is adequate can only be determined as the final result of our whole enquiry; all that is at present maintained is, that there is nothing in the nature of our intelligence which makes it hopeless to attempt a characterization of the nature of God, since the rejection of any given determination is legitimate only if we have already reached a higher determination.

That this is the only defensible conclusion may be shown in another way. When the predicates by which we seek to characterize the nature of God are said to be true in an analogical but not in a literal sense, it is implied that, somehow or other, we are capable of comparing them with the predicates which actually and precisely characterize his nature. It is therefore assumed that the human intelligence has in some sense before it both the inadequate and the adequate categories. For, in every case of comparison, both terms must be present. If, for example, a photograph is declared to be a good likeness, obviously the picture and the original must both be known. No one can say that the photograph is either good or bad unless he knows what the person represented looks like. Similarly, if it is said that personality is a good representation of the actual character of God, he who says so must have some knowledge of what that character is. If he knows nothing of the nature of God, how can he tell whether personality is a more or less adequate determination than force or substance, or indeed whether any of these predicates has the remotest likeness to the attributes of God? If God is truly beyond our knowledge, for us he is perfectly destitute of all positive attributes. Even granting therefore that in himself God has a determinate nature, this does not help us to show that we can prove him to have such a nature. It thus seems illegitimate for anyone who denies that we have a positive knowledge of the nature of God to claim that he has even an analogical apprehension of that nature. He must therefore be content to say that there is a God, without affirming that the predicates by which the nature of God is made intelligible to himself have the remotest resemblance to the actual nature of God. And when we have reached this stage, it is only another step to the denial of God altogether, since that of which nothing more can be said than that it is, cannot be distinguished from that of which nothing whatever can be said. The assumption of an absolutely incomprehensible God is at the most merely an index of the ineradicable tendency of our nature to refer all modes of being to a single all-comprehensive unity; but the assertion that this unity is indefinable is inconsistent with the assertion that it exists. We must, therefore, conclude that, neither in its unqualified, nor in its qualified form, can the absolute limitation of the human intelligence be defended.

Now if the human intelligence is not infected with an absolute limit, it cannot be shown that the universe is in any sense irrational. The two propositions, indeed, imply each other. Intelligence is not an abstract power which works in a vacuum, but essentially consists in the comprehension of reality. If we suppose the universe to be infected with irrationality, the intelligence, if it proceeds upon the assumption that the universe is completely rational, is bound to find itself checked and frustrated in its effort to comprehend that which, as irrational, must be absolutely incomprehensible. We cannot set up an absolute limit in the universe without virtually dividing it into two opposite and irreconcilable sections: one in which every element is supposed to be combined in the transparent unity of the whole, and the other which is the absolute negation of that unity. So long as, beyond the region which is radiant in the light of reason, there falls a perfectly dark and opaque region, of which nothing can he said but that it is absolutely inscrutable; so long the human intelligence must be subjected to the absolute limit which is implied in the absolutely unintelligible. Such an external limit to the intelligence necessarily implies a limit in the intelligence itself; for the intelligence can only exist, and possess the nature of intelligence, provided that it is consistent with the total nature of things. A perfectly rational intelligence cannot exist in a partially rational universe; and therefore the complete rationality of the universe is the indispensable condition of an intelligence free from any absolute limit. The rational and the real must coincide: if we cannot show that the real is rational, it is certain that we cannot prove the rational to be real; and the truth of both propositions is the indispensable condition of a philosophy of religion.

I have attempted to summarize the arguments for a rational universe and for the possibility of its comprehension, partly because these two principles seem to me indispensable as the basis of a theology, and also because there is a class of thinkers who claim, in the most emphatic way, that all such attempts are foredoomed to failure from the very nature of the case. We are precluded, it is held, from making any absolute statements in regard to the ultimate nature of things by the very nature of our experience. The Critical Philosophy, as we know, because of its distinction between phenomena and noumena was led to deny that we can have any knowledge of reality as it is in itself, though it also contended that we can reach an assured faith in God, freedom and immortality through the moral consciousness. And radical empiricism, as advocated by late Professor James and endorsed by his followers, goes refusing to admit that we can even refer all things to a single principle. If this last contention can be successfully defended, there can be no philosophy of religion as I understand it; and therefore it seems necessary to ask whether its main contention may not rest upon a false and indefensible foundation.

Radical empiricism, we are informed, denies that the universe can be shown to be a rational whole, maintaining that the character of our experience does not justify such an inference. Holding this view, it naturally refuses, we are told, to admit that any of the conceptions by which we seek to introduce order into our experience can be regarded as more than “working conceptions,” liable to be superseded at any moment. It is therefore opposed in principle to rationalism, understanding by this term the doctrine that the universe is an intelligible whole. There is nothing in our experience, it is maintained, which entitles us with certainty to go beyond particular facts. On the other hand, we cannot in consistency exclude any element that is actually experienced, and therefore, since “conjunctive relations” are experienced, they must be accounted as real as anything else. The recognition of conjunctive relations, Professor James assures us, is the great superiority of the new empiricism over the old. “Ordinary empiricism, in spite of the fact that conjunctive and disjunctive relations present themselves as being fully co-ordinate parts of experience, has always shown a tendency to do away with the connections of things, and to insist most on the disjunctions.”1 Radical empiricism, however, does not, like rationalism, treat these relations as being true “in some supernal way,” as if the unity of things and their variety belonged to different orders of truth and vitality altogether.

Now, nothing seems at first sight more reasonable than the contention that we must not go beyond what the facts of experience warrant. In the infinity of particulars ever crowding upon us, we are in many cases forced to be content with mere “brute fact,” without being able to show why it should exist. Who will prove to us the rationality of a thunderstorm or an earthquake? And if we cannot show the rationality of all the facts that we experience, how shall we prove the rationality of the whole? Does the rationality of the whole demand that there should be pain and sorrow, crime and guilt? Must we have a Catiline and a Borgia? Is it not more reasonable to say that, while these are undeniable facts, we cannot explain them? Must we hold that whatever occurs is consistent with the rationality of the universe? Is it true, in Hegel's famous phrase, that “whatever is rational is real, and whatever is real is rational”? Why should we run counter to the plain facts of experience? Why should there not be “chance” “accident,” “matter,” or by whatever term we choose to express the seeming unintelligibility of things?

I do not think that reasoning of this kind can be met by urging the probability that, as many things have been explained which once seemed inexplicable, there is good ground for believing that all things could be explained, if only our experience were more ample. This hardly meets the difficulty. The progress of science has no doubt in many cases revealed the particular conditions under which certain events occur, conditions of which we were at first ignorant, and to that extent has done away with our first impression that they simply happened, or were due to chance or accident. But, while this is true, I do not think that we can base an inviolable law upon any accumulation of particular instances; and therefore in the end we seem forced to admit, that the conception of inviolable law assumption that we cannot or analogy is necessarily the argument from probability or analogy is necessarily weak, because an objector may always retort that, as we have never been able to get beyond a limited number of instances, it is just as legitimate to argue that there are things which by their very nature are inexplicable, or at least can never be shown by us not to be inexplicable, as to maintain that they are capable of explanation. I think we must therefore attack the problem in another way.

The first thing to notice is that the plausibility of the argument against the complete intelligibility of the universe implies that it is partially intelligible. Now, I think it may be shown that even partial intelligibility either implies complete intelligibility or the absence of all intelligibility. The changes which objects undergo are capable of explanation just in so far as they occur, not in an irregular way, but in a fixed and inexorable order; and if this order is denied, there is no longer anything that admits of explanation. Suppose for a moment that our experiences were so discrepant that there was absolutely nothing in any two of them that we could call identical; suppose, for instance, that the pavement should suddenly get up and hit one on the head; and what would be the logical result? The logical result would be that no judgment whatever could be framed, since judgment rests upon a recognition of something identical in our experience. But the hypothesis of an experience in which there is no identity whatever is absurd, for the simple reason that the minimum of experience involves at least the distinction of “this” from “that,” and such a distinction is impossible unless there is something identical in “this” and “that”—whether it is identity in extension, or in time, or in quality, or in some other mode. An absolutely chaotic experience, in fact, is no experience at all; to my mind indeed it is simply nonsense: some fixity or order there must be, even if one is only to buy a pair of boots or to match two pieces of silk. Now fixity or order, from the nature of the case, is not something which can be limited in its application: it must be true absolutely and without any exception. It cannot be established by any accumulation of particular instances; for, unless we could perform the impossible feat of summing up an infinite series, we should never in this way attain to the universality of a law. Anyone who seeks to base the regularity and order of our experiences upon a supposed summation of particulars, can give no reason why at any moment all order should not disappear, leaving us weltering in an absolute chaos. The supposition, therefore, of an experienced world absolutely destitute of order, or absolutely unintelligible, is one that cannot be entertained without self-contradiction; it is an hypothesis which, by making all experience impossible, makes itself impossible. We must either postulate the complete intelligibility of the universe, or deny that we can have any experience whatever. That of course is very different from saying that we can explain every fact of our experience. What is asserted is, not that we can show in detail that the universe is completely intelligible, but that on no other supposition than its complete intelligibility can we make any assertion whatever, not even the assertion that it is not completely intelligible. This distinction seems to meet the difficulty, that we are asserting absolutely that which we only prove relatively. Such a judgment as “water rusts iron” we affirm, not on the ground that we have observed all the cases in which water rusts iron, but on the ground that, without presupposing the universal principle, we cannot have the particular experience. How, on any other supposition, can we make even so simple a judgment as “that is green”? Admit even that the judgment means no more than that “I am conscious of something green”—without determining whether it is a mere illusion of my own, or something capable of Presenting itself to anyone with normal eyesight—even then the judgment itself is at least an actual judgment, which every mind under precisely the same circumstances would make. But, in the assumption that every mind must recognize the certainty of the fact as a fact of consciousness, we have tacitly assumed it to be an unchangeable law, that under precisely the same conditions the same judgment must be made. And this involves the tacit assumption that nothing can be real for any mind except that which is compatible with the fundamental law of all mind, that no judgment can be true which affirms and denies in the same breath.

Radical empiricism rests upon the assumption that experience presents us with an assemblage of facts, the relations of which to one another we are capable of partially discovering, while yet the facts do not warrant the inference that we live in a completely intelligible world. This doctrine obviously presupposes the reality of the facts, while denying the inference that the world is one. Now, what is a fact? It is something which presents itself within the experience of this or that individual, and so presents itself that he cannot but admit its reality. The assertion of the reality of what falls within the experience of the individual is in truth the ground upon which the denial of Monism is based. The character of the facts, it is argued, is such that they do not warrant the inference to the complete rationality of the world as a whole. If therefore the facts were not indubitable, the negative conclusion drawn from their character would be invalid. Now, the facts are not momentary states of this or that subject, but an identity in the experience of the same subject, or of different subjects at different times. Deny this identity, and there is no fact, and indeed nothing that we can speak of as an “experience.” “The sun shines” does not mean: “There is in my consciousness a sensation of light at this moment”; what it means is: “I am conscious at this moment of a fact that I call the sun shining,” a consciousness which every other mind would have under precisely the same conditions, inward and outward. The simplest everyday fact of experience thus presupposes that very intelligibility which radical empiricism affects to deny. The fixed and unalterable nature of a fact—and if not fixed and unalterable, how can it be a fact?—is meaningless, if the systematic connection of all facts is denied. Thus the unity and intelligibility of the world is first tacitly assumed by radical empiricism, under the guise of particular facts, and then plausibly denied just because it has been assumed, Not only do we admittedly always assume that the world is intelligible, but our assumption is one that justifies itself.

Rationalism, however, we are assured, “tends to emphasize universals and to make wholes prior to parts in the order of logic as well as in that of being. Empiricism, on the contrary, lays the explanatory stress upon the part, the element, the individual, and treats the whole as a collection and the universal as an abstraction.”2

Without entering into an historical enquiry into the legitimacy of this characterization of rationalism, I shall only say that I do not think we are bound to accept either radical empiricism or what here is characterized as rationalism. I for my part distinctly reject the doctrine that “wholes are prior to parts,” just as I emphatically deny that “parts” are “prior to wholes,” or that “wholes” are simply a “collection” of parts. It is not true that “wholes” are “prior to parts,” and just as false that “parts” are “prior to wholes.” Neither is prior to the other, either “logically” or “really.” Take the simple case of such a judgment as “the sun warms the stone.” This is a particular case of a causal connection, and as such it is meaningless if causal connection is denied. No one can truly say that “the sun warms the stone,” who dies not presuppose the universal and inviolable law, that every change occurs under certain fixed conditions. Deny the principle, and you deny the particular instance of the principle. Thus the particular does not precede the whole, but the whole is involved in it. Nor does the whole precede the particular; for the principle has no existence except as manifested in this and other particulars. To speak of the whole as a “collection” is as absurd as to suppose that a heap of stones is an organism. The whole is always in the part, and were it otherwise the part would be meaningless.

This principle of the essential correlativity of whole and part, universal and particular, is violated by all forms of empiricism, and not least by this new form of it. As we find it in the school of Locke, empiricism, starting from the assumed opposition of real things and states of the individual mind, seeks to explain how from these states a knowledge of real things may be obtained. Such knowledge, it is maintained, can only come to the individual through impressions of sense, while universals or conceptions derive their meaning from what is discovered by a comparison of impressions with one another. Now, the objection that may fairly be made to this doctrine is that it attempts to explain knowledge by assuming that things as known are independent of the conscious subject, and only act upon him in an external way. Thus perception is supposed to be the purely passive apprehension of an object, which is for the subject a sensation or idea produced in his mind without the exercise of any activity on his part, or at least of any activity that affects the impression he receives from the external thing. Such a doctrine falls into the fatal mistake of explaining knowledge by eliminating from it every vestige of thought, and seeking to account for it in a purely external and mechanical way. In contrast to this doctrine we maintain that things are what they are for us as thinking beings, and that to speak of knowledge as given in mere sensation is to commit oneself to a theory which must empty the known world of all meaning. From sensation as defined not even the appearance of knowledge could arise. It is only by identifying this supposed sensation with perception, and treating perception as if it did not differ from what is called sensation, that we seem to exclude all universality. Nor can any modification of empiricism alter its essential nature, so long as the fundamental postulate is retained, that objects are known in a purely passive apprehension. To say that the individual subject does in point of fact find before him, not unrelated feelings, but feelings which are conjoined to one another, does not overcome the fundamental fallacy of empiricism, which consists in viewing the mind as if it were on the same level as other things, and might therefore be treated, if not as a series of events, at least as a “stream,” in which a new thought, internally complex it may be, perpetually displaces the old. For, though it is no doubt true that an absolutely simple feeling can yield no knowledge, the mere complexity of feeling will not explain the universality of judgment, without which knowledge of reality is impossible. So long as the separate individuality of mind is assumed, and each mind is resolved into a temporal succession of states, the difficulty remains, that there is nothing in a mind as so defined which accounts for its comprehension of the actual nature of things. Conceiving of reality as a collection of separate things, and applying the same idea to mind, empiricsm is forced to reduce the individual mind, either with the older empiricists to a “succession,” or, with radical empiricism, to a “stream” of feelings. Whether we take the one view or the other is of very little importance, so long as the mind is practically defined to be a collection of elements only related to one another by the superficial bond of time. A disconnected “series” of feelings and a “stream” of feelings have this in common, that the elements of either are merely particular, and as such can yield no universal judgment. What is really characteristic of mind is that it cannot live in the particular, but always in some way grasps the universal in the particular.

These considerations seem to show that this supposed “new way of ideas” is not free from the fundamental incoherences which led to the bankruptcy of the older empiricism in the scepticism of Hume. Attempting to build up a theory of knowledge on the basis of an accumulation of particulars destitute of universal significance, it is able to give plausibility to its denial of the unity and rationality of the world only by tacitly assuming an order that by its principles it ought to deny. What it supposes to be mere particulars really involve universal principles; and if it were not so, there could be no law whatever in things. The “facts” which it assumes are not merely particular experiences, valid only for the subject experiencing them, at the moment of experiencing them, but fixed conditions under which experience takes place. Thus, in the simplest fact there is already involved that very unity or rational system which radical empiricism denies. To admit the reality of anything whatever, however apparently insignificant, is to assume the reality of a rational and intelligible universe. This is the sense in which we may rightly claim that the reality of God is involved in the reality of the simplest object of experience. Such a doctrine naturally can only be made convincing by a comprehensive review of experience in all its forms; but I hope to convince you, in the immortal words with which Spinoza concludes his “Ethics,” that, “if the way to the blessed life is hard and steep, it yet is not impossible of attainment. Hard it must be, or it would not be followed by so few, all that is of great value being as difficult as it is rare.”

Meantime, it will make our progress somewhat easier if we first ask what must be the general character of a universe which is at once rational and intelligible. The first and most obvious characteristic of a rational universe is that it must be an absolute unity. And this unity must be understood in the strictest possible sense. It must not, for example, be confused with an aggregate, or a mere assemblage of particulars. An aggregate is not a unity, because it implies the separate and independent existence of particulars which have no necessary connection with one another. A number of points in space is not a unity, when these points are regarded as distinct and separate from one another: it is a perfectly arbitrary collection, dependent upon the mind of the conscious subject, who chooses for his own purposes to regard the collection as a whole. If the points were really a whole, they would not be separate and independent, but would be so indissolubly bound up with one another that no power could separate them. And, constituting a whole, they could neither be increased nor diminished, whereas it is the very idea of number that it can be increased or decreased at will, every element in it being capable of separation from every other. If therefore the universe is a true unity, it cannot be of the nature of an aggregate: its elements must be so adapted to one another as to be incapable of existing apart, so that to remove any single element would be to destroy the whole. Hence its unity cannot be dependent upon the caprice or arbitrary choice of a conscious subject but must belong to the object itself. Every element in it must be intimately and indissolubly connected with every other No assemblage of points in space or moments in time can constitute a true unity. Nor can the elements of an absolute unity be “parts,” in the sense that they have a nature of their own in no way dependent upon the unity. For this would immediately degrade the unity into an aggregate or collection being, standing outside of the parts, and only externally related to them. In that case the so-called unity would not; for nothing can really unify except that which would not unify; for nothing can really unify except that which includes within itself all the elements that it unifies. On the other hand, there is no unity which is not differentiated into distinguishable elements. For a unity which contains or implies no differences, is a mere abstraction. To be a real unity, the universe must therefore have two aspects: it must be absolutely one, and it must be absolutely many. And the two aspects must be correlative: a unity that is not differentiated is essentially exclusive of all differences which are independent are exclusive of unity. If the universe is truly a unity, it must comprehend within itself all possible differences. It will, then, be a perfect whole—not in the sense that, as a matter of fact, it contains in itself all differences, but in the sense that no differences can possibly exist independently of it. As an absolute whole, it must be infinite, not as excluding all finite determinations in the way of quality or quantity, but because it is an absolutely complete whole, beyond which there is nothing, and can be nothing. It is not enough, therefore, to say that a rational unity is a perfect or complete whole, but we must add that it is the only perfect or complete whole. The conception of Leibnitz, for example, that the universe is “the best of all possible worlds,” implies that there must be a perfect or complete whole; but, in admitting the possibility of other wholes, it virtually assumes the imperfection and irrationality of the whole that is held to exist. Moreover a truly rational unity must be not only completely differentiated, but it must be absolutely perfect. This does not mean that it can undergo no change. To suppose so, is to predicate of it a stiff unbending inflexibility, which is incompatible with the reality of its differences. The unity cannot be a dead unchanging identity, but, on the contrary, it must express itself in an infinity of changes. These changes, however, must be due to nothing but itself. When Plato argues that the divine cannot change, because it must change either from better to worse, or from worse to better, what he is thinking of is that the divine being by changing must lose its absolute unity and perfection. But Plato overlooks, or seems to overlook, the other and equally essential truth, that a unity which excludes all process is an abstract or dead identity, which cannot possibly exist, because it is nothing. What is true is that the process involved in the absolute unity must not be confused with a transition from lower to higher, or from higher to lower the absolute unity must be equally perfect in all its phases.

The second main characteristic, therefore, of a rational unity is its self-differentiation. That is self-differentiated, which is in no way dependent for its differences upon anything else. As all reality or being is contained within the one unity, obviously there is nothing on which it could be dependent; and indeed dependent being necessarily implies self-dependent being. Granting that there is an absolute unity, it is therefore illogical to regard reality as made up of parts only externally and arbitrarily virtue of its very nature must differentiate itself in its parts, and this differentiation is therefore no accident, but the expression of what it is and must be. Not that the Absolute is the only possible self-differentiating unity, but Only that all other self-differentiating beings must be subordinated to the Absolute, and cannot be absolutely self-differentiated. To this point we shall have to return, when we come to deal with man as a free or self-determining being.

Lastly, a rational universe must be a coherent system. Every element in the whole must be related to every other; so that any change in one element will involve a correspondent change in all. This close connection of all the parts is indispensable to a true conception of the whole; for if a change may occur in one part which in no way involves t correspondent change in other parts, we must suppose that the part which changes is in no way affected by the others, while they in turn are not affected by it; and this is inconsistent with the unity and self-differentiation of the whole, which demands a consistent system of changes. System or coherence is, therefore, an essential mark of a rational whole; so that we may legitimately argue, from the appearance of disconnection and arbitrariness in our experience, that we have not truly comprehended the nature of things.

The real, then, must be an absolute unity, it must be self-differentiating, and its differentiations must form a perfect system. These seem to be indispensable features in a rational universe. Whether they can be shown to be inferences necessitated by the character of our experience considered in its totality, is the main problem of a philosophy of religion, to which we must address ourselves in our next lecture.