Louis Berkhof

Systematic Theology:

V. The Attributes of God in General


The name "attributes" is not ideal, since it conveys the notion of adding or assigning something to one, and is therefore apt to create the impression that something is added to the divine Being. Undoubtedly the term "properties" is better, as pointing to something that is proper to God and to God only. Naturally, in so far as some of the attributes are communicable, the absolute character of the proprium is weakened, for to that extent some of the attributes are not proper to God in the absolute sense of the word. But even this term contains the suggestion of a distinction between the essence or nature of God and that which is proper to it. On the whole it is preferable to speak of the "perfections" or "virtues" of God, with the distinct understanding, however, that in this case the term "virtues" is not used in a purely ethical sense. By so doing we (a) follow the usage of the Bible, which uses the term arete, rendered virtues or excellencies, in 1Pe_2:9; and (b) avoid the suggestion that something is added to the Being of God. His virtues are not added to His Being, but His Being is the pleroma of His virtues and reveals itself in them. They may be defined as the perfections which are predicated of the Divine Being in Scripture, or are visibly exercised by Him in His works of creation, providence, and redemption. If we still continue to use the name "attributes," it is because it is commonly used and with the distinct understanding that the notion of something added to the Being of God must be rigidly excluded.


Scholastics in their attempt to construct a system of natural theology posited three ways in which to determine the attributes of God, which they designated as the via causalitatis, via negationis, and via eminentiae. By the way of causality we rise from the effects which we see in the world round about us to the idea of a first Cause, from the contemplation of creation, to the idea of an almighty Creator, and from the observation of the moral government of the world, to the idea of a powerful and wise Ruler. By way of negation we remove from our idea of God all the imperfections seen in His creatures, as inconsistent with the idea of a Perfect Being, and ascribe to Him the opposite perfection. In reliance on that principle we speak of God as independent, infinite, incorporeal, immense, immortal, and incomprehensible. And finally, by way of eminence we ascribe to God in the most eminent manner the relative perfections which we discover in man, according to the principle that what exists in an effect, pre-exists in its cause, and even in the most absolute sense in God as the most perfect Being. This method may appeal to some, because it proceeds from the known to the unknown, but is not the proper method of dogmatic theology. It takes its starting point in man, and concludes from what it finds in man to what is found in God. And in so far as it does this it makes man the measure of God. This is certainly not a theological method of procedure. Moreover, it bases its knowledge of God on human conclusions rather than on the self- revelation of God in His divine Word. And yet this is the only adequate source of the knowledge of God. While that method might be followed in a so-called natural theology, it does not fit in a theology of revelation.

The same may be said of the methods suggested by modern representatives of experimental theology. A typical example of this may be found in Macintosh's Theology as an Empirical Science.18 He also speaks of three methods of procedure. We may begin with our intuitions of the reality of God, those unreasoned certitudes which are firmly rooted in immediate experience. One of these is that the Object of our religious dependence is absolutely sufficient for our imperative needs. Especially may deductions be drawn from the life of Jesus and the "Christlike" everywhere. We may also take our starting point, not in man's certainties, but in his needs. The practically necessary postulate is that God is absolutely sufficient and absolutely dependable with reference to the religious needs of man. On that basis man can build up his doctrine of the attributes of God. And, finally, it is also possible to follow a more pragmatic method, which rests on the principle that we can learn to a certain extent what things and persons are, beyond what they are immediately perceived to be, by observing what they do. Macintosh finds it necessary to make use of all three methods.

Ritschl wants us to start with the idea that God is love, and would have us ask what is involved in this most characteristic thought of God. Since love is personal, it implies the personality of God, and thus affords us a principle for the interpretation of the world and of the life of man. The thought that God is love also carries with it the conviction that He can achieve His purpose of love, that is, that His will is supremely effective in the world. This yields the idea of an almighty Creator. And by virtue of this basic thought we also affirm God's eternity, since, in controlling all things for the realization of His Kingdom, He sees the end from the beginning. In a somewhat similar vein Dr. W. A. Brown says: "We gain our knowledge of the attributes by analyzing the idea of God which we already won from the revelation in Christ; and we arrange them in such a way as to bring the distinctive features of that idea to clearest expression."19

All these methods take their starting point in human experience rather than in the Word of God. They deliberately ignore the clear self-revelation of God in Scripture and exalt the idea of the human discovery of God. They who rely on such methods have an exaggerated idea of their own ability to find out God and to determine the nature of God inductively by approved "scientific methods." At the same time they close their eyes to the only avenue through which they might obtain real knowledge of God, that is, His special revelation, apparently oblivious of the fact that only the Spirit of God can search and reveal the deep things of God and reveal them unto us. Their very method compels them to drag God down to the level of man, to stress His immanence at the expense of His transcendence, and to make Him continuous with the world. And as the final result of their philosophy we have a God made in the image of man. James condemns all intellectualism in religion, and maintains that philosophy in the form of scholastic theology fails as completely to define God's attributes in a scientific way as it does to establish His existence. After an appeal to the book of Job he says: "Ratiocination is a relatively superficial and unreal path to the deity." He concludes his discussion with these significant words: "In all sincerity I think we must conclude that the attempt to demonstrate by purely intellectual processes the truth of the deliverances of direct religious experiences is absolutely hopeless."20 He has more confidence in the pragmatic method which seeks for a God that meets the practical needs of man. In his estimation it is sufficient to believe that "beyond each man and in a fashion continuous with him there exists a larger power which is friendly to him and to his ideals. All that the facts require is that the power should be other and larger than our conscious selves. Anything larger will do, if it only be large enough to trust for the next step. It need not be infinite, it need not be solitary. It might conceivably even be only a larger and more godlike self, of which the present self would then be the mutilated expression, and the universe might conceivably be a collection of such selves, of different degree and inclusiveness, with no absolute unity realized in it at all."21 Thus we are left with the idea of a finite God.22

The only proper way to obtain perfectly reliable knowledge of the divine attributes is by the study of God's self-revelation in Scripture. It is true that we can acquire some knowledge of the greatness and power, the wisdom and goodness of God through the study of nature, but for an adequate conception of even these attributes it will be necessary to turn to the Word of God. In the theology of revelation we seek to learn from the Word of God which are the attributes of the Divine Being. Man does not elicit knowledge from God as he does from other objects of study, but God conveys knowledge of Himself to man, a knowledge which man can only accept and appropriate. For the appropriation and understanding of this revealed knowledge it is, of course, of the greatest importance that man is created in the image of God, and therefore finds helpful analogies in his own life. In distinction from the a priori method of the Scholastics, who deduced the attributes from the idea of a perfect Being, this method may be called a posteriori since it takes its starting point, not in an abstract perfect Being, but in the fulness of the divine self-revelation, and in the light of this seeks to know the Divine Being.


The question of the classification of the divine attributes has engaged the attention of theologians for a long time. Several classifications have been suggested, most of which distinguish two general classes. These classes are designated by different names and represent different points of view, but are substantially the same in the various classifications. The following are the most important of these:

1. Some speak of natural and moral attributes. The former, such as self-existence, simplicity, infinity, etc., belong to the constitutional nature of God, as distinguished from His will. The latter, as truth, goodness, mercy, justice, holiness, etc., qualify Him as a moral Being. The objection to this classification is that the so-called moral attributes are just as truly natural (i.e. original) in God as the others. Dabney prefers this division, but admits, in view of the objection raised, that the terms are not felicitous. He would rather speak of moral and non-moral attributes.

2. Others distinguish between absolute and relative attributes. The former belong to the essence of God as considered in itself, while the latter belong to the divine essence considered in relation to His creation. The one class includes such attributes as self- existence, immensity, eternity; and the other, such attributes as omnipresence and omniscience. This division seems to proceed on the assumption that we can have some knowledge of God as He is in Himself, entirely apart from the relations in which He stands to His creatures. But this is not so, and therefore, properly speaking, all the perfections of God are relative, indicating what He is in relation to the world. Strong evidently does not recognize the objection, and gives preference to this division.

3. Still others divide the divine perfections into immanent or intransitive and emanent or transitive attributes. Strong combines this division with the preceding one, when he speaks of absolute or immanent and relative or transitive attributes. The former are those which do not go forth and operate outside of the divine essence, but remain immanent, such as immensity, simplicity, eternity, etc.; and the latter are such as issue forth and produce effects external to God, as omnipotence, benevolence, justice, etc. But if some of the divine attributes are purely immanent, all knowledge of them would seem to be excluded. H. B. Smith remarks that every one of them must be both immanent and transeunt.

4. The most common distinction is that between incommunicable and communicable attributes. The former are those to which there is nothing analogous in the creature, as aseity, simplicity, immensity, etc.; the latter those to which the properties of the human spirit bear some analogy, as power, goodness, mercy, righteousness, etc. This distinction found no favor with the Lutherans, but has always been rather popular in Reformed circles, and is found in such representative works as those of the Leyden Professors,23 Mastricht and Turretin. It was felt from the very beginning, however, that the distinction was untenable without further qualification, since from one point of view every attribute may be called communicable. None of the divine perfections are communicable in the infinite perfection in which they exist in God, and at the same time there are faint traces in man even of the so-called incommunicable attributes of God. Among more recent Reformed theologians there is a tendency to discard this distinction in favor of some other divisions. Dick, Shedd, and Vos retain the old division. Kuyper expresses himself as dissatisfied with it, and yet reproduces it in his virtutes per antithesin and virtutes per synthesin; and Bavinck, after following another order in the first edition of his Dogmatics, returns to it in the second edition. Honig prefers to follow the division given by Bavinck in his first edition. And, finally, the Hodges, H. B. Smith, and Thornwell follow a division suggested by the Westminster Catechism. However, the classification of the attributes under two main heads, as found in the distinction under consideration, is really inherent in all the other divisions, so that they are all subject to the objection that they apparently divide the Being of God into two parts, that first God as He is in Himself, God as the absolute Being, is discussed, and then God as He is related to His creatures, God as a personal Being. It may be said that such a treatment does not result in a unitary and harmonious conception of the divine attributes. This difficulty may be obviated, however, by having it clearly understood that the two classes of attributes named are not strictly co-ordinate, but that the attributes belonging to the first class qualify all those belonging to the second class, so that it can be said that God is one, absolute, unchangeable and infinite in His knowledge and wisdom, His goodness and love, His grace and mercy, His righteousness and holiness. If we bear this in mind, and also remember that none of the attributes of God are incommunicable in the sense that there is no trace of them in man, and that none of them are communicable in the sense that they are found in man as they are found in God, we see no reason why we should depart from the old division which has become so familiar in Reformed theology. For practical reasons it seems more desirable to retain it.

QUESTIONS FOR FURTHER STUDY: What objections are there to the use of the term attributes as applied to God? Do the same objections apply to the German "Eigenschaften" and the Holland "eigenschappen"? What name does Calvin use for them? What objection is there to the conception of the attributes as parts of God or as additions to the Divine Being? What faulty conceptions of the attributes were current in the Middle Ages? Did the Scholastics in their search for the attributes follow an a priori or an a posteriori, a deductive or an inductive method? Why is their method inherently foreign to the theology of revelation? What classifications of the attributes were suggested in addition to those mentioned in the text? Why is it virtually out of the question to give a faultless division? What division is suggested by the Westminster Catechism?

LITERATURE: Bavinck,Geref. Dogm. II, pp. 100-123; Kuyper, Dict. Dogm., De Deo I, pp. 268-287; Honig, Geref. Dogm., pp. 182-185; Hodge, Syst. Theol. I, pp. 368-376; Shedd, Dogm. Theol. I, pp. 334-338; Thornwell, Collected Works, I, pp. 158-172; Dabney, Lectures on Theol., pp. 147-151; Pieper, Christl. Dogm. I, pp. 524-536; Kaftan, Dogm., pp. 168-181; Pope, Chr. Theol. I, pp. 287-291; Steenstra, The Being of God as Unity and Trinity, pp. 89-111.