How Should We Distinguish God's Attributes?
There are many ways to think about God's attributes. They can be divided between those that affirm something about God and those that deny something about God, those that are communicable to creatures (i.e. can exist in or be communicated to them somehow) and those that are incommunicable, and those that are expressed ad intra (within the Godhead) and ad extra (toward creatures), among other possible divisions. Today, I will consider a related issue: Whether the divine attributes are distinct in God--as opposed to being one and the same thing--and how we ought to describe that distinction, if indeed it exists.
It may seem odd to ask such questions, but this has been an important issue for theologians throughout Church history. It became particularly controversial during the medieval period, continued to be debated throughout the Reformation era, and has made its way down to the present day. At issue is the way we think about God and his relationship with us, so it is a matter of no small importance.
Why is the distinction of divine attributes problematic or difficult?
Perhaps the most natural reaction for the uninitiated person when coming across this debate is to ask, "Why does it matter?" The average churchgoer accepts easily enough that God has the attributes mentioned in Scripture (holiness, wisdom, eternity, etc.) and doesn't give them deeper consideration until one or more attributes--say, love and justice--seem to conflict with one another. The way that theologians distinguish the attributes in seminary classrooms may be of little interest to most, or at the very least it may not be intuitively obvious to outsiders why this process should be controversial.
In short, the issue is difficult because it hinges on a kind of paradox. On the one hand, Reformed Christians and many belonging to other traditions confess that God is simple, not in the sense of being easy to understand, but in the sense of being indivisible and immutable, a perfect existence of unchanging being, in whom the fullness of all that is good eternally resides--and in whom there is no composition of separate parts (A longer explanation of divine simplicity can be found here.) See, for example, the wording of the Westminster Confession of Faith.
"There is but one only, living, and true God, who is infinite in being and perfection, a most pure spirit, invisible, without body, parts, or passions; immutable, immense, eternal..."
Because the Reformed churches confess that God is without "parts," the theologians in this tradition have almost universally held that it is not possible for God's attributes to be really different things in him. (Notice here that I say "in him" and not "in his works" or "in our minds." The distinction becomes important.) God's attributes belong to his being, and that being cannot be divided. There was no collection of parts that had to come together to make God, as argued by the following theologians:
"From the simplicity of God it follows that God and His attributes are one. The attributes cannot be considered as so many parts that enter into the composition of God, for God is not, like men, composed of different parts. Neither can they be regarded as something added to the Being of God, though the name, derived from ad and tribuere, might seem to point in that direction, for no addition was ever made to the Being of God, who is eternally perfect. It is commonly said in theology that God's attributes are God Himself, as He has revealed Himself to us." - Louis Berkhof 
"Each attribute is identical with God's being: he is what he possesses. In speaking of creatures we make all sorts of distinctions between what they are and what they have. A person, for example, is still human even though he or she has lost the image of God and has become a sinner. But in God all his attributes are identical with his being. God is light through and through; he is all mind, all wisdom, all logos, all spirit, and so forth." - Herman Bavinck 
That is one end of the paradox--an extreme from which we must steer clear. However, there is another end that has been equally concerning to some Reformed theologians: If we deny any distinction between the attributes of God, we rob them of meaning and turn them into what Geerhardus Vos called "empty sounds." Consider what he has to say here:
"We may be content to say that all God's attributes are related most closely to each other and penetrate each other in the most intimate unity. However, this is in no way to say that they are to be identified with each other. Also in God, for example, love and righteousness are not the same, although they function together perfectly in complete harmony. We may not let everything intermingle in a pantheistic way because that would be the end of our objective knowledge of God."
Yes, those who have pressed for a closer identification between God's attributes have often been accused of pantheistic tendencies. Some have pushed back against that designation, but the point is that it is a cause for concern among certain theologians, such as Charles Hodge:
"If in God eternity is identical with knowledge, knowledge with power, power with ubiquity, and ubiquity with holiness, we are using words without meaning when we attribute any perfection to God. We must, therefore, either give up the attempt to determine the divine attributes from our speculative idea of an infinite essence, or renounce all knowledge of God, and all faith in the revelation of Himself, which He has made in the constitution of our nature, in the external world, and in his Word."
Attempting to hold together the simplicity of God and the different meanings of his attributes is the difficult task of the Reformed theologian.
What types of distinctions have been proposed?
Several solutions have been offered for this puzzle in the form of different types of distinctions. Originating in the medieval period of Christianity and much debated among scholastic theologians, these distinctions have come down to us with Latin names and definitions that can be difficult to understand. Adding to the difficulty is the fact that theologians do not always mention every type of distinction in their relevant works, and they sometimes refer to the same type of distinction by different names.
For the sake of simplicity and clarity, I will stick with the distinctions found in Richard Muller's Dictionary of Latin and Greek Theological Terms, as this is a widely respected work that seems to cover all the options. Muller outlines the following types of distinction mentioned by the Reformed scholastic theologians of the 16th and 17th centuries.
- Real Distinction (distinctio realis, distinctio absoluta, distinctio adaequata)
- Modal Distinction (distinctio modalis)
- Formal Distinction (distinctio formalis, distinctio formalis a parte rei)
- Distinction by reason of analysis, i.e. a distinction by way of rational analysis but with its foundation in the thing (distinctio rationis ratiocinatae, distinctio rationis ratiocinatae quae fundamentum in re), more commonly called a Virtual Distinction
- Distinction by reason reasoning, i.e. a purely rational distinction with no foundation in the thing (distinctio rationis rationans), more commonly called a Nominal Distinction
Among the great medieval theologians, the formal distinction was most associated with Duns Scotus, the virtual distinction with Thomas Aquinas, and the nominal distinction with William Ockham. The real distinction, as exists between two separate things or two parts of a composite thing, has always been rejected as a descriptor for God's attributes by those who hold to divine simplicity--at least, by those who do so consistently. The modal distinction, which Muller says "can be identified either between a thing and the various modes of its subsistence or among the various modes or ways in which a thing exists," has likewise had minimal impact among Reformed theologians in this context and is not associated with Scotus, Aquinas, or Ockham.
Reformed theologians have typically gravitated toward the other three distinctions, and they are therefore of the greatest interest to us. By no means were these three preeminent metaphysicians of the medieval period--Scotus, Aquinas, and Ockham--the only theologians to hold such views, but they have come down to us as the most prominent promoters of these ideas.
The formal distinction between God's attributes attempts to maintain divine simplicity by making a metaphysical argument: The attributes are different in form, but create no division of substance or essence in God. According to James Dolezal, this distinction...
"...holds that though God's attributes may be really identical in him they must be formally distinct from each other since the form of any attribute cannot be identical with the form of any other attribute...This formal distinction (or formal non-identity) is not merely in the mind, but is in objects themselves, including God. Underlying this formal distinction is Scotus's commitment to univocism in which attributes are said univocally of God and creatures with the difference being located in God's infinite mode of being and the creature's finite mode."
The two objections that Reformed theologians have typically raised in regard to the formal distinction are both mentioned in Dolezal's explanation: 1) the distinction between God's attributes is found formally within the Godhead or in something foundational beyond the divine essence, and, 2) Scotus' theory is closely linked with his view that attributes can be univocally predicated of God and creatures, meaning that when we say God is "good" and creatures are "good," we use the word in exactly the same way rather than in an analogical manner, the latter being the position of Thomas Aquinas.
Furthermore, Dolezal sees a problem in the fact that these "forms" of God's attributes are not identical with each other or with God's being. He writes, "Since the 'forms' are not 'things'...Scotus can still say that God is not composite; but this is a weakened non-composition in which the Godhead itself is not the most basic sufficient explanation of the divine attributes."
Against this formal distinction of Scotus, we must consider the virtual distinction of Aquinas, which goes even further in attempting to maintain divine simplicity:
"Thomas Aquinas explains the distinction between God's attributes as a virtual distinction (distinctio virtualis), which is a distinction of reason with an extramental foundation in its object (distinctio rationis ratiocinatae quae habet fundamentum in re). As close as this may appear to Ockham's later nominalism, it is a decidedly different position inasmuch as the foundation of each attribute is not a distinct concept underlying it, but is, rather, the divine essence itself."
Aquinas therefore argues that the divine attributes are identical in God but distinct in our minds (i.e. by reason of analysis). However, he states that these distinctions do have a basis in God himself (i.e. "an extramental foundation in its object"). That is, when composite, finite, fallible creatures attempt to understand a simple, infinite, perfect Creator, we may focus on one aspect of him at a time. This is not a false impression of God, but merely an incomplete one caused by our limitations. We can never know God perfectly in his essence, nor fully comprehend his being. If that seems like a bit of a cop-out or an appeal to mystery, then you may be understanding it correctly, because there is a certain extent to which Aquinas argues that we know it can't be option A or option B, so while we don't know all the details about option C, it must be the correct one.
Muller describes the virtual distinction in this manner, using its Latin description:
"Since this distinction is neither between things nor in a thing, it is purely rational; yet it is claimed as a distinction expressive of extramental reality since it is grounded on the thing and therefore preserved from being merely a product of the mind. In other words the distinctio rationis ratiocinatae represents no distinction in the thing but a truth of reason concerning the thing."
Finally, we need to consider the nominal distinction of William Ockham. Those who hold to nominalism in this context teach that there is no difference between the divine attributes in God, nor any foundation for such a difference in him, but the separation comes about in human reasoning. Muller describes it as "a merely rational distinction resting only on the operation of the reason and not on the thing." Perhaps you can see how this hits upon one of the extremes Reformed theologians hope to avoid: The idea that the Scriptures do not tell us things that are really true of God, but merely provide "empty sounds" to suit our minds.
Which type of distinction has been historically favored by Reformed theologians?
The degree to which Aquinas, Scotus, and Ockham respectively influenced those in the Reformed tradition is a matter of no little debate at the present time. Some would prefer that Reformed thinkers steer clear of all three of them, but that is a rather impossible task given what I have described in this article: There is a spectrum of options for how we describe the distinctions between God's attributes, and everyone has to plant their flag somewhere. Going all the way to the other end of the spectrum and arguing for a real distinction between the attributes in God would bring an end to divine simplicity.
Perhaps no historian of Reformed thought has given this issue more consideration than Richard Muller, whose research is still ongoing. His analysis, though not universally accepted in its entirety, has served to set the tone and parameters of the current debate.
In both his Dictionary and Post-Reformation Reformed Dogmatics, Muller draws conclusions as to how the early Protestant theologians, and particularly those in the Reformed tradition, approached this issue. He writes that the Reformed orthodox "tended to accept the Thomist and nominalist arguments against Scotus, but remained dissatisfied with the nominalist and, at times, with the original Thomist solution."  He then adds,
"The influence of Thomist, Scotist, and nominalist thought on the orthodox doctrine of the attributes was, thus, of enough importance that the orthodox needed to define their position over against that of the great medieval systems with great care: they accepted the Scotist and nominalist assumption of a disproportion between the finite and the infinite, but they were at the same time unwilling to press the distinction to the point that the principle of analogy utterly disappeared -- as also they were unwilling to press the argument over the precise manner of distinction to be found between the divine attributes to the point of insisting categorically that the attributes must be eminently or rationally, rather than formally, distinct or, indeed, rather than distinct only ad extra."
After reading that explanation, one may be tempted to think that there was little or no consensus among the Reformed on this matter, but the picture becomes clearer when examining other portions of Muller's work. While he has written in a general sense that "all of the Reformed orthodox assume the simplicity of the divine essence, and all understand the attributes as in some sense distinct," he says elsewhere that, "Most of the Protestant scholastics reject the formal distinction and accept the distinctio rationis rationcinatae," providing a more specific picture. (Distinctio rationis rationcinatae is synonymous with the virtual distinction of Aquinas.) Elsewhere in his Dictionary, Muller notes,
"This distinctio rationis ratiocinatae, or distinction of ratiocinative reason, also called a distinctio virtualis, is taught by nearly all the Lutheran and Reformed scholastics. It is, incidentally, the solution to the problem of predication proposed by Thomas Aquinas and Henry of Ghent. The Scotist concept of a distinctio formalis, or formal distinction, is occasionally used by the Reformed orthodox to explain the distinction of attributes in the divine work, ad extra."
Here we see that the Reformed orthodox have occasionally made a division between how they distinguish God's attributes in his being ad intra and in his works ad extra. To speak of his works ad extra--that is, what God does in linear time in relation to creatures, or how they are impacted by him--as opposed to the divine being ad intra is, in fact, to draw very nearly the type of distinction that Aquinas allows for in his virtual distinction between the attributes. There is a slight difference of nuance that has caused Reformed theologians to reach varying conclusions, but in general, Muller argues that both the Reformed and Lutheran theologians have been closer to the virtual distinction than formal or even the nominal one. In summary, he divides the opinions of the early Reformed orthodox into three categories when it comes to the distinction of God's attributes:
"...first, the attributes are essentially one in God, but known to reason as distinct in their operation ad extra; second, the attributes are essentially one in God, are understood by us as distinct ratio rationcinata cum fundamento in re, but are also recognized to be distinct in God eminenter or virtualiter and, according to some of the Reformed writers, distinct realiter in their effects; third, in a slight variant or elaboration of the second, the attributes are essentially identical in God, externally distinct rationaliter or formaliter, as known in their operations, and distinct in the Godhead itself eminenter or virtualiter."
Reviewing this explanation, the common thread is that all the attributes are essentially identical in God. The differences of description come about in the works ad extra and their effects. I therefore conclude that Dolezal is correct when he states that, "In keeping with this virtual distinction, Thomas and the Reformed argue that the divine attributes are really identical in God." There is some question in my mind as to whether Vos and Hodge would have embraced that definition, given their quotes earlier in this article, but it certainly seems to have been the opinion of Turretin, Owen, Bavinck, and Berkhof. And we must remember that some of the comments which stress a lack of identity between God's attributes are simply pressing back against nominalism, which is seen to be dangerous in other areas of theology as well.
The virtual distinction of God's attributes is therefore the most typical of Reformed thought, as it steers the best middle course between two extremes, maintaining both divine simplicity and the reality of biblical descriptions of God's character. Though no explanation will ever be perfect, and variations have been offered, the Reformed seemed to have followed Aquinas in the main on this particular issue over and against Scotus and Ockham, even if they have done so begrudgingly at times.
 1646 Westminster Confession of Faith 2.1
 Berkhof, Louis. Systematic Theology, Kindle edition (Escondido, CA: Ephesians 4 Group, 2017), 25-6.
 Bavinck, Herman. Reformed Dogmatics, Vol. II - God and Creation, Ed. John Bolt, Trans. John Vriend (Grand Rapids, MI: Baker Academic, 2004), 118.
 Vos, Geerhardus J. Reformed Dogmatics, Volume One: Theology Proper, Trans. and ed. Richard B. Gaffin Jr., Kindle edition (Bellingham, WA: Lexham Press, 2014), location 106. He is speaking in this case about the names of God, but the context suggests that we should take them in the same way as the divine attributes: descriptions of God's character.
 Vos, 135-48.
 Hodge, Charles. Systematic Theology: The Complete Three Volumes, Kindle edition (Lousiville, KY: GLH Publishing, 2015), location 7053.
 Muller, Richard A. Dictionary of Latin and Greek Theological Terms, Second Edition (Grand Rapids, MI: Baker Academic, 2017), 95-6.
 Dolezal, James. God Without Parts: Divine Simplicity and the Metaphysics of God's Absoluteness, Kindle edition (Eugene, OR: Pickwick Publications, 2011), 130-1.
 Dolezal 132-3.
 Dolezal, 133.
 Muller, Dictionary of Latin and Greek Theological Terms, 95-6.
 Muller, Dictionary, 96.
 Muller, Richard A. Post-Reformation Reformed Dogmatics, Volume Three - The Divine Essence and Attributes, 2018 reprinting with minor revisions (Grand Rapids, MI: Baker Academic, 2003), 287.
 Muller, PRRD, 288.
 Muller, PRRD, 297.
 Muller, Dictionary, 95-6.
 Muller, Dictionary, 43-4.
 Muller, PRRD, 289.
 Dolezal, 135-6.
Amy Mantravadi holds a B.A. in Biblical Literature from Taylor University. She is an active member of Patterson Park Church in Beavercreek, Ohio. You can read her blog at www.amymantravadi.com or follow her on Twitter @AmyMantravadi.
"A Cross-Shaped View of God's Attributes" by Aaron Denlinger
"Where's the Love? Polanus on Divine Goodness" by Ryan McGraw
The Identity and Attributes of God by Terry L. Johnson