Grudem on the Doctrine of God (I) – Communicable and Incommunicable Attributes

The treatment of the doctrine of God in Grudem’s Systematic Theology follows a traditional arrangement, covering the existence and knowability of God (in two short chapters which necessarily repeat some ground), then working through God’s attributes divided into incommunicable and communicable, then turning to the Trinity. Coming to Trinity only at the end of this section no doubt has dismayed some of his readers, but if there is a failing in this it’s not one of neglecting to follow the tradition, in which this is a (the?) common order.

The incommunicable attributes are described as being those attributes which God does not share with creatures (infinity, eternity etc), and the communicable those which he does, in some sense (love, mercy, justice etc). Grudem, however, claims that his distinction should not be preserved strongly. Instead, we should think of the attributes of God as only more or less communicable (p156-7). He has two arguments for this:

1. Of the “communicable” attributes, all are only analogously shared with creatures
   (there is a fundamental distinction between e.g. God’s love and ours)
   
2. Of the “incommunicable” attributes, all have some very weak echo in the creature
   (e.g. God is unchangeable, but we change somewhat, but not totally)

To me, this is a mistake. I wholeheartedly agree that the communicable attributes are only analogously shared by creatures, and I laud Grudem for making this vital point. However, the incommunicable attributes are expressed as absolutes, particularly absolute negations of some creaturely quality. God is infinite, meaning he has no limits whatsoever, eternal in that he has no time-duration whatsoever. One cannot have some measure of infinity – you’re either infinite or not! Similarly, we don’t share immutability simply because we’re not in a state of utter flux, nor does it make sense to say that we weakly share God’s omnipresence because we are present somewhere (the entire definition of omnipresence being to be present at all places and without any place where you’re not).

On the contrary, then, there is good reason to say the incommunicable attributes involve a qualitative, not merely quantitative difference. Turretin nails this:

In this sense, those attributes can properly be called incommunicable strictly and in every way, which are so proper to God that nothing similar or analogous, or any image and trace can be found in creatures. Such are the negative attributes which remove from him whatever is imperfect in creatures (such infinity, immensity, eternity, which are such that every creature is either without them or has their contraries). But others are not badly termed communicable (of which there is some appearance or certain faint vestiges in creatures) and by simple analogy of name and effects. Such are the affirmative attributes which are attributed to God by way of eminence or causality.

Institutes of Elenctic Theology, III.6.3 (vol 1 p190).

This is one of those places in which Grudem may be drawing, consciously or not, on Bavinck, who similarly doubts that the division between incommunicable and communicable is absolute.¹ For Bavinck, however, the argument is that if the incommunicable were totally incommunicable, they would be unknowable and therefore saying nothing. This is a rather stronger argument, but doesn’t seem to me to undermine the validity of preserving as a separate category those attributes that involve an absolute negation. The question of the ability of human language to speak of God’s essence at all is a separate issue, though clearly related.

Another interesting feature of these chapters is the position of invisibility. At first I wondered what Grudem had been smoking when he decided that invisibility was a communicable attribute. I was further confused by how he seemingly defends this with a discussion of God’s modes of revealing himself to us. In what way does it make sense to say God’s invisibility is communicated to us because he sometimes manifests himself to us despite his invisibility? But, once again, Bavinck is in the background: he does exactly the same in listing invisibility as a communicable attribute and following it with a discussion of the way in which God is/isn’t “seen”. Reading through Bavinck’s discussion, however, it seemed to me that Bavinck was simply treating the much-debated question of whether the divine essence can ever be seen by created natures under this locus – not using this discussion to defend the assignation of invisibility as communicable. I’d been confused because Grudem’s chapter generally follows this pattern (attribute -> why this attribute is incommunicable/communicable), but in this case if he is doing the same thing it doesn’t quite work.

Reading back through other systematics, I can’t find much explicit treatment of invisibility, but where it does appear it is generally treated as a corollary of spirituality. Because spirituality is communicable (human beings are partly spiritual), there is a sense in which invisibility is too (we have a spiritual component, which is invisible). This is the implication of Van Mastricht’s discussion, for example (Theoretical-Practical Theology 1.2.6.7, vol 2. p132). Bavinck seems to be following this pattern in having spirituality and invisibility as the first two communicable attributes – the inclusion of the latter being thereby implicitly justified.

These criticisms, therefore, are not specific to Grudem. There seems to be a lack of precision, possibly downstream of Bavinck, possibly always there, over the communicable/incommunicable division and which attributes belong where. Is there not a similar problem when Michael Horton describes omniscience and omnipotence as communicable attributes? (Horton ST 259). One is tempted to suggest we need a dash more scholastic precision. A more consistent approach would be to use incommunicable solely for absolute apophatic attributes (infinity, immutability, simplicity), and stick with descriptions of the communicable that don’t focus on the unique divine mode of having them (knowledge, not omniscience; presence, not omnipresence etc). Though invisibility, perhaps, remains an awkward customer.

Addendum: some other divisions

Since posting the above, I’ve been stimulated to think further on the way in which the classic incommunicable/communicable distinction can seem to foreground abstract qualities of God over the concrete characteristics which Scripture most exults in: goodness, truth, wisdom etc.

Perhaps a better way, then, is shown us by the Westminster Shorter Catechism, which declares that

God is a Spirit, infinite, eternal, and unchangeable in his being, wisdom, power, holiness, justice, goodness, and truth.

WSC, A4

In this way of framing the divine attributes, the negative ones are lenses through which to view the positive ones. God is wisdom and goodness, positive qualities we see reflected (analogically) in creatures. We understand God’s wisdom, however, by removing from the positive concepts that we see in creatures all elements that are creaturely: the function of the negative qualifiers. God’s wisdom, unlike ours, is infinite, eternal, unchangeable etc.

This approach might signal that the communicable/incommunicable division, while defensible as getting at a real distinction, is not the best categorisation. For a (much!) more thorough set of distinctions based on medieval categories, this is a stimulating lecture from Ryan Hurd worth careful attention.

1. I would guess consciously given how closely his divisions of the communicable attributes into essential, intellectual, moral follows Bavinck’s. Also in his inclusion of invisibility as communicable, on which more below. ↩︎