The extent of Anselm’s revision of Augustine doesn’t end with his substitution of a divine locutio for the Augustinian divine ideas. William Mann, for example, points out that Anselm’s particular choice of locutio to describe the Son through whom God conceives and makes all things
is initially curious; one might have expected verbum for logos. In fact, Anselm also uses verbum. But locutio conveys more clearly than does verbum an aspect of the Son that is important to Anselm. A locutio… is a speech act. But Anselm includes in the notion of a speech act something that may be surprising to modern sensibilities. Thinking is a kind of speaking, an inner speaking: concepts function as inner words in the language of thought.
Like Augustine in his De Trinitate, Anselm saw creation’s pre-existence in God as occurring within the divine verbum, yet it seems it was specifically the spoken character of this verbum that Anselm particularly wanted to stress through his use of the term locutio. This may be seen in one of the earliest critics of Anselm’s theory of the divine utterance, the thirteenth-century theologian Robert Kilwardby. As Kilwardby points out in his commentary on Lombard’s Sentences, in which he alludes to an implicit contrast between the verbal and the visual model of cognition found within Augustine’s De Trinitate,
[t]o speak is to point to or refer to something, but understanding is a kind of vision. Making a comparison with seeing, speaking has to do with the thing already seen, and understanding with what is going on in the person doing the seeing…. [L]ike external speaking, inner speaking has the character of an action… Thus it seems that understanding and inner speaking are opposites by definition; and if they are, then Anselm is wrong to say that […] in the supreme Spirit, this kind of speaking is nothing but grasping by thinking (cogitando intueri)…
According to Kilwardby, in short, there is a referentiality, intentionality, transitivity, and implied alterity to the act of speaking that is not (or at least not obviously) the case in the act of mere understanding. As Mary Sirridge has aptly put it in her analysis of Kilwardby’s critique of Anselm, “speaking is active and understanding passive.” One reason for Kilwardby’s disagreement with Anselm, moreover, is his belief in the derivative nature of the speech-act in relation to understanding: speech presupposes a prior knowledge or vision of that which is spoken. In terms of our present thesis, for Kilwardby it is an act of (passive) divine visio that is the prior possibility of any subsequent divine locutio.
 Mann, “Anselm on the Trinity,” 264.
 Robert Kilwardby, 1 Sent., q. 36, 372-386, cited in Surridge, “Utrum idem sint dicere et intelligere,” 258.
 Surridge, “Utrum idem sint dicere et intelligere,” 253.