Words, Thoughts, and Things
Linguistic Analysis, Method, and Pedagogy We do not know how Anselm might have reconnected what he has distinguished in the Fragments. He walks to the edge of language that seems to drop off into equivocation, where what is not can be described as being, and what does nothing can be said to do or cause. In Anselm's other work, his impulse is to distinguish in order to unite, to resolve paradox into necessity. In some ways De grammatico and these fragments represent only the first movement, creating distinctions and paradox out of ordinary and noncontroversial language, not union of differences or the resolution of paradox into necessity. We see only the very beginnings of the tendency toward unity in the fragments, in their ordering of the improper to the proper, the many to the one, and in De grammatico's creation of distinctions that let both halves of an apparent contradiction stand. The trilogy of dialogues beginning with De veritate takes up this task, uniting all truths in the One Truth but also exploring division on a deeper level, division in the will causing separation from God.
Words, Thoughts, and Things Where does this reading of these two works leave us on the relationship between words, thoughts, and things for Anselm? Many have taken this as the question of whether Anselm a realist or not. Do words signify things or concepts directly? If things have universal natures, do universals exist only post or also ante rem? Depending on which texts are taken as central, De grammatico, or the Proslogion and Monologion, Anselm has been taken to be everything from a naïve realist to a moderate realist to a contemporary analyst on the subject of universals.113 Marilyn Adams argues that Anselm is a realist in the Monologion and Proslogion but pulls back to a more moderate and nuanced position in the De grammatico.114
However, a careful look at the evidence shows that Anselm does not have a different theory of language in De grammatico as opposed to the Monologion or elsewhere; rather, De grammatico and the Fragments simply focus on the distinctions between words, things, and understandings.
The famous passage on which Anselm is sometimes judged is not in these logical, grammatical works but in his refutation of Roscelin's views on the Trinity. This response makes Anselm sound, at least in terms of the categories of the twelfth century if not the eleventh, like a real113. On the range of these views, see Jasper Hopkins, "The Anselmian Theory of Universals,"
57–68. See also Katherin Rogers, The Neoplatonic Metaphysics and Epistemology of Anselm of Canterbury,
Studies in the History of Philosophy 45 (Lewiston, N.H.: The Edwin Mellen Press, 1997), 114–17.
114. Marilyn Adams, "Was Anselm a Realist? The Monologium," Franciscan Studies 32 (1972):
5–14.
Linguistic Analysis, Method, and Pedagogy ist. Roscelin, Anselm complains, essentially holds the view that either there are three gods or the Father and Holy Spirit became incarnate with Christ. According to Anselm, such errors spring from Roscelin's view words are only flatus vocis. Anselm's well-known dismissal of Roscelin as both logician and theologian bears repeating:
For someone who does not yet understand how several men are one man in species, how will he comprehend the way in which in that highest and most mysterious nature several persons, each one of which individually is perfect God, are one God? And someone whose mind is too darkened to distinguish between his horse and its color, how will he distinguish between the one God and his multiple relations? Finally, someone who cannot understand anything to be human except an individual will in no way understand a human being to be anything except a human person. . . . How, then, will he understand that man was assumed by the word not a person, that is another nature not another person was assumed?115
This passage has received notoriously different readings. For some, Anselm's almost naïve realism is evident in his assertion of a human nature distinct from individual human beings. This view can arguably be traced through the rest of Anselm's corpus. Just as the name of God in the Proslogion necessarily refers to a reality, so too, some would argue, the names of the perfections of God correspond to those realities in God, and of which the less than divine instances are imperfect reflections.
The Monologion, on this view, asserts the Platonic view that all goods are good through that which is good-in-itself in order to prove God's existence.116 On the other hand, Henry argues that in this passage Anselm is merely criticizing Roscelin for failing to make distinctions in the order of words. Hence, he argues, Anselm is asserting the importance of distinguishing between "quidditative" and "referential" uses as Anselm does in De grammatico.117 Henry's reading, however, seems to go against the obvious sense of the passage.118 For Anselm is clearly asserting something about being and not just words; he claims that human beings are united in the species, human nature, and that individuals and the nature they inhabit are distinct. The distinction between person and nature makes it possible for Christ to assume human nature without assum-
115. EDIV 1, S II, 10, 4–13.
116. See Maurice de Wulf, History of Medieval Philosophy, vol. 1, translated by Ernest C. Messenger (New York: Dover, 1952), 165. Cited in Rogers, Neoplatonic Metaphysics, 116. In a different way,
Marilyn Adams also argues that the Monologion commits Anselm to realism.
117. Desmond P. Henry, "St. Anselm and the Linguistic Disciplines," Anselm Studies 2 (1988):
329.
118. Hopkins makes the same point. See his "The Anselmian Theory of Universals," 63.
Linguistic Analysis, Method, and Pedagogy ing human personhood but retain divine personhood, while the unity of Christ's divinity with human nature makes it possible for Christ to substitute for and make restitution for Adam and all human beings. Thus
Anselm's Christology requires a distinction between natures, which are not mere fictions, and individuals. However, Anselm does not specify the exact status of those natures, either as subsisting universals, as ideas in God's mind, or as the product of abstraction from individuals.119
Anselm, like virtually all thinkers in the Latin medieval tradition before Ockham, holds that there are ideas in the mind of God which are in some important sense the source of created natures. The Monologion clearly refers to ideas or exemplars in the mind of God. At issue, then, are not universals in the sense of divine ideas. However, Anselm does not have to hold the view that the horse and its color (his example in the passage on Roscelin previously quoted) can exist separately but only that the distinction between them conceptually corresponds to a real distinction in those realities. Anselm is, we might say, though still perhaps in language more appropriate to the twelfth century than the eleventh, committed to the reality of natures in things. If what is meant by realism is the view that things really have natures that are designated by the common nouns we apply to them, then Anselm is a realist, but so is
Aquinas and even Abelard (though the latter with some qualification).120
However, Anselm need not be an "extreme realist"; he could be a kind of Aristotelian, holding that the forms are in individuals without existing separately.121
It is important to remember the view Anselm is arguing against here:
Roscelin's.122 Roscelin is not, as Mews and Jolivet have pointed out, easily 119. Hopkins maps out clearly the ways in which scholars have argued that Anselm is realist or Thomist on the question of universals; see his "The Anselmian Theory of Universals," 58–69.
Hopkins's own conclusion is agnostic and aporetic, concluding that Anselm simply does not say what secondary substances are, just that they are not mere vocal sounds. Hopkins, 95.
120. For a discussion of Abelard see Eileen Sweeney, Logic, Theology and Poetry in Boethius, Abelard and Alan of Lille: Words in the Absence of Things (New York: Palgrave Macmillan, 2006), 70–71.
121. This does not preclude Anselm being more Platonic than Aristotelian in terms of how he thinks of the relationship between divine ideas or exemplars and existing, material individuals.
As is clear in the Monologion, Anselm finds more "reality" in the divine idea than the material instance. Hopkins is rightly skeptical not only of terms like "realist" or "nominalist" in discussing
Anselm but also "Platonic" or "Aristotelian," though he reluctantly has to use some kind of label, concluding, "I tend to agree with Henry that Anselm is no "crude realist," while at the same time agreeing with Grabmann that Anselm is dimly aligned with Boethius and Aquinas." Hopkins, 95.
Cf. Iwakuma Yukio, "The Realism of Anselm and His Contemporaries," in Anselm, Aosta, Bec and Canterbury, edited by Luscombe and Evans (Sheffield : Sheffield Academic Press, 1996), 120–35, for an account of Anselm as Boethian and Aristotelian.
122. For a full discussion of Anselm's Epistola de incarnatione Verbi and his disagreement with Roscelin on the Trinity, which prompts the comment quoted above, see chapter 6.
Linguistic Analysis, Method, and Pedagogy categorized as a nominalist; if he is not, then we might have to reevaluate what Anselm is holding out as the alternative to Roscelin as something different from realism. Jolivet argues that Roscelin's account of words is governed by reference, each word referring to a thing.123 This fits with Roscelin's letter to Abelard, the only text we have which scholars are sure is his. Roscelin concludes that three names and three persons means three things that are "eternal, just as a plurality of eternal things."124
Roscelin's view is a kind of literalism rather than nominalism, based on the common sense (but wrong) notion that there is a direct relationship between names and things; hence he claims that if there are three persons there are three things, that is, three substances or essences.125
If there are not different things, then the different names which might be used of that same thing do not really name anything different. From this we can understand Anselm's frustration with what he characterizes an immersion in sense and imagination in Roscelin and other dialecticians. They let the physical material world be the final arbiter of what can be meaningfully said. In contrast, Anselm holds that the substance and its qualities are distinct without claiming they are subsisting things or conceding that they are just different names for the same single, onedimensional thing. Anselm's rejection is of the flatness of the views of the dialecticians he criticizes. In insisting on the distinction between nature and individual or person, Anselm is insisting on another layer to multilayered being consisting of substance, quantity, quality, etc.
We might be able to characterize Anselm, strangely enough, as Jolivet does Abelard, as a combination of "nonrealism" and "Platonism."126 He is a nonrealist in the sense that he does not take the orders of words and things to be symmetrical, does not take it that there are "things" which correspond to words, but Platonic in the sense that there are exemplars in the mind of God (identical to the divine substance). For Anselm there
123. Jean Jolivet, "Trois variations médiévales sur l'universel et l'individu: Roscelin, Abélard,
Gilbert de la Porrée," Revue de métaphysique et de morale 97 (1992): 111–55.
124. Roscelin, letter to Abelard, cited in Constant Mews, "The Trinitarian Doctrine of Roscelin of Compiègne and its Influence: Twelfth-century Nominalism and Theology Re-considered," in Reason and Belief in the Age of Roscelin and Abelard (Burlington, Vt.: Ashgate, 2002), IX, 357, n. 33. Mews cites large portions of the letter and translates a number of passages in this essay, 355–
58. For the full Latin text see the edition edited by J. Reiners, in Der Nominalismus in der Frühscholastik. Ein Beitrag zur Geschichte der Universalienfrage im Mittelalter, Beiträge zur Geschichte der Philosophie und Theologie des Mittelalters 8.5 (Münster: Aschendorff, 1910), 62–80.
125. Roscelin, "Letter to Abelard," 72, cited in Mews, "Trinitarian Doctrine of Roscelin," IX,
355–56; cf. Mews, IX, 357.
126. Jean Jolivet, "Non-réalisme et platonisme chez Abélard: Essai d'interprétation, " in Abélard et son temps, Actes de Colloque international organisé à l'occasion du 9e centenaire de la naissance de
Pierre Abélard (May 14–19, 1979) (Paris: Les Belles Lettres, 1981), 175–95.
Linguistic Analysis, Method, and Pedagogy are real natures or essences in things, which, though not as separately existing objects, can be discussed meaningfully, independently of the particular existing things and in abstraction from their other features.
Though it may be unfair to Roscelin, Anselm seems to want to cast him as his opposite, a combination of realism and non-Platonism, convinced of a kind of direct relationship between words and things but rejecting any way of using words that maps on to ideas (whether in the human mind or the mind of God) rather than concrete individuals.
Some scholars who reject the view that Anselm is an extreme realist, have then tried to see him in a line with Boethius and Aquinas, both as moderate realist and as holding what they see as an Aristotelian account of knowledge as deriving from sense experience and as opposed to an Augustinian or illuminationist view).127 Like nominalism vs. realism, these views as distinct and mutually exclusive postdate Anselm by at least a century. Anselm speaks both of understanding as a kind of presence of or illumination by the divine light and of coming to know things in and through our physical encounters with them.128 Moreover, Anselm does not take up in any detail the question of how it is that humans come to know. What is clear is that Anselm distinguishes strongly between sensation and knowledge and speaks of God and divine light as the source of intelligibility and, though he does not specify how, of human knowledge.
This much Anselm shares not just with Augustine but also Boethius and Aquinas, though we could say that for Anselm over against Abelard (and to a lesser degree over against Aquinas and Boethius) there is a more direct relationship between things, words, and divine ideas. On universals,
Anselm shares with Boethius, Aquinas, and even Abelard the notion that meaning is not reference, that words are linked in one way to ideas and concepts and in another to individual existing objects. This is the point of De grammatico, that appellation and meaning, signification per se and per aliud are distinct, and it is the same point Anselm makes as he insists on the distinction between natures and persons in the De incarnatione.
127. On Anselm as abstractionist, see F. S. Schmitt, "Anselm und der (Neu-)Platonismus," in vol. 1 of Analecta Anselmiana: Untersuchungen über Person und Werk Anselms von Canterbury, edited by F. S. Schmitt (Franfurt: Minerva GMBH, 1969), 48, 62–64. In order to acquit Anselm of the charge of being a Neoplatonist, Schmitt defines Neoplatonism in fairly narrow and technical terms. See
Schmitt, "Anselm und der (Neu-) Platonismus," 70–71. On this debate among Anselm scholars see
Rogers, Neoplatonic Metaphysics, 159–62. Rogers argues, against Schmitt and others, that Anselm's views are illuminationist. She argues against the view that knowledge comes from innate ideas or "a created light in the human mind which is caused and constantly supported by God," but argues that for Anselm knowledge comes about by direct illumination by God, who gives not just to the ability to make judgments about sense knowledge but gives the actual content. See Rogers, Neoplatonic Metaphysics, 164–84.
128. Rogers, Neoplatonic Metaphysics, 164–74.
Linguistic Analysis, Method, and Pedagogy Though the question about the degree of Anselm's realism is anachronistic, the suspicion that Anselm is a realist is, in the end, a response to something that is really present in his thought. Anselm aspires to find the words that deliver the thing, arguments that are beyond doubt.
In other words, he seeks the perfect correspondence between words and things realism also promises. In order for language to have the power to propel us to necessary conclusions about things, language must be more than mere sounds, human custom, or even just roughly on to something.
It has to be able to name what is real about things. However, readers arguing for Anselm's realism neglect the other strong element in Anselm's writing we have seen in De grammatico and the Fragments also present in his speculative works: his sophisticated awareness of the gaps between word, thing, and understanding. This caution is evident both within the grand arguments of the theological and philosophical treatises and in the careful analysis of the grammatical,or logical works. Anselm is not worried about the "problem of universals." What he is worried about are the ways in which ordinary language connects and fails to connect to the world in more complicated ways than his students and readers suspect.
The goal of Anselm's linguistic analysis is to clear the ground, to create a launching pad in order to mount up from this world to God powered by words. The results of the logical works help make the grandest instances of this quest, the necessary and indubitable conclusions of the Monologion, Proslogion, and Cur Deus homo possible, by rooting out apparent but false identities, by explaining the differences (and links) between similar names and predications, uses, and meanings. The thrust of Anselm's thinking is to envelop all truths in the One Truth, God, as he does in De veritate. But he knows the dangers of discrediting his whole project by making identical things that are different. As Southern points out, the fool addressed in the Proslogion has not learned the lesson of De grammatico: "it is not the use of the same word in the major and minor premises of a syllogism which binds the argument together, but the use of the same word with the same sense."129
What is most significant about the passage on Roscelin and the dialecticians is not, then, whether it shows Anselm to be realist or not. Rather the passage indicates how grammar and logic impact philosophical and theological issues for Anselm. Anselm upbraids Roscelin for failing to make the right distinctions and connections between individuals and species, substance and accident, and nature and person. We can see in this a reason for Anselm to go back, if indeed he did, from the sublime
129. Southern, Portrait in a Landscape, 130.
Linguistic Analysis, Method, and Pedagogy heights of the Monologion and Proslogion to the mundane explorations of De grammatico and the fragments.130 In any case, Roscelin's views seem to have forced Anselm's articulation of how, as Aquinas (citing Aristotle citing Plato) puts it in his own exploration of the problem of universals, "a small mistake at the beginning can lead to a large mistake at the end."131
Southern makes a similar point about the teaching of De grammatico that could apply to all Anselm's reflections on language: "The great lesson that he [Anselm] impresses on his pupil is that he cannot advance unless he penetrates the surface of words and appearances to reach the reality to which they refer."132
Anselm combines a great skepticism and a great deal of sophistication about language (which may be the same thing) with the naïve desire to move seamlessly from being to Being, from words to the Word.
As we shall see, De veritate consists of both the movement to distinguish between different kinds of truth and the movement to unite them in the One Truth. And as we shall see in De veritate and beyond, those who classify Anselm as an illuminationist are, like those calling him a realist, led by something in Anselm: that all that is as well as all that is true and is known finds its source in God.
The logical and theological works in Anselm are related in a way that does not make logic the mere servant of apologetics nor reduce theology to an area for making linguistic distinctions. The kind of logical and linguistic analysis in these texts is present throughout his corpus and is at the core of a profoundly metaphysical and theological project. They come together without one being reduced to the other because both the logical and theological works and the logical and linguistic analysis of theological problems are concerned with the paradox of being and language found in Parmenides.133 That problem is the apparent impossibility of making predications of either being or nonbeing. On the one hand, one cannot make predications of nonbeing without seeming to attri130. On the dating of De grammatico, see p. 78.
131. Thomas Aquinas, De ente et essentia, in Über Seiendes und Wesenheit, translation and commentary by Horst Seidl (Hamburg: Meiner, 1988), 2. Cf. Plato, Cratylus, 436d and Aristotle, De Caelo
271b9.
132. Southern, Portrait in a Landscape, 64.
133. For another, complementary way of connecting Anselm and Parmenides, see Michael P.
Slattery, "Parmenides: Anselm Eminenter," Anselm Studies 2 (1988): 229; and Vincent J. Ferrarra, "Some Reflections on the Being-Thought Relationship in Parmenides, Anselm and Hegel," in vol. 3 of Analecta Anselmiana: Untersuchungen über Person und Werk Anselms von Canterbury, edited by F. S. Schmitt (Frankfurt: Minerva GMBH, 1972), 95–111. Ferrarra argues that in the ontological argument Anselm is concerned, as Parmenides is, with the identification of being and thought. The claim here is that Anselm is concerned in the Fragments as well as in De veritate and the Proslogion with the problem that gives rise to Parmenides's reflection: how being can be many.
Linguistic Analysis, Method, and Pedagogy bute being to nonbeing in the "is" that joins subject and predicate and in the predicate that attributes a property to a nonexistent object. This is the issue that is ultimately behind the most basic questions of the Lambeth Fragments. The student asks about ability and inability, possibility and impossibility, and freedom and necessity. In part, the problems are problems of language, requiring only the distinction between verbal form predicating ability, possibility, or freedom and the real predication of the opposite. But at a deeper level, the questions are metaphysical, about how, for example, nonexistent things can have abilities or possibilities. The student grasps both the verbal and metaphysical problem when he reasons toward the paradox: since "that which does not in any respect exist has no ability; it does not have either the ability to exist or the ability not to exist."134 Being, of course, presents no less a difficulty than nonbeing (or, as we shall see, becoming). For one cannot predicate anything of being except something other than being, but the only thing other than being is nonbeing. In the medieval context, this becomes the problem of theological language: how can one say anything about God without contradicting the divine nature? The Fragments touch on these issues when raising questions about the nature of divine freedom and causation, but they are, of course, taken up more directly in the famous arguments of the Proslogion and Monologion.
What the logical works show us is Anselm's awareness that the asymmetry between language and things takes two forms. First, language taken to signify the full possession of what is predicated is ill suited to name that which is becoming or nonbeing. Second, language derived from and appropriate to the incomplete being of creatures is ill suited to describe the pure and absolute being of God. Plumbing the depths of linguistic complexity, exposing in as many ways as possible the mismatch between words and things, is the equivalent, then, of beginning from the depths of sin in the prayers. From this tangled and incomplete correspondence between words and things, the project is to move toward words which clearly and indubitably capture their objects, from division and difference to unity.
The approach to philosophical questions about the nature of reality approached through an analysis of language is, as many have remarked, more prominent in the Latin medieval world before the advent of the full
Aristotelian corpus (the works in natural philosophy, ethics, metaphysics) in the West. Language is, in this context, the main tool for solving problems. Anselm chooses to probe language deeply and at the dramat-
134. LF 23, 5–6.
Linguistic Analysis, Method, and Pedagogy ic extremes, language of pure being and absolute nonbeing. And at both extremes, he confronts the possibility of utter failure as he strives for complete success of language, in a drive toward linguistic integrity. This desire is evident in the logical works, which work to uncover and clarify ambiguities and inconsistencies in ordinary language. The same drive toward the perfect identity of word and thing is exemplified in the Prologion's search for the name that predicates being of being, and mirrored in the search for union with God in the prayers.