- John Scottus Eriugena and Anselm of Canterbury
- John Scottus Eriugena and Anselm of Canterbury Stephen Gersh INTRODUCTION by John Marenbon John Scottus Eriugena came from Ireland, as his name indicates (‘Scottus’ meant ‘Irishman’ in the Latin of this period, and ‘Eriugena’, a neologism invented by John himself, is a flowery way of saying the same thing). He worked on the Continent, however, under the patronage of Charles the Bald. The first mention of him, in a letter of 851 or 852 about the predestination controversy, is as ‘an Irishman at the royal court’. After the disastrous reception of his own contribution to this dispute, On Predestination (discussed in Chapter 5), it seems to have been Charles’s protection which saved Eriugena from punishment and ensured he could continue his work. Glosses survive by Eriugena on Martianus Capella’s On the Marriage of Mercury and Philology, a late antique handbook of the seven liberal arts widely studied in the ninth century, and it is likely that these represent some of his teaching at the palace school in the late 840s.1 Already his comments show some of the characteristic themes of his thought. For instance, a reference by Martianus to the myth of Orpheus, who tries to rescue his wife, Eurydice, from the underworld, is glossed in terms of the relation between the beauty of sound (represented by Orpheus) and the art of music ‘in its profoundest reasons’ (represented by Eurydice), which the musician must seek by descending into the depths of his discipline. Eriugena’s intellectual horizon was greatly enlarged in the 850s when Charles commissioned him to translate from Greek the writings which had been issued as (and were taken to be) by Dionysius, the learned pagan converted by St Paul, though they were in fact the work of a fifth-century Christian deeply influenced by the Neoplatonism of Proclus. The manuscript of pseudo-Dionysius had been sent as a present by the Byzantine emperor to Charles’s father, Louis the Pious. An obscure translation had been made at the time by Hilduin, Abbot of St Denis. Eriugena had taught himself Greek much better and succeeded, not merely in producing a comprehensible translation which would be used for the next three centuries, but also in absorbing the ideas he found in the text. He went on to translate various other Greek Christian texts, by Gregory of Nyssa and the seventh-century Maximus the Confessor. All these influences, along with his wide reading of the Latin fathers (especially Ambrose and Augustine) and his enthusiasm for logic (especially as found in the pseudo-Augustinian Ten Categories), are combined in his masterpiece Periphyseon (About Nature’; it is also sometimes known as De divisione naturae, ‘On the division of nature’), written in the 860s. The Periphyseon has been seen by some as continuing a tradition of Greek Neoplatonic thought, and by some as anticipating nineteenth-century German Idealist philosophy; whilst other scholars have concentrated on placing the work within the context of Carolingian thought.2 Yet other approaches, too, are possible (as Stephen Gersh’s discussion below will illustrate)—a diversity of interpretation encouraged by a text of remarkable breadth and audacity, where bold strokes of the imagination sometimes stand in for rigour of argument and suggestiveness of imagery for clarity of thought. The Periphyseon begins by setting out a fourfold division of universal nature—discussed below in greater depth by Stephen Gersh—into: (1) that which is not created and creates, (2) that which is created and creates, (3) that which is created and does not create, and (4) that which is not created and does not create. God, as creator, constitutes (1); the primordial causes—which are both like Platonic Ideas and the Stoic seminal reasons Eriugena learnt about in Augustine’s Literal Commentary on Genesis—make up (2); (3) is the created world of men, animals and things and (4), like (1), is identified with God, but God as the Final Cause to which all things return. The underlying course of universal history, seen as the progress from (1) to (4), is described in the five books of the work, which takes the form of a dialogue between master and pupil. Book I is mainly devoted to showing that God does not belong to any of Aristotle’s ten categories. Drawing on pseudo-Dionysius’ negative theology, Eriugena argues that God does not even belong to the first category, that of ousia (substance or essence) as Augustine had held. The remaining four books are structured round an exegesis of the story of creation and fall in Genesis, in which Eriugena discovers not only an account of divisions (2) and (3) but also that of the return of all things at the end of time to the uncreated and uncreating God of (4). Unusual positions abound: that (following Maximus) sexual differentiation arose only as a consequence of the fall; that the nothing from which God created all things is God himself who, being beyond all description, is nothing rather than something; that (continuing the line of thought from On Predestination, but hedging it around with qualifications and even contradictions) there will be no Hell, at least in the ordinary sense. Eriugena also composed a commentary on pseudo-Dionysius’ Celestial Hierarchy, a homily on the prologue to John’s Gospel and the beginning of a commentary on that Gospel. The homily provides a short and beautifully written summary of some of the main themes of his later work. Anselm was born at Aosta in Italy in 1033. He became a monk of Bec in Normandy in 1059, where he was taught by Lanfranc, whom he went on to succeed as Abbot (in 1078) and as Archbishop of Canterbury (in 1093). He died in 1109, after a stormy tenure of the archbishopric in which he tried to assert the power and independence of the Church. Anselm did not begin to write his theological and philosophical works until he was over 40. From then up almost until his death he produced a series of writings distinguished by an extraordinary elegance of thought and clarity of purpose. Unlike almost every other medieval thinker, Anselm makes no parade of philosophical or theological authorities, although he clearly knew as well as anyone of his time the logical texts of Aristotle, Porphyry and Boethius then available, and he had studied deeply Augustine’s more philosophical writings.3 Anselm’s two earliest monographs, the Monologion (1076) and the Proslogion (1077–8), are both concerned to provide rational arguments for the existence and attributes of God, although he assumes that his readers will be Christians who already accept by faith the truth of the assertions which he is setting out to prove. The Monologion uses a variety of arguments designed to show that there exists a triune God. The Proslogion uses a single line of argument and does not attempt to argue for triunity, but restricts itself to the not specifically Christian divine attributes such as omniscience, omnipotence, perfect goodness and eternity. The piece is built around the notion of that-than-whichnothing- greater-can-be-thought: what, for simplicity’s sake, may be called the notion of a ‘maximal being’. Most of the work is devoted to showing that, in the case of each presumed divine attribute, it must belong to a maximal being because, without it, the being would not be maximal. But this would merely show that, if it existed, a maximal being would be omnipotent, omniscient and so on. By far the greatest attention, in Anselm’s time and ever since, has been given to the argument placed at the beginning (often called Anselm’s ‘ontological proof’) to demonstrate that a maximal being does actually exist. Anselm advances two premisses: (1) that a maximal being does at least exist in thought, and (2) that to exist in reality and thought is greater than to exist in thought alone. He considers (1) to be proven by the fact that even someone who denies the existence of a maximal being (such as the fool of Psalm 14, who denies that God exists) has the mental concept of such a being; and he takes (2) for granted. He then argues that it must be false to claim that a maximal being A exists in thought and not in reality, because such a being would be less great than a being B exactly like it except that it existed in thought and also in reality, and so A would not be a maximal being. Therefore, given that a maximal being exists in thought, it must exist in reality too. The classic objection to this argument, that existence is not a predicate, is not very powerful, since Anselm’s argument is based on the contrast between ways of existing, in thought and in reality. His premiss (2) may not be convincing, but it is not obviously false or meaningless. Modern reworkings of the ontological proof usually adapt premiss (1) to read: ‘“God exists” is possibly true’, and, in order to make a plausible argument, they need to add another claim (3), that if a maximal being exists, it must exist in such a way that it cannot not exist: it must exist necessarily. (3) is found in the next chapter of the Proslogion, but as a further argument rather than as an additional premiss to the proof that a maximal being exists. It remains a matter for dispute among philosophers whether any version of the ontological argument, strengthened in this way, is sound.4 Besides writing a detailed reply to the criticisms raised by Gaunilo, a monk of Marmoutier, to his ontological proof, Anselm went on to write, among others, works On Truth, On Free Will and on the compatibility of grace and divine prescience with human freedom. His Cur Deus homo (Why God became man, 1094–8) is especially ambitious: basing himself on Scripture, but only on that part of it accepted by Jews and Muslims as well as Christians, Anselm tries to show that God needed to become incarnate if he was to remain just but also maintain the benevolent purpose of his creation. Two works of Anselm also survive which are more purely philosophical in content: De grammatico, an intricate logical discussion, following on from Aristotle’s Categories and Boethius’ commentary, of the semantics of denominative words such as grammaticus (‘literate’), and the ‘Philosophical fragments’, which examine modal notions and sketch out a philosophy of action.5 QUESTIONS OF METHOD Like any other object of critical analysis, the literary production of those writers of the ninth to eleventh centuries who are usually styled ‘philosophers’ is approachable from various perspectives. One such viewpoint, dominant in medieval philosophical scholarship until quite recently, has been that of orthodox Thomism. However, the notion that pre-thirteenth-century intellectual figures should primarily be valued for their tentative movements towards certain doctrines of high scholasticism is nowadays losing its appeal. There is obviously neither the socio-political pressure nor the metaphysical conviction to sustain it. John Marenbon’s survey, published in 1983 ([6.33]), makes the prescholastics speak, at least to a degree, in an idiom intelligible to a late twentieth-century audience. That he should emphasize their preoccupation with problems of language is therefore perfectly understandable. This is documented by their elaboration of the doctrine in Aristotle’s On Interpretation about words, thoughts, and things ([6.33] 21–2, 32–3, 101–2) and by their rediscovery of the distinction between sense and reference of terms ([6.33] 102–3, 106 ff.).6 One only needs to adopt a more comprehensive notion of the linguistic— including the structural element and the overlap with the semiotic—in order to see such preoccupation in greater relief. However, that he should limit their claim to be called ‘philosophers’ is perhaps too drastic. A careful reading of Early Medieval Philosophy reveals its author’s personal conviction about the nature of philosophy. For him, it is primarily methodological in a sense opposed to ontological realism and system building ([6.33] 6, 10, 15–16, 81). The methodology consists of arguments from premisses to conclusions (pp. 4, 58), the premisses being generally open to doubt but ideally self-evident either to observation or reason rather than textually given, the conclusions being unknown in advance (pp. 4, 12). Philosophy also employs terms which are literal rather than metaphorical and univocal rather than equivocal in its discussions (pp. 5–6, 9–10). Since these criteria define a discipline recognizable to Bertrand Russell but not to early medieval writers, Marenbon is left with relatively few illustrations of genuine philosophy before the twelfth century. Although the traditions of logic and of logic’s application to theology represented by certain passages in Augustine and Boethius are to be excepted (pp. 10, 47–8), a substantial portion of the late antique and early medieval literature fails to meet one or more criteria. The Latin translation of Plato’s Timaeus is too metaphorical (pp. 5–6), the Latin Platonic material of late antiquity too much concerned with system building and metaphorical expression (pp. 9–10, 15–16). Likewise, Eriugena’s thought involves too much system building (p. 81), too many premisses derived from texts (p. 58), and too much equivocal language (pp. 65–9), Anselm of Canterbury’s too many conclusions known in advance (pp. 95–7). Despite the persuasiveness of this discussion, a different approach to the philosophical writing of the ninth to eleventh centuries is possible. This would involve equal attention to the linguistic component but— since history shows this term to imply not universality but family resemblance—fewer prior assumptions about the meaning of ‘philosophy’. What follows is an attempt to investigate samples of Carolingian and post-Carolingian philosophical literature from such a viewpoint. I shall suggest that these materials, in their concern for systematic construction, pre-existing textual data, and the polysemy of etymology and metaphor, exhibit not intellectual weaknesses but intellectual strengths. ERIUGENA In some respects, Western medieval philosophy can be viewed as beginning with the brilliant and controversial ninth-century thinker John Scottus Eriugena.7 Marenbon values him for his ability to reason abstractly yet criticizes his tendency to system building. However, it is Eriugena’s notion of structure which perhaps makes him closer to modern writers than to other medieval ones. Few would deny that a particular concept of ‘structure’ is one of the intellectual paradigms of our era.8 This involves a priority of relation to related terms, such relations being either of opposite to opposite where one opposite exists through or is understood through the other, or else of whole to part where the whole exists through or is understood through the part, or vice versa.9 Originating in linguistics, where it determined both the phonological and semantic spheres—for example as the Saussurian concept of ‘value’,10 the theory regarding presence (+) or absence (-) of distinctive features elaborated by Trubetzkoy and Jakobson,11 and the Hjelmslevian notion of ‘form’12 —it has passed into the currency of historical, anthropological, literary, psychoanalytic, and other studies. Although avoiding the term ‘structure’ itself, Eriugena builds his metaphysical system with identical components. Priority of relation is underlined by his discussion of the Aristotelian categorical doctrine in Periphyseon I where the category of ‘relation’ (relatio, ad aliquid) or of ‘condition’ (habitus) is found to be present in all the other categories.13 Contrast of opposite with opposite is a recurrent theme of Eriugena’s writing, as instanced by the negative and affirmative predicates applied to God (I. 458A–462D, II. 599B–600A, III. 684D– 685A, etc.) and the five dichotomies constituting nature (II 529C–545B); contrast of whole with parts is only slightly less frequent, an instance being God’s status with regard to created things of which man’s is the microcosmic reflection (IV. 759A–B. Cf. II. 523D–524D). Strict relatedness is clearly the writer’s underlying assumption in such cases, since each binary term is said to be dependent ontologically and epistemologically on its counterpart (V. 953C–954A, V. 965A–B). Eriugena exploits the notion of structure in developing his own variant of the classical Platonic Theory of Forms. The expression of this doctrine, acquired through intermediary Greek and Latin patristic sources, combines ontological and semiotic criteria. From the ontological viewpoint,14 there exists a set of transcendent i.e. atemporal and non-spatial principles. These are termed ‘reasons’ (rationes) in Latin, and ‘Ideas’ (ideai), ‘prototypes’ (pr)totypa), ‘predestinations’ (proorismata), or ‘divine volitions’ (theia thel%mata) in Greek.15 They possess a metaphysically intermediate status since they depend upon a prior cause: God (the technical term for such dependence being ‘participation’ (participatio)), while subsequent terms, created objects, depend on them.16 According to Eriugenian textual exegesis, when the Bible describes God as making heaven and earth ‘in the beginning’, it means that the first principle establishes the reasons or Ideas of intellectual or sensible creatures within its Word.17 Examples of the transcendent principles are Goodness, Being, Life, Wisdom, Truth, Intellect, Reason, Power, Justice, Salvation, Magnitude, Omnipotence, Eternity, and Peace (II. 616C–617A). From the semiotic viewpoint,18 Eriugena proposes an analysis of the term ‘nature’ (natura) using a combination of traditional logical principles like the square of opposition19 and the division of genus into species versus the partition of whole into parts.20 Within nature, four ‘differences’ (differentiae) are posited: creating (A), not created (D), created (B), and not creating (C), these combining to form four ‘species’ (species): creating and not created (1), both created and creating (2), created and not creating (3), and neither creating nor created (4).21 The relations between 1 and 3 and between 2 and 4 are described as ‘opposition’ (oppositio), those between A2 and Al, between B3 and B2, between C3 and C4, and between D4 and D1 as ‘similarity’ (similitudo), and those between B2 and D1, between C3 and A2, between B3 and D4, and between C4 and Al as ‘dissimilarity’ (dissimilitudo) (I. 441A–442A, II. 523D–528B). This semiotic analysis is applied to metaphysics when species 1 is identified with God as the beginning of the cosmic process, species 2 with the reasons or Ideas, species 3 with the effects of the reasons or Ideas, and species 4 with God as end of the cosmic process.22 By endorsing the thesis that there is an analogy between the cosmos and a book, Eriugena can pass easily from assumptions about the structure of reality to assumptions about the structure of texts.23 That he has a systematic approach to texts is suggested by the possibility of dissolving Periphyseon into a mosaic of citations.24 Of course, he presents no formalized theory concerning the relations between a literary text, its reader, and antecedent texts comparable with those developed in connection with modern fiction by Bakhtin, Kristeva and others.25 Nevertheless, the combination of quotations in his writing indicates several interpretative strategies. Among Eriugena’s citations,26 a considerable number come from the Greek Fathers. Taking them in chronological order of authorship, there are lengthy passages from Origen on the end of the world (V. 929A–930D), from Gregory of Nyssa on man as the image of God (IV. 788A–801C), from pseudo-Dionysius the Areopagite on the divine names and on the celestial hierarchies,27 and from Maximus the Confessor on the fivefold division of nature (II 529C–542B). Two Latin Fathers contribute textual materials of importance: Augustine on miscellaneous questions28 and Ambrose on the interpretation of Paradise (IV. 815B–816C). Among Eriugena’s further quotations, a large group comes from Latin secular authors. Considering these also in chronological order of authorship, there are substantial extracts from pseudo-Augustine on the ten categories,29 from Martianus Capella on the measurement of the cosmos,30 and from Boethius on the nature of number.31 The incorporation of all these antecedent texts into Periphyseon reflects one paramount exegetical purpose. This is to make them agree in meaning so that, when two texts are perceived to disagree on the denotative level, agreement must be sought in some connotative meaning;32 and when they are seen to disagree on the connotative level, the denotative meaning of one text should be accepted, its selection being founded on a hierarchy of socio-political value.33 The application of this exegetical principle can be documented by many examples. Latin Christian and Greek Christian writings are held to agree when Augustine and pseudo-Dionysius discuss the divine ignorance beyond knowledge (II. 597C–ZZ598A), Latin secular and Greek Christian when ‘Plato’, Virgil and Gregory of Nyssa describe the four elements,34 Latin secular and Latin Christian when ‘Plato’ and Augustine interpret the world soul as principle of life (III. 727C–728D), and Latin secular, Greek Christian and Latin Christian when ‘Aristotle’, pseudo-Dionysius and Augustine discuss the ten categories.35 Disagreement on the denotative level overcome by shifting to the connotative level of one or both texts is instanced among Latin Christian and Greek Christian authors when Ambrose, Augustine, pseudo-Dionysius and Maximus describe the indirect perception of God through theophany;36 disagreement on the connotative level overcome by concentrating on the denotative level of one text only is illustrated among Latin Christian and Greek Christian authors when pseudo-Dionysius and Maximus but not Augustine discuss the threefold division of the soul into substance, power and activity,37 and among Latin Christian and Latin secular authors when Pliny and Martianus Capella but not Augustine calculate the measurements of the cosmos (III. 719A, 721C. Cf. III. 724A–C). Eriugena obviously exploits the notion of multiple meanings in texts. That this is in the late twentieth century part of the definition of literariness would hardly be questioned,38 and that it is nowadays also a fundamental problem in philosophy is the legacy of Gadamer, Derrida and others.39 But it is important to find the exact coordinates of Eriugena’s position. Of the theoretically possible views of meaning which are relevant here, one would connect the polysemy of individual texts with an ultimate monosemy—metaphysical truth—and establish a limit for hermeneutical activity and a distinction between denotation and connotation.40 This was the attitude of medieval theologians.41 Another view would connect the polysemy of individual texts with an ultimate polysemy—a linguistic ‘reality’—and establish no limits for hermeneutical activity and no distinction between denotation and connotation. Such is the position of modern deconstruction.42 A careful study of Eriugena’s philosophical methodology reveals him supporting neither the first nor the second viewpoint exclusively but oscillating between the two: a most unusual approach for a Western medieval thinker. The evidence for Eriugena’s concept of polysemy consists primarily of various statements about thought and language.43 Clearly the notion that polysemy is a property to be exploited rather than a defect to be overcome in the pursuit of philosophy requires a fusion rather than a separation of the cognitive and the verbal. Eriugena explicitly advocates such a fusion in several instances while commenting on Martianus Capella and Maximus the Confessor. Among Eriugena’s comments on the text of Martianus Capella, those dealing with the meaning of its initial allegory are particularly relevant. This narrative depicts the god Mercury’s search for a bride, culminating in his choice of the mortal Philology, and then the preparations for the marriage of Mercury and Philology, including a ritual of Philology’s deification. Since Eriugena quite plausibly interprets Mercury and Philology as figures of language and reason respectively, the marriage of the two protagonists for him indirectly signifies the fusion of discourse and thinking.44 Naturally, this represents a primary rather than exclusive meaning of such an inherently polysemous text.45 Among Eriugena’s developments of Maximus the Confessor’s teaching, those concerned with a threefold psychological process are particularly important. Here, Eriugena sometimes contrasts two inner cognitive functions: intelligence and thinking with an outer expressive function: sensation=sign-manipulation,46 but sometimes describes three inner cognitive and expressive functions: intelligence=noninterpretation, reason=expression, and interior sensation=quasi signmanipulation. 47 The shift between the first and second formulations— tantamount to replacing the traditional contrast of thought and language with a more unusual combination of the two—results from the contextual pressure of a Trinitarian analogy in the latter case.48 Just as God expresses himself to himself and to creation through his Word, so does man reflect the same processes on a lower level of being. In order to appreciate these developments, one should pause momentarily to recall Aristotle’s theory in On Interpretation that spoken words are signs—symbola or s%meia—of mental affections and that, although mental affections are identical for all mankind, spoken words are different.49 Thanks to Boethius’ translation and commentary on this text, the radical cleavage between thinking and language which it advocated became a medieval commonplace.50 However, modern linguistic theory in the tradition of Saussure’s Cours de linguistique générale would insist that the acoustic image—the signifier, and the concept—the signified, are inseparable components of one wholly arbitrary linguistic sign.51 Further evidence for his concept of polysemy is provided by the writer’s practice in connection with etymology. Here, Eriugena follows the doctrine, established by the Stoics and transmitted to the Latin West by Isidore of Seville, that study of the forms and derivations of words leads to knowledge of the things which they represent.52 The Periphyseon contains numerous examples of simple etymologies exploited in this way. Because metoch% (‘participation’) is composed of meta (‘after’) plus echein (‘to have’), it indicates the derivation of an essence from a superior one (III. 632B) and because stere)ma (‘firmament’) is composed of ster% (‘solid’) plus hama (‘together’), it indicates the common boundary of all corporeal things (III. 694B). Similarly the noun ousia (‘substance’) comes from the verb eimi (‘I am’) and therefore signifies subsistence of each thing in its transcendent causes whereas the noun phusis (‘nature’) comes from the verb phuomai (‘I am born’) and therefore signifies the generation of each thing in some material substratum.53 When Eriugena alternates etymologies of a single term, the fusion of real and verbal begins to predominate over the separation of the two.54 For example, the word theos (‘God’) is derived both from the verb the)r) (‘I see’), so that God is the one who sees all things in himself, and from the verb the) (‘I run’), so that he is that which itself runs through all things. (I. 452B–C). The word angelos (‘angel’) is connected both with the preposition engus (‘near’), meaning that angels are the creatures immediately after God, and with the verb engigno (‘I engender’), meaning that they are the creatures who transmit divine illuminations.55 When Eriugena connects etymologies of different terms, the fusion of real and verbal completes its ascendancy over the separation of the two. Because bonitas (‘goodness’) comes from the verb bo) (‘I call’), while bo) is synonymous with kal), from which comes the adjective kalos (‘beautiful’), the God who is both goodness and beauty can be understood as calling all created things from nonexistence into existence.56 The writer’s practice in connection with metaphor provides yet more evidence for his concept of polysemy.57 For Eriugena, ‘metaphor’ (metaphora/translatio) represents the application to something of a name normally applied to something else (see I. 458C, 461C, 463B, 464D, 512B–D, 522A, etc.). This is a notion derived from such textbooks as the pseudo-Ciceronian To Herennius, although Eriugena does not specify the ground of this transference of names in the perceived similarity between the objects concerned.58 ‘Metonymy’ (met)numia) is defined as a more specific version of the above, involving the application to the contained of a name normally applied to the container (I. 480B. Cf. To Herennius IV. 32. 43). ‘Synecdoche’ (sunekdoch%) is a more specific version of the above, involving the application to the part of a name normally applied to the whole, or else the application to the whole of a name normally applied to the part (II. 560A–B, III. 706B, IV. 744C. Cf. To Herennius IV. 33. 44). When Eriugena advocates such transferences of terms either between a created thing and God59 or between one created thing and another, he remains within the traditional theory. When he treats these transferences as simultaneously metaphors, metonymies and synecdoches (I, 480B, III. 706B) he is perhaps metaphysically rationalizing certain imprecisions in that established teaching. But when he understands such transferences of terms not as unilateral between a literal and a figurative sense but as bilateral between two literal-figurative senses,60 he passes beyond the traditional doctrine. In fact, the writer seems to have developed this notion of ‘reciprocal metaphor’ (reciproca metaphora) (III. 706A) against a twofold background. Within his theory of divine names, a given term e.g. ‘goodness’ can be applied to the creator but is normally applied to the creature while that same term can be applied to the creature but is ultimately grounded in the creator.61 In connection with his theory of the Incarnation, a certain term e.g. ‘air’ can be applied to a higher element but is normally applied to a lower one, while another term e.g. ‘light’ can be applied to a lower element but is normally applied to a higher one. This example is particularly interesting since air and light are already metaphors of human and divine respectively.62 Also in connection with his theory of the Incarnation, a certain term relating to salvation e.g. ‘flesh’ may be applied to the redeemed but is normally applied to the fallen, while another term e.g. ‘spirit’ may be applied to the fallen but is normally applied to the redeemed.63 The understanding of metaphor emerging from such texts moves away from that implying comparison of two spheres of meaning, and associated with the classical tradition from Aristotle to Quintilian and beyond, towards that based on fusion of two spheres of meaning and advocated by Richards and other modern critics.64 Such a viewpoint has one important consequence which Eriugena intuitively grasps: that the traditional distinction between the verbal and the real is becoming questionable. This is because the metaphorized and metaphorizing terms are no longer contrasted as verbal and real but as equally verbal-real. The same viewpoint has another consequence which he explicitly states: that the habitual distinction between ‘figurative’ and ‘literal’ language is almost unworkable (see III. 705Aff.). It is because of this deliberate rather than accidental role of polysemy in his thought that we should be less ready than some have been to accuse Eriugena of philosophical confusion. For example, Marenbon finds serious fault in the handling of substance ([6.33] 65–70). He rightly notes that Eriugena’s substance is primarily universal but, since he has confused two distinct types of universal: (a) classes of things where whatever distinguishes their members is present wholly in each one, and (b) universal qualities where whatever is characteristic of individuals is present to different degrees in each, concludes that this substance is a notion vitiated by ambivalence. However, it is also reasonable to see deliberate polysemy rather than unconscious confusion here.65 Eriugena’s ‘substance’ is simply a lexeme whose semantic properties enter into numerous configurations, forming a simple structure where it is opposed to non-substance (I. 461 A–464A)—the affirmative and negative theologies. It forms a more complex structure where the opposition of universal and particular is discovered within it and it is opposed to accident (I. 467D–468B, 470Dff.)—the Aristotelian categories. It forms the most complex structure where it is combined with form, opposed to quality in combination with form, metaphorically fused with ‘dry land’, and opposed to quality metaphorically fused with ‘water’ (III. 698Cff.), the exegesis of Genesis 1:9–10. Any structure may actualize semantic properties logically inconsistent with those of other structures. That inconsistencies are an ineradicable feature of natural languages and of all literature and philosophy derived from them is a fact which Eriugena perhaps saw more clearly than did most of his contemporaries and successors. It had always been assumed by nineteenth-century historians of philosophy that Eriugena exercised little influence over later thinkers. Although various attempts have been made to counter this negative assessment in recent times,66 the only hitherto undiscovered influences to be brought to light have been those on the immediately subsequent generation. Thus, Eriugena’s studies of the Latin Fathers are known to have influenced one set of Carolingian glosses on Augustine’s De Musica (edited by Boeuff [6.45]) and his studies of Latin secular authors of various glosses of the same period on the pseudo-Augustinian Ten Categories.67 These latter glosses have been extensively discussed in recent scholarship. From passages now published it is possible to see that various commentators had grasped the semiotic ramifications of Eriugena’s work. Indeed, certain glosses recall the structural preoccupations of his thought in elaborating the notion of ‘nature’,68 and others its polysemic tendency by applying ideas concerning homonymy, synonymy, and paronymy or etymological arguments to metaphysics.69 If Eriugena had exercised influence over later thinkers, it would undoubtedly have run counter to the norm of medieval intellectual development. In general, writers of this period went back directly to antique sources for their material, and during the tenth and eleventh centuries this meant primarily Boethius, whom Eriugena had only partially exploited.70 For example, Notker Labeo makes extensive use of Boethius’ translation of Aristode’s Categories (see [6.63]), Abbo of Fleury of the Boethian monographs on logical division and on various kinds of syllogism (see [6.64]), Gerbert of Aurillac of Boethius’ first commentary on Porphyry’s Isagoge, etc. Gerbert is arguably the most important member of this group.71 His treatise De rationali et ratione uti (On ‘rational’ and ‘to use reason’) is a discussion of logical problems surrounding the extension of the two predicates ‘rational’ and ‘using reason’ ([6.16] 1. 299) more interesting for the ideas arising en route to the solution than for the solution itself. Here, Gerbert reveals the structural preoccupation of a typical Platonist in establishing three ‘semiotic’ categories:72 act without potency, act with potency, and potency without act, which are applied to hierarchies of physical and metaphysical principles,73 yet a desire to reduce polysemy more characteristic of the re-emergent Aristotelianism ([6.16] 9. 304). ANSELM OF CANTERBURY The next major figure in the Western intellectual tradition and the dominant thinker of the late eleventh century is Anselm of Canterbury.74 Marenbon arrives at an ambivalent judgement in his case, on one hand denying him the title of ‘philosopher’ because his argumentation does not arrive finally at its conclusions but assumes them from the outset, and on the other conceding it in recognition of his contributions to the study of the language—thought relation and of the logic of possibility and necessity. Yet it is possible to reevaluate Anselm’s philosophical contribution under the three headings proposed earlier: structure, text and polysemy (see p. 125). Anselm exploits the notion of structure in developing a variant of the classical Platonic Theory of Forms during the early chapters of Monologion which combines ontological and ‘semiotic’ criteria. The ontological viewpoint is clearly indicated when he describes a set of transcendent i.e. atemporal and non-spatial principles, each of which is termed an ‘exemplar’ (exemplum), ‘form’ (forma), or ‘rule’ (regula) (Monologion 9, 24. 7–20). It is either present in the divine mind or an aspect of the divine essence,75 and is somehow the cause of lower i.e. spatio-temporal things.76 The semiotic viewpoint is adopted implicitly when Anselm introduces the set of transcendent principles with a discourse based on semantic permutation.77 In the first place, there is an argument in the abstract. This is founded on the following inventory of semantic elements: two terms—the plurality of things having property x (a1,a2…) and the single property x (b); two relations constitutive of terms—effect of another (R→) and effect of itself (R←); two terms constituted by relations—the plurality of things having property x through another (aa1, aa2…) and the single property x through itself (bb); and three relations—greater than (R), less than (R<), and equal to (R=). The inventory is activated gradually as the argument proceeds through six stages: 1 There are things having property x [a1, a2…]; 2 A thing having property x to greater, lesser, or equal degree than another thing having property x has this through the property x 3 The property x is itself x [b R←]; 4 Things having property x are things having property x through another [a1, a2…R=aa1, aa2…]; 5 The property x is the property x through itself [b R=bb]; 6 The property x through itself is greater than things having property x through another [bb R> aa1, aa2…]. In the second place, the argument is applied to three concrete instances: where property x is identified with ‘good’ sensed or understood, ‘great’ sensed or understood, and ‘existent’ sensed or understood respectively.78 Important features of Anselm’s philosophical method are revealed here. For example, it seems that there is less an alternation of premisses and conclusions—as in formal logic—than a permutation of semantic properties. In fact, the whole discourse can be understood in semantic terms with the exception of the idea (point 3 above) that the property x is itself x. This is purely ontological in character, since it makes no sense to say that the semantic property x has the semantic property x.79 Furthermore, it appears that the permutation of semantic properties follows a largely symmetrical pattern, the clearest indication of a writer’s thinking in structural terms. It would be inappropriate to seek the relation to textual authorities here which was apparent in Eriugena. The difference between the two philosophers seems extreme, given that Anselm’s works—especially Monologion and Proslogion—are attempts to construct a discourse ‘by reason alone’ (sola ratione) without explicit dependence on sources.80 Nevertheless, Anselm’s relation to textual authorities is different from that of his predecessor rather than non-existent. Although numerous Latin patristic sources are mentioned in the extant letters, the only authority cited in the treatises themselves is Augustine. But this citation is of overwhelming interpretative significance. In the preface to Monologion, the writer diverts potential criticism that he is advocating novel or false teachings by stressing the complete agreement between the doctrines of his book and those of Augustine’s On the Trinity ([6.11] I:8. 8–14). Some modern scholars would interpret this as the typical statement of a medieval writer endeavouring to conceal the novelty of his thought behind a declaration of traditionalism. However, Anselm’s remarks are more than a rhetorical commonplace. This becomes clear on analysing the Monologion into an assemblage of Augustinian materials reorganized according to the structural principles described above. Anselm’s relation to textual authorities is even indicated by the Proslogion, which cites no source at all. This treatise contains a famous passage where a premiss that God is ‘something than which nothing greater can be thought’ (aliquid quo nihil maius cogitari possit) is postulated as self-evident, the premiss then being used as the starting point for an argument allegedly proceeding by the application of reason alone to the conclusion that God exists ([6.11] I:101. 1–4, 104. 7). But even if one were to concede the premiss to be self-evident—a dubious point in itself—one could not consider it independent of textual background. In fact, the premiss corresponds to a definition of God found in Christian texts like Augustine’s On the Customs of the Catholic Church and those of the Manicheans ([6.11] I:11. 24) and Boethius’ On the Consolation of Philosophy ([6.11] I:10, 57–8), and in secular works like Cicero’s On the Nature of the Gods ([6.11] I:77) and Seneca’s Natural Questions81 to name only the most obvious parallels. So Anselm’s purpose was perhaps to recommend the faith to non-Christians by deducing it from a premiss stated by Christian and non-Christian authors alike. Anselm obviously does not exploit the notion of multiple meanings in texts; indeed, the ideal of univocity would seem more consistent with his method. Nevertheless, some of his ideas about signification, had they been extended in a different direction, would have supported the exploitation of polysemy. One suggestive idea is the distinction between appellation and signification elaborated in the treatise De grammatico. Here, he argues that in statements like ‘the horse is white’, the adjective is ‘appellative’ (appellativum) of the white thing but ‘significative’ (significativum) of its possession of the property ([6.11] I:159. 12–15, 161. 21). Since he stresses that what is appellated is an existent object but what is signified is not, the distinction seems to approximate that between reference and sense in modern linguistic theory.82 However, any Platonist would maintain that in the statement ‘the horse is x’, the x signifies a transcendentally existent x-ness in which Socrates participates. This is the viewpoint which also seems to underlie the argument about divine attributes in Monologion 1–4.83 Another aspect of Anselm’s theory of signification conducive to the systematic exploitation of polysemy is his notion of a ‘speaking’ (locutio) within the divine nature. By explaining that the exemplar in the divine mind according to which all things are created is a speaking (see p. 133), he follows traditional patristic teachings regarding the Word as second person of the Trinity.84 However, the use of the term ‘speaking’ also requires a rational justification. Anselm therefore proposes to distinguish three ways of speaking about an object:85 1 Speaking of things by employing sensible signs in a sensible manner e.g. signifying a man by using the word ‘man’—such signs being unmotivated and non-universal;86 2 Thinking by employing sensible and external signs in an insensible and internal manner e.g. silently thinking the word ‘man’—these signs also being unmotivated and non-universal; 3 Speaking things themselves by employing sensible signs in neither a sensible nor an insensible manner e.g. perceiving a man either by imagining his sensible shape or by thinking his universal essence ‘animal, rational, mortal’—such signs being motivated and universal.87 It is the third type of speaking which can be attributed to the divine mind.88 The exemplar in the latter, according to which all things are created, can therefore be described as a thinking process coextensive with rather than anterior to the manipulation of signs.89 With this argument, Anselm points towards that elimination of the distinction between cognitive and verbal characteristic of post-Saussurian linguistic theory albeit from a restricted theological perspective (cf. pp. 128–9). Another suggestive idea is the application of metaphor to philosophical method underlying the entire Monologion. Towards the end of that text Anselm raises an important question: given that the divine nature surpasses human understanding and is accessible only through words whose meaning is transformed, how true are all the inferences constructed from such words in respect of the divinity?90 He answers that there is a certain truth in things signified ‘not properly but through some likeness’ (non proprie…sed per aliquam similitudinem). The passage should be noted by those modern scholars who agonize over the cogency of Anselm’s arguments about God, since he shows clearly that the ‘logic’ which they contain is intended to be not the embodiment but only the reflection of truth.91 Apparently, logical metaphor is to logic in the Monologion what arithmetical metaphor was to arithmetic in Eriugena’s exposition of the divine names. NOTES 1 See Leonardi, ‘Glosse eriugeniane a Marziano Capella in un codice leidense’, in Roques [6.57]. 2 For Eriugena and Greek Neoplatonism, see esp. Beierwaltes [6.44] and Gersh [6.49]. Dermot Moran [6.54] explores the connections with German Idealism; cf. also W.Beierwaltes, ‘Zur Wirkungsgeschichte Eriugenas im deutschen Idealismus. Ein kurze, unsystematische Nachlese’, in [6.44] 313–20. Accounts more directed to the historical context will be found in Jeauneau [6.51], Marenbon [5–75] and Schrimpf [6.59]. 3 On Anselm’s knowledge of logic, see Henry [6.69] and his editions of De grammatico [6.14 and 6.15]. 4 See Bibliography [6.75–6.82] for some modern treatments of the ontological proof. 5 On De grammatico, see the works by Henry listed in n. 3 above; on Anselm’s theory of modality and philosophy of action, see Serene [6.73]. 6 In addition, Marenbon stresses the relation between logic and language in general explored by Fredegisus (p. 51), Gottschalk (pp. 55, 105), ninth-century writers at St Gall (p. 105), the anonymous eleventh-century glossator of Priscian (p. 106ff), etc. 7 The most useful books providing a general introduction to Eriugena’s life and works are Cappuyns [6.24] and Moran [6.54]. See O’Meara and Bieler [6.55], Allard [6.38], Beierwaltes [6.42] and [6.43], Jeauneau [6.51], for essays on specific aspects of his thought. 8 This is true not only of the original ‘structuralists’ but also of the semioticians and even the deconstructionists who have followed them. 9 On these criteria see Lévi-Strauss, C., Structural Anthropology, English trans., New York, 1964, pp. 279–80 and Greimas, A.J., Structural Semantics, English trans., Lincoln, Neb. 1983, pp. 18ff. 10 See Saussure, F. de Course in General Linguistics, English trans., New York, 1959, pp. 114–15. 11 This theory is conveniently summarized by Barthes, R., Elements of Semiology, English trans., London, 1984, pp. 135ff. 12 See Hjelmslev, L., Prolegomena to a Theory of Language, English trans., Madison, Wis., 1961, p. 23. 13 Eriugena, Periphyseon I. 466A–467C. References to Eriugena’s work give the column numbers of Floss’s edition [6.1] which are reproduced in the modern editions and translations and so provide a standard form of reference. Because of his interpretation of pseudo-Augustine: The Ten Categories, Eriugena allows the separate Aristotelian categories of relation and condition to coalesce. On Eriugena’s theory see Flasch [6.48]. 14 In discussing both Eriugena’s and Anselm’s notions of structure, I shall distinguish ‘ontological’ and ‘semiotic’ components. By the former is meant any aspects of the metaphysical system stated in the texts, by the latter those aspects corresponding to elements in the notion of structure described earlier. Of course, neither Eriugena nor Anselm could have made such a distinction. 15 II. 529A–C. Elsewhere, Eriugena calls these ‘primordial causes’ (causae primordiales). See III. 622Bff. 16 II. 616B. ‘And they are said to be the principles of all things since all things whatsoever that are sensed or understood either in the visible or invisible creation subsist by participation in them, while they themselves are participations in the one cause of all things: that is, the most high and holy Trinity’. Cf. III. 630A–C, III. 644A–B, III. 646B–C, III. 682B–C. 17 II. 546A–B. ‘But on considering the interpretations of many exegetes, nothing strikes me as more probable or likely than that in the aforesaid words of Holy Scripture—that is, within the meaning of “heaven” and “earth”—we should understand the primordial causes of the entire creature which the Father had created before the foundation of all other things in his only begotten Son who is designated by the term “beginning”, and that by the word “heaven” we should hold the primal causes of intelligible things and celestial essences to have been signified, but by the word “earth” those of the sensible things in which the entire corporeal world is completed’. 18 See note 14. That Eriugena was aware of the linguistic even if not semiotic starting point of his analysis is suggested by his reference to nature as a ‘generic term’ (general nomen) rather than as a generic entity. See Cristiani, M., ‘Nature-essence et nature-langage. Notes sur l’emploi du terme “natura” dans le “Periphyseon” de Jean Erigène’, Miscellanea Mediaevalia 13/2: Sprache und Erkenntnis im Mittelalter, Berlin and New York, 1981, pp. 707–17. 19 The square of opposition was a classificatory schema applied by Greek writers of late antiquity to (a) substance and accident and (b) the numbers 1–10. Thus, in (a) four terms: of a subject (A), not in a subject (D), in a subject (B), not of a subject (C) are grouped into four combined terms: of a subject but not in a subject (1), both in a subject and of a subject (2), in a subject but not of a subject (3), neither of a subject nor in a subject (4) where 1=universal substance, 2=universal accident, 3=particular accident, 4=particular substance. See Porphyry, On the Categories 78, 25ff. In (b) four terms: generating (A), not generated (D), generated (B), not generating (C) are grouped into four combined terms: generating but not generated (1), both generated and generating (2), generated but not generating (3), neither generating nor generated (4) where 1 =the numbers one, two, three, and five, 2=the number four, 3=the numbers six, eight, and nine, 4=the number seven. See Theo of Smyrna, Exposition of Mathematical Matters 103. 1–16. Such schemata were repeated in Latin texts and thereby transmitted to Eriugena and others: see Marius Victorinus, To Candidus 8. 1–21, Macrobius’ commentary on Cicero’s Dream of Scipio I. 5. 16, Martianus Capella, On the Marriage of Mercury and Philology VII. 738, Boethius, Commentary on Aristotle’s Categories I. 169Bff. The square of opposition in antiquity has been discussed by Hadot [6.31] 148ff., Libera, A. de, ‘La sémiotique d’Aristote’, in Structures élémentaires de la signification, ed. F.Nef, Brussels, 1976, pp. 28–55. The square of opposition in Eriugena has been examined most recently by Onofrio [6.56] and Beierwaltes [6.43] 17–38. An analogous schema applied to propositions was also traditional and certainly known to Eriugena; see Martianus Capella, On the Marriage of Mercury and Philology IV. 400–1. 20 See Martianus Capella, On the Marriage of Mercury and Philology IV. 352–4. 21 I. 441A–442B. Eriugena himself seems to envisage a diagram in the form: The notation A, B…1, 2…is not provided by Eriugena. 22 I. 442A–B, II. 525A, II. 526C–527A, II. 527C. The fourfold schema is repeated later in Periphyseon but with no additions to the basic doctrine. Cf. III. 688C– 689A, IV. 743B–C, V. 1019A–B. 23 See Eriugena, Homily on the Prologue to John [6.9] 14, 291B–C. The analogy between the cosmos and a book was derived from Maximus the Confessor, Ambigua 1245A–1248A. See Duclow [6.47] 131–40. 24 Eriugena is here elevating a standard Carolingian literary practice—illustrated by Alcuin, Hrabanus Maurus, Ratramnus of Corbie, etc.—to a more philosophical level. 25 For example, see Kristeva, J., S%mei)tik%. Recherches pour une sémanalyse, Paris, 1969, pp. 143ff., 181–2, etc. 26 A complete inventory can be found in Madec [6.53]. 27 I. 509B–510B, II. 617A–620A. Cf. Eriugena, Commentary on pseudo-Dionysius’ Celestial Hierarchy, passim. 28 For the Augustinian citations see Madec [6.53]. These are peculiar in being (a) extremely frequent, (b) generally brief, and (c) somewhat oblique. 29 I. 463Aff. This text is paraphrased rather than quoted. Eriugena associates the material with ‘Aristotle’, and tends not to quote secular authors verbatim. 30 III. 716B–719A. Paraphrase only. 31 III. 654A–655C. Paraphrase only. For Eriugena, naturally, Boethius ranks among the Christian authors. However, his On Arithmetic—the only text cited in Periphyseon—is thoroughly secular in character. 32 On connotation and denotation in Eriugena see below. 33 The hierarchy is as follows: Greek Christian writers are preferred to Latin Christian writers, and Christian writers to pagan writers. 34 I. 476C–477B. ‘Plato’ may be considered a Latin author, since Eriugena knew only Calcidius’ Latin translation of the Timaeus. 35 I. 458Aff. ‘Aristotle’ may be treated as a Latin author, since Eriugena relied entirely on Aristotelian testimonia in pseudo-Augustine and others. 36 I. 446A–451C. A ‘theophany’ is an appearance of God. Eriugena held that God is never cognized directly, but only in theophanies. 37 I. 486B–D, II. 567Aff. Cf. II. 602D–603C, 610B–611A. It is highly significant that the references to the Greek Fathers are made by the ‘Teacher’ and those to the Latin Fathers by the ‘Student’ in the Periphyseon’s dialogue. 38 For example, see Barthes, R., SZ, English trans., New York, 1974, pp. 1–16. 39 For example, see Derrida, J. Margins of Philosophy, English trans., Brighton, 1982, pp. 209ff. 40 I shall follow the predominant usage of modern semantic theory where the ‘denotation’ of a term is a primary meaning, the ‘connotation’ a secondary one. In realist semantics, where denotation can be associated with a term’s ‘reference’ to an object and connotation with its ‘sense’—using Frege’s nomenclature—the distinction between denotation and connotation is easy to maintain; but in strict nominalism where denotation cannot be associated with a term’s ‘reference’ to an object, the distinction between denotation and connotation becomes problematic. 41 Given that early medieval theologians assume (a) that a spiritual meaning resides behind the literal meaning of biblical texts and (b) that the spiritual meaning is the ultimate truth underlying the derivative truth of the literal meaning, they share one important assumption with the realist semantic theory discussed above: that there is an ontologically grounded primary meaning. On the relation between medieval exegesis and polysemy see Eco, U., Semiotics and the Philosophy of Language, London, 1984, pp. 147–53. 42 See Eco, Semiotics, pp. 153ff. 43 Eriugena’s contribution to the understanding of this question—and therefore to medieval semantic theory in general—has not been studied to date. However, there are some useful comments in Beierwaltes [6.41], 44 Eriugena, Commentary on Martianus Capella [6.2] pr. 3, 16–22. ‘Wishing to write about the seven liberal arts, he invented a certain story about the marriage of Philology and Mercury. And this was not without the display of a most subtle intelligence, for Philology represents the love of reason and Mercury the eloquence of speech. If these have come together as though by a certain marriage in the souls of those pursuing the study of wisdom, it is possible to arrive without any difficulty at knowledge and possession of the liberal arts.’ 45 The impact of the polysemous tendency initiated by Martianus Capella on medieval writers has gone largely unnoticed. Thus Kristeva, S%mei)tik%, pp. 168–9, contrasts a ‘Menippean’ polysemy with the theocentric monosemy of the medievals. Yet On the Marriage of Mercury and Philology is one example of ancient Menippean satire which became standard reading in medieval schools. 46 I. 454B ‘For our intellect, too, before it enters into thought and memory is not unreasonably said not to be. It is invisible in itself and known to nobody besides God and ourselves. But when it has entered into thoughts and acquires form in certain phantasies, it is not undeservedly said to come into being. For it comes to be in the memory when it acquires certain forms of things, sounds, colours, and other sensibles, having had no form before it entered into memory. Then it receives a kind of second formation when it is formed in certain signs of forms or sounds— I mean letters which are signs of sounds and figures which are signs of mathematical forms—or in other sensible indicators by which it can be introduced into the senses of those who are sentient.’ 47 II. 572C–573B ‘There are three universal motions of soul of which the first is according to mind, the second according to reason, and the third according to sense. The first is simple, above the nature of the soul itself, and devoid of interpretation: that is, knowledge of that around which it moves. “Through it, the soul moves around the unknown God but, because of his excellence, in no way has knowledge of him derived from anything which exists” as to what he is—that is, it cannot find him in any essence or substance or in anything which can be said or understood, for he surpasses everything which is or is not and cannot be defined in any manner as to what he is. The second motion is that by which the soul “defines the unknown God as being the cause” of all things. For it defines God to be cause of all things, this motion being within the nature of soul. It is that “through which the soul moved naturally imposes on itself through the activity of knowledge all the natural reasons formative of all things which subsist as having been eternally made in him who is known only causally”—for he is known because he is cause: that is, it expresses them in itself through its knowledge of them, this knowledge itself being born in the second motion from the first. The third motion is “the composite one through which the soul comes into contact with eternal things and reforms the reasons of the visible in itself as though through certain signs.” It is described as composite not because it is not simple in itself as the first and second motions are simple but because it begins to know the reasons of sensible things not through themselves.’ In this passage, a good example of Eriugena’s intertextual method, the words of Maximus appear between quotation marks. 48 The Trinitarian analogy will be more explicit in Anselm of Canterbury’s development of the same theme. See p. 135. 49 Aristotle, On Interpretation 1, 16a1ff. See Kretzmann, N. ‘Aristotle on spoken sound significant by convention’, in J.Corcoran (ed.) Ancient Logic and its Modern Interpretations, Dordrecht, 1974, pp. 3–21; Lieb, H. ‘Das “semiotische Dreieck” bei Ogden und Richards. Eine Neuformulierung des Zeichenmodells von Aristoteles’, in H.Geckeler (ed.) Logos Semantikos, Berlin, 1981, pp. 137–56; and Weidemann, H. ‘Ansätze zu einer semantischen Theorie bei Aristoteles’, Zeitschrift for Semiotik 4 (1982): 241–57. 50 This influence is documented in standard works on the history of medieval semantics. See especially, Kretzmann, N., ‘Semantics, History of, in P.Edwards (ed.) The Encyclopedia of Philosophy, vol. 7, New York, 1967, pp. 362–3, 365ff.; Pinborg [6.36] 29ff. and Eco [6.29]. 51 See Saussure, Course in General Linguistics, pp. 65–70, 111ff. The same fusion occurs in the semiotic theory of Peirce. See Peirce, C.S., Collected Papers, vol. 5, ed. C.Hartshorne and P.Weiss, Cambridge, Mass., 1931–58, p. 314, etc. 52 See Klinck [6.32] for the medieval tradition in general. 53 V. 867A–B. These etymologies are all based on the Greek. However, Eriugena also explores Graeco-Latin etymologies at III. 697A (ouranos/caelum), V. 954D– 955A (aid%s/infernus). An etymology based on the Latin occurs at I. 494D– 495A. 54 This situation is naturally conducive to polysemy. Fusion of real and verbal parallels and complements the fusion of cognitive and verbal described on p. 128. 55 III. 668C–D. Cf. Commentary on ‘Celestial Hierarchy’ 4. 314–25. 56 II 580C–581A. The passage is particularly interesting when combined with III. 624A–625A. Since this states that the order of the divine names is—according to Eriugena’s philosophical idealism—partially dependent on the human mind’s perception, the etymological activity of II. 580Cff. must be not only the discovery but also the positing of ‘reality’ itself. There is another complex etymology at V. 1003B–D. 57 The precise nature of metaphor is a matter of controversy. However, it clearly represents a specific application of the concept of polysemy where the primary meaning of a metaphorized term is the secondary meaning of the metaphorizing term and vice versa. 58 See To Herennius IV. 34. 45, Quintilian, Institutes of Oratory VIII. 6. 4ff., Martianus Capella, On the Marriage IV. 359–60, Isidore of Seville, Etymologies I. 37. 5, etc. 59 See I. 458C, I. 461C, I. 463B, etc. This application of metaphor is discussed by Beierwaltes, W., ‘Negati affirmatio. Welt als Metapher. Zur Grundlegung einer mittelalterlichen Ästhetik durch Johannes Scotus Eriugena’, Philosophisches Jahrbuch 83 (1976):237–65. 60 Traditionally, synecdoche occurs in two forms—transference from whole to part and transference from part to whole—and is therefore already bilateral. See To Herennius IV. 33. 44. 61 I. 459C. ‘But since the divine significations which are predicated of God by transference from the creature to the creator in Holy Scripture—if indeed it is rightly said that anything can be predicated of him (which we should consider elsewhere)—are innumerable and cannot be discovered or collected together in the smallness of our reasoning, only a few such divine names should be set down by way of illustration’; I. 461C. ‘For the statement “It is Truth” does not affirm that the divine nature is Truth in a proper sense but that it can be called by such a name in a metaphor from the creature to the creator. It clothes the divine essence which is naked and devoid of all proper signification with such words’; I. 463B–C. ‘But as we have said above, just as almost all things which are properly predicated of the nature of created things can be said metaphorically of the creator of things in order to signify, so also the significations of the categories which are discerned properly in created things can be uttered not absurdly concerning the cause of all—not to signify properly what it is but to suggest in a transferred mode what we should reasonably think about it when investigating it in some fashion’; I. 480B. ‘So if all things which are are rightly predicated of God not properly but by a kind of transference since they derive from him, why is it surprising that all things which are in place—since they seem to be enclosed everywhere by greater things—can be called places although none of them is properly a place but is contained within what is place in its proper nature?’ Cf. III. 624A–625A. 62 I. 480B–C. ‘We see that those things which are contained are named after the things which contain them through metonymy—that is, transferred naming— although they are not so contained by them that they are unable to subsist in their natural limits without them. It is the common practice of mortals to call the wife or the family a “house” although these things are naturally distinct. For it is not the house which confers substantial existence on the wife or the family but the place of their own nature. Yet because they possess that existence in the house they are accustomed to be named after it. Likewise the things which contain are named after the things which are contained. For example: air contains light, and so air which is illuminated is called “light”; the eye is called “sight” or “vision” although according to its proper nature it is neither sight nor vision’. Cf. I. 450A– B, I. 515B–C, V. 876A–B, V. 1021B. 63 III. 706A–B. ‘Not unreasonably, given that it is the most common practice of Holy Scripture to signify the natural subsistences and reasons of invisible things with words signifying visible things, in order to train pious philosophers. And this is not surprising, since the same practice has the very frequent custom of suggesting corporeal and sensible things with the names of spiritual and invisible things. Since there are many and innumerable examples of this reciprocal metaphoricity and they are very well known to all those trained in Holy Scripture, it would appear to be a lengthy and superfluous task to amass them in the present discussion. However, let us use a few illustrations: “That which is born of flesh is flesh”—here the entire man born in original sin is called by the name “flesh”— “And that which is born of the spirit is spirit”—the entire man reborn through regeneration in Christ is expressed by the term “spirit”. And if somebody says that it is not the entire man but only the flesh of a man that is born of flesh, I shall reply that it is therefore not the entire man but only the soul that is born of spirit and if so it follows that there is no grace to benefit the baptized bodies. But if the entire man, namely soul and body, is reborn in Christ and becomes spirit, then necessarily the entire man is born of flesh in Adam and is flesh, from which it is concluded that the flesh is called spirit and the spirit flesh. The Word of God is called flesh and flesh the Word, and there are similar cases where both synecdoche and metaphor are understood simultaneously.’ 64 See Richards, I.A., The Philosophy of Rhetoric, New York, 1936, pp. 89ff. The distinction between ‘comparison’ and ‘fusion’ theories of metaphor owes something to Black, M., Models and Metaphors, Ithaca, NY, 1962, pp. 25ff., who sets out a complete typology consisting of substitutive, comparative, and inter-active approaches. 65 The problems associated with polysemy were formally discussed in pseudo- Augustine, Ten Categories 9. 135, 13ff., a text with which Eriugena was particularly familiar. Cf. Martianus Capella, On the Marriage IV. 355–7. 66 The volume of essays, Beierwaltes [6.42], setting out to prove that Eriugena exercised significant influence over later medieval thinkers, has not achieved the desired result. In fact, the following conclusions now seem to have been established: (1) Eriugena’s influence was considerable for one or two generations after his time (the evidence: Heiric of Auxerre, Remigius of Auxerre, and other glossators); (2) In the eleventh century there are only a few traces of his influence, e.g. in Hrotsvitha of Gandersheim; (3) Eriugena’s influence becomes more noticeable from the beginning of the twelfth century but only in certain respects: (a) His ideas influence many in a negative sense (the evidence: copies of Eriugenian MSS, polemic against him), (b) He is influential as translator of pseudo-Dionysius, (c) His ideas influence a few in a positive sense (the evidence: Honorius Augustodunensis, ‘Marius’, On the Elements). See also Lucentini [6.52]. 67 Edited by Marenbon [5.84] 173ff. Eriugena also influenced the gloss tradition on Martianus Capella in a manner now difficult to describe precisely; see Schrimpf [6.58]. 68 Gloss I in Marenbon’s edition. 69 Gloss IIIb in Marenbon’s edition. 70 On the Boethian logical tradition in the Middle Ages see van de Vyver [6.37] and Minio-Paluello [6.34]. 71 On Gerbert’s work in general see the collection Gerberto, Scienza, Storia e Mito [6.61]. This includes papers by Riché [6.62]—stressing the important of Boethius— and Frova [6.60]. 72 On the term ‘semiotic’ see note 14. 73 Gerbert, [6.16] 6, pp. 301ff. Gerbert here systematizes material in Aristotle, On Interpretation 13. 23a 21–5. 74 The most useful book providing a general introduction to Anselm’s life and works is Hopkins [6.22]. Among other modern studies, Kohlenberger [6.70] and Evans, G.R., Anselm and Talking about God, Oxford, 1978 should be mentioned. 75 The first interpretation predominates at 9, 24. 7 to 10, 25. 27, the second at 1, 13. 1 to 4, 18. 3. Both are perfectly standard in the Augustinian tradition which Anselm represents. 76 The type of causality (efficient) is discussed at 6, 18. 18 to 7. 22, 10. 77 The semiotic always implies the semantic even though the reverse is not the case. 78 [6.11] I: 14. 5ff. ‘It is therefore easy for someone to say to himself silently: Since the goods are so numerous whose great diversity we both perceive through the bodily senses and discern by the reason of the mind, should we believe that there is one thing through which alone whatever things are good are good or are things which are good good through one another? But it is absolutely certain and clear to all those willing to pay attention that whatever things are called something in such a way as to be called this in greater or lesser or equal degree in respect of one another, are called this through something which is understood not differently in different things but the same in each case, whether it be considered as equally or unequally present in them… Therefore, since it is certain that all good things, if compared to one another, are either equally or unequally good, it is necessary that all good things are good through something which is understood as the same in different things, although sometimes good things seem to be called good through one another… But who would doubt that that through which all good things are good is a great good? So it is good through itself, since every good is good through it. Therefore it follows that all other goods are good through something other than that which they are themselves, and that only this other is good through itself. But no good which is good through another is equal to or greater than that good which is good through itself. So that alone is supremely good which is only good through itself, for that is supreme which so excels others that it has neither an equal nor a superior.’ 79 This is one feature reinforcing the picture of Anselm as a Platonic realist. That he was moving away from this position was argued by Schmitt [6.72]. However, the only evidence for such an interpretation is an apparently non-realist handling of abstract terms to be discussed on p. 135. Anselm’s position as a Platonic realist is examined by Flasch [6.67] and Adams [6.65]. 80 In Anselm’s writing, the term ratio itself has a multiplicity of connotations given by the earlier textual tradition: ontological, theological, epistemological, psychological and logical. See Gersh [6.68]. 81 [6.11] I, pr. 13 On the textual background to Anselm’s argument see Audet [6.66] and Nothdurft [6.35]. 82 The modern discussion seems to have begun with Frege about 1892. See Frege, G., ‘On sense and reference’, pp. 118–40 and Russell, B., ‘On denoting’, pp. 143– 58, both in F.Zabeeh (ed.) Readings in Semantics, Urbana, Ill., 1974. 83 Cf. note 79. To the question whether Anselm saw any inconsistency between these two positions the answer is uncertain. However, since he probably viewed the signifieds of De grammatico but apparently not the transcendent properties of Monologion as universals in the logical sense, the philosophical problems raised by the two treatises were more easily separated for him than they are for his modern reader. The issue of universality is first raised at Monologion 27, [6.11] I. 45. 1–22. 84 See Augustine, On the Trinity X. 1ff., XV. 10–16. On the history of this theory see Colish [6.28] 50–1, 99. 85 The threefold division in this text: sensible signs+sensible manner, sensible signs+insensible manner, sensible signs+neither sensible nor insensible manner, juxtaposes semiotic categories in a manner recalling Eriugena. See n. 19. 86 Anselm does not himself employ the terms ‘unmotivated’ and ‘non-universal’ here. However, he clearly views the first type of sign as defined negatively with respect to the third type. The latter will be specified as motivated and universal. 87 Anselm says that the third type of sign is ‘natural’ (naturalis) apparently meaning that it is motivated. In modern linguistic theory, a motivated sign is one whose signifier and signified are related analogically. See Barthes, SZ pp. 114ff. 88 Monologion 10 [6.11] 1:24. 29ff. ‘It is noted in common usage that we can speak of a single thing in three ways. We speak of things either by using sensible signs— that is, signs which can be perceived by bodily senses—in a sensible manner; or by thinking the same signs which are sensible externally in an insensible manner within ourselves; or by neither using these signs in a sensible nor an insensible manner but by speaking of the things themselves inwardly in our mind through imagination of the bodily or through a rational understanding in place of the diversity of things themselves. For I speak of the man in one way when I signify him with the name “man”, in another when I think the same name silently, and in another when my mind contemplates that same man either through an image of the bodily or through reason. It is through an image of the bodily when the mind imagines his sensible shape, but it is through reason when it thinks his universal essence which is “animal, rational, mortal”. Of these three ways of speaking each consists of its own kind of words. However, the words of that speech which I have posited as third and last—when they are of things which are not unknown— are natural and the same among all races.’ In this passage, Anselm develops the theory which he found in Aristotle’s On Interpretation; see p. 129. 89 Anselm states unambiguously that even the third type of speaking constitutes sign-manipulation of a sort. 90 Monologiom 65 [6.11] I:75. 17–65, 77. 3. The reference to transformation of meaning indicates metaphoricity. 91 It is possible to treat the statement at Proslogion 15 [6.11] I:112. 12–17 that God is ‘something greater than can be thought’ (quiddam maius quam cogitari possit) as a correction of the famous premiss of Proslogion 2 [6.11] I:101. 4–5. If so, Anselm is pointing out that the ontological argument is in the final analysis only an image of the truth. BIBLIOGRAPHY Original Language Editions Eriugena 6.1 Floss, H.J. (ed.) Joannis Scott of era quae supersunt omnia (MPL 122), Paris, 1853. 6.2 Lutz, C.E. (ed.) Annotationes in Marcianum, Cambridge, Mass., 1939. 6.3 Jeauneau, E. (ed.) Commentary on Martianus, Book I, in Oxford, Bodleian Auct. T. II. 18, in Quatre thèmes érigéniens, Montreal and Paris, 1978. 6.4 Madec, G. (ed.) De praedestinatione, (CC c.m. 50), Turnhout, 1978. 6.5 Sheldon-Williams, I.P. (ed.) Periphyseon I–III with facing English translation, Dublin, 1968–81 (Scriptores latini hiberniae 7, 9, 11). (On this problematic edition, see P.Lucentini, ‘La nuova edizione del “Periphyseon” dell’Eriugena’, Studi medievali, 3a serie, 17(1), 1976.) 6.6 Jeauneau, E. (ed.) Periphyseon IV with facing English translation by J.J.O’Meara and I.P.Sheldon-Williams (Scriptores latini hiberniae 13), Dublin, 1995. 6.7 ——(ed.) Periphyseon I (CC c.m. 161), Turnhout, 1996. 6.8 Barbet, J. (ed.) Expositiones in Ierarchiam Coelestem (CC c.m. 31), Turnhout, 1975. 6.9 Jeauneau, E. (ed.) Homélie sur le Prologue de Jean with parallel French translation (Sources chrétiennes 151), Paris, 1969. 6.10 ——(ed.) Commentaire sur Jean with parallel French translation (Sources chrétiennes 180), Paris, 1972. Anselm 6.11 Schmitt, F.S. (ed.) Opera omnia I, II, Edinburgh, 1946. (These volumes contain all the philosophical works except for those in [6.12].) 6.12 Philosophical fragments, edited in F.S.Schmitt and R.W.Southern, Memorials of St Anselm, Oxford, 1969. 6.13 Charlesworth, M.J. Proslogion with parallel English translation and commentary, Oxford, 1965. 6.14 Henry, D.P. The De Grammatico of St Anselm: The Theory of Paronymy, Notre Dame, Ind., 1964. 6.15 ——Commentary on De grammatico: the Historico-logical Dimensions of a Dialogue of St Anselm’s, Dordrecht and Boston, 1974. (Both [6.14] and [6.15] contain the text, translation and detailed commentary on this dialogue.) Gerbert 6.16 Olleris, A. (ed.) Oeuvres de Gerbert, Clermont-Ferrand and Paris, 1867. English Translations Eriugena 6.17 Sheldon-Williams, I.P. (but issued under the name of J.J.O’Meara), Periphyseon (The Division of Nature), Montreal and Washington, 1987. 6.18 Homily on Prologue to John, translated in J.J.O’Meara, Eriugena, Oxford, 1978. Anselm 6.19 Hopkins, J. and Richardson, H. Anselm of Canterbury I–IV (complete philosophical and theological works), London, Toronto and New York, 1974–6. Bibliographies, Concordances and Handbooks 6.20 Brennan, M. A Guide to Eriugenian Studies, Fribourg and Paris, 1989. 6.21 G.H.Allard (ed.) Periphyseon—indices générales, Montreal and Paris, 1983. 6.22 Hopkins, J. A Companion to the Study of St Anselm, Minneapolis, Minn., 1972. 6.23 Evans, G.R. (ed.) A Concordance for the Works of St Anselm, 4 vols, Millwood, NY, 1984. Biographies 6.24 Cappuyns, M. Jean Scot Erigène. Sa vie, son oeuvre, sa pensée, Louvain and Paris, 1933. 6.25 Southern, R.W. (ed. and trans.) The Life of St Anselm, Archbishop of Canterbury, by Eadmer, London, 1962. 6.26 —— St Anselm and his Biographer: a Study of Monastic Life and Thought 1059–c. 1130, Cambridge, 1963. 6.27 —— Anselm, a Portrait in a Landscape, Cambridge, 1990. General and Background Studies 6.28 Colish, M. The Mirror of Language: a Study in the Medieval Theory of Knowledge, 2nd edn, Lincoln, Neb., 1983. 6.29 Eco, U. ‘Denotation’, in [6.30] 43–77. 6.30 Eco, U. and Marmo, C. (eds) On the Medieval Theory of Signs, Amsterdam and Philadelphia, Pa., 1989. 6.31 Hadot, P. Porphyre et Victorinus, Paris, 1968. 6.32 Klinck, R. Die lateinische Etymologie des Mittelalters, Munich, 1970. 6.33 Marenbon, J. Early Medieval Philosophy (480–1150): an Introduction, 2nd edn, London, 1988. 6.34 Minio-Paluello, L. Opuscula: the Latin Aristotle, Amsterdam, 1972. 6.35 Nothdurft, K.D. Studien zum Einfluss Senecas auf die Philosophie und Theologie des zwölften Jahrhunderts, Leiden and Cologne, 1963, pp. 192–7. 6.36 Pinborg, J. Logik und Semantik im Mittelalter. Ein Überblick, Stuttgart and Bad Cannstatt, 1972. 6.37 Vyver, A. van de ‘Les étapes du développement philosophique du haut moyen âge’, Revue belge de philologie et d’histoire 8 (1929): 425–52. Studies on Eriugena 6.38 Allard, G.-H. (ed.) Jean Scot écrivain, Montreal and Paris, 1986. 6.39 Beierwaltes, W. ‘Negati affirmatio. Welt als Metapher. Zur Grundlegung einer mittelalterlichen Ästhetik durch Johannes Scotus Eriugena’, Philosophisches Jahrbuch 83 (1976): 237–65; repr. in Beierwaltes [6.44], 115–58. 6.40 ——(ed.) Eriugena. Studien zu seinen Quellen, Heidelberg, 1980. 6.41 ——‘Language and object: Reflections on Eriugena’s valuation of the function and capacities of language’, in [6.38] 209–28. 6.42 ——(ed.) Eriugena Redivivus: Zur Wirkungsgeschichte seines Denkens im Mittelalter und im Übergang zur Neuzeit, Heidelberg, 1987. 6.43 ——(ed.) Eriugena: Begriff und Metapher, Heidelberg, 1990. 6.44 ——Eriugena. Grundzüge seines Denkens, Frankfurt, 1994. 6.45 Boeuff, P. le ‘Un commentaire érigénien du “De musica”’, Recherches Augustiniennes 22 (1987): 271–309. 6.46 Cristiani, M. ‘Nature-essence et nature-langage: notes sur l’emploi du terme “natura” dans le “Periphyseon” de Jean Erigène’, Sprache und Erkenntnis im Mittelalter (Miscellanea Mediavalia 13, 2), Berlin and New York, 1981, pp. 707–17. 6.47 Duclow, D.F. ‘Nature as speech and book in John Scotus Eriugena’, Mediaevalia 3 (1977): 131–40. 6.48 Flasch, K. ‘Zur Rehabilitierung der Relation: Die Theorie der Beziehung bei Johannes Eriugena’, Philosophie als Beziehungs—Wissenschaft (Festschrift Schaaf), Frankfurt-On-Main, 1971, pp. 1–25. 6.49 Gersh, S. From lamblichus to Eriugena, Leiden, 1978. 6.50 ——‘Omnipresence in Eriugena: Some reflections on Augustino-Maximian elements in Periphyseon’, in Beierwaltes [6.40] 55–74. 6.51 Jeauneau, E. Etudes érigéniennes, Paris, 1987. 6.52 Lucentini, P. Platonismo medievale: Contributi per la storia dell’Eriugenismo, Florence, 1979. 6.53 Madec, G. ‘Jean Scot et ses auteurs’, in Allard [6.38] 143–86. 6.54 Moran, D. The Philosophy of John Scottus Eriugena, Cambridge, 1989. 6.55 O’Meara, J.J. and Bieler, L. (eds) The Mind of Eriugena, Dublin, 1973. 6.56 Onofrio, G. d’ ‘Über die Natur der Einteilung: Die dialektische Entfaltung von Eriugenas Denken’, in Beierwaltes [6.43] 17–38. 6.57 Roques, R. (ed.) Jean Scot Erigènt et l’histoire de la philosophie, Paris, 1977. 6.58 Schrimpf, G. ‘Zur Frage der Authentizität unserer Texte von Johannes Scottus’ “Annotationes in Martianum”’ in [6.55] 125–38. 6.59 ——Das Werk des Johannes Scottus Eriugena im Rahmen des Wissenschaftsverständnisses seiner Zeit. Eine Hinführung zu Periphyseon (BGPTMA 23), Münster, 1982. Studies on Gerbert and his contemporaries 6.60 Frova, C. ‘Gerberto philosophus: II De rationali et ratione uti’, in [6.61] 351–77. 6.61 Gerberto, Scienza, Storia e Mito, Atti del Gerberti, Symposium, Bobbio, 1985. 6.62 Riché, P. ‘L’enseignement de Gerbert à Reims dans le contexte européen’, Gerberto. Scienza, Storia e Mito [6.61] 51–69. 6.63 Rijk, L.M. de ‘On the curriculum of the arts of the Trivium at St Gall from c. 850–c. 1000’, Vivarium 1 (1963): 35–86. 6.64 ——‘Les oeuvres inédites d’Abbon de Fleury’, Revue bénédictine 47 (1935): 125–69. Studies on Anselm 6.65 Adams, M.M. ‘Was Anselm a realist? The Monologium’, Franciscan Studies 32 (1972): 5–14. 6.66 Audet, T.A. ‘Une source augustinienne de l’argument de saint Anselme’, in J.Maritain et al. Etienne Gilson, philosophe de la Chrétienté, Paris, 1949, pp. 105–42. 6.67 Flasch, K. ‘Der philosophische Ansatz des Anselm von Canterbury im Monologion und sein Verhältnis zum augustinischen Neuplatonismus’, Analecta Anselmiana 2 (1970): 1–43. 6.68 Gersh, S. ‘Anselm of Canterbury’, in P.Dronke (ed.) A History of Twelfthcentury Western Philosophy, Cambridge, 1988, pp. 255–78. 6.69 Henry, D.P. The Logic of St Anselm, Oxford, 1967. 6.70 Kohlenberger, H., Similitude und ratio. Überlegungen zur Methode bei Anselm von Canterbury, Bonn, 1972. 6.71 Koyré, A. L’idée de Dieu dans la philosophie de saint Anselme, Paris, 1923. 6.72 Schmitt, F.S. ‘Anselm und der (Neu-)Platonismus’, Analecta Anselmiana 1 (1969): 39–71. 6.73 Serene, E. ‘Anselm’s modal conceptions’, in S.Knuuttila (ed.) Reforging the Great Chain of Being (Synthese Historical Library 20), Dordrecht and Boston, 1981. 6.74 Vuillemin, J. Le Dieu d’Anselme et les apparences de la raison, Paris, 1971. Studies of the Ontological Argument 6.75 Gale, R.M. On the Nature and Existence of God, Cambridge, 1991, pp. 201–37. 6.76 Gombocz, W. ‘Zur neueren Beiträgen zur Interpretation von Anselms Proslogion’, Salzburger Jahrbuch für Philosophie 20 (1975): 131–5. 6.77 Henry, D. ‘The Proslogion proofs’, Philosophical Quarterly 5 (1955): 147–51. 6.78 Hick, J. and McGill, A.C. The Many-Faced Argument, New York, 1967. 6.79 La Croix, R.R. Proslogion II and III: a Third Interpretation of Anselm’s Argument, Leiden, 1972. 6.80 Lewis, D.K. ‘Anselm and actuality’, in Philosophical Papers I, New York, 1983. 6.81 Malcolm, N. ‘Anselm’s ontological arguments’, Philosophical Review 69 (1960): 41–62. 6.82 Plantinga, A. ‘God and necessity’, in The Nature of Necessity, Oxford, 1974.
Routledge History of Philosophy. Taylor & Francis e-Library. 2005.