Abelard’s investigations into the philosophy of language are of great interest not only with respect to the history of philosophy, but also with respect to systematic considerations. These investigations, however, are not readily accessible. They offer nothing to a reader who wants to glean information quickly from them. A thorough study is required, and this itself requires extraordinary patience. The purpose of this chapter is to contribute to the project of making Abelard’s investigations into the philosophy of language accessible to the general philosophical community.
Unlike contemporary philosophers, Abelard does not conceive of philosophy of language as a distinct or separate branch of philosophy. Indeed, as he conceives of philosophy, it is a genus with just three species, namely, logic, physics, and ethics (LNPS 506.18–19). If we want to identify his contribution to what we now recognize as issues in philosophy of language, therefore, we must extract his views from his discussion of that branch of philosophy in which they are embedded, namely, logic or, as he also refers to it, dialectic (hereafter I shall use the terms “logic” or “dialectic” synonymously).1
In the foreword to his discussion of Porphyry’s Isagoge, Abelard explains what he understands logic to be. As logic must be considered as a part of philosophy or as its tool, the route to a definition of logic attains its goal through the definition of philosophy. Philosophy is defined by Abelard as knowing how to discern (discernendi scientia).2 By this Abelard means a type of knowledge obtained through reflection as opposed to other ways of knowing, which he refers to as knowing how to perform (cf. LNPS 505.13–506.10). Whoever is in possession of this knowledge is capable of furnishing reasons for something.
In light of this, Abelard proceeds to a definition of Logic itself. Here again he makes reference to knowing how to discern:
Logic is not the knowledge of the use or construction of arguments, but rather the knowledge of discerning and judging them correctly, namely, why some are strong [valid] and others weak [invalid]. No one is in possession of logic who does not have the power of judging every single argument.
(LNPS 506.26–507.2)3
It is important to recognize that for Abelard logic is no purely formal discipline. Indeed, as Abelard conceives of it, logic is not only a species of philosophy but also that discipline which, together with grammar and rhetoric,4 constitutes a science of human language. Of these three disciplines, Abelard takes logic or dialectic to be the central, most important component of the science of language. Everyone, those leading an active life as well as those of a scholarly occupation, makes use of language. The dialectician reflects upon languages; he investigates their syntactic and semantic rules. He does not do this by distancing himself from the given language and constructing a metalanguage,5 but by means of reflection.
Abelard’s manner of investigating issues in logic or dialectic is not the empirical method of contemporary linguistics. He is not concerned with the syntax and with the semantics of a particular language, but with the syntax and semantics of language in general. Language is, of course, experienced by the logician, as by every other human being, as a given language. But the logician wishes to explain the given language. Abelard the philosopher of language considers language as a meaningful human construct;6 he investigates the rules that constitute meaning in general.
Words and word sequences, that is, grammatically well-formed combinations of words, have meaning. One cannot grasp their meaning if one is not familiar with the things being named and described. As Abelard sees it, therefore, the logician must also inquire into the nature of things. But the logician does this “for the sake of the imposition of a linguistic expression upon something,” that is, in order to “distinguish the meaning of linguistic expressions according to the particularities of things” (Dial. 286.38; 287.2). The modern reader accustomed to the sharp division of various disciplines may be surprised by the lack of a sharp division in Abelard between logic, linguistics, and applied sciences. Indeed, such a reader may even be irritated by the fact that this author allows logical, ontological, and psychological considerations to cross over into each other.7 But Abelard’s procedure has its own standard of rightness and its own epistemological value, as I shall attempt to show.
My discussion in what follows is divided into two main parts. In the first part (§I), I provide some general orientation to those texts and issues that are most important for understanding Abelard’s views in philosophy of language. As we shall see, these particular texts and issues lead Abelard to place questions of semantics at the very center of his logical studies.8 In this respect Abelard’s approach to logic differs from that of contemporary logicians, for whom questions of possible semantic interpretations are typically secondary to the syntactic construction of a calculus. In the second, and longer part of the chapter (§§II–V), I turn to those aspects of Abelard’s philosophy of language that constitute his most enduring legacy, namely, his views about the semantics of terms and propositions. In the course of examining these views, we shall see that Abelard’s distinctive approach in logic is determined in large part by “metaphysical issues such as the status of universals as well as psychological topics such as the role of images in thought.”9
In all his logical writings, Abelard comments upon the canonical texts of the so-called logica vetus – that is, Porphyry’s Isagoge and Aristotle’s Categories and De interpretatione, together with a few of Boethius’s logical works.10 The particular procedure adopted in each of Abelard’s commentaries varies significantly depending on the purpose for which they were written. For example, in the commentaries which have received the name Introductiones parvulorum, the main purpose is mere explanation; hence they are primarily suitable for acquainting students with the content of a text which they do not have at hand. The logical texts entitled Ingredientibus and Nostrorum petitioni sociorum, on the other hand, contain not only explicit, line-by-line commentary, but also argumentative sections in which Abelard enlarges upon the subject matter contained in the text and thinks through questions posed by the texts in an independent manner.11 Finally, there is the Dialectica, which is also very important for understanding Abelard’s views in logic (and hence philosophy of language). In this work, as in the others, one can see clearly that the canonical texts are presented in a canonical sequence. In contrast to his other logical writings, however, Abelard does not follow the texts of the logica vetus sentence for sentence in the Dialectica. Rather, he arranges the material systematically, indicating the subjects in which he is particularly interested by the headings of individual subsections.
For a long time Abelard scholars were virtually unanimous in assuming the Dialectica to be a late work, or at any rate to have been composed after the commentaries of which the Logica “ingredientibus” is comprised. Constant Mews has shaken the traditional confidence in this assumption.12 His arguments are purely internal: Mews notes the absence in the Dialectica of topics and doctrines which play a great role in the Logica “ingredientibus.” De Rijk has examined Mews’s suggested dating and warned of exaggerating the differences.13 De Rijk suggests that both works be considered as roughly contemporaneous,14 though he still argues in a manner which assumes the old dating to be correct.15 Clanchy has recently put forth a suggestion which shows that one can never be sure with respect to hypotheses concerning the development of lines of thought. Instead of assuming a development “from the simple to the complex,” he inverts this valuation. He finds it plausible to assume that Abelard, in the process of his confrontation with William of Champeaux, had to “proclaim the most complex problematics of logic”; later, when he “no longer had an academic rival to impress,” he could afford a very simple style.16
I believe that Clanchy’s argument is clever and instructive, and thus provides an antidote to much traditional prejudice. Nonetheless, it seems to me extremely unlikely that the so-called Introductiones parvulorum is the work of a mature author concentrating upon the essentials of logic. I am convinced, moreover, that Abelard was an author who, his whole life long, sought out rather than avoided complexity and its associated problems. His most cutting criticism is directed against overly simple answers to difficult questions, and he is habitually combating simplifications in an inventive, shrewd, and often sarcastic manner. What he seeks to attain is comprehensibility, and not simplicity, with respect to theoretical constructs. In my view historians of philosophy – not only in the case of Abelard, but also in general – have all too often been one-sidedly concerned with tracking down lines of development with respect to a particular author’s thought. As a compensation for such one-sidedness, especially with respect to the development of Abelard’s logical views, we must take into account the audience to which Abelard’s texts address themselves. In the case of the Dialectica in particular, Abelard himself reports that he wrote it for his brother Dagobert.17 Dagobert seemingly requested a primer which could be useful in the education of his sons. Abelard remarks at the beginning of the second book that the work is growing too voluminous; he almost regrets having promised it (Dial. 146.21–29). One can well imagine that Dagobert, on the other hand, was shocked when his brother sent him thick manuscripts, one after another. They were probably of little use in the instruction of his sons. Nevertheless, this is the purpose for which the work was intended. In the case of the Logica “ingredientibus” and Nostrorum petitioni sociorum, on the other hand, we can see that it is Abelard’s own advanced students that are being addressed.18 Indeed, in reading the digressions or excurses, we can even begin to understand what it was about this teacher that fascinated the students so much that they followed him wherever he went. In these works, Abelard temporarily puts aside the task of commentary, freely develops a problem, and seeks a satisfactory answer, all the while allowing his listeners to take part in his investigative, conscientious, and inventive research.
Abelard was well acquainted not only with the textual corpus of the logica vetus and the research both of Boethius and of his own contemporaries on the subject, but also with Priscian’s grammar and the relevant contemporary research. It is to be assumed that the influence of contemporary grammarians on Abelard is great and that many of his discussions implicitly refer to theses put forth by grammarians. This assumption is difficult to establish in particular cases, however, both because Abelard, like other medieval authors, typically cites his contemporaries without mentioning names and editions, preferring instead the standard locution, “certain people say” (quidam dicunt) – and also because many treatises from this period have not yet been edited. There is still much work to be done in this area.19
During Abelard’s time there were authors and schools that attempted to separate grammar and logic from each other, considering them utterly different with respect to their investigative and theoretical procedures. In such cases, these two disciplines were commonly compared according to the scheme “the grammarians (secundum grammaticos) say . . . whereas the dialecticians (secundum dialecticos) say . . .” There are passages in Abelard’s works where he appears to contrast grammatical and logical theory in just this way. When these passages are examined more closely, however, it can be seen without exception that far from contrasting them as utterly separate disciplines, Abelard is in fact indicating the way in which logic builds upon grammar, by introducing terminological clarifications. To give just one example:
What the substantive is to the grammarians is something other than the substantial is to us, and what the adjective [is to them] is something other than the accidental [is to us]. For the grammarians call all words fixed [with respect to gender] “substantives,” including those which are taken from accidents, such as “woman,” “man,” and “soldier.” We, on the other hand, speak of no names derived from accidents as “substantials.” Again, they call only those names “adjectives” which are of themselves attributed to other names (namely, the substantives), such as “white man,” “rational animal.” We, on the other hand, consider “rational” to be a substantial rather than an accidental. However, we call all names derived from accidents accidentals, including substantives like “man,” “woman.”
(LI De in. 3.05.78, G 384.31–385.1)20
In this passage, Abelard is contrasting morphology and semantics, whereas elsewhere he contrasts the surface and deep structure of propositions with one another. Abelard makes it clear that he thinks the difference between grammar and logic has to do in large part with the fact that grammar is taught before logic in the curriculum. Hence, the grammarians have to make themselves understandable to the young students, and it should be left to the more subtle discipline of logic or dialectic to inquire more deeply into the nature of language and, insofar as it is directed at advanced students, to correct the oversimplifications presented to them during their beginners’ lessons (Dial. 140.23–29).21
During Abelard’s time, the logic curriculum was oriented towards the study of classical texts, with Porphyry’s Isagoge being read first. This text, written as an introduction to the Categories, is understood by Abelard to be an introduction to logic as a whole. Here, concepts with which the logician must work are explained: genus, species, differentia, proprium, and accidens (cf. LI Isag. 2.21–36). However, whereas Porphyry rushes on to the investigation of these logical concepts, Abelard explains at length the questions which Porphyry merely mentions and does not pursue: namely, questions concerning the nature and ontological status of universals. Thus, according to Abelard, logic begins with a tract on universals.
The further structure of the logic, as Abelard sees it, appears simple:
When writing . . . a logic the following arrangement is necessary: as arguments are joined from propositions and propositions are joined from words (ex dictionibus), he who writes a logic in a complete manner must first write of simple words (de simplicibus sermonibus),22 and then of propositions, and must finally bring the goal of logic to perfection in arguments. Our chief [example], Aristotle, also did it thus. He wrote the Categories (Praedicamenta) on the subject of words (ad sermonum doctrinam), Peri hermeneias on the subject of propositions, and the Topics and the Analytics on the subject of arguments.
(LI Isag. 2.8–15)23
It is the principle of grammatical composition that determines this structure.24 One begins with words; out of words propositions are constructed, and out of propositions arguments are constructed. But the matter is not as simple as it is here presented. The title of Aristotle’s first logical work, Praedicamenta, by itself shows that it is concerned not only with words, but words that can be predicated of others. And already in the second chapter he introduces distinctions pertaining to the theory of predication. The theory of words, therefore, contains anticipations of the theory of predication and the theory of propositions. Abelard is not concerned with avoiding such anticipations for the sake of a stringent adherence to a simple structure. In studying grammar one may exercise the capacity to differentiate words according to type and meaning. And in logic the view towards propositions already determines the manner in which words are considered. This attitude can also be recognized in the beginning chapters of Aristotle’s work on sentences or propositions (Peri hermeneias). Aristotle writes chapters on nouns and verbs before coming to propositions composed of nouns and verbs. The verb, however, is simultaneously introduced as a “statement word” (LNPS 506.18–19). Here, too, Abelard expands upon, rather than restricts, these anticipatory remarks.
There is a reason why Abelard proceeds in this manner. The principle of composition is a syntactic principle. Abelard’s main interest, however, is in questions of semantics. Our understanding of words is affected by the context in which words are used. In particular, the employment of a word in subject or predicate position in a sentence influences how we understand a word.25 Questions of semantics are connected to questions of syntax and logical function.
In the passage cited above, Abelard describes argument as the “goal of logic.” It seems to me, however, that the theory of propositions is plausibly taken to be at the center of Abelardian logic. The anticipations of the theory of propositions are indispensable for the theory of words. It belongs to the theory of propositions to treat dicta as components of compound propositions (if–then sentences). Anticipations of arguments and demonstrations are infrequent, and when they are to be found, they are not indispensable.
The question as to what Abelard saw as his main task with respect to the philosophy of language, as well as to his specific achievement in this regard, has not yet been settled in Abelard scholarship. In some contributions to this subject only certain of Abelard’s teachings are emphasized, namely, those of particular interest to modern academics for various reasons.26 In such cases, Abelard is seen as a logician working on a consistent theory of argumentative language with particular explanatory force. In other contributions, however, Abelard is considered rather as an analyst of language. According to this view, his main concern is with the development of a rich interpretative arsenal and his particular strength lies in the ability both to demonstrate the need to interpret particular sentences and to develop distinctions that are decisive for determining the truth or falsity of such sentences.27 Representatives of both viewpoints are able to find statements of Abelard to support their claims. Friends of systematic construction can point to the emphasis with which Abelard plays off position against position in his discussions and shows his own view to be superior. Friends of linguistic analysis find confirmation of their views in Abelard’s conception of logic as knowing how to discern and in the undogmatic variability with which Abelard carries out his distinctions. It seems to me unnecessary, however, to decide which interpretative attitude is correct. Both are fruitful. Abelard is neither a logician in the modern sense nor a linguist in the modern sense. He reflects upon the language of arguments with a view towards the question of their normativeness.
With these preliminaries in mind, let us turn to some of Abelard’s specific views in philosophy of language.
According to Abelard, all the things we find in the world are, in each case, determinate singular things. (The word “thing” is to be understood here in the widest possible sense, corresponding approximately to the word “something.” Thus, a color that we see, or a sound that we hear, is also a determinate singular “thing.”) As Abelard recognizes, in some cases we have proper nouns at our disposal for the naming of singular things, as for example when we name a determinate singular human “Socrates.” Indeed, it is precisely the function of the proper noun to mark a singular thing as such: the proper noun makes known who or what is meant, and over and above this it has no further meaning. Thus, when proper nouns are used repeatedly – that is to say, for naming different persons or things – one should not seek a common property of the persons or things named. In such a case, proper nouns represent, as we might put it, a multiplicity of things by way of equivocation (cf. LI Isag. 16.28–35; Spade 1994, 65).
The case is different, however, when we name singular things by means of a descriptive term in conjunction with a demonstrative pronoun: “this substance,” “this body,” “this animal,” “this man,” “this whiteness,” “this white thing.” Here we understand not only who or what singular thing is meant, but also something about that singular thing, namely, that it is a substance, or a body, or an instance of whiteness. By denoting the singular thing in this way, we direct our attention upon it with reference to a quite definite aspect (cf. LI Isag. 27.24–29; Spade 1994, 141).
When we describe particular things we employ words “which are apt to be predicated of many.”28 With each such descriptive word we highlight one and only one aspect of the thing being described; each descriptive word is, however, also applicable to things other than the thing presently being described. The descriptive word is in this sense universal. The obvious questions as to whether something universal exists which is denoted by descriptive or universal words, and, if so, what the nature of its existence is, have come to be known as the problem of universals.29
We can make clearer what is actually being asked here, especially as it would have been understood during Abelard’s time, if we consider the communicative function of language. A speaker, A, denotes a determinate singular thing by means of a descriptive word. (In Abelard’s terminology A imposes a word on the thing.30) A hearer, B, does not, let us suppose, know precisely which singular thing A is speaking of. Nonetheless, he understands the word, and by means of his understanding, is capable of imposing the word on another singular thing, namely, a singular thing other than the one to which A is referring. What does B understand when he understands the word? (Cf. LI De in. 3.01.91–92, G 324.29–325.11)31
Now there are authors, during Abelard’s time, who postulate the existence of universal things (res universales) to serve as the denotation of “universal words” (simplices sermones universales) and assume that universal things are denoted by such words in precisely the same manner in which singular things are denoted by “singular words” (simplices sermones singulares).32 Abelard criticizes and rejects every variant of such realism with which he was acquainted.33 He does not admit the existence of any universal things; all things, he says, are singular things (LI Isag. 18.10; Spade 1994, 78). It thus “remains to ascribe . . . universality only to words (restat ut . . . universalitatem solis vocibus adscribamus)” (LI Isag. 16.21–22; Spade 1994, 63).34
Instead of the word vox, Abelard’s later commentaries on Porphyry employ the word sermo as the subject of universality (LNPS 514.8; 522.10–13). The question as to the importance of this change in formulation cannot be addressed here.35 In any event the formulation with the word sermo is more to the point, as linguistic expressions are not universal, but singular; rather it is their contents, which are understood, that are universal.
The realistic theory provides a straightforward answer to the question as to what is denoted by universal words. The vocalistic or sermocinalistic theory must, if it is to succeed in its rejection of this realistic answer, provide a better answer. Abelard’s answer is the following:
Single men, who are discrete [i.e., wholly distinct] from one another since they differ both in their own essences and in their own forms . . ., nevertheless agree in that they are men. I do not say they agree in man, since no thing is a man unless it is discrete. Rather they agree in being a man (esse hominem). Now being a man is not a man or any other thing . . ., any more than not being in a subject is a thing, or not admitting contraries, or not admitting of greater and less. Yet Aristotle says (Cat. 5, 2a12–4b19) all substances agree in these respects . . . Thus Socrates and Plato are alike in being a man, as a horse and an ass are alike in not being a man.
In addition to speaking of agreement in “being a man,” Abelard also employs the formulation “agreeing in the status of man” (LI Isag. 20.3–4);37 both formulations are equivalent.38
It is clear from which position Abelard is distancing himself. He argues cleverly in favor of the notion that being-F is not a thing by taking into account the negation “not-being-F.” No one, he thinks, would want to maintain that not-being-F is a thing. But singular things are nonetheless capable of agreeing insofar as they are not F (as in the case of a horse and an ass, which are alike “in not being a man”). But if it is not necessary to postulate a thing for the explanation of this agreement, then there is also no reason to postulate a thing for agreement in being-F.
But does Abelard’s account provide an explanation of the intelligible content of universal words? Or are his remarks too trivial and scanty to provide a genuine theory of universals?39
Every theory is an attempt to solve some problem. And a good theory is one that offers a solution to the problem it addresses without introducing assumptions that are indefensible. In the critical part of his investigation, Abelard takes himself to have shown that all the realistic solutions available to him make use of assumptions that cannot be defended. In the expository part of his investigation, therefore, he changes the problem that needs to be addressed. That is to say, he modifies the ontological questions posed by Porphyry (Isag. 1.9–13) and taken up by Boethius in his second commentary (In Isag. maior 1.10 159.3–167.20) into questions of a decidedly semantic type (LI Isag. 8.11–23; Spade 1994, 9–11).40 In fact he investigates primarily his own questions (LI Isag. 18.4–27.34; Spade 1994, 76–142), addressing Porphyry’s questions only in an appendix (LI Isag. 27.35–32.12; Spade 1994, 143–175).
Abelard’s discussion focuses on two questions. The first question deals with “the common cause of the imposition of universal names” and asks, “in what respect . . . the individuals of a particular genus or species [agree], so that it is justified . . . to impose universal names . . . on them.”41 As we have seen, this question is answered through his account of status or agreement of individuals with respect to being-F. The second question asks about “what . . . one actually [understands], when one understands the sense of a universal name.”42 This question is answered by means of a theory of concept formation. The understanding leading to the use of universal words is an act of abstraction (LI Isag. 24–27; Spade 1994, 102–153). Abelard describes this as an act of selective attention to a particular characteristic of a thing; this characteristic is seen clearly, as it were (cf. LI Isag. 25.6–16; Spade 1994, 127),43 while the thing as a whole is seen only vaguely.44 By means of abstraction, therefore, a thing is conceived “otherwise than as it subsists”; but the resulting conception is not, by virtue of this fact, false or empty, since it does not represent the things and its characteristic as actually existing in separation from one another. As Abelard puts it: “The understanding attends [to things] by abstraction ‘dividedly,’ not as ‘divided’” (cf. LI Isag. 25.15–26.3; Spade 1994, 128–132).
The grammarians of Abelard’s time taught a semantics, taken from Priscian, based on word classes in which fields of meaning were assigned to the following word classes: “substantive,” “adjective,” and “verb.” Substantives were taken as words for things, adjectives as words for properties, and verbs as words for activity and passivity.45 Abelard distances himself from word-class semantics. Although he thinks it often does reflect the sense of words, there are too many exceptions to provide a solid foundation for semantic investigations. Grammatical and ontological categories are more or less in agreement, but never precisely.46
Abelard thinks it is particularly important to refute what the grammarians take to be the semantic difference between nouns and verbs.47 On the one hand, he emphasizes that “[the term] ‘universal’ not only includes names (nomina), but also verbs and infinite names [i.e., names formed with the prefix ‘non-’]” (LI Isag. 18.1–2; Spade 1994, 75). On the other hand, he says, every name can be verbalized.48 This takes place by means of the addition of “is” or temporal modifications of “is.” “Is” functions as a tense indicator.49 Thus, the difference between names and verbs consists solely in the fact that verbs have a temporal meaning in addition to their central meaning (consignificatio temporis). Indeed, as Aristotle had already taught (De in. 3), verbs are temporal words.50 The difference between words which have a simultaneous temporal meaning and words which do not must then be considered when Abelard – following Aristotle’s Peri hermeneias – considers the proposition, where the question arises as to how the temporal meaning and the proposition-building predicative function are connected with each other.
Instead of differentiating between fields of meaning assigned to names on the one hand and verbs on the other, Abelard differentiates between dimensions of meaning,51 which he thinks all categorematic words have, whether nouns or verbs:
Names . . . and verbs are signs in two ways – on the one hand, they are signs for things; on the other, they are signs for understandings. Indeed, they are signs for things insofar as they constitute the understandings associated with them, i.e. the understandings by which one attends to an essential feature or a peculiar characteristic of things.
(LI De in. 3.00.4, G 307.26–30)52
We can understand this text as a summary of that which has been said about universals. It must now be added, however, that Abelard thinks the distinction between the two kinds of signification – the signification of things vs. the signification of understandings – should be used in the construction of a logical system. Thus, in the Categories, he says, simple words are treated as signifying things, whereas in the Peri hermeneias they are treated as signifying understandings (ibi . . . secundum significationem rerum, hic secundum significationem intellectuum) (LI De in. 3.00.11, G 309.14–17).53 This attempt at correlation leads to problems. According to Abelard, things can only be signified or referred to on the basis of an understanding. Thus, in a passage quoted at the beginning of his commentary on Peri hermeneias, he has the following to say about the signification of understandings: “as far as the cause of the invention of words is concerned, it is primary” (LI De in. 3.00.11, G 309.17–18). According to Abelard’s logical curriculum, however, the Categories is to be taught before the Peri hermeneias. But since Abelard understands the Categories to be investigating the signification of things and the Peri hermeneias to be investigating the signification of understandings, this implies that the signification of things is discussed before the signification of understandings in the curriculum, despite the fact that the latter is supposed to be primary. Abelard is not confused here. As he explains, there is a sense in which each signification, that of things and that of understandings, can be called primary. One must simply keep the senses of “primary” straight.
At the beginning of the commentary on the Categories Abelard reaffirms that “as far as the cause of the imposition of a name is concerned,” the “signification of an understanding” is primary.54 However, he simultaneously introduces another priority: “By their very nature things are prior to understandings, . . . thus whoever has invented a word had first observed the nature of a thing and imposed the name in order to indicate the thing’s nature” (LI Cat. 112.33–36).55
The two kinds of signification we have been discussing are not to be distinguished from one another in such a way that they may be treated entirely independently of each other. What Abelard distinguishes, therefore, are different emphases. The Categories is concerned primarily with the things that lie behind speech, whereas the Peri hermeneias is concerned primarily with the manner in which things are understood. And it makes sense to investigate that which is understood through thought and speech in the curriculum before reflecting upon the manner in which it is understood (modus concipiendi) (LI De in. 3.00.7, G 308.28–29).56
Now as Abelard sees it, the analysis of the signification of things does not include an investigation into the manner of understanding, but it does include an investigation into the manner of signifying (modi significandi). Central to Abelard’s interpretation of the Aristotelian Categories, moreover, is a distinction between a thing’s being signified in essentia and its being signified in adiacentia.57
When the logician reflects upon that which is denoted linguistically, he does this in a most general fashion; particular words with their particular significations serve only as examples for similar words with similar denotations (cf. LI Cat. 111.18–20, and 193.11–16). The most general distinction with respect to that which can be denoted by words is the distinction between substance and accident. In Abelard’s discussion of the Categories, this distinction is mirrored on the level of the signification in a characteristically transformed manner. Thus, he speaks not of words which signify independent things and of words which signify properties, but of words which signify something as subsisting (in essentia) and words which signify something as existing in a subsisting thing (in adiacentia) (LI Cat. 113.5).58 Reflection on Abelard’s examples makes evident why he chooses such a complicated manner of expression. Examples of words which signify something as subsisting are “man” and “whiteness” (albedo), whereas an example of a word signifying something as existing in a subsisting thing is “white” (album) (LI Cat. 113.9–11). Why is the substantivized or abstract property-word specially mentioned? Why should a distinction between whiteness and white influence his entire commentary on the Categories?59
In the Categories, Aristotle aims at developing a theory of the highest genera. The members of these genera should be clearly distinguished from one another; they must not overlap. This criterion is fulfilled, however, only when the members of accidental categories are signified by abstract nouns such as “color,” “whiteness,” etc. The concrete nouns “[something] colored,” “[something] white,” etc., indicate, on the other hand, an interweaving of the categories; without further specification, they signify some independent thing having the property of being colored or white, and thus unite, in a contingent manner, the categories of substance and quality.60 In the theory of categories, therefore, abstract constructions are decisive for the proper signification of things in the categories of accidents. The members of the category of substance, on the other hand, are clearly distinguished from accidents, even when they are referred to by concrete names. For this reason there is no discussion in Abelard of “man” and “manness” analogous to the discussion of “white” and “whiteness.” Only in the case of the accidental categories are the concrete forms semantically “named after” the abstract forms.61 Words such as “color” or “whiteness,” therefore, indicate something accidental, but they do not indicate it with respect to its being-in-something, but rather as taken by itself. Thus, when we wish to attribute an accident, such as color or whiteness, to something, or to predicate it of a substance in which it inheres, we must use the concrete name “[something] colored” or “[something] white.”62
In his commentary on Peri hermeneias, Abelard suggests that the distinction “in essentia vs. adiacentia” is a distinction belonging to the theory of predication (LI De in. 3.03.94, G 360.15–361.3).63 The passage has often been discussed in the secondary literature,64 because it appears to conflict with what he says in LI Cat. about the distinction between signifying in essentia vs. in adiacentia. A short discussion of the passage must be undertaken here.
The questionable passage, in which the construction of propositions with the help of the word “is” is explained, runs as follows:
Because [the substantive verb (verbum substantivum) – i.e., a form of the verb “to be”] signifies something as subsisting, it never lacks the conjunction associated with subsistence, because in every case it expresses that something else is. This is true even when it is conjoined with adjectives, as for example when one says “that thing over there is white.” For even though, according to the intention of the one making the statement, whiteness is alone conjoined, and for this reason it alone can be said to be predicated, nonetheless because of the force of the substantive verb the foundation (subiectum) of whiteness [i.e., something white] belongs to Socrates in the manner of subsistence (essentialiter) . . . Two things are conjoined with Socrates through the predicate “white,” namely, whiteness as conjoined (in adiacentia), and something white (i.e. that which is affected through whiteness) as subsisting (in essentia). Nonetheless, only whiteness is predicated because it is intended only to conjoin whiteness.
(LI De in. 3.03.94, G 360.15–25)
It is confusing that, in this passage, Abelard characterizes the concrete term, “[something] white,” as in essentia and the abstract term, “whiteness,” as in adiacentia. This, of course, is precisely the opposite of what he said about these qualifications in his Categories commentary. A little reflection, however, shows that the difference in Abelard’s terminology is perfectly consistent.
In the proposition “Socrates is white,” an accident is attributed to a substance. Whoever puts forth such a proposition does not intend to place Socrates among the white things; rather, he intends to say of Socrates that whiteness is in him – in other words, that he is affected by whiteness. The form of the proposition, however, is diametrically opposed to the speaker’s intention. “Whiteness” in the predicate position cannot be conjoined with “Socrates” in the subject position. In order to do justice to the speaker’s intention, says Abelard, whiteness must be predicated, and predicated as conjoined (where being predicated is, of course, the “conjoining form” denoted by the word) (LI De in. 3.03.94, G 361.1). Since the form of the proposition, however, demands “[something] white” as a predicate of “Socrates,” the predicate of a proposition such as this one makes reference to something subsisting, which is then conjoined with the subject through “is,” which always – even as a copula – signifies something subsisting. It is thus conjoined as something subsisting. “The foundation indicated by the word is conjoined in the manner of subsistence” (LI De in. 3.03.94, G 361.1–2).
The point of all this is that the signification of words depends in large part upon context. When one asks what an isolated word such as “whiteness” or “[something] white” indicates, the correct answer is “whiteness” signifies something as subsisting65 whereas “[something] white” denotes something as existing in a subsisting thing.66 When one asks how predicates refer to a subject, however, then the correct answer is that the intention of the speaker is to predicate “whiteness” as conjoined, but the propositional form leads to the conjunction of “[something] white” as subsisting. The accidental form of whiteness is referred to as subsisting, but it is predicated as conjoined.
One might ask whether there could be a propositional form that would correspond to the speaker’s intention. Abelard is interested, however, not in inventing an ideal language, but rather in analyzing language as it is actually used. Moreover, he thinks the suggestion of a propositional form such as “Socrates has whiteness” leads to greater shortcomings than could be comfortably accepted. “Has” is not as empty of content as “is”; the manner of having would need to be precisely determined, as the word is used in a plurality of senses. Better, therefore, simply to rely on “is” as an auxiliary verb in the function of the copula.
The brief discussion undertaken here indicates something of fundamental importance for Abelard’s philosophy of language. As he sees it, the employment of words in context, in particular the employment of a word in subject or in predicate position of a proposition (or of a dictum) is important in semantic analysis.67
What is a proposition? And what, generally speaking, is signified by a proposition? In order to answer these questions satisfactorily, Abelard thinks extensive investigations are necessary.68 To begin, he argues, we must undertake an investigation into the signification of thoughts or understandings – that is, into the manner in which we grasp intellectual contents. Then he thinks it is necessary to discuss a series of distinctions in order to lay the proper groundwork for an answer to these questions and in order to avoid mistakes along the way.
The Peri hermeneias is organized in such a manner that Aristotle first treats of the noun (De in. 2) and then of the verb (De in. 3), and only then, in the following section, of the proposition, which he says is composed of at least one noun and one verb (De in. 4). In his commentary, Abelard emphasizes that Aristotle’s thematic selection of words (dictiones) was made in advance with reference to the theory of the proposition (LI De in. 3.00.3, G 307.17–23).69 From the same perspective he explains why the signification of understandings is important in the Peri hermeneias, while the signification of things can be neglected (LI De in. 3.00.6–10, G 308.19–309, 13).70
In his first argument, he explains that a name such as “running” and a verb such as “runs” can only be distinguished according to the understandings signified by them, since they signify the very same thing.
The different manner of grasping brings a change to the understanding; for with the former running is shown as subsisting (in essentia), with the latter as conjoined (in adiacentia); in the case of the former, it is shown without a temporal distinction, in the case of the latter with a temporal distinction.
(LI De in. 3.00.7, G 308.23–33)71
In the second argument, which anticipates later explanations, Abelard formulates a central thesis of his propositional semantics. A proposition consists of at least one noun and one verb. A proposition is understood by conjoining the understandings of the parts of speech with one another. Abelard closes off the path to speaking of any thing signified by a proposition, which would arise through the conjunction of the things signified by the parts of speech with one another: “The proposition has no thing which is subjected to it” (LI De in. 3.00.8, G 308.34–40).72
One can, according to the third argument, have a lasting understanding of words even when the words refer at one point in time to a thing and at a later point in time do not. Meaningful words assembled in a meaningful fashion result in a proposition. For this, however, the signification of things is not required (LI De in. 3.00.9–10, G 309.1–13).73
The understanding signified by a particular word (dictio) is simple. But when particular words are combined in a grammatically correct manner to form a word sequence (oratio), the understanding signified by the sequence itself is compound. Now it is important to realize that a simple understanding is not the understanding of something simple, but the grasping of something in a simple manner. Abelard makes this clear in a skilful way. The simple act of understanding “man” and the compound act of understanding “rational mortal animal” have the same meaningful content. The former is one single act of understanding. The latter is a compound act of understanding in which several single acts of understanding are related to one another and conjoined with one another (LI De in. 3.01.94, G 325.17–37; cf. also LI De in. 3.01.117, G 330, 15–18).74
In addition to distinguishing between simple and compound understandings, Abelard also distinguishes between understandings that are sound and understandings that are empty (LI De in. 3.01.97, G 326.30). Several words, he says, each of which refer to something, can lead to an empty understanding when they are conjoined; this happens, for example, whenever the individual understandings signified by the words are in contradiction with one another (LI De in. 3.01.117, G 330.11–15).75
Keeping these distinctions in mind, let us turn to propositions. It is certain that each proposition is a sequence or string of words, but not every word string is a proposition. Is it possible to specify the sort of word string that has propositional character? As a first step, Abelard suggests, we might draw a further distinction between incomplete word strings and complete word strings or sentences (oratio imperfecta vs. oratio perfecta). The criterion for this distinction, as Abelard understands it, is psychological or epistemological and is related to the hearer. In the case of an incomplete word string or utterance the hearer expects an additional element – that is, something additional to be said (LI De in. 3.05.4, G 373.6–8).76 In the case of a complete word string or sentence, on the other hand, the speaker no longer expects any such empty place to be filled in. From this initial standpoint, Abelard poses the question whether propositions are semantically distinguishable from other types of sentence. After emphasizing how important the signification of understandings is for Peri hermeneias, it is natural to suppose that Abelard would attempt to distinguish propositions in terms of the truth or falsity of the understandings signified by them.77
As it turns out, however, Abelard rejects any attempt to distinguish propositions in terms of an analysis of understandings which corresponds to them. For according to him, in the case of “a man runs” we understand exactly the same thing as in the case of “a running man”; there is nothing capable of distinguishing them (LI De in. 3.01.98, 3.05.4, G 327.1–13; 373.1–6).78 And orders, wishes, questions, and propositions again can have precisely the same intelligible content (LI De in. 3.05.13, G 374.29–30).79 Propositions are true or false, however, and this distinguishes them from incomplete word strings as well as from orders, wishes, and questions. Evidently, therefore, in order to clear up the sense of the expressions “true” and “false,” we must somehow proceed from the signification of understandings to the signification of things. But how is this supposed to work? There is no procession from the distinction “sound vs. empty” to the distinction “true vs. false.” For one can construct true propositions in which a concept referring to nothing occurs in subject position. This is clear in the case of negative propositions and “if, then” statements (LI De in. 3.04.22, G 366.28–29); but even in the case of sentences such as “Socrates is Socrates” some distinction must be made between the question whether the naming term denotes something real and the question whether the sentence is true (LI De in. 3.04.23, G 366.33–40).
In his theory of the proposition, Abelard attempts to preserve both of the intuitions suggested above, namely, the intuition of arriving at the linguistic element “proposition” via linguistic construction, as well as the intuition of demanding a particular relation to reality for true propositions. Moreover, he attempts to preserve these intuitions without succumbing to the problems which accompany an overly simplistic formulation of these intuitions. A proposition, he says, is an assertion whose structure can be analyzed as follows: “It is true/false that . . .” The propositional content, or the content asserted by the that-clause (in the original Latin, an “accusative-infinitive” construction is most often used), is accompanied by a certain sort of evaluative judgment, one concerned with truth or falsity. (There are judgments other than those concerned with truth or falsity; judgments can also have the sense of a recommendation [It is good/bad that . . .] or a regulative sense [It is commanded/forbidden/permitted that . . .].) The distinction between propositional content and evaluation or judgment is also maintained in the case of conditional sentences: neither the antecedent nor the consequent is itself asserted but only the whole conditional (cf. LI De in. 3.04.19–21, G 366.5–27).80
Abelard refers to propositional content as “that which is said by the proposition” (dictum propositionis) (LI De in. 3.04.26, G 367.9–16).81 He paraphrases the sense of the judgment or evaluation “it is true that . . . ,” “it is false that . . .” by “it is the case in reality” (ita est in re), “it is not the case in reality” (non est ita in re) (LI De in. 3.04.26, G 367.16–20; cf. 3.01.100, G 327.18–21).82 Propositions of the form “it is true/false/possible/necessary/good/bad that . . .” are irreducibly impersonal.83 Propositional contents or dicta are related to propositions in the same way that things are related to names; but they are not themselves things (LI De in. 3.04.26, G 367.9–13).84 The discovery of irreducibly impersonal propositions is an important achievement; and it is through this discovery that Abelard distances himself from the Aristotelian fixation upon propositions of the form “something is said of something.”85 The semantic distinction between personal and impersonal propositions is particularly important with respect to the analysis of modal propositions. Modal propositions de dicto (or de sensu) and modal propositions de re are not merely distinguished through the position of the their operator – whether in sensu compositionis (per coniunctionem) or in sensu divisionis (per divisionem). Rather de dicto modal propositions are irreducibly impersonal, whereas de re modal propositions address a capacity or a determination of a thing in a personal fashion (LI De in. 3.06.15, G 392.33; LI De in. 3.12.20, MP 14.20ff.).86
For Abelard, the discovery of irreducibly impersonal propositions is an extension of subject-predicate logic. Subject-predicate logic remains the foundation. All propositions investigated by Abelard have a subject and predicate in their unanalyzed form. And this is true of even of dicta in their analyzed form.
According to Abelard, propositions or propositional contents must be analyzed, since propositions that are identical with respect to their grammatical form may have completely different truth-conditions.87 In Latin, which has no articles, this is even clearer than in a language, such as English, where the interpretation of a noun can be indicated through the addition of definite or indefinite articles. A few examples will suffice to demonstrate this point. Take, for example, the proposition “Homo sentitur” (“Man is perceived”). This is true only if, at the point in time at which the proposition is put forth, there is in fact a man of whom it can be said that he is being perceived. The sentence is thus to be translated as “A man is being perceived.” By contrast, a proposition of the same form, “Homo intelligitur” (“Man is understood”), does not, according to Abelard, concern a definite man who is present; rather it is a proposition concerning a concept (since only concepts are understood). Using quotation marks, which medieval authors did not have at their disposal, it can be translated most perspicuously as: “The concept ‘man’ is understood.”88
Again, with respect to “fire,” “hot” is a specific difference: “fire is hot” is a proposition in which an essential property is asserted of a subject. With respect to water the same predicate is the expression of an accidental determination: “This water is hot.”89
The predicate determines how the subject of the sentence is to be comprehended. When a predicate is combined with another adjective or qualification, then the qualification can itself in many, but not in all, cases be predicated of the subject in its own right. One can proceed from “This is a wooden statue” to “This is wooden” or “This is wood”; but not from “This is a good musician” to “This is a good” or “This is good (simpliciter).”90
As Abelard recognized, intransitive predications,91 in which something is said of something in an affirmative or negative manner, are linguistically encountered in two variants: (a) the predicate is expressed nominally and joined to the subject by means of a copula (as, e.g., in “Socrates is running”); (b) the predicate is expressed verbally and joins itself directly to the subject (as, e.g., in “Socrates runs”). Aristotle had taught that this grammatical variation is no indication of semantic difference (De in. 12, 21b9–10). But while Aristotle, and with him almost every medieval logician, considers (a) the tripartite form of predication (that is, the dipartite form of the predicate “is . . .”) as a clarification and explanation of (b) the dipartite form of predication (that is, the monopartite form of the predicate), Abelard opts in several places to conceive of the tripartite constructions of (a) – in Latin, with est and a complement – as mere substitutions for the dipartite constructions of (b) in cases where there are missing verbal forms which have not been invented for a specific semantic content (LI De in. 3.03.25–28, G 348.21–349.17; Dial. 122.22–123.5).92 The unprejudiced manner in which Abelard suggests this and the systematic interest of this reversal of the usual direction of thought have led different scholars, including myself, to consider the dipartite form “F(a)” with the tensed predicate “F-t(a)” as the Abelardian theory of predication.93 Other authors emphasize the fact that Abelard’s arguments are only applicable to definite propositional contents, namely, for tensed predicates. “Socrates will walk” does not, of course, mean that Socrates will at some future point in time – erit – be a present walker – ambulans – but rather that there will come a point in time at which Socrates is walking – ambulabit (cf. LI De in. 3.03.28–30, G 348.37–350.3).94 And tensed propositions are to be converted in such a manner that the entire tensed predicate is in subject position: “Every old man was once young” – “Some of those who were once young are (now) old” (Dial. 139.82–140.14; cf. LI De in. 3.03.33, G 350.29–39).95
In cases such as these, however, Abelard was probably not working upon a particular theory of predication which would be valid for all possible predications, but rather showing an interpretative possibility for certain sentences.96 As Abelard sees it, the purpose of research into the theory of predication is not to discover an ideal form of all possible predication, but to find reliable distinctions between different manners of predication. This task was already proposed by Aristotle, when in the second chapter of the Categories he differentiated between (1) something which is “said of” a subject and (2) something which is “in” a subject (Cat. 2, 1a20–b9).97 Abelard has no doubt that one must speak of predication in both of these cases. Indeed, he construes the Aristotelian distinction between (1) and (2) as a distinction between essential predication, in which a subject concept is analyzed into that which belongs to its definition, and (2) accidental predication.98 According to Abelard, moreover, there is essential predication in all categories (e.g., “Whiteness is a quality”), for it is not only substantial concepts that are defined, but accidental concepts as well (LI Cat. 127.8–10).99 In the case of accidental predication, however, the above-mentioned tension between a speaker’s intention and linguistic execution occurs: “[something] white” is conjoined with “Socrates,” but “whiteness” is what the speaker intends to predicate of “Socrates.”100 In the case of accidental predications the predicate is typically tensed.
Investigations into the sense of the word “is,” in its various different uses, belong to the theory of predication.101 In the discussions to which Abelard contributes, it is generally assumed that the word can be used as a main verb (secundum adiacens) – as, e.g., in “Socrates is,” which just means Socrates exists. In such cases, “is” denotes the subsistence of something. From this function, a name for the word “being” is derived in the terminology of the time: verbum substantivum. The topic of discussion then becomes whether the word “is,” when it is used as a copula (tertium adiacens) maintains its primary sense. Abelard is particularly interested in propositions in which one cannot derive a proposition of the form “a is” from a proposition of the form “a is F.” These are propositions in which a term whose denotation does not exist – “chimaera” (cf. LI De in. 3.03.98–100, G 361.12–25; Dial. 162.16–18)102 – or no longer exists – “Homer” (cf. Dial. 135.9–138.26) – is in subject position. Abelard’s thesis is that the word “is,” when used as a copula, loses its existential meaning. In such cases, we use it as a means of conjunction. It can “additionally” signify a time at which what is said in the proposition is true. The copula, he says, resembles other signs, for example conjunctions, which taken by themselves mean absolutely nothing, but nonetheless cause us to relate the meaningful signs to which they are attached in a certain fashion (cf. LI De in. 3.02.28–29, G 339.24–32).103
Abelard was a teacher who was followed by his students wherever he went, and he writes in the manner in which he taught. His strength lies in his perceptive discussion of viewpoints and examination of particular theses. He reviews arguments in favor of a given thesis and constructs arguments against it. Often his investigations break off without a conclusive result, showing that, for him, the investigation is an end in itself – it is nice if it has a conclusive result, but not bad if it has none. For this reason, editions of Abelard’s works with commentaries are an immediate necessity, and the most important task of the commentator is to make Abelard’s line of thought transparent by suggesting a division of his text into subsections.
In his semantic analyses, Abelard works with example sentences. Many of them are alien to the normal use of language, and thus can give the impression of being all-too trivial or all-too artificial. One has to be clear, therefore, about the purpose for which Abelard works with such example sentences: when semantic distinctions are to be drawn, he thinks as little as possible should be left to feeling and intuition. Semantic distinctions are to be made verifiable by presenting them in terms of true–false distinctions. An example sentence is thus put forth as true or false according to how it is comprehended. This method of presenting semantic distinctions is not unique to Abelard. Later medieval logicians too proceed in this manner. They execute their investigations by constructing and analyzing so-called sophismata.
Abelard was a great logician. But his genius develops in a particular historical environment. Abelard’s place is the school. We still know too little about the schools at the beginning of the twelfth century in Paris and in other places. What sort of people considered it meaningful to dedicate many years of their lives to the study of logic and the philosophy of language? In Paris many schools were in competition. At that time they taught grammar and logic, and for advanced students there was theology. They distinguished themselves through learned opinions and theses which they stubbornly defended. There is now scholarship which makes visible the entire spectrum of discussion in the case of particular lines of questioning.104 But in this area of research much more work is necessary. We need detailed investigations and we need, above all, further editions of the primary sources. In this process it is not of decisive importance that a treatise or commentary be attributable to an author known by name. Anonymous works can, in a philosophy which is bound to the schools, be just as informative as works whose authors are known to us from other sources.105,106
1. Cf. LNPS 506.21–23: “Logicam . . . idem dicimus quod dialecticam et indifferenter utroque nomine in designatione utimur eiusdem scientiae.” Relevant considerations are not, of course, to be found exclusively in his logical writings, but also in his theological texts and in the Coll. Jean Jolivet, David Luscombe, John Marenbon, and others have emphasized precisely this aspect of the fruitful applications of Abelardian philosophy of language.
2. Cf. LNPS 506.4–5: “Sola . . . discernendi scientia philosophia dicitur.” Cf. Jacobi 2000.
3. “Non enim est logica scientia utendi argumentis sive componendi ea, sed discernendi et diiudicandi veraciter de eis, quare scilicet haec valeant, illa infirma sint. Nemo enim aliter logicam habet, nisi uniuscuiusque argumenti vim diiudicare valeat.” The qualification furnished here was worked out by Abelard when he was in the process of commenting upon Boethius’s definition of logic as ratio disserendi. Cf. LI Top. 209.5–211.37.
4. On the topic of a grammar presumably written by Abelard but not preserved for posterity, and of a planned, but presumably unwritten rhetoric, cf. Mews 1985, 92–93.
5. Such an account of Abelard’s procedure can be found in Pinzani 1986, 182.
6. Cf. Mews 1987, 40.
7. Cf. de Rijk 1967, I.186–189 with reference to Dial. 286.31–287.4, and I.196–197. Cf. also Jolivet 1982, 28, 55, 85, 93.
8. Cf. de Rijk 1981b, 19: “Whoever . . . speaks of semantics cannot ignore the questions of ontology and metaphysics.” Tweedale 1998, 8: “The objective grounds for logical relationships lie in what these expressions signify.”
9. Tweedale 1998, 8.
10. In the middle third of the twelfth century, further texts of Aristotelian logic, De sophisticis elenchis, the Topics, and the Analytics, were rediscovered and translated into Latin. Abelard was not particularly interested in these discoveries.
11. See Jolivet 1982, 175–183, esp. 180, and Jacobi and Strub 1995. Abelard was not the only person repeatedly to comment upon the same texts, at times giving cursory explanations, at other times giving explanations in the context of questions of research. Boethius had already proceeded in a like manner, as Abelard knew, and Averroes proceeds in a similar manner as well.
12. Mews 1985, 73–104, 130–132; 1987. Cf. the careful summary of the current state of research in Marenbon 1997a, 40, 43–44, 48, 50. Jacobi, Strub, and King 1996, 36, 39, and n. 39 warn of the necessity for caution in questions of dating: it is quite possible that Abelard held his lectures repeatedly, revising them upon each occasion.
13. See de Rijk 1986, 103–108.
14. See de Rijk 1986, 107.
15. See de Rijk 1986, 116–118, 123; 1996. Cf. also Pinzani (1995, esp. 156) who continues to base his investigations on the old chronology.
16. Clanchy 1997, 103–104.
17. Dial. xiv; cf. also the references given in the Index nominum, s.v. “Dagobertus frater.”
18. The word ingredientibus (in Logica “ingredientibus”) does not mean that this logic is intended for beginners, but that the logic, or course of logic, begins with the Isagoge, that is, with the commentary upon this work (LI Isag. 1.3: “Ingredientibus nobis <!> logicam . . .”).
19. Irène Rosier-Catach has published a series of articles in the last few years which are impressive in terms of their wealth of knowledge and thoroughness. Further investigations by her of the manner in which the study of grammatical tracts influenced Abelard’s philosophy of language are in progress. Cf. Rosier-Catach forthcoming-a,-b, and -c. Pinzani 1995 treats of the relationship between grammar and logic in Abelard in a different, systematic, fashion, and not with an orientation towards historical relations between texts.
20. “Est autem aliud substantivum apud grammaticos quam substantiale apud nos et aliud adiectivum quam accidentale. Illi enim substantiva vocant omnia fixa, etiam illa quae sumpta sunt ab accidentibus, ut femina, vir, miles, nos vero substantialia nulla sumpta ab accidentibus dicimus; et illi adiectiva dicunt tantum ea quae aliis, id est substantivis, per se adiunguntur, ut homo albus, animal rationale, nos vero rationale magis substantiale dicimus quam accidentale. Accidentalia vero omnia sumpta ab accidentibus dicimus, etiam ea quae substantiva sunt, ut vir, mulier.” On the interpretation of this passage, see Kang 1999, 52–53. For nomen sumptum Kang suggests the translation “Gebrauchswort”; cf. 47–60.
22. Abelard uses the word sermo when he is thinking of the meaning and use of spoken expressions.
23. “In scribendo . . . logicam hic ordo necessarius est, ut quoniam argumentationes ex propositionibus iunguntur, propositiones <ex> dictionibus, eum qui logicam perfecte scribit, primum de simplicibus sermonibus, deinde de propositionibus necesse est scribere, tandem in argumentationibus finem logicae consummare, sicut et princeps noster Aristoteles fecit, qui ad sermonum doctrinam Praedicamenta perscripsit, ad propositionum Peri hermeneias, ad argumentationum Topica et Analytica.”
24. Cf. Rosier-Catach 1999.
25. Cf. de Rijk 1980, 143.
26. Cf. for example Tweedale 1976, part IV; Jacobi 1983; Jacobi 1986; and Pinzani 1995.
27. Cf. for example Rosier-Catach 1999. Marenbon 1997a also appears to be inclined toward this evaluation.
28. Aristotle, De in. 7, 17a38–b1. My translation follows that in Spade 1994, x.
29. Cf. the concise presentation in Weidemann 1999, 50.
30. Concerning the concept of impositio, cf. Marenbon 1997a, 178, 183–184, 190–194, 198; de Libera 1981, 70–71; Tweedale 1988, 211; Shimizu 1999; Rosier-Catach forthcoming-c, 2.1. The phrase “to impose a word on a thing” is used throughout this article in order to express more clearly the sense of the Latin.
31. Cf. also Shimizu 1999, 193–194, and Rosier-Catach forthcoming-c, 2.2.
32. Concerning terminology, cf. LI Isag. 16.22–17.28; Spade 1994, 64–72. Abelard is here subtly comparing the relationship of the aforementioned logical terms to the terms used by the grammarians, namely “naming word” (nomen appellativum) and “proper noun” (nomen proprium).
33. Cf. LI Isag. 10.17–13.17; Spade 1994, 23–40. It is advisable to employ the translation in Spade 1994. Spade cleverly divides the discussion into sections by means of paragraphs and subheadings, which greatly facilitates the understanding of an otherwise opaque textual structure. Cf. also LNPS 515.10–522.9.
34. Cf. also LI Isag. 10.6–7; Spade 1994, 21; Jolivet 1982, 89; and Rosier-Catach forthcoming-c, 2.2.
35. De Rijk 1980, 139 and 1981a, 1–3 emphasizes the importance of the change; Mews 1985, 80 considers it simply as an explanatory improvement (cf. also Mews 1987, 16–20). Similarly Jolivet 1982, 25, 27 and n. 52, 69–71; and Marenbon 1997a, 176. Shimizu 1999 distinguishes “stages of theoretical construction” in his well-considered depiction.
36. I have followed Spade’s translation.
37. A suitable translation is “the condition of being a man”; cf. Rosier-Catach forthcoming-c, 2.2.
39. Cf. Weidemann 1999, 54–56.
40. Cf. Jolivet 1982, 94, and Weidemann 1999, 51–52.
41. Weidemann 1999, 52. For Abelard’s answer, cf. LI Isag. 52–54.
42. Weidemann 1999, 52. For Abelard’s answer, see Jacobi 1999 and 2000. Cf. also LI Isag. 18.12–17 (Spade 1994, 79): “Since universal names are certainly not imposed on things according to the difference in their discreteness (for then they would not be common but singular names), and again since universal names cannot name things as agreeing in any thing (for there is no thing they agree in), therefore universal names seem to bring about no signification of things – especially since they establish no understanding of any thing.”
43. Cf. Jacobi 1981, 51–57; Tweedale 1988, 216–217; Marenbon 1997a, 166–167; Rosier-Catach forthcoming-c, 3.3.
44. Cf. LI Isag. 21.28–22.2 (Spade 1994, 102–103), 27.20–24 (Spade 1994, 141), and especially 29.3–7 (Spade 1994, 153): “Universal names . . . name what is discrete, yet they do not do so discretely and determinately.”
45. The terminology of “thing-word” (Dingwort), “property-word” (Eigenschaftswort), and “activity-word” (Tätigkeitswort) was still in use during my time at school.
46. Cf. Marenbon 1999, 202–203.
47. Cf. de Rijk 1986, 87; Jacobi 1986, 149–150. On the critique of the maintained semantic difference between substantives and adjectives see §I.2 above and Rosier-Catach 1999.
48. Cf. de Rijk 1986, 88: “For each and every sememe, e.g. the sememe common to ‘white’ (albus, -um), ‘whiteness’ (albedo) or ‘be-white’ (albere), you may form a noun (whether substantival or adjectival) or a verb. Therefore, Abelard argues, the criterion of lexical meaning for distinguishing between the noun and the verb is unsound.”
49. Cf. de Rijk 1986, 89; and Jacobi 1986, 150.
50. This terminology is familiar to me from school as well. Of course, I only learned later that very different theories lie behind the terms “activity-word” (Tätigkeitswort) and “temporal-word” (Zeitwort), and that from medieval logicians.
51. Cf. Weidemann 1999, 58–59.
52. Cf. Jolivet 1982, 67, 71–72, 76; Nuchelmans 1973, 139–140; Jacobi 1981, 46–52; and Rosier-Catach forthcoming-c, 2.3.
53. Cf. Jolivet 1982, 104, 124, 128.
54. See LI Cat. 112.37–40: “Quantum . . . ad causam impositionis nominis prima et principalis significatio intellectus dicitur, quia scilicet ideo tantum vocabulum rei datum est, ut intellectum constitueret.”
55. Cf. Rosier-Catach forthcoming-c, 2.3.
56. Cf. Nuchelmans 1973, 142; Pinzani 1995, 35, 37, 126; and Kang 1999, 224.
57. On the these descriptions, cf. the very thorough Freiburger dissertation of Sang-Jin Kang (Kang 1999). On the differentiation between an analysis incorporating modi significandi and an analysis which incorporates modi concipiendi, see 219, 224.
58. Cf. Jolivet 1982, 42, 51.
59. Cf. LI Cat. 113.4–11; 115.19–116, 17; 122.29–123, 21; 127.27–40; 138.19–26; 138.31–139.17; 144.28–39; 145.6–24; 150.5–27; 260.23–30.
60. Cf. Kang 1999, 202, 209, 224.
61. According to Abelard’s terminology, the concrete names are “taken from” the abstract (nomina sumpta). Cf. Jolivet 1987, 223; Jolivet 1982, 51, 107–108 and n. 290; Marenbon 1997a, 140–141; Kang 1999, 234 et passim. The logical theory runs here in a direction opposite to the grammatical theory. According to the formation of words, the abstract forms are “named after” the concrete forms; cf. Priscian, Inst. l. IV.
62. Cf. Kang 1999, 42, 76; and Pinzani 1995, 158.
63. Cf. Jolivet 1982, 43, 107, 124.
64. Cf. de Rijk 1981a, 21–24; 1981b, 33–35; 1986, 108–113; Jacobi 1980, 170; 1981, 80; Mews 1985, 86; and Rosier-Catach forthcoming-b, 14, 16, 20, 24–25.
65. Cf. Pinzani 1995, 45.
66. Cf. Pinzani 1995, 129, 158.
68. Cf. Nuchelmans 1973; de Rijk 1975; de Libera 1981; Jacobi 1981, 1983; Jacobi, Strub, and King 1996; and Marenbon 1999.
69. Cf. Jacobi 1981, 52, and Jacobi, Strub, and King 1996, 16.
70. Cf. Jacobi 1981, 52–54, and Jacobi, Strub, and King 1996, 16–17. But see §IV below for an analogy of the meaning of a thing which is indispensable for the theory of the proposition.
71. In the last line, the sequence “without . . ., with . . .” has been corrected. Cf. Rosier-Catach forthcoming-c, 2.3.
74. Cf. Nuchelmans 1973, 142–143. Jacobi 1981, 65–66; and Jacobi, Strub, and King 1996, 18.
75. Cf. Jacobi 1981, 67–68.
76. Cf. Jolivet 1982, 29–30; and Nuchelmans 1973, 144.
77. Cf. Jacobi, Strub, and King 1996, 24–25, 34–38.
78. Cf. Jacobi 1981, 68–73; Jacobi 1983, 90–91; and Jacobi, Strub, and King 1996, 19–20.
79. Cf. Nuchelmans 1973, 144–150.
80. Cf. Tweedale 1976, 234–237; and Rosier-Catach forthcoming-c, 3.1, 3.4.
81. Cf. Jolivet 1982, 77–85; Nuchelmans 1973, 150–156; de Libera 1981; and Rosier-Catach forthcoming-c.
82. Cf. Tweedale 1976, 216–220; 1988, 218–219; and Jacobi, Strub, and King 1996, 31–34.
83. Cf. Tweedale 1976, 237–272; and Jacobi 1985.
84. Cf. Jolivet 1982, 80; Tweedale 1976, 273–281; Jacobi 1999, 86–89; and Lewis 1987, 84–85.
85. Cf. Tweedale 1998, 8.
86. Cf. Jacobi 1985, 12, 23, 36–37; Marenbon 1997a, 221–222; Martin 2001, 98–99, 114–124; and Astroh 2001, 79–81.
87. On the distinction “quantum ad constructionem . . . quantum ad sensum,” cf. for instance Pinzani 1995, 101–105, 109; and Jacobi 1985, 24.
88. Cf. de Rijk 1980, 147; and Jacobi 1999.
91. Cf. Rosier-Catach forthcoming-b, 16.
92. Cf. Jacobi 1980 and 1986, 151, 163, 166.
93. Cf. Tweedale 1976, 285–302; 1982, 145–147; 1988, 218; 1998, 10; de Rijk 1981b, 27–28, 30; 1986, 103; Jacobi 1980, 168, 171–172; 1983, 89–90; and 1986, 150–152.
94. Tweedale 1976, 285–287, 290; and Jacobi 1980, 172.
95. Tweedale 1976, 298–302.
96. Cf. Jolivet 1982, 59–60; Rosier-Catach 1999, 150–152, 155–156; and Marenbon 1999, 199–200, 210–215.
97. More precisely, the matter construes a matrix in which all possible combinations occupy a place: (1) and not (2), (2) and not (1), (1) and (2), neither (1) nor (2).
98. Cf. Kang 1999, esp. 32, 67.
99. Cf. Kang 1999, 67, and 71–72, where he takes the not completely satisfactory phrase in substantia praedicari for essential predication.
100. See III.2.2 above.
101. Cf. Jacobi 1986; and Marenbon 1999.
102. Cf. Jolivet 1982, 57–58.
103. Cf. Nuchelmans 1973, 140–141; Tweedale 1976, 231–232; Jacobi 1986, 146–147; de Rijk 1992, 293–294; and Rosier-Catach forthcoming-b, 19–20.
104. De Rijk 1967 represents an important beginning. Further investigations were presented at the 7th Symposium for Medieval Logic and Semantics, 17–22 June 1985 in Poitiers. The records were published in Jolivet and de Libera 1987. A great step forward was taken by a conference, brilliantly prepared and organized by William Courtenay, on “Twelfth Century Nominalism,” which took place from the 3rd–5th October in Madison. The proceedings of this conference have been published as volume 30 (1992) of the Journal Vivarium. Particularly worthy of mention is the collection of source texts contributed by Yukio Iwakuma and Sten Ebbesen (Iwakuma and Ebbesen 1992). Further scholarship has been presented by Iwakuma or is in preparation. Cf. Iwakuma 1999. Irène Rosier-Catach has published several studies – cf. Rosier-Catach 1999 and the bibliography presented there. Further articles have been made available to me as prepublished texts. Cf. the anonymous commentary on De Interpretatione mentioned in Jacobi, Strub, and King 1996.
105. Cf. again de Rijk 1967. Peter King is in the process of editing an edition that will be published in connection with the new edition of Abelard’s commentary on this text prepared by Jacobi–Strub for the series “Corpus Christianorum, continatio medievalis” (Jacobi and Strub forthcoming). Further editions are being prepared by Corneille Henri Kneepkens and Yukio Iwakuma.
106. The translation of this chapter was undertaken by Dr. Richard Sembera, Ottawa. I am sincerely indebted to him for both his careful attention to detail and his collegiality.