Aristotle on Homonymy: Dialectic and Science

by Julie K. Ward

Overview

In this book, Julie K. Ward examines Aristotle's thought regarding how language informs our views of what is real. First she places Aristotle's theory in its historical and philosophical contexts in relation to Plato and Speusippus. Ward then explores Aristotle's theory of language as it is deployed in several works, including Ethics, Topics, Physics, and Metaphysics, so as to consider its relation to dialectical practice and scientific explanation as Aristotle conceived it.

Product Details

ISBN-13:	9780521128476
Publisher:	Cambridge University Press
Publication date:	02/04/2010
Edition description:	New Edition
Pages:	232
Product dimensions:	5.90(w) x 8.90(h) x 0.70(d)

About the Author

Julie K. Ward is associate professor of philosophy at Loyola University, Chicago. The editor and co-editor of Feminisms and Ancient Philosophy and Philosophers on Race: Critical Essays, she has also contributed to Ancient Philosophy, Polis, and Hypatia.

Read an Excerpt

Cambridge University Press
978-0-521-87486-1 - Aristotle on Homonymy - Dialectic and Science - by Julie K. Ward
Excerpt

Introduction

This work examines homonymy, a topic that lies within Aristotle’s theories of language and predication. In Aristotle’s work, the idea of homonymy is paired with that of synonymy, and in fundamental ways, rests upon it. To English speakers, homonymy is known as a grammatical category referring to the case in which the same word has different meanings, and synonymy, the case in which different words have the same meaning. In contrast, Aristotle finds homonymy and synonymy to be concerned not merely with words, but also, and primarily, with things. As he explains in Cat. 1, synonymy refers to the situation in which two or more things have the same name, or term, and the same defining character (cf. Cat. 1a6–7). For example, we use “animal” synonymously of a human being and an ox in the sense that the term picks out the same characteristic feature in each thing; human and ox share what it is to be animal (cf. Cat. 1a7–8). In contrast, homonymy refers to the case in which two or more things have the same name, but not the same defining character (cf. Cat. 1a1–2). Re-iterating Aristotle’s example, we use “animal” homonymously of a human being and a drawing of one in the sense that they are not animal in the same way, and so, the term “animal” picks out differentproperties in each (cf. Cat. 1a1–4). One way of taking Aristotle’s distinction in Cat. 1 is that homonymy signifies the absence of an identical common character, F-ness, across all instances of things called “F,” whereas synonymy requires it. But drawing the distinction in this fashion leaves open a conceptually interesting possibility, that of an intermediate range between synonymy and its absence. The things lying between synonymy and non-synonymy would include those that have a common term and some, but not all, common characteristics. For, as long as this group did not share all the features in virtue of which they had the name in common, they would prove to be non-synonymous. At the farthest extreme, we would find things having only a common name “F” and no shared characteristic signified by “F.” While this latter group would belong to what Aristotle considers accidental homonymy,¹ and hold no special interest to him, the previous group of things that have the same term and overlapping but not identical defining characteristics he finds worthy of particular philosophical consideration. To this group belong those things that are systematically ambiguous or, as some have it, that fall under the concept of core-related homonymy, which includes such diverse notions as good, one, nature, cause, life, and others.² In addition, as is well known, it is upon the possibility of such a kind that Aristotle founds metaphysics, or the science of being.

   In Western scholarship on Aristotle, ancient commentators such as Alexander and Simplicius as well as medieval thinkers like Albert and Zabarella found Aristotle’s discussions of synonymy, homonymy, and analogy to be of great philosophical interest.³ Later Scholastic and Renaissance philosophers, including Aquinas and Cajetan, gave special consideration to the concept of analogy, which they distinguished as two types:  analogia attributionis and analogia proportionalitatis.⁴ The former kind, “analogy by attribution,” is thought of as holding among things with a common name but having different relationships to the term (cf. Cajetan, De Nom. An. Ⅱ, 8) and, in this regard, approximates Aristotle’s notion of systematic homonymy. In fact, Aristotle’s examples of the medical and the healthy from Meta. Gamma 2 and elsewhere are mentioned throughout as central cases of analogy of attribution by the Scholastics.⁵ The second type, known as “analogy by proportion,” names what Aristotle considers “geometrical” analogy, expressible in an arithmetic proportion, A:B::C:D (cf. EN 1131b12–14).⁶

   Subsequent to the largely theological interest paid by medieval philosophers, systematic ambiguity and homonymy emerged as topics in their own right in mid-twentieth century, culminating with studies by Owen (1960), Lezsl (1970), Ferejohn (1980), Irwin (1981, 1988), and most recently, Shields (1999). Among these contributions, Owen’s work remained pivotal in the sense that, while not always in congruence with its conclusions, other works took their bearing from it.⁷ Typically, modern scholarship devoted to the topic of pros hen relation or systematic homonymy sought to address two conclusions that Owen drew. The first was concerned with a developmental claim that what Aristotle calls pros hen relation was a later invention, absent in the earlier, logical works and appearing only in conjunction with Aristotle’s conception of metaphysics in Meta. Gamma 2. The second concerned the nature of Owen’s interpretation: from his baptism of Aristotle’s notion as “focal sense,” it was clear this was to be understood primarily as a linguistic claim about the inter-relation of meanings. Scholarly work, post-Owen, has tried to show where Owen’s line of thought proved inadequate. In particular, the contributions by Irwin (1981, 1988) and Shields (1999) moved the discussion forward in central, but different, areas of discussion. Shields’ book provided critical impetus by unifying Aristotle’s various discussions about homonymy to a single, comprehensive account, advancing explanations about the ways in which different kinds of homonymy function. Here Shields develops two notions of special interest, core-dependence, and a causal analysis, which he demonstrates by application to central philosophical concepts, such as being, body, life, and one.

   The present book on homonymy seeks to augment recent discussions, particularly aspects of Irwin’s and Shields’ work, by furthering the investigation in some areas and initiating study in others. In brief summary, the present chapters fall into three areas: (1) Aristotle’s account of homonymy in Cat. 1 and its possible precursors, (2) the utility of homonymy for refining premises in scientific arguments, and (3) the application of homonymy to specific concepts. Let me say more about these topics in turn. First, in chapter 1, the question about possible historical antecedents is conjoined with a discussion about homonymy in Cat. 1. For one of the notable features of Cat. 1 concerns its truncated appearance, allowing for speculation that it originally had an introductory section.⁸ The absence of mention about rival accounts to Aristotle’s classification naturally leads to speculation about the origin of the terms. While some terminological similarities suggest connections to Speusippus, Plato’s nephew, more substantial evidence supports a link to Plato, whose influence is evident even in Aristotle’s rejection of his theories about naming and participation in Meta. Alpha 6 and 9, for example. With regard to the second topic, that relating homonymy to scientific investigation, parts of chapters 2, 3, and all of 6 are concerned with setting out the relation of homonymy to dialectical practice and how this activity is conjoined with scientific inquiry and knowledge. These chapters focus on the underlying unity of homonymy, distinguishing what I find to be the positive and negative aspects of homonymy. Across this discussion, homonymy begins with the postulation of to posachōs, asking in “how many ways” a word is being used, as the first step in the inquiry. So construed, I consider homonymy as a means for investigating terms and concepts arising from the method of dialectical argument, which is also applicable to the process of scientific inquiry. Finally, with respect to the third topic, the middle chapters of the work, chapters 4 and 5, consider how the method of homonymy is used with regard to some lesser discussed philosophical notions. More specifically, the work takes up an examination of nature, being, and friendship, topics that have not been adequately dealt with in other studies of homonymy.

   The present discussion about the truth-seeking function of homonymy inevitably leads to the broader question concerning the relationship between dialectic and science. The ongoing scholarly debate has focused on the nature and aim of the two disciplines and, specifically, whether dialectic can aid us in providing scientific knowledge (epistēmē ). The objects of scientific knowledge are invariant, necessary, and real, and the exposition of such knowledge is demonstration (apodeixis, cf. An. Po. 71b15, 73a22). As he elaborates in An. Po. A 2–4, demonstration is a kind of inference with true, non-dialectical premises,⁹ and it describes “what cannot be other than it is” (An. Po. 74b5–6). Yet since demonstration is not an investigative tool, it seems to fall to dialectic to provide the means for obtaining the premises for such arguments. However, Aristotle’s descriptions of dialectical practice, both in Topics and Rhetoric, appear to render it of no use to scientific inquiry. One construal of the problem posed by Irwin (1988) depends on two types of dialectic, what he terms “weak” and “strong,” as distinguished by their functions.¹⁰ On his view, weak dialectic is used to frame preliminary definitions of a thing based on endoxa, “what is accepted.” Since weak dialectic lacks a method of winnowing true from false endoxic propositions, it cannot establish their truth independently; it only shows which endoxa are self-consistent.¹¹ As this method leaves a gap between dialectic and scientific knowledge, Irwin supposes that texts like Meta. Gamma 2 employ another kind, “strong” dialectic, which possesses a rigorous means for screening endoxa so as to achieve a correct subset of endoxa for use in scientific arguments.¹² The study of homonymy presented here does not aim at directly answering the conflict of method that Irwin and others have posed. Instead, it aims to provide the lines of continuity between the negative method of homonymy in the earlier works and the positive, synthetic method utilized in the later, mature works. Depending on the nature of the work in which it appears, rhetorical or metaphysical, I find that the method of homonymy can be put to use for purely argumentative or for philosophical ends. However, there is a common thread across these various contexts in which homonymy appears: Aristotle’s practice of asking “how many ways” (posachōs) something is being said is present in both the negative and positive uses of homonymy. Having the benefit of Shields’ analysis concerning core-dependent homonymy, the present work examines the way in which systematic homonymy constitutes an ampliative method of inquiry, able to illuminate meaningful inter-relations among characteristics referred to by a common term.

   The plan of investigation for the work consists in opening the consideration of homonymy by examining the account in Cat. 1, where Aristotle introduces the concept as part of a three-way distinction including synonymy and paronymy. The extreme brevity of Cat. 1 and the lack of discussion concerning its terminology invite speculation concerning the possible sources of influence for the three-way division. As I noted, two contemporaries of Aristotle, Speusippus and Plato, recommend themselves as likely sources of influence. Speusippus’ division of names is preserved for us through Simplicius’ commentary on Aristotle’s Categories, where we find that Speusippus made use of the same three terms Aristotle employs. Following an examination of Aristotle’s account of synonymy and homonymy in Cat. 1, we examine Speusippus’ account of names, and then turn to consider Plato’s views of naming and participation, centering on his theory of eponymy in the middle dialogues where certain features suggest Aristotle’s own theory of core-related homonymy.

   The study proceeds in chapter 2 with an examination of the nature and uses of homonymy from the standpoint of the dialectical practice described in Topics. In addition to discussing the notion of a truth-seeking aspect of dialectical practice that Aristotle raises in Top. Alpha 2, I also examine certain standing objections about the nature of dialectic in relation to science. I find it significant that Aristotle uses a specific tool of dialectic, homonymy, as the means for the refinement of terms and testing of ambiguity in Top. Alpha 15. Developing this notion in chapter 2, I raise the possibility that this kind of concept refinement may be utilized for substantive ends, such as constructing overlapping accounts of homonymous things, as well as for discerning simple ambiguity among common terms. Taking up this line of thought, subsequent chapters move to consider the positive role of homonymy outside Topics. Chapter 3 is primarily concerned with discussing what is termed “core-dependent” homonymy using the standard cases of the medical and the healthy from Metaphysics Gamma and Zeta. Chapters 4 and 5 investigate the extent to which core-dependent homonymy may be seen at work in the discussion of Aristotle’s central philosophical concepts, including nature, friendship, and being.¹³

      Overall, the mid-section of this study, chapters 3 through 5, has two aims: to set out and modify Shields’ account of how the causal analysis of core-dependent homonymy works and to utilize the causal framework of core-dependent homonyms so as to examine some concepts central to Aristotle’s philosophical discussions.¹⁴ With regard to the latter aim, the present study adds to Shields’ recent work by extending the scope of the concepts under investigation and by coming to different conclusions about the homonyms he investigated, as is the case in the core-dependence of being.

   Finally, the analytical method of homonymy typified in the arguments of Top. A 15 allows us to find the continuity with its use in philosophical contexts. Chapter 6 is concerned with examining how homonymy functions in relation to scientific investigation, specifically, the search for and formulation of scientific premises. By attempting to make evident the investigative and testing aspects of homonymy, the chapter suggests that homonymy can make a positive contribution to such inquiry. For, homonymy as method provides core-dependent analyses of homonymous things which elucidate their systematic organization. The description of their organization may then be used to advance our understanding about how the features signified by the terms are inter-related. In this way, we can refine our use of the common terms, or predicates, employing them in arguments in such a way as to preserve the natural systematization of things.

   Two other issues are raised in the course of this work. The first is about whether homonymy should be understood as a theory about words or about extra-linguistic things. While Cat. 1 distinguishes synonymy and homonymy by reference to ways that “words” (onomata) signify in different ways and Topics often refers to homonymy as a device for analyzing terms and their meanings, these examples are misleading. For Aristotle holds that spoken or written words ultimately signify things, and not other words (De Int. 16a3–8). Similarly, when Aristotle employs the method of homonymy in Meta. Gamma 2 in relation to being, it is clear that he is drawing conclusions about the ways in which extra-linguistic items, such as substances, qualities, or relations, are metaphysically, not semantically, related. Thus, the detection of homonymy as well as its use across philosophical discussions has to be regarded as implying that the things so named as homonymous are real, extra-linguistic entities. Since Aristotle holds that spoken or written signs ultimately refer to extra-linguistic items, it should occasion no surprise to find that homonymy is used sometimes in relation to language and other times in relation to the investigation of extra-linguistic, or real, entities, some of which are the subject matter of scientific study.

   The second, and larger, issue here concerns the question about developmentalism in Aristotle’s work. It is perfectly reasonable to hold that Aristotle’s works reflect theoretical changes across texts where the same topics appear; nonetheless, the purely negative aspect of a developmental approach is not being stressed in this work. So, although the appearance and function of homonymy as a method are not the same in Topics and  Metaphysics, they are not unrelated either. Rather, I find that Aristotle’s account of homonymy constitutes a single theory, one that undergoes refinement and extension. Consequently, instead of posing a discontinuity of theory, this work will attempt to show where the method of homonymy begins as a testing device for non-synonymy and gets developed into a full-fledged tool for philosophical speculation. In this regard, I find that the account of homonymy in the early works constitutes the first stage of the full method appearing in later texts. In brief, I find that homonymy as it appears in Topics is primarily a tool for negative dialectical practice, and this tool develops into a method for mapping related instances of the same concept for scientific, philosophical ends.¹⁵ As well, it allows Aristotle the theoretical framework to criticize Platonic metaphysics and to avoid regress problems such as the Third Man.¹⁶ In this respect, the full method of homonymy is adequately plastic so as to allow its use in varied contexts, from detecting the lack of synonymy in Topics to framing a science of being in Metaphysics.

---

1

The Theory of Homonymy in Categories 1 and Its Precursors

1.0 Overview

Aristotle makes use of the concepts of homonymy and synonymy in various contexts, such as detecting ambiguity in dialectical arguments, formulating preliminary scientific definitions, and describing relations among existing things.¹ But he comes to the point of giving definitions of the concepts only once, and this occurs in the first chapter of the Categories. It seems reasonable, then, to begin the examination of homonymy with the account Aristotle gives in Cat. 1. The chapter, running only fifteen lines in the Greek text, proposes a tripartite distinction among homonymy, synonymy, and paronymy. Overall, the aim is to differentiate the ways in which a common term or a set of related terms can be said to signify something.² Put briefly, homonymy refers to things having the same name and different definition; synonymy, to things having both the same name and the same definition; and paronymy, to terms related by their inflected ending. The bare bones of the three-way account may be set down easily. Yet certain interpretive issues concerning the precise lines of the account, as well as the overall scope and nature of Cat. itself, remain subjects of debate.

   One long-standing problem arising from the Greek commentators concerns whether Categories is primarily a logical or metaphysical work.³ The question about the nature of the work as a whole bears on the interpretation of Cat. 1 in the sense that taking one position rather than another would incline us to a more (or less) expansive understanding of homonymy or synonymy. For example, if we were to consider the Cat. as concerned with marking logical distinctions, the account of homonymy in Cat. 1 would more likely be regarded as linguistic in nature. In this case, the account would appear less germane to the mention of homonymy in relation to being in Meta. Gamma 2 or Zeta 1, for example. However, the original dichotomy between a logical and metaphysical discussion is unnecessary, as subsequent chapters, such as Cat. 2, make clear. There, Aristotle’s distinctions among things that are “said of” a subject and those that are “in” a subject subsume items in both categories.⁴ As the work thus resists being classified under a single kind, we may avoid the dichotomy suggested by the ancient commentators. By taking Cat. as offering metaphysical, as well as logical, insights about words and their referents, we find a basis for comparison with similar theories in the mature works, such as the Physics or Metaphysics. By providing a basis for the comparison, we may then move to consider the more central issue about the nature of Aristotle’s theory of homonymy, specifically, whether the theory changes in fundamental ways across his works.

   Briefly said, we may ask whether Aristotle’s account of homonymy sketched in Cat. 1 and utilized in early works, such as Top. Alpha 15, differs fundamentally from that which is articulated in relation to to on, or being, in Meta. Gamma 2 and Zeta 1. My own hypothesis is that what appears in the later works is, at base, a more complex version of what appears in the Organon. So, in what follows, that which we shall come to call systematic homonymy (based largely on its appearance in Meta. Gamma and Zeta) should be regarded as a refinement of the simple account offered in Cat. 1 and Topics. Thus, I find Aristotle’s later utilization of homonymy to be an extension, not a reversal, of the theory in Cat. 1.

   The continuity hypothesis running behind the theory might be seen as resting on two different interpretive principles about how to approach Aristotle’s writings. While the traditional ordering of his works into an earlier and later group is being accepted, it also seems reasonable to look for continuities between them, ignoring the difference in their dating. Since the present works are based on redactions of works likely to have been revised over Aristotle’s lifetime, it seems perfectly justified to observe the whole corpus in order to make comparative observations about the theories, particularly that of homonymy. Moreover, in what follows, I make theoretical continuity across the earlier and later works a priority insofar as the theory of homonymy as a method is concerned.

---

© Cambridge University Press

Table of Contents

1. Aristotle's theory of homonymy in Categories and its precursors; 2. Homonymy in The Topics; 3. The concept of related homonymy; 4. The homonymy of being; 5. Physis, Philia, and homonymy; 6. Homonymy and science.

Editorial Reviews

Julie Ward's volume is a valuable study, on a subject that needs further inquiries. One cannot but be engaged with Ward's fine and thorough analysis of Aristotelian homonymy. —BMCR

From the Publisher