261 92 2MB
English Pages [367] Year 1998
Page i
Peirce, Pragmatism and the Logic of Scripture Charles Peirce, founder of the American philosophic movement of pragmatism, offered the first logical critique of all the philosophies of "modernism" as failed attempts to replace the philosophic—religious systems of medieval Europe with systems of reason, alone. For Peirce, contemporary philosophy's task is not to replace societal norms, but to disclose and clarify modern society's deeper resources for correcting itself. In his later work, Peirce discovered that, against his own principles, his earlier doctrine of pragmatism had, itself, sought to replace rather than help correct the misguided reasonings of modern philosophy. He coined the term "pragmaticism" for his concluding studies of how to restore pragmatism to its task of helping repair, rather than recreate, the norms of modernity. This book is the first attempt to interpret Peirce's pragmaticism as a method of correctively rereading his earlier philosophic writing. Peter Ochs reinterprets pragmatism itself as a method of correctively rereading the errant philosophic writing of the great modern thinkers. As corrected by the pragmatists, the task of modern philosophy is, through writing, to diagram the otherwise hidden rules through which modern society repairs itself. Peirce labeled this elemental writing "enscribing" or "scripture." Redescribing Peirce's pragmatism as "the logic of scripture," Ochs suggests that Christians and Jews may in fact reread pragmatism as a logic of Scripture: that is, as a modern philosopher's way of diagramming the Bible's rules for repairing broken lives and healing societal sufferings.
Page iii
Peirce, Pragmatism and the Logic of Scripture Peter Ochs Edgar Bronfman Professor, The University of Virginia
Page iv
PUBLISHED BY CAMBRIDGE UNIVERSITY PRESS (VIRTUAL PUBLISHING) FOR AND ON BEHALF OF THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE
The Pitt Building. Trumpington Street, Cambridge CB2 1RP, United Kingdom CAMBRIDGE UNIVERSITY PRESS
The Edinburgh Building, Cambridge CB2 2RU, United Kingdom 40 West 20th Street, New York, NY 100114211, USA 10 Stamford Road, Oakleigh, Melbourne 3166, Australia © Peter Ochs 1998 This edition © Peter Ochs (Virtual Publishing) 2001 This book is in copyright. Subject to Statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1998 Printed in the United Kingdom at the University Press, Cambridge Typeset in Lexicon No 2 8.75/13.25 pt (from the Enschedé Font Foundry), in QuarkXPress [SE] A catalogue record for this took is available from the British Library Library of Congress cataloguing in publication data Ochs, Peter, 1950– Perice, pragmatism, and the logic of Scripture / Peter Ochs. p. cm. Includes bibliographical references and index. ISBN 0 521 570417 (hardback) 1. Peirce, Charles S. (Charles Sanders), 1839–1914 2. Pragmatism—History. 3. Hermeneutics. 4 Bible Hermeneutics. 5. Ethics in the Bible. I. Title. B945.P44024 1988 144.3—dc2l 9853323 CIP 0521570417 hardback
ISBN
eISBN 0511003978 virtual (netLibrary Edition)
Page v
for Vanessa Ochs tzadkah mimeni
Page vii
CONTENTS List of Abbreviations
viii
Part I Peirce's Pragmatic Writing
1
1 Introduction: Reading Peirce's Pragmatism
3
2 Pragmatic Methods of Reading and Interpretation
20
3 Problems in Peirce's Early Critique of Cartesianism
51
4 Problems in Peirce's Early Theory of Pragmatism
74
5 Problems in Peirce's Normative Theory of Pragmatism, 1878–1903
104
6 A Pragmatic Reading of Peirce's Lectures On Pragmatism
128
Part II Peirce's Pragmaticist Writing
159
7 Irremediable Vagueness in Peirce's Pragmaticist Writings: A PlainSense Reading
161
8 Pragmaticism Reread: From CommonSense to the Logic of Scripture
246
Notes
326
Index
358
Page viii
ABBREVIATIONS The following abbreviations are used for some references in the main body of the text and in the endnotes.
(4.656)"
Collected Papers of Charles Sanders Peirce, Charles Hartshorne, Paul Weiss, and A. Burks, eds. (Cambridge, Mass.: Harvard University Press, 1935, 1958). References are to volume and paragraph number of this series; thus "4.564" is volume 4, paragraph 564.
"Consequences"
"Consequences of Four Incapacities," Journal of Speculative Philosophy, 2 (1868): 140–57. Repr. 5.264–317.
"Fixation"
"The Fixation of Belief," Popular Science Monthly, 12 (1877): 1–5. Repr. 5.358–87.
"Grounds"
"Grounds of Validity of the Laws of Logic: Further Consequences of Four Incapacities," Journal of Speculative Philosophy, 2 (1868): 193–208. Repr. 5.318–357.
"How to"
"How to Make Our Ideas Clear," Popular Science Monthly, 12 (1877): 286– 302. Repr. 5.388–410.
M
The Charles S. Peirce Papers, microfilm edition (Harvard University Library, Photographic Service, 1966). References numbered according to the system developed by Richard S. Robin, Annotated Catalogue of the Papers of Charles S. Peirce (Amherst: University of Massachusetts Press, 1967), and as supplemented by Robin in "The Peirce Papers: A Supplementary Catalogue," Transactions of the Charles S. Peirce Society, 7 (1971): 37–57. For example, "M28: 2" means Robin's catalogue listing 28, page 2.
Page ix
"Questions"
"Questions Concerning Certain Faculties Claimed for Man, "Journal of Speculative Philosophy, 2(1868): 103–4. Repr. 5.213–63.
Trans.
Transactions of the Charles S. Peirce Society (journal).
W
Writings of Charles S. Peirce: A Chronobgical Edition, Max Fisch et al., eds. References are abbreviated as "W" followed by volume and page numbers (for example, W 3.54). Vol. 1—1857–1866, Max Fisch, gen. ed. (1982) Vol. 2—1867–1871, Edward C. Moore, gen. ed. (1984) Vol. 3—1872–1878, Christian J.W. Kloesel. gen ed. (1986) Vol. 4—1879–1884, Christian J.W. Kloesel. gen ed. (1986) Vol. 5—1884–1886, Christian J.W. Kloesel. gen ed. (1993)
Note to the Reader Peirce delivers some of his most dramatic arguments through some of his more technical explorations in mathematics and logic. For the sake of interested readers, I reproduce and analyze some of those explorations, beginning in chapter 5, but I mark the more technical sections with asterisks and present them in a smaller font. More general readers may want to pass through these sections, getting a sense of their contributions without getting bogged down in too many details.
Page 1
PART I— PEIRCE'S PRAGMATIC WRITING
Page 3
Chapter 1— Introduction: Reading Peirce's Pragmatism Even though many of his contributions to the discipline may still appear idiosyncratic, Charles Peirce's name no longer needs introduction to an audience of professional philosophers. For many years after his death in 1914, Peirce's largely unpublished works were known only to a band of devotees—or known only secondhand through the works of those he influenced, such as William James, John Dewey, Charles Hartshorne and Paul Weiss. Now, however, Peirce's innovations in the theory of language and of signs (he called it semeiotic 1 ), in phenomenology, in the logic of relations, in the philosophy of mathematics and in pragmatism have been the subject of dozens of scholarly books, hundreds of essays, and have begun to enter into the standard curricula of graduate schools in philosophy. Beyond the profession, an increasing number of literary scholars, hermeneuts, theologians, and postmodern theorists of various descriptions find themselves surveying the secondary literature, and also venturing into Peirce's labyrinthine corpus. I address this study of Peirce, first, to members of this latter group. I imagine they may have turned to his work for some of the reasons I first did, twenty years ago, when I was studying the philosophy of rabbinic thinking, as displayed in the literature of classic Judaism: the Talmud and the midrash in particular (the collections of rabbinic interpretations of biblical narrative and biblical law that were compiled, roughly, in the 1st6th centuries, in what were then Babylonia and Palestine). I found it difficult to map out meaningful units of rabbinic discourse with the instruments offered up by modern textbooks in the philosophy of religion. The most helpful guide was Max Kadushin, a littleknown rabbinic scholar at the Jewish Theological Seminary tutored in early twentiethcentury "organicist" approaches to social science and in Mordecai Kaplan's Jewish pragmatism. Kadushin's analyses of the quasilogic of rabbinic thinking were vague enough to adjust to rabbinic discourse but too vague to be fully convincing. In one tantalizing footnote, however, he also cited Peirce's contribution to his thinking. That was enough to send me off to graduate work on Peirce and related thinkers. I brought with me interests I imagine other readers may share: a
Page 4
concern to find reasonably precise ways of talking about imprecise things without losing the meaning of the imprecision itself; a belief that phenomena of everyday language, including the everyday practices of religion, are among those things; a love of critical reasoning but an acquired distrust of criticism that has lost sense of having a purpose; disillusionment with ''modern" or Enlightenment attempts to make a metaphysics—and also a religion—out of the rational critique of inherited traditions of knowledge and practice; a conviction that postEnlightenment antirationalism—including romanticisms, emotivisms and a variety of totalizing ideologies of power, history, experience and so on—may prove, logically, to be the other side of the rationalist coin; fascination with the irrevocable contextuality—temporal, historical, linguistic, social, biological—of rational judgments, but also with their persistent "rationality" (to be identified through this study with such powers as "the capacity to be diagrammed in a system of Existential Graphs" and as "the capacity to guide the repair of problematic modes of conduct"). Readers who share these concerns may expect Peirce's various inquiries to contribute significantly to their work. If, however, they have gone past the secondary studies of Peirce, they may also discover how difficult it is to understand his contributions. So much formal detail and all those tantalizing generalizations about history, meaning, truth and relativity, but couched, in their details, in abstruse mathematical allusions! And, for those who have gotten past the details, all those misleading changes in Peirce's arguments from one period of work to the next! All those apparent contradictions, vague concepts, incomplete arguments, and unfinished essays! In the face of these challenges, my goal is to offer readers who share the interests I have just mentioned a way of reading through the contradictions and vagueness of Peirce's writings, to hear his responses to their concerns. Toward this goal, I have written about the single thread of inquiry that responds most generally to all the concerns I have mentioned: Peirce's "pragmatism." This term will be redefined through the study, but, for now, it will refer to his theory about how to correct inadequate—because overly precise—definitions of imprecise things. For Peirce, the prototypically imprecise things are matters of fact. The prototypically inadequate way modern philosophers define matters of fact is to define them with too much precision. Philosophers may also sin in the other direction, however, by overreacting against philosophic precision and concluding that matters of fact cannot really be defined at all. Differently put, the sin of overprecision is rationalism; the sin of underprecision is irrationalism. Peirce's response is to offer pragmatic definitions of imprecise things. My thesis is that pragmatic definition is not a discrete act of judgment or classifica
Page 5
tion, but a performance of correcting other, inadequate definitions of imprecise things. Pragmatic reasoning is thus a different sort of reasoning than the kind employed in defining things precisely. It is a corrective activity. Pragmatic reasoning displays the pragmatic rationality of things in the way it corrects our abstract definitions of the rationality of things. My thesis is therefore not a thesis in the usual sense. Since my claim is that to define pragmatically is to correct and that to correct is to read, my "thesis" is better named my "corrective reading." But that is not quite right, either, since my claim is that reading cannot be done "in general," or "for anyone," but only for someone: for some community of readers. My thesis about Peirce's pragmatism is thus a corrective reading of his pragmatic writings for some community of readers. And this is not to correct Peirce per se but to correct problems in the way Peirce would be read by a given community. The point is not that Peirce is wrong and I can see better! Not at all. Only that his pragmatism can show itself to another thinker only in the way that thinker acquires the practice of corrective reading. The only way to learn pragmatism from Peirce's own writings is thus to read those writings correctively. It will take us many chapters to share a sense of what "correction'' means. To exhibit the meaning of pragmatism will therefore be to perform someway of correcting the meaning of pragmatism. For this study, I read Peirce's writings on pragmatism as his corrective performance of pragmatism, and I offer the following chapters as one way of pragmatically and thus correctively studying his performance. Since pragmatic reading can be done only for some community of readers, I cannot presume that my study will make sense or be of interest to everyone. I therefore address different levels of reading to different communities of wouldbe readers. In this chapter, I describe my method of reading in a way that may be helpful, in particular, to literary and scriptural text scholars who are not specialists in philosophy. In chapter 2, I redescribe my method, at greater length, for philosophers and Peirce specialists. The Pragmatic Method of Reading: A Rabbinic Analogue In short, this book offers a textualrhetorical study of Peirce's four published series of writings on pragmatism (from 1867–8, 1877–8, 1902–3 and 1905–6). My method of study is to read through problems in these writings with my intended readers until I can imagine our having gained some shared sense of how Peirce would correct himself and, thus, of what pragmatism means. For text scholars, it may make most sense to describe my pragmatic method of reading as if it were modeled on the classic methods of rabbinic scriptural interpretation. On one level, I offer a straightforward explication of what I call the "plain sense" of Peirce's pragmatic writings. I take the term "plain sense" from
Page 6
the exegetical practice of medieval Jewish scholars, for whom the "plain sense" (peshat) of a text is its meaning within the rhetorical context of some body of received literature. Here, "plain sense" is contrasted with "interpreted sense" (derash), much in the way we might contrast textual exposition with hermeneutical or performarive use of a text—provided that we do not grant epistemological authority to one sense over the other. This epistemological model comes from the premedieval scholars or "rabbis" of the Talmud, for whom, as I read them, the lived meaning of a scriptural text will be found in its derash, but only when the derash is itself performed within the grammatical, philological and semantic rules of the peshat: as the Talmud says "the scriptural text must not be deprived of its plain sense.'' 2 To identify the plain sense of Peirce's writings, I assume that he says what he means, and I explicate his arguments according to the contextual meaning of his philosophic terminology.3 On the level of plainsense reading, however, I also find that each of Peirce's arguments on behalf of pragmatism is equivocal, confused, or contradictory. I interpret these burdens as warrants for conducting a secondlevel interpretation of Peirce's writings. I call this secondlevel interpretation "pragmatic reading," or the philosophic equivalent of the rabbis' derash. In the Talmudic literature, the rabbis tend to move from plain sense to interpretive reading only when something burdensome in the plain sense stimulates them to do so: some apparent contradiction (stira) or textual difficulty (kashia). Because the plain sense in question is the plain sense of Scripture (torah) as God's revealed word, the rabbis assume that the textual burden is merely apparent and that the nonburdensome meaning of a given passage will be disclosed through further "searching out" (derasha: "interpreted meaning," or the result of "searching out"). Unlike ancient allegorical or esoteric readers—such as the Essenes or Christian Gnostics or Jewish Platonists—the rabbis do not believe that any single activity of searching out could disclose a meaning that was not apparent in Scripture's plain sense. The nonevident meaning of a burdensome passage would be disclosed only through the indefinite giveandtake of past, present and future readings. Otherwise put, a particular interpretive meaning for a particular passage could be identified only for a particular reader in a particular context. New contexts of interpretation would disclose new aspects of meaning. On this model, I search after a pragmatic interpretation of Peirce's writings only when burdens in the plain sense of his texts stimulate me to do so. In these contexts, I believe that pragmatic reading will reveal a meaning otherwise hidden in the plain sense, but I do not believe that any one activity of interpretation will suffice to reveal that meaning. My own pragmatic reading, for example, will at best disclose the meaning of some burdensome text of Peirce's for some particular context of inter
Page 7
pretation. This reading would not preclude other readings for other contexts, although it could also contribute to them. The success of the reading would be judged by how well it resolves the given problem in Peirce's writings for a given community of interpreters. Part I of this study treats three series of Peirce's strictly "pragmatic," as opposed to "pragmaticist" writings on pragmatism. Part II treats in more detail one reconstructed series of Peirce's post1905 writings on what he then called "pragmaticism,'' as opposed to his earlier "pragmatism." Peirce introduced his pragmaticism as a way of correcting what he considered his peers' (and his own) misinterpretations of pragmatism. By the middle of part II, I will suggest that Peirce's pragmaticism may be interpreted as a pragmatic method for rereading and thus correcting equivocal, confused or contradictory texts of pragmatism—Peirce's and others'. According to that rereading, I will suggest that Peirce's pragmatism may be interpreted more generally as a program for rereading equivocal, confused or contradictory philosophic texts. I will add that pragmatic arguments may appear confused when they are not interpreted in the context of such programs. In terms of the rabbinic model of interpretation, I am thus depicting Peirce's pragmatism as a hermeneuticalandperformative interpretation of burdensome philosophic texts; pragmatism is relabeled "pragmaticism" when the texts in question are texts of pragmatic philosophy. This analogy between rabbinic and pragmatic modes of interpretation has strengths and weaknesses. One strength is that the rabbinic model counters persistent tendencies among some philosophic scholars to reread pragmatism as yet another epistemological theory or model: one of several, competing pictures of how humans know the world. The counter reading is to suggest that, like the rabbinic interpreter of Scripture, the pragmatic reader is not in the business of constructing possible models, but only of offering corrective readings of existing claims, read as texts. The readings are guided by interpretive principles that are shared by some particular community of interpreters for whom these texts offer significant, but somehow confused instructions about how to conduct some practice. According to these interpretive principles, the texts' confusions or contradictions are to be reread as signs of recondite or suppressed meanings whose disclosure would both clarify the texts' meanings and contribute in some significant way to the life of the community. Since there is, however, no single way to reread the confused texts, the rereading may amplify but never supersede the original text, from which this or other communities may derive other readings in the future. A second strength of the analogy is that it may have a prototype in early Hellenistic hermeneutics. As I will discuss at the end of part II, however, I can refer to such a prototype only in
Page 8
the context of offering a communityspecific, pragmatic interpretation of Peirce's lifetime of pragmatic inquiries. One weakness of the analogy is that the rabbinic interpretation is addressed to a community of rabbinic practitioners, for whom a scriptural text offers instructions about how to criticize and correct all dimensions of everday practice. Peirce's pragmatic readings, on the other hand, are addressed to a community of empiricist cumpragmatic philosophers, for whom pragmatic writings offer instructions only about how to criticize and correct the empiricist philosophic practices. The pragmatist's work is thus more circumscribed. Furthermore, the rabbinic interpreter claims to search out the intentions of a text's divine author, which is to exhibit the foundational principles of a particular people's created order. The pragmatic reader, on the other hand, claims to search out the intentions of a text's human author, which is to exhibit the principles of interpretation according to which one person attempted to make sense out of and refashion a given world. In this sense, the pragmatic reading of a text is also a study of human character and its relation to a public practice. We cannot generalize the work of pragmatic reading too far beyond the scope of the public practice of western, academic philosophy. Peirce belonged to an empiricist practice of philosophy. This means that his community of interlocutors were empiricists (Wright, James, Mill and so on) and that he addressed traditional philosophic problems as they were posed by the tradition of inquiry that links Descartes to Locke, Kant, Mill etc. Within that practice, he was dissatisfied with a tendency of empiricist argumentation he labeled variously Cartesianism, nominalism, seminary thinking, a priorism and individualism. Each of Peirce's writings on pragmatism is an attempt to identify and correct this errant tendency. What we call "pragmatism" refers to any of his attempts to define the method of correction. The problem is that Peirce displayed these errant tendencies, as well, in his own philosophic inquiries and in his pragmatic critique. As a result, his pragmatic writings challenge not only his peers' but also his own intellectual proclivities. These writings are therefore remarkably complex and, paradoxically, unpragmatic. Each of them may be read as both a philosophic critique and a document of Peirce's developing philosophic selfunderstanding. I say "developing," because Peirce was neither lucidly selfaware nor immune to his own criticism. Arranged chronologically, each of the writings may be read as a critical commentary on the previous one, identifying as a philosophic error one of the previous writing's dominant principles of interpretation. Each of the writings therefore displays the progressive development of Peirce's own pragmatic tendencies and the progressive weakening of his unpragmatic tendencies. 4 The culmination of this development is
Page 9
Peirce's post1905 series of writings on pragmaricism: reinterpreted in this study as his attempt to reinterpret pragmatism as a method of correctively rereading the confused, equivocal or contradictory texts of empiricist philosophy. In part II, I argue that these writings are no longer burdened by such confusion, but only by a persistent vagueness that calls for yet another kind of pragmatic reading. This is a species of definitional, rather than of corrective reading: one that, for some particular community on some particular occasion, defines the meanings of inveterately vague symbols. I suggest that, in the tradition of great wisdom literatures, Peirce's pragmaticist writings defer the activity of completing their definitions or meanings to some other occasion: prototypically, this means the occasion of some community's reading them for some particular purpose. According to the method of pragmatic reading I employ and examine in this study, vague symbols may be defined only with respect to particular contexts of interpretation ("interpretants," in Peirce's vocabulary). The resulting definitions display, nonetheless, a species of indeterminacy that would enable interpreters to draw lessons from one context to another. This phenomenon is significant, since it discloses a way of generalizing the results of pragmatic inquiry without transgressing the limits of contextspecific interpretation and, thus, without recourse to the universalisms (or "foundationalisms") the pragmatists criticized. In chapter 7,I examine Peirce's attempts to write a logic of this species of indeterminacy: what he calls, variously, the logic of vagueness or of relations or of signs, and what he frames, variously, in terms of a predicate calculus or of a "mathematical" system of existential graphs. My descriptions of these logics belong to a plainsense reading of Peirce's pragmaricist writings. The plain sense of these logics is also irremediably vague, however, which means that it also stimulates a pragmatic reading that gives it definition within the reader's context of interpretation. I tend to identify the latter with the reader's "community of interpretation." At the close of part II, I offer a series of such pragmatic readings of Peirce's logics of vagueness, and so on, appropriate to a series of increasingly defined and thus narrow communities of interpretation. I assume my plainsense readings of Peirce's writings are appropriate to any community of Peirce readers, but that my pragmatic readings will interest only those who find themselves included in my characterizations of the communities of readers I address. These are not necessarily selfdescribed communities but, rather, communities of wouldbe readers, who would read Peirce with respect to given sets of assumptions. The only way I atttribute some pragmatic reading to "Peirce'' (as author of his writings) is to attribute it to some community of wouldbe readers within which a given proportion of scholars would tend to include Peirce. I assume
Page 10
this proportion would be greater for more broadly defined communities (such as the "pragmatic empiricists") and less for more narrowly defined communities (such as "Christian (or Jewish) Scriptural pragmatists"). There are literaryhistorical warrants for including within such communities the authors of a series of philosophic hermeneutic writings which, if placed in reverse chronological order, would link certain of Peirce's writings to those of primordial communities of Bible scholars. I make no historical claims about such hermeneutical linkages, however, but only pragmatic claims: that such linkages may help certain small communities of contemporary pragmatic scholars solve certain philosophichermeneutic problems. I also claim that the domain of readers who would read a certain, vague text in a certain way defines the generality or breadth of that text's vagueness. This means that the way I refer various pragmatic readings of Peirce's vague texts to various domains of readers illustrates the logic of vagueness that Peirce presents in those texts. What Texts I Am Examining, to What Degree of Precision Going against the thrust of recent Peirce scholarship, I examine the more famous essays that Peirce published in his lifetime and/or that have been since published in the Collected Papers of Charles Sanders Peirce. For Peirce scholars, this topically arranged selection of writings, edited by Paul Weiss and Charles Hartshorne, has been recently superseded by the Peirce Edition Project's Writings of Charles S. Peirce: A Chronological Edition, founded by Edward Moore, with Max Fisch as general or, currently, senior editor, and Christian Kloesel as current director. 5 Hartshorne and Weiss published six volumes of edited selections on all of what they considered Peirce's major areas of philosophic and mathematical/logical inquiry (in 1958, two volumes were added on scientific writings and on correspondence and reviews, edited by Arthur Burks). The Peirce Edition Project, on the other hand, is publishing a comprehensive collection, without editorial changes or topical rearrangements, according to the chronology of Peirce's work. Both collections draw from the manuscript collection of the Charles S. Peirce Papers, housed at the Houghton Library of Harvard University and now generally available on microfilm. The vast majority of Peirce's writings were unpublished in his lifetime. The Peirce Edition Project will remedy that situation, but not yet: it has currently completed volume 5, 1884–6, with completion of all its approximately fortyseven volumes anticipated in about twenty years (volumes 26–7, including the 1905 Monist essays on pragmatism may be out in about ten years). Current Peirce scholarship tends therefore to draw on the manuscript collection itself, but there are several reasons why I am not (with some exceptions6 ) basing this pragmatic reading on the manuscripts.
Page 11
Addressing this book, first, to sophisticated readers of Peirce who may not be specialists in the discipline, I want to comment on writings that these readers would most likely read and, in doing so, to contribute to their practice of reading Peirce, rather than to the specialized scholarship on Peirce. Since the Peirce Edition Project (the primary source of the future) will not for some time get to the 1902–8 essays that have most influenced my study, and since most general readers will not be reading the microfilmed manuscripts, I focus primarily on essays that appear in the trusty old Collected Papers. I realize this means my work will soon be dated, but I also acknowledge that temporal and other sorts of finitude must accompany a pragmatically nonfoundational effort. In this spirit, another aspect of pragmatic reading is appropriate vagueness and a resistance to the perfectionism that may accompany more ambitious readings. Historicalcritical expositions ofPeirce's manuscripts are as essential to a pragmatic reading of Peirce's writings as they are to other forms of Peirce scholarship. For a pragmatic reading, however, it is also essential to avoid what I will label "textualist" attempts to reduce the study ofPeirce's writings to a study of "what Peirce actually wrote at a certain time in his life" and, then, to avoid historicist attempts to reduce this to a study of "what, on the evidence of this writing, Peirce the man must have thought at a certain time in his life." In the terms ofPeirce's semiotic, I understand both textualist and historicist readings to be attempts to refer Peirce's writings, as symbols, to their indexical objects (or what the theological hermeneut Hans Frei calls their "ostensive referents'') considered independently of the particular interpretive interests that textualist and historicist scholars bring to their scholarship. While these readings have value, the value is displayed precisely only with respect to these interests and, thus, only once these interests are identified. I do not, therefore, assume that the latest critical reading of the most recently edited manuscript ofPeirce's is going to reveal something fundamental about "what Peirce really thought." I trust such readings will contribute to plainsense study by expanding our sensitivities to the semantic range of Peirce's writing. But, beyond that, I assume manuscript or documentary study constitutes its own subfield of inquiry with its own pragmatic interests; this study may, but need not, illuminate other pragmatic readings. I thus find that philological, rhetorical, historical, biographical, psychological, logical, and other forms of inquiry all contribute to the work of pragmatic reading, but none in a privileged way. Pragmatic readings are made precise only by the way they speak directly to the vaguely defined interests that lead them.
Page 12
Some of the Scholarly Contexts of This Study Developmental Studies of Peirce's Work This study of Peirce's pragmatism is developmental in two senses. On the level of plainsense reading, I examine Peirce's four major publications on pragmatism in chronological order, noting how each one comments on its antecedents and anticipates subsequent developments. Here, Peirce appears to have generated a series of pragmatisms, each one responding to something troublesome in the previous one. The fourth and last pragmatism, or "pragmaticism," occupies a privileged place in this sequence, since it articulates rules of inquiry that both appear in and correct each of the previous three. Characterizing these rules as rules of pragmatic reading, I reread Peirce's pragmatic writings, pragmatically, as various tokens of a dynamic or developing habit or Rule of conduct (a "genuine symbol"), exhibited now in the context of these issues and problems, now in the context of those, and so on. As plainsense reader, I examine what we might call several different Peirces, identified with the authors of several different pragmatisms. As pragmatic reader, I examine one complex and vaguely defined Peirce, identified with the author whose pragmatism emerges in the contexts of now this and now that inquiry. In between the two, one might say that I examine two different Peirces: the pragmatist of part I, who struggles, without ultimate success, to separate his own philosophic practice from the burdened practices he seeks to correct (Cartesian, nominalist, foundationalist, and so on); and the pragmaticist of part II, who exhibits his pragmatism most clearly in the rules according to which he has corrected his own burdened practice of philosophy. The classic, developmental study of Peirce is Murray Murphey's The Development of Peirce's Philosophy. 7 I have drawn much inspiration from Murphey's work, particularly in examining ways in which innovations in Peirce's logical theory have stimulated developments in his pragmatic inquiry. My study differs from but complements Murphey's work in the way it focuses on Peirce's pragmatism and on a pragmatic way of studying it. Murphey examines Peirce's philosophy as a developing series of arguments, grounded, successfully or not, in a series of logical systems. In this scheme, Peirce's work improves as he makes newer, better, more refined discoveries in the technology of logic or of related inquiries; his work fails when his philosophic claims overreach or misrepresent the scope or power of his logical technology. According to the present study, Peirce's philosophic writings represent various commentaries or corrective readings of antecedent philosophic writings (his or others'): they succeed to the extent that they repair whatever problems in those antecedent writings stimulated them. Murphey therefore takes more seriously than I do Peirce's
Page 13
efforts to build his philosophy systematically or "architectonically." According to these efforts, Murphey concludes that Peirce's ultimate system remains a mere "castle in the sky." I ascribe tensions in Peirce's work to conflicting tendencies, either to foundational system building or to pragmatic criticism. I conclude that the latter tendency dominates in the final period of Peirce's work, leading Peirce away from system building to another model of philosophic conduct: the philosophic healer, or one who compassionately employs the tools of mathematical imagination and logical rigor to repair, or at least respond to, the various species of intellectual suffering that complement what Peirce called Cartesianism and others, more recently, call "modernism.'' Among other more recent scholars, Richard Bernstein and Sandra Rosenthal have offered developmental studies of Peirce that come closer to raising the pragmatic and hermeneutical issues that are central to this study. In a study of Peirce's theory of perception, Bernstein writes, My purpose is to try to get inside Peirce's philosophic dialectic, to see what were the main problems that he was struggling with and to show how his different emphases are responses to different facets of the problems. By approaching Peirce in this way, I believe that we will not only come to understand the coherence of his views concerning perception, but we will gain a deeper insight into [his] entire philosophic endeavor. 8
My approach is similar to this, only narrower in focus and more labored: an attempt to formalize the method through which one "gets inside Peirce's dialectic," identifying each step in a progressive interpretation of the elements of the dialectic and of its resolutions. In a comparable study of perception, Rosenthal displays particular sensitivity to the teleological character of the development of Peirce's thought: Peirce's array of writings on the topic of perception can most profitably be understood as an attempt to clarify what he at first saw only obscurely and hesitatingly, and that this is so true that his later writings are to a large extent indispensable to a correct understanding of his earliest writings.9
My attempt to study the early Peirce through the lens of a later Peirce reiterates Rosenthal's attention to both teleology and chronology in Peirce's corpus.10 In brief, my developmental approach is as follows.11 I first order Peirce's pragmaticwritings chronologically, according to date of publication or of writing, in case of manuscripts. Among contemporaneous writings, I then collect together as sets those writings that I believe exhibit common Thesistypes. In this way, I divide Peirce's pragmatic writings into four sets, presented chronologically as follows: in chapter 3, Peirce's critique of Cartesianism (the journal of Speculative
Page 14
Philosophy series); in chapter 4, his early pragmatic writings, the "Illustrations of the Logic of Science" (published in the Popular Science Monthly); in chapters 5 and 6, what I call his "Normative Theory of Pragmatism," as displayed in essays appearing between 1878 and 1903, and culminating in the 1903 Lectures on Pragmatism; in chapters 7 and 8 (part II), his pragmaticist writings, including those published in 1905 in The Monist and after. These chapters are of increasing length, corresponding to the accumulating layers of Peirce's pragmatic readings: as each one correctively interprets its antecedents. I place Peirce's writings in chronological order to reflect their hermeneutical rather than historical sequence. As I argue, the meaning of Peirce's pragmatism lies in the movement of his interpretations from one writing to the next, where each writing is seen not merely as a particular event, but also as token of a movement from one interpretive thesis type to the next. Each of the four sets of writings thus represents a Type of pragmatic analysis, adopting each preceding set as its subject matter. In the first half of each chapter, or pair of chapters, I offer a plainsense reading of a set of writings, thereby isolating a basic lesson about a Type of pragmatism. In the second half of each chapter or pair of chapters, I offer pragmatic readings of what happens to that lesson as it is reread in the next lesson. As pragmatic interpreter of pragmatic readings, I imagine that a reason for each interpretive movement may be located, metaphorically, ''between" the text and its movement. This reason is that Peirce finds something disturbing in each lesson and that the interpretive movement of his reading is to respond to what is disturbing. He does not find the lessons wholly wrong, however; he claims, in fact, that the prototypical error of Cartesianism was to imagine just that—that antecedent, scholastic lessons were wholly wrong, or wholly unreliable as sources of the norms that would guide corrective reading. Each lesson therefore has something undisturbing in it, as well as something disturbing, and the purpose of each successive lesson is to identify the former as the source of norms for doing something about the latter. The central claims of this study concern how to account for Peirce's moving from a disturbing lesson to a response that is itself based in that lesson. He does not say explicitly that his essays belong to a given type or period, nor that each type corrects a previous one, nor that problems in one therefore lead in some identifiable way to the next. He does not say these things explicitly, because whatever he says explicitly belongs to his lessons per se, and not to the way in which they correct their antecedents. In place of a given lesson (1), he offers simply lesson(2).It is only I as pragmatic reader who claims that lesson(2)comes as a way of correcting the disturbing aspects of lesson (1). Peirce does, in fact, admit his errors more than most philosophers, and he does much more than
Page 15
most do to refine his earlier work. But, more often and more boldly, he claims to move on to new lessons because(a)he has made some new technical discoveries in his work; or (b) he has come upon new adversaries worthy of correction. He never portrays himself, as I do, as his own principal adversary. I argue, however, that my claims on his behalf are not extraneous to what he says, but, rather, represent reasonable ways of defining clearly, with respect to particular contexts of definition, what his work exhibits only vaguely. My study builds on the work of my doctoral dissertation, "Charles Peirce's Metaphysical Conviction". In that work, I trace a metaphysical orientation that persists throughout Peirce's philosophic writings and that achieves consistent, selfconscious definition only in the last stage of his inquiry, after 1906.1 divide his work as a whole into four periods: (1) Early Idealism and AntiCartesian Polemics,1860s to early 1870s;(2)The Law of Mind Period, 1880s to early 1890s; (3) Pragmatism Revisited, Cantor Rejected, 1903 and after; (4) Material Continuity, The Vague Replaces the General,1906 and after. I have kept the first and last stage pretty much the same, but the different titles and different inner divisions in my current study reflect ten years of a different kind of study, focused now on pragmatism, then on metaphysics. Influenced at that time more by Peirce's 1902–3 idealism than by the post1905 pragmaticism that leads the present study, I offered a more architectonic view of his philosophic system. At the same time, I also offered a more detailed analysis than I do now of the dialectical tensions and movements in what I call Peirce's Normative Logic of Inquiry of 1902–3. 12 Studies of the One Peirce and the Many Peirces One consequence of reading Peirce's writings developmentally is that different writings then appear to display different tendencies of interpretation, or what postmodern scholars may call different "authors." From the early years of Peirce scholarship, in the 1940S, to today, scholars have tended to read Peirce's corpus as if it displayed the work of one, two, three or more authors. To identify only one author means to claim, for example with James Feibelman, that all of Peirce's corpus displays a single, dominant tendency of interpretation—even if it takes a little work to see the one tendency in the many different writings.13 To identify two Peirces means to claim, for example with Thomas Goudge, that the writings display two competing tendencies.14 As in Goudge's prototypical case, this claim may carry with it the message that one of the two tendencies is the better one, or the "real" or "rational" Peirce. I cannot say who identifies simply ''many" authors in Peirce's work—or more than three—but I imagine that this could be the consequence of a postmodern reading that identifies each major
Page 16
claim of Peirce's, or each thesistype, with a different author. My own pragmatic reading implies that there are two or three Peirces, depending on how one divides up Peirce's thesistypes. While dividing Peirce's pragmatic writings into four sets, I also refer all those sets to only two or three "leading tendencies" of interpretation. I read the problems in Peirce's earlier writings as signs of contradictions between two logically contrary leading tendencies. I read the unproblematic aspects of those early writing as tokens of another leading tendency that will eventually guide his correcting those tendencies and generating the relatively unproblematic essays of his later, pragmaticist period. There are thus three Peirces, if one adds up the two contrary tendencies and the one corrective tendency. There are two Peirces, if one treats the contrary pair as consequences of a more general, intrinsically conflictual tendency, to be opposed to the pragmaticist tendency. Then again, I also divide the latter into three subtendencies. Of course, these interpretive machinations are useful only if one maintains the appropriate sense of humor. To lack a sense of humor, perhaps, is to read a philosopher's text as if it directly expressed the explicit intentions of some single subject to have some single or general reader understand some single subject matter in a single way. On Peirce As Pragmatist and Semiotician My approach to Peirce's pragmatism still reflects the different but complementary interests of the two philosophers with whom I first studied Peirce: John E.Smith, my dissertation advisor; and Rulon Wells, advisor for my subsequent postdoctoral study. In chapters 7 and 8,I label the two complementary poles ofPeirce's mature pragmaticism "mathematicist and logicist," or "diagrammatic and hermeneutical," and so on. When I do this, I think of Wells' studies of pragmatism as exemplifying the former pole of pragmatic inquiry, and Smith's studies of pragmatism as exemplifying the latter—indicating, of course, the effect these studies have had on my work, rather than any description of their own philosophic projects. I have received Smith's work on Peirce—and more generally on American philosophy and the philosophy of religion—as a model of what I will label common sense or hermeneutical pragmatism. 15 This is, for one, a philosophical inquiry that selfconsciously emerges from and returns to its context in the everyday life and everyday experiences of some human community. This context refers, more specifically, to the sets of historically and communally conditioned problems that have stimulated the philosophic inquiry. The inquiry thus emerges as a method of problemsolving, drawing on the resources of rigorous philosophic inquiry to contribute, ultimately, to correcting some form of practice. I have received Wells' work on Peirce
Page 17
—and more generally on American philosophy, semiotics and the logic of language use—as a model of what I will label both mathematical and critical pragmatics. 16 By this I mean a family of logical analyses that generates and then critically evaluates diagrams, or formal descriptions, of the elements of pragmatic inquiry. Among these elements are linguistic forms and expressions of everyday practices; semiotic forms and expressions of the logic of corrective inquiry; and ultimate criteria or procedures for evaluating this logic. Readers may recognize these two approaches to pragmatic inquiry in my efforts to integrate formal analysis and performative interpretation in my readings of Peirce's pragmatism. Beyond these two sources, my approach to Peirce's pragmatism remains indebted to Manley Thompson's classic study17 as well as to the following types of more recent pragmatic inquiry: careful text readings, exemplified, for example, in Vincent Colapietro's recent essays;18 performative studies which includes feminist inquiries in pragmatism as well as politically focused inquiries, such as Cornell West's, and ethicoreligiously focused inquiries (see below);19 hermeneutical studies, including Richard Bernstein's classic work (discussed above) and, beyond the field of pragmatism per se, suggestive studies by students of Gadamer, and of Derrida, by David Griffin and other "nonrelativistic postmodern philosophers," and by a collection of Jewish and of Christian philosophers I will mention below.20 My method of reading pragmatism reflects the literary focus in postmodern philosophic inquiry more generally: pragmatics, logical semiotics and formal or mathematical semiotics, including Paul Grice on pragmatics; Marcello Pagnini and others on the pragmatics of literature; Umberto Ecoonlogical semiotics; Michael Shapiro and others on Peirce's semeiotic; and Kenneth Ketner and others on Peirce's "Existential Graphs" and related innovations in formal logic and formal semiotics.21 On Peirce As Theosemiotician and Hermeneut While my plainsense studies of Peirce's writings reflect the kind of orthodox exegesis one tends to find in the Transactions of the Charles Peirce Society, my pragmatic interpretations are addressed to progressively narrower communities of interpreters: from the communities of pragmatists mentioned in the previous section to the narrower communites of theosemioticians and, then, biblical pragmatists to whom I address my final interpretations. The broader of these narrow communities are those philosophers who find postfoundational use for metaphysics and, among them, those who classify metaphysics as in some sense a theological inquiry. Of the latter, I am indebted, among others, to the work of Michael Raposa, who read through the book manuscript carefully
Page 18
and offered crucial advice. Raposa has coined the term "theosemiotics" for Peirce's theologic, based as it is in his semiotic language of logical analysis and on his philosophically realist conception of the significance of logic. John E. Smith's work remains a guide in this direction. 22 I am indebted also to Robert Corrington's many studies of theosemiotics as a form of theistic or "ecstatic" naturalism, as well as to his sense of the place of biographical and psychological study in pragmatic hermeneutics.23 My work has also been enriched by discussions with former Drew graduate students, among them Kathy Hull, who writes broadly on Peirce's philosophy; Todd Driskill, who writes on Wittgenstein and theology; William Elkins, who writes on Ricoeur, pragmatism and what some call the "Yale school" of scriptural theology; Roger Badham, who writes on hermeneutics, ethics, and theology; and Terry Baker, University of Virginia graduate student, who writes on Kierkegaard, Levinas and the philosophy of religious language. Mr. Baker offered muchappreciated help with the final editing and glossary. I recently edited a collection of what I call Jewish and Christian "postcritical scriptural interpreters," including the "Yale school" theologians and their complements in recent Jewish thought.24 Among the Christian members of this group I number the late Hans Frei, George Lindbeck, Stanley Hauerwas, Kathryn Tanner, Bruce Marshall, Gregory Jones, James Buckley and many others; among the Jewish members are the scriptural or rabbinic text scholars Steven Fraade, Michael Fishbane, David Weiss Halivni, Moshe Greenberg, Daniel Boyarin and many others. In chapter 8,I describe all these scholars as "scriptural pragmatists," since I believe that their studies of scriptural commentary exemplify, in three ways, the kind of pragmatic inquiry that emerges in Peirce's later writing. For one, they portray scriptural reading as itself a pragmatic response to crises in human communities. For two, their own inquiries respond, pragmatically, to crises in certain communities of scriptural readers. For three—and here is one of the more controversial and speculative claims in this study—their inquiries belong to a tradition of scriptural commentary that may generate the most reliable prototypes we have of Peirce's pragmatic inquiry interpreted as a method of corrective reading. My sense of what it means to read and to interpret emerges largely from the work of a community of Jewish philosophers who offer both formal evaluations of scriptural inquiries like these and performative evaluations of philosophic inquiries like Peirce's. Of these thinkers, Robert Gibbs commented extensively on the entire book manuscript, working both as its primary philosophic and stylistic editor and as my chevruta, or regular study partner, on Talmudic texts read in light of Peirce's and Rosenzweig—Levinas' semiotics; I am indebted to his insightful contributions.
Page 19
His studies in Jewish philosophy bring pragmatism and semiotics into dialogue with major currents in Continental thought. Through commentaries on my work and through shared study, Norbert Samuelson explored with me various relations between mathematical and Jewish philosophical/textual studies and has inspired significant parts of the study. Eugene Borowitz, David Novak, David Weiss Halivni, Edith Wyschogrod, Michael Wyschogrod, Steven Kepnes, Yudit Greenberg and Michael Rosenak have contributed to my work in Jewish pragmatism more broadly. 25 Vanessa Ochs, to whom I dedicate this book, has provided the most detailed models I have seen of corrective, pragmatic inquiry as it bears on everyday life in the Jewish community and as it is communicated through writing.26 In the conclusion to part II, I mention various prototypes of theosemiotics and of pragmatic reading in Hellenistic hermeneutics—in patristic and rabbinic hermeneutics in particular. These explorations were nurtured by the community of philosophic and scriptural theologians at the Center of Theological Inquiry in Princeton, including Center Director Daniel Hardy, David Ford, Wesley Kort, Garrett Green, Kathleen McVey, Corby Finney and Craig Koester. I am grateful, finally, for persistent encouragement from Alex Wright and for trustworthy editorial guidance by Ruth Parr.
Page 20
Chapter 2— Pragmatic Methods of Reading and Interpretation This is a technical chapter, for specialists in epistemology. General readers may prefer to move ahead to chapter 3, to take up the actual story of Peirce's intellectual struggles. This chapter addresses concerns the epistemologist may have about how I go about telling this story. Am I presenting an intellectual biography or a philosophic analysis? Why have I chosen this method of analysis? How are my claims about the development of Peirce's thought warranted? In responding to these questions, I will describe a pragmatic method of reading and interpreting a modern philosopher's writings. The rest of the book should exemplify the method, as applied to Peirce's writings. I argue that this method is a token ofPeirce's pragmaticism, but I will be able to defend my argument only at the end of the book, after having performed it. In other words, this study itself belongs to what I will label a Method of Interpretation and Repair: which is to interpret a philosopher's thought according to a critically refined principle of his or her own thinking. Pragmatic Reading in the Cartesian—Kantian Tradition of Epistemology Peirce addressed his writings on pragmatism primarily to communities of epistemologists. In this chapter, I also describe the pragmatic method of reading primarily in the language of what I call the Cartesian—Kantian tradition of epistemology. Peirce presented his earliest work in pragmatism as a critique of what he called "Cartesianism" and as a revision of Kant's critical philosophy. I argue that this was a selfcritique, however, since Peirce's pragmatic critique applies first to his own Cartesian search for philosophic foundations and, then, to his own Kantian manner of reevaluating that search. Peirce's pragmatism addresses a particular sub community of epistemologists concerned, specifically, with empirical knowledge and, more specifically, with empirical inquiry in the natural sciences, especially the laboratory sciences. Nonetheless, Peirce identified the errors of his empiricist peers with the errors of Cartesian epistemology more generally. This study follows suit: describing pragmatism in the
Page 21
context of general epistemology, which I will come to label more generally the context of modern philosophy, or philosophic modernism. In this section, I discuss features of the pragmatic method that may be of greatest interest to epistemologists. A NonGeneral Method Against the fashion in modern epistemoloy, I claim only a limited generality for the pragmatic method of reading. The method is developed, specifically, for interpreting the works of epistemologists in the modern, or CartesianKantian practice, 1 in which Peirce is to be included. The method may have other applications as well, but I do not claim that it does. I would expect it to apply elsewhere only to writings from analogous practices of thought. For the pragmatists, in short, CartesianKantian inquiry introduces a rationalistic method of problemsolving. The nonrationalistic method is to undertake rational inquiry in response to some behavioral problem of some sort, attempting first to identify the problem and then to solve it. The fundamental rules (or "leading tendencies") of this inquiry belong to the same tradition of thought and practice to which the problem belongs. Should the problem be identified with a defect in some of those rules, inquiry will be guided by other, more basic rules from the tradition. Scientific reports—including philosophic reports in the empiricist tradition—may usually be classified as written records of such attempts at problemsolving. To read these records, the nonrationalist attempts to identify the tradition to which both problem and inquiry belong; to understand the problem as described in those records; to identify the rules of inquiry employed in solving it; to observe how the inquiry applies those rules; and, finally, to evaluate the inquiry's success or failure in solving the problem. CartesianKantians describe their works, in this same manner, as attempts to solve certain problems. Attempts to read them in this manner, however, do not succeed. According to the pragmatists, this reading fails because the problems described in CartesianKantian texts do not effectively substitute for the problems that have, "in fact," stimulated the CartesianKantian inquiries—or that may be effectively described as having stimulated it.2 On the one hand, CartesianKantians offer their work to us, urgently, as making claims upon our actual behavior in this world and as responding to profound problems in antecedent traditions of thought and practice. On the other hand, they also claim that their inquiry founds a new tradition of thought discontinuous with and irreducible to the interpretive categories of the antecedent ones. The pragmatists theorize, however, that these are competing claims. They offer their theory
Page 22
in order to account for the fact that the CartesianKantians' "new" tradition shows an inveterate tendency to generate indefinitely selfreplicating inquiries that fail to solve the problems they intend to solve. The inquiries may fail, say the pragmatists, if these are not actual problems, but only feigned problems, and the inquiries are not attempts at actual problemsolving, but attempts to construct whole systems of ideas that are admired for various intrinsic properties, rather than for their efficacy in bringing about change in any system of actual practice. The pragmatists call these systems "rationalistic" to indicate that they are not merely rational, but rational only according to their own rules of reason, and to suggest that these rules are most likely abstracted from the traditions of thought and practice from which the Cartesian Kantians claim independence. This brings us back to the CartesianKantians' competing claims. The pragmatists theorize that the CartesianKantians' claim to "found a new tradition" may be read as a mark of their having abstracted their reasonings from an antecedent tradition. Such abstraction would be an innocent affair if the CartesianKantians presented it as such and called their inquiry "merely speculative—an exercise in hypothesismaking.'' Their "urgent claim on our behavior," however, may be read as a claim about the world of actual practice. If their inquiry is abstractive, then the only actual world they have referred us to is the world of that antecedent tradition of practice from which they have abstracted their claims. For the pragmatists, in sum, the CartesianKantians' "urgent" claim may be meaningful only with respect to that antecedent tradition. There remain, therefore, two ways for pragmatists to receive the CartesianKantians' competing claims. One way is to ignore the claim to urgency and read Cartesian Kantian inquiries as displays of various powers of human reasoning, interesting for their intrinsic properties. Another way is to criticize the claim to novelty and independence and read CartesianKantian inquiries as asyetundeciphered messages about some urgent problems in some antecedent traditions of practice. Pragmatism makes use of both of these approaches to CartesianKantian philosophy. The "pragmatic method of reading" refers to the second approach, as the particular focus of this study and as complemented by the first approach. The pragmatic method is a method of reading for the simple reason that the CartesianKantians present their work by way of individually authored writings. To get at the truthvalue of the work, the pragmatist must read these writings in a way that respects the intentions of their individual authors and in a way that respects the authority of those writings themselves to deliver information independently of the author's intentions. I am not suggesting that all philo
Page 23
sophic writings should be doublyread in this fashion, only that the peculiar character of the CartesianKantian writings demands it. A Method Developed by and for Members of the CartesianKantian Practice The pragmatic method of reading emerges from the same, single practice it serves. Pragmatists, in other words, are nurtured in the CartesianKantian practice. They learn from it love of the abstract sciences and concern about the errors of antecedent traditions. The practice eventually disappoints them, however. The various definitions of pragmatism reflect the attempts of various pragmatists, from Peirce to James to Dewey to their many disciples, to locate the source of the disappointment and repair it. In each case, this source is described as an unwarranted effort to divorce the epistemological practice from the systems of thought and behavior which inform it and from the empirical conditions which underlie its emergence as a distinctive practice. For each pragmatist, therefore, the practice is repaired by making it respond to these conditions and by redefining it with respect to these systems. 3 Different pragmatists attend to different dimensions of these conditions and systems. James is most attentive to biological and psychological dimensions, referring the practice of modern philosophy, therefore, to the empirical conditions that motivate the work of individual thinkers. Dewey is more attentive to sociological dimensions, placing that work in its social, as well as biological and psychological setting. Peirce, finally, is most attentive to semiotic and historical dimensions, placing the social setting, in turn, within a history of evolving systems of symbols. To articulate Peirce's perspective, I try both to attend to the psychological and sociological dimensions of analysis he fails to emphasize and to draw out some undeveloped implications of his own semiotic/historical perspective. In particular, joining a contemporary school of philosophic hermeneuts, I consider behavioral and literary traditions to be essential to the general semiotic/historical processes that Peirce discusses.4 Later in this study, I argue at length that, in the interests of his own project of inquiry, Peirce ought to attend more to the particular behavioral traditions that inform his work and, thus, less to ungrounded speculations about "semiosis in general."5 This means that I subject Peirce's work to the pragmatic method of reading it recommends: a sure sign that I consider Peirce to be a member of the CartesianKantian practice he both criticizes and serves. What is true of Peirce may be true of all pragmatists: that they criticize epistemology from within its practice and that what they call the errors of epistemology may be reflected in their own work as well. In other words, they may be unprepared to practice fully what they preach. Until shown
Page 24
otherwise, I will therefore assume that all pragmatic studies, including the present one, may require a pragmatic reading. The Method, Formally Stated To read pragmatically is to receive a CartesianKantian epistemologist's written text as, at once, two kinds of sign: an explicit statement (or iconic symbol) about characters that may be attributed to a possible world as envisioned by its author, and an implicit index (or indexical symbol) of events in the actual world from which the statement emerged. Interest in the implicit text is stimulated only by the pragmatist's conviction that the explicit text delivers some manner of misinformation: in this case, that the text is, in part, a product of rationalization. Reading the implicit text is a way of correcting the misinformation and identifying the truthvalue of the explicit text. The pragmatic method of reading is a method for discovering an implicit text within the explicit text. In the following outline of the method, formal terms drawn from Peirce's or another author's semiotics will be more precisely defined at some later point. Stage 1— In the Method of Reading Is to collect the explicit texts as, on their own terms, collections of particular arguments. The general form of such arguments in the CartesianKantian practice is simply "s(x)," where (s = (~f), s + f=U]. That is, the CartesianKantian text presents itself as a set of clear and distinct arguments, "s(x)," about some character of the actual world, "x," such that ''s" is the logical contrary of "f"—identified as the argument of some antecedent tradition, for example, a scholastic argument, to be corrected by the CartesianKantian one. (To say that "s" is the logical contrary of "f" is to say that the two of them, "f + s," exhaust a universe (U) of possible arguments on some topic.) The problem is that, for any two such arguments, "s(x)" and "t(x)," the set of arguments logically implied by one would include the logical contraries of the set of arguments logically implied by the other: S1,= ~T1, where S1[ ]s;T1[ ]t. In different terms, the text's claims may be mapped semiotically according to the dyadic distinction between sign and object, as drawn from de Saussure's distinction of signe and signifié: "x" is a sign of some x in the world; and "s(x)" is a sign of some state of affairs s(x) in the world. According to the pragmatic analysis that will follow, however, the text's claims may be reduced to this distinction, because the arguments it delivers are abstracted from the pragmatic dimension of sign reference, as formalized in Peirce's distinction of sign, object, and interpretant. The arguments assimilate interpretant to object or, in another vocabulary, reference to sense: this means that meaning in these texts is exclusively
Page 25
intentional meaning, or what I will call "rationalized" meaning. The text's set of contradictory implications is a sign of this rationalization and the reduction it entails. The next step in the analysis is to show that there is a lack of proportionality between the intentional object and the sign. The object belongs exclusively to a universe of ideal objects, a possible world, but the sign carries with it references to a universe of actual objects, an actual world. Different CartesianKantian authors respond in different ways to this lack of proportionality. From Cusanus on, the practice tends towards Kant's solution: that the epistemological project emerges in the discovery that our language's reference to the actual world is unreliable. All reference is intentional; therefore, the distinction between reference and sense is not helpful. In place of that distinction, new criteria are introduced within the category of sense: our language is more or less reliable depending on its clarity and distinctness, its coherence, or, with Kant, the universality and necessity of our assent to its use. A pragmatic analysis of CartesianKantian texts suggests that Kant's solution idealizes what the epistemologists are actually doing. Kant claims that the distinction between reference and sense belongs to the principles to which we give assent, universally and necessarily. But these modalities are appropriate only to intentional meaning, which means that Kant's distinction reinforces, rather than resolves, the CartesianKantian tendency to assimilate reference to sense. If the use of ordinary language tends to distinguish reference and sense, then, we might say, CartesianKantian argumentation separates itself from ordinary language use. One claim of pragmatic analysis is that reference carries with it the particularizing, or indexical, features of argumentation: to refer to the actual world is to refer with respect to some particular community of speakers with a particular set of histories and expectations. If this claim were valid, then we could account for the equivocal character of CartesianKantian arguments. On the one hand, the intentional objects of these arguments are separated from any particularizing references: the arguments are offered on behalf of anyone, in general. On the other hand, since reference has been assimilated to sense rather than explicitly bracketed, the arguments also bear, inexplicitly, marks of their particularizing references to the actual world. We might say that it is on behalf of these marks that CartesianKantians urge their arguments, confusedly, without purposefully showing us the ground of the urgency. The pragmatic method of reading is designed to resolve contradictions in the arguments of CartesianKantian epistemologists by rereading contradictory (or otherwise equiviocal) arguments as signs of both their explicit objects and their inexplicit references to the actual world. At the same time—and here the
Page 26
method displays its difference from the various hermeneutics of suspicion—the pragmatic method offers its readings fallibilistically: referring the epistemological project only to some particular world of experience, meaningful to some particular community of inquiry, not to some generic portrayal of "the world itself." To introduce the question of reference into CartesianKantian texts is thus to replace their underlying semiotics (built on a distinction between signe and signifié) with Peirce's semiotics (informed by the distinction among sign, object and interpretant). Stage 2— In the Method of Reading Is to reconstruct the texts of equivocal CartesianKantian arguments in a way that could conceivably separate the senses and references of the arguments. To accomplish this, the pragmatic reader analyses the explicit texts according a set of artificial categories that faciliate the work of raising reasonable hypotheses about the texts' references. The purpose of the following set of categories is to stimulate and structure hypotheses about the leading tendencies of inquiry that could conceivably have led certain CartesianKantian thinkers to produce the equivocal theses they produced. I have found this particular set of categories useful for rereading equivocal texts, but I can also imagine other ways of going about the job. Themes: topics of inquiry, such as "perception," or "the logic of relatives." Later on in the method, the pragmatist will need to introduce a more precise definition: noting, for example, that particular themes are designated as such only within particular practices, belonging to particular traditions. Problems: within a given theme, the particular difficulties whose solution is to be found in the text. Again, these problems will be seen, later, to belong to particular practices and traditions. Theses: the fundamental response that an explicit text offers to its problems, as guided by its leading tendencies. It is in presenting theses that the explicit text functions as a general sign of a possible world. For the purposes of analysis, this category is further divided: Thesistoken: the explicit text itself, as a particular written expression of the author's theses.Tokens may appear as essays, books or merely incomplete fragments. They appear as ordered collections of statements bound by principles of coherence and consistency. This means that, by definition, a given token cannot, without being subject to revision, deliver contradictory claims. The token can, however, deliver its claims by way of a number of different methods of analysis. The juxtaposition of different methods may lend an appearance of contradiction where, in fact, there is none. Thesistype: the most general statement of a text's claims to which the principle of noncontradiction applies. Such a statement is definite (nonvague) but
Page 27 indeterminate with respect to token. This category serves best as an analytic tool if it is limited to the language of one method of analysis. It is unlikely that a thesistoken will give expression to only one thesistype, but this is not impossible. Thesistypes are generally reconstructed through precisive abduction from a collection of thesistokens.
Methods: specific procedures for investigating any problem, displayed by way of specific languages of analysis that are used to apply leading tendencies to the production of theses. Leading tendencies are articulated only through the agency of such methods. Leading tendencies: statements of the elemental rules of inquiry that, through the agency of methods and as applied to resolving particular problems, guide the production of particular theses. The pragmatic reader regards each thesistype as a token of the final interpretant that would be exhibited with respect to a particular habit of interpretation, or practice. Since such practices are identified only by their sensible effects, or by their tokens, the reader can characterize them only by describing the thesistypes. At the same time, the reader can, per hypothesis, conceive of a practice of which both a given thesistype and other possible thesistypes would be tokens. To construct this hypothetical collection of thesistypes, the reader must already have applied the pragmatic method of reading to a significant number of the author's texts, ideally to all texts that employ the methods of inquiry exhibited in the text under examination. In Peirce's case, I have found it helpful to group his texts chronologically and to reconstruct practices that may be exhibited in all the texts of a given period of his work. To examine the development of his work, I then reconstruct a chronologically ordered series of practices. To identify these practices, I construct a general thesistype whose tokens would be the hypothetical collection of all thesistypes exhibited in the period of Peirce's work I am examining. I say the general thesistype represents a leading tendency of Peirce's practice, and I describe it within the terms of that method of Peirce's that is most pertinent to my interpretation of a given text. A pragmatic reading is successful if the reader can offer evidence that the leading tendencies exhibited in a given text are also exhibited in comparable texts by the same author, and if the reader can refer all contradictions among the text's theses to contradictions among these leading tendencies. The reader then reconstructs the confused text as a conjunction of two or more clear texts, each one of which exhibits a single leading tendency or a collection of tendencies of which no two are mutually incompatible. Conflicting leading tendencies may be symptoms of profound conflicts within the author's thinking, or within the practices in which the author participates. 6 In the latter case, these conflicts would be exhibited, as well, in texts written by other authors who share the same
Page 28
practice. The task of the pragmatic intellectual historian would be to examine the etiology of such a practice. 7 The notion of leading tendency employed here is a variant of Peirce's notion of "leading principle": It is of the essence of reasoning that the reasoner should proceed, and should be conscious of proceeding, according to a general habit, or method, which he holds will either (according to the kind of reasoning) always lead to the truth, provided the premisses were true; or, consistently adhered to, would eventually approximate indefinitely to the truth . . . The effect of this habit or method could be stated in a proposition of which the antecedent should describe all the possible premisses upon which it could operate, while the consequent should describe how the conclusion to which it would lead would be determinately related to those premises. Such a proposition is called the "leading principle" of the reasoning. (In Baldwin's 1902 Dictionary of Philosophy and Psychology, 2.588–9; emphasis mine)
As employed here, a"leading tendency" is like a"leading principle," except that, at this point, we need not make any ontological claims about the habits of reasoning that inform inquiry. For this stage of the reading, leading tendencies are no more than analytic constructs. Like thesistypes, they are reconstructed through a procedure I have labeled "precisive abduction," in this case from a collection of thesistypes. Precisive abduction corresponds to the method of transcendental regress employed in various phenomenologies, defined here, specifically, as a method of analyzing texts. I define the procedure by linking aspects of Peirce's definitions of abduction and of precisive abstraction. Most simply put, abduction is the process of forming an explanatory hypothesis. It is the only logical operation which introduces any new idea . . . Abduction merely suggests that something may be. Its only justification is that from its suggestion deduction can draw a prediction which can be tested by induction, and that, if we are ever to learn anything or to understand phenomena at all, it must be by abduction that this is to be brought about. (5.171:1903)
As a form of inference, Peirce illustrates the rule of abduction in this way: The suprising fact, C, is observed; But if A were true, C would be a matter of course, Hence, there is reason to suspect that A is true. (5.189:1903)
"A," it is to be observed, is not itself inferred, but is postulated, and postulated only within a conditional proposition, of which it is the antecedent. For Peirce, this implies, first, that the abductive conclusion, or hypothesis, must already be present, as he says, "in the mind habitually" (8.65) and, second, that it must be present as an inference.
Page 29
Peirce defines precisive abstraction as follows. First, abstraction has the etymological meaning of "drawing one element of thought (namely, the form) away from the other element (the matter), which last is often then neglected" (2.428:1893). There remain, however, two different procedures for abstraction. The first is subjectal, or hypostatic abstraction, which Peirce defines as the transformation of a relative predicate into a subject: for example, "it is light" into "there is light here," or "white" into "whiteness'' (4.235:1902). The second is precisive abstraction, or "dissection in hypothesis" (5.449:1905), "where the subject prescinded is supposed (in some hypothetical state of things) without any suppositions . . . in respect to the character abstracted" (1.549 n.1: 1911). Peirce founds his phenomenology on the claim that perceptual judgments are acritical inferences which we can best understand on the model of abductive inference. Unless we deny the possibility of observing anything surprising, however, this acritical inference must also perform a selective, or abstractive function. This may be illustrated by the following syllogism, which displays only our approximately understanding the protological process involved in the perceptual judgment "this is blue": 8 This is abcp' p". (the perceptual elements abcp' p" are observed) But Blue is p' p". (in the mind habitually) (i) This is . . . p' p" (precisive abstraction: neglect of elements treated as if unknown) (ii) This is Blue. (abductive conclusion)
For the phenomenologist, this understanding of a protological process is the archetype of a logical procedure: S (a class of experiences, e.g. "of surprise") is observed to display the classcharacters F (e.g., "feeling"), I ("interruption of expectation), and so on. But E (a class of possible experiences, such as "sensations of ego overagainst nonego") is assumed to display the classcharacter I. (i) S displays the classcharacter I. (precisive abstraction) (ii) By hypothesis, S is a case of E. (abductive conclusion)
Since E is general with respect to S, this inference is possible only if E is present "in the mind habitually," or, as is more likely the case, E is, through the same procedure, referred in turn to D, D to C and so on to A. If there is no category available more general than A, A is termed an indecomposable element of experience, or a phenomenological Category. The whole procedure presupposes a formal logic (for example, Peirce's mathematical logic of relatives), through which a collection of most general categories is displayed as a possible resource for phenomenological reductions. To identify the categories of thesistype and leading tendency, the method of
Page 30
reading applies this phenomenological procedure to the observation of an explicit text. First, the reader selects a given problem addressed by the text, such as "errors in the Cartesian theory of perception." Second, the reader collects some domain of thesistokens pertinent to this problem. The range of this collection is normally suggested by the problematic of concern to the reader, who, for example, may be concerned to correct scholarly misinterpretations of an author's published text, or, as in this study, of the entire direction of an author's thinking. Third, through a process of precisive abduction, the reader identifies the methods of analysis and the thesistypes displayed in each thesistoken. The first task is to form a collection of possible methods. The author's own claims provide one, not necessarily trustworthy, source of such a collection. Other sources are critical studies of the author's work, general studies of the logic of inquiry and of the history of philosophy and science, and a bit of guesswork. For a study of Peirce's analyses of perception, for example, any of these methods might appear in a given thesistoken: phaneroscopy, logical analysis of judgment, logic of quantity, semiotics, logic of inquiry, pragmatic analysis of controlled inquiry, metaphysical analysis. The next task is to construct individual thesistypes, by conceiving of the most general statement of an author's claims which, when enunciated in the language of a given method of analysis, would be displayed within the collection of thesis tokens. The construction is informed by the following rules. Generally, a thesistype expresses the author's applying one method to one problem. No method may determine contradictory or mutually exclusive thesistypes. (If a selected method does this, it may simply be a more general form of the method desired. The method should be redefined more restrictively.) Finally, each thesistoken will most likely display a number of different thesistypes, and thus a number of different methods. (In my study of Peirce's analyses of perception, for example, I found that his 1868 essay "Some Consequences of Four Incapacities" displays four thesistypes, among them this one: "There are no images, or absolutely determinate representations, in perception." In this case, I found that the thesistypes belong to two contradictory sets, which I then attributed to contradictory leading tendencies.) This stage of reading is completed when every statement of every thesistoken can be described as an instance of a constructed thesistype. Fourth, through a process of precisive abduction, the reader identifies the leading tendencies displayed in each thesistype. The first task is to form a collection of possible leading tendencies, using the same sources employed in identifying the methods of analysis. In this case, the author's own claims may be particularly untrustworthy, because authors tend to employ their ultimate pre
Page 31
suppositions uncritically. Authors who seek to identify the leading tendencies of their own inquiry will tend to hypostatize those leading tendencies as what each will call "his or her philosophy." (Such authors may then attempt, selfconsciously, to guide the future development of their inquiry according to this hypostatized norm. This effort will tend to systematize their future work, serving, on the positive side, to eliminate previously undisclosed contradictions in their work and, on the negative side, to inhibit their inquiry's responsiveness to unanticipated discoveries.) To supplement an author's own claims, the reader must be imaginative in constructing possible statements of the leading tendencies, considering every possible source of influence on the author's work. Intellectual/cultural history, biography and the logic of inquiry are the most important tools in this endeavor. The next task is to identify which leading tendencies could, when applied by way of a given method of analysis to a given problem, produce the thesistypes already identified. The following rules guide this effort. Each thesistype is particular to a given problem and is, in most cases, informed by several leading tendencies; it cannot be precisively abstracted from method and leading tendencies. Each method is applicable to many different problems and can serve several, or in the ideal case, all available, leading tendencies; it cannot be abstracted precisively from leading tendencies. In most cases, each leading tendency guides inquiry into only some aspect of a given problem or theme, but is served by several, or all available methods. In practice, thesistypes, methods and leading tendencies are only abstracted from the explicit texts; they are no more neatly separable than are Peirce's phenomenological categories. The following disclaimer of Peirce's, therefore, applies equally to the method of reading: "Finally, though it is easy to distinguish the three categories from one another, it is extremely difficult accurately and sharply to distinguish each from other conceptions so as to hold it in its purity and yet in its full meaning" (1.353: c. 1880). To return to that illustration from Peirce's analyses of perception: I found that contradictions in Peirce's 1868 thesis on perception can be referred to contradictions between a leading tendency I label "Kantian conceptualism" and an integrated set of three I label "tychism, anancism and agapism." According to the pragmatic method of reading, the appearance of contradiction in a thesistoken is a sign that the thesistoken needs revision. If the leading tendencies abstracted from the token do not contradict one another, the text may be revised simply by correcting the manner in which those principles are applied to methods and those methods to the problem at hand. If the leading tendencies are themselves contradictory, then the reader needs to proceed to a third stage of pragmatic reading.
Page 32
Stage 3— In the Method of Reading Is to distinguish between two kinds of leading tendency and, thereby, to apply the working hypothesis that underlies the method of reading. This first hypothesis in the method of reading is that contradictions will appear between leading tendencies introduced by the Cartesian—Kantian method of rationalization and those that are adopted independently of that method. The Cartesian—Kantian principles instruct inquirers to construct their own methods of inquiry, or to adopt methods which have been constructed in this manner. They therefore instruct inquirers to refer both the object of inquiry (what the explicit text signifies) and the problem that motivates inquiry (that to which the text refers) to the single universe of possible worlds that can be constructed out of their chosen method of inquiry. This precludes the appearance of any incommensurability between problem and inquiry and, thereby, constitutes what the pragmatist calls a "rationalization" of actual problems. Non Cartesian—Kantian principles have their source in the practices that belong to various traditions of behavior; at this stage, there is no need to identify them further. According to this hypothesis, contradictions in Cartesian—Kantian texts may be resolved as follows. First, the reader must distinguish Cartesian—Kantian leading tendencies from nonCartesian—Kantian principles. Then, the reader may reconstruct the text's argument by applying the nonCartesian—Kantian principles to the text's theme and problem, by way of those methods of inquiry that do not also presuppose the Cartesian—Kantian leading tendencies. This procedure should produce a noncontradictory thesistoken. The token remains abstract, however, since the reader has not yet reassigned the leading tendencies that govern it to the tradition of practice in which they operate. The reader may also reconstruct a noncontradictory token by applying only the Cartesian—Kantian principles to the text's theme and problem. This would produce a strictly rationalized argument: both abstract and unverifiable, since its leading tendencies would bear no necessary relation to the actual world in which the argument would be verified. In sum, the third stage of reading provides a working model of how the Cartesian—Kantian text would look if it remained within its antecedent and unrepaired tradition of practice. Precisively abstracted from the explicit text, this model is only a formal sign, or icon, of the implicit text which the reader is now prepared to reconstruct. Stage 4— Of the Method of Reading Is to distinguish between the explicit Cartesian—Kantian text and the implicit text of which it is a sign. The explicit text defines itself as a definite and general (or indeterminate) sign of that possible world of which it makes some
Page 33
claim. (For readers interested in more formal analysis, we will designate the explicit text Ax, the world it signifies Wx, and the claim [Ax [&!;] c(Wx)] where c is some character or function of the world.) Initial reading of the explicit text, however, suggests, on one level, that the text offers contradictory information about that world [Ax [&!;] c(Wx)+ ~ c(Wx)] (here,[&!;] refers to an iconic [clear] sign of logical implication). We hypothesize that the contradiction lies in the complex character of the explicit text: that along with explicit claims about a possible world, it also delivers implicit information about an actual world and that this information either restricts or contradicts the claims [Ai [&!;] ~ c(Wx) or ~{c(Wx)}]. To deliver such information, the implicit text must, unlike the explicit text,be an indexical sign of that actual world it interprets, or, in Peirce's semiotic, of which it is an interpretant. Of such a world, the implicit text will be what Peirce calls an indefinite or vague sign [Ai » Wa] (here, » refers to an indexical [vague] sign of material implication; "a" indicates that this is the world as characterised in the implicit text A [which defines an otherwise vague signifier]). Finally, according to the pragmatic method, we must understand this actual world to be a tradition of practice (of thought and behavior) whose errors alone stimulate inquiry. The implicit text, then, must be considered a sign of some particular problem in that actual world, which problem gives rise to the sign [Ai » p (Wa)] (where p refers to problem in the world) and which might, ordinarily, find its solution in the corrective inquiry(s) which the sign, in turn, indicates [Ai [&!;] s{p (Wa)} = Wa]. From the perspective of this stage, the work of pragmatic reading is to explicate the implicit text, transforming it from an indefinite sign of some problem in some world to a general sign that recommends to an interpreter methods of solving that problem. The first task of this reading has been accomplished in stages 2 and 3: referring the explicit text to its leading tendencies, discovering that those principles contradict and resolving the contradiction by distinguishing Cartesian—Kantian from non Cartesian—Kantian principles. (In terms of the formal analysis, the explicit text has been relabeled a thesistoken [Ax = t]. The claim offered by that token has been identified with a collection of thesistypes [c = {T1, T2, . . .}] which, in turn, have been redescribed as functions of sets of leading tendencies [Tn=f(L1 _n)]. In sum, we have redefined the explicit text in this way: [t