Chaïm Perelman was only 21 years old when he secured a subvention from the Université libre de Bruxelles to publish his 1933 âDe lâarbitraire dans la connaissanceâ [On the Arbitrary in Knowledge], which was published by the Maurice Lamertin publishing house based in Brussels.1 âDe lâarbitraireâ follows Perelmanâs brief 6-page article on âLe statut social des jugements de véritéâ [The Social Status of Truth Judgments], which appeared in a 1933 issue of Revue de lâInstitut de Sociologie Solvay and which we consider as an initial formulation of âDe lâarbitraireâ.2 Harvardâs T. H. Costelloâs 1934 review of Perelmanâs article in the Journal of Philosophy asserts that Perelman had not developed his postulates, ânor has he succeeded in expounding them very wellâ.3 Yet for a Harvard professor to offer a reviewâeven one that was negativeâin one of the premier philosophy journals of a young Belgian studentâs overview on reason and logic suggests that the article was noteworthy and that it had earned attention in philosophical circles. Other, present-day scholars, more appreciative of Perelmanâs work, place âDe lâarbitraireâ in the context of Perelmanâs fifty-two-year body of scholarship. We believe that this article offers the reader insight into the evolution of his thought.
The article is also prescient, anticipating Quine and the collapse of the fact/value dichotomy, and beyond Quine, the intellectual roots of Perelman and Olbrechts-Tytecaâs New Rhetoric Project. Many of the issues and concepts Perelman introduces and populates in this 1933 article are fleshed out by 1958 in the Traité.4 At the outset of âDe lâarbitraireâ, Perelman announces his theme: the putative opposition between reality judgments and value judgments accepted by most philosophers, and especially by those in the nineteenth century. This opposition rested on the claim that reality judgments (later truth judgments) are necessary, and value judgments are arbitrary. In the first section, Perelman shows that reality judgments and truth judgments ultimately hinge on the arbitrary. He demonstrates this through an analysis of several patterns of reasoning: deduction; induction; inductive reasoning in science, classical logic, and law; and finally, reasoning by analogy and by appeals to authority. He concludes that all methods of verification are in some sense arbitrary: all judgments are value judgments; all truth claims presuppose the arbitrary.
In the second section, Perelman examines the subject-object problem and the problem of error. The subject-object correspondenceâthat the perception of an object corresponds to our cognition of itâdepends on the arbitrary assumption that all people over time share the same perception. At this point, Perelman returns to the idea of necessity, and concludes, drawing on the work of Eugène Dupréel, that what is necessary is what appears to be necessary. Perelman acknowledges that it is part of our human nature (part of what he calls âuniversal reasonâ) to posit some universal âtruthsâ as necessary. However, universal reason is not posited as necessary on logical grounds but on psychological grounds; that is, human nature requires that we all posit the existence of universal rules. Perelman places himself with those who consider truth claims as ârationalâ and thus âexplicable by something other than itselfâ,5 but points out that this something is âlogically precariousâ, not logically necessary. Perelman considers the problem of error in the context of pragmatism. For pragmatists, what is true becomes so as it is shown to be useful to the person(s) who believes it. This makes âtruthâ âan individual affairâ. If Perelman agrees with the pragmatists that truth cannot be logically necessary in absolute terms, he nevertheless argues that it cannot be individual or private as it can be for pragmatists. âTruthâ is grounded in a relative necessity, which is constituted by arbitrary but accepted rules. All statements accepted as trueâincluding moral rules as well as rules set forth as means of verificationâdepend on a conformity to group-shared assumptions. Therefore, as Perelman writes, âin order to communicate [truth], we must take someone elseâs point of viewâ. As a consequence, and it is here that Perelman concludes, it follows that âtolerance between groups, all of which are established by means of value judgments, is the most immediate practical consequence of our theoryâ.
âDe lâarbitraireâ is an audacious attempt by the very young Perelman to seek a rapprochement between logical positivism and sociology, one that sought to identify the roles played by logic, rationality, and reason in social life. In this article, Perelman both embraces and confronts the positions taken by the main proponents of logical positivism, who dominated philosophical thought in the time period in which Perelman wrote the article.
Logical positivists aimed at developing a system of science that could escape the province of the social, which they viewed as the source of irrational, emotive, and metaphysical impulses undermining rationality. In this article, Perelman takes an âintermediate positionâ to bring reason into alignment with the social world. In developing this realignment, he focuses on the questions of the arbitrary, necessary, and absolute in judgments of truth and value. Perelman maintains that all modes of reasoning and judgments of truth and value must at some point rely on arbitrary and necessary premises. The arbitrary and the necessary, however, need not be absolute; they can be modified through experience and an appreciation of the social contexts that produce human understanding. All groups, Perelman concludes, use value judgments that may be necessary but not absolute to define themselves and others: âIt is the arbitrary that is the soul of a group; it is because of this very arbitrariness that there are multiple groups: the arbitrary leads to pluralismâ.
âDe lâarbitraireâ and Perelmanâs scholarship of the 1930s have been in the main neglected by historians of rhetoric, although there is scholarship on Perelmanâs intellectual trajectory offered by French (Guillaume Vannier) and Italian (Laetitia Gianformaggio) scholars that captures some of the nuances we identify here.6 This neglect was encouraged by Perelman himself: he rarely cited articles from this period as he matured as a scholar, and in his histories of his intellectual trajectory, he would either begin with his first postwar book, On Justice, eliding the 1930s, or quickly dismiss his work of the 1930s as that of a logical positivist in despair about the possibility of reasoning about values. Indeed, Perelman later reported that after having finished his book on justice in 1945, he could not see how value judgments âcould have any foundation or justificationâ in logic or reason.7
During this period, Perelman labeled himself both a logical empiricist and a logical positivist. Logical empiricism/positivism was in part a response to the horrors of World War I, which positivists believed were the result of Hegelian German idealism, impulses, and uncontrolled passions. Perelman had been a student of the leading logicians at the center of the empiricist positivist movement of the 1930s: Gottlob Frege, Kurt Gödel, and Alfred Tarski. His 1938 dissertation focused on Fregeâs system of mathematical logic;8 he contested Gödelâs proof in a 1936 article that created some controversy,9 and he was Tarskiâs student for a year in Poland.10
In the brutal aftermath of World War I, prominent scientists and philosophers concluded that âcultures were incapable of the necessary reform and renewal because people were in effect enslaved by unscientific, metaphysical ways of thinkingâ.11 Logical empiricism identified mathematical logic and experience as the appropriate sources of reasoning, rejected metaphysical thinking, and endorsed the fact/value dichotomy.12 Values were deemed âmeaninglessâ because they lacked an anchor in reason or logic. While working from within the assumptions of logical positivism and empiricism (or in Uebel and Richardâs terms, âneopositivismâ13), Perelman, beginning with his first article in 1931, sought to challenge the assumption that values were, by nature, meaningless and beyond the pale of reason.
Scholars of the trajectories of Perelmanâs thought often miss the diachronic consistency in his scholarship that targeted, from the beginning of his intellectual journey, a rapprochement between reason and values, rationality and the empirical world. Carlin Romano, in his positive review of Perelmanâs philosophy, writes that Perelmanâs âearly training and writings screamed âlogical empiricistââanother of those âwe canât argue about valuesâ types influenced by A. J. Ayer and the Vienna Circleâ.14 Gross and Dearin, in the most complete survey of Perelmanâs intellectual trajectories in English, Chaïm Perelman, concur with Romano and observe,
[f]or the most part, Perelmanâs writings at the end of the 1930s remained strongly anchored in the intellectual currents of that era: Cartesian rationalism, logical positivism, and empiricism. His analysis of several logical paradoxes and antinomies in law had shaken his faith in these doctrines, but his attachments to the orthodoxy of his age had not been completely severed.15
Very few scholars seem to have read Perelmanâs pre-1947 scholarship, and it is clear, based on our translations and commentaries, that Romano is no exception. But even those scholars who have read this pre-1947 work, including Gross and Dearin, do not detect the effort to create a rapprochement between logical positivism and sociology in his scholarship of the 1930s.
We do not hear Perelman âscreamingâ in celebration of logical empiricism in our reading of the fourteen articles and doctoral dissertation that he wrote during the 1930s. The volume and tenor of his scholarly voice is in the main carefully modulated: he strongly endorses the value of reason and finds worth in logical empiricism, while interrogating the assumptions undergirding the doctrine and straining against the intellectual currents of his era in order to expand the range of reason to include values. The Université libre de Bruxelles has now archived Perelmanâs writings and notebooks, and our extensive review of the materials in the archives and a close reading of Perelmanâs articles published prior to 1947 suggest that Perelmanâs postwar turn to rhetoric was more of a progression than a conversion, despite how he and others portray his ârhetorical turnâ.
Perelman did not so much reject logical empiricism as broaden its reach; moreover, he remained loyal to the central premise of logical empiricismâthe commitment to reason and logic. Our work may be seen as complementing that of such scholars as Guillaume Vannier and Laetitia Gianformaggio. Vannier identifies two phases in Perelmanâs response to positivism: the criticism of particular dimensions of logical positivism (tied to the Vienna Circle, and Grzegorczykâs irrational emotivism) prior to 1945, and the criticism of all forms of positivism tout court thereafter. Gianformaggio, on the other hand, identifies four phases in Perelmanâs philosophyâthe pluralist (1933â1945), the emotivist (1946â1948), the dialectical (1948â1950), and the rhetorical (1950 forwards) (429â50). In contrast, we see, beginning with âDe lâarbitraireâ, a dominant diachronic pattern in Perelmanâs response to the problem of logical positivism and rationality, which is distinguished by his effort to construct a link between reason and the social world.
Perelmanâs writings in the 1930s document his efforts both to understand and to critique the doctrines of his time. These early intellectual journeys allowed him to navigate between and among the major dominant schools of thought, beginning with his mentors Marcel Barzin, the mathematician, and Eugène Dupréel, the sociologist. His studies with the major Polish scholars of logic and his dissertation on the German mathematician Frege also allowed him to command the language of mathematics and the tools of logical positivism and empiricism.
Throughout this period, it is clear he remained true to the agenda he had outlined in his 1931 inaugural article, in which he argues for a logic of values, advancing that social groups could coexist in a spirit of tolerance, and that reason could, if properly defined and understood, inform human action. His criticism of logical positivism was never cast as a rejection of its commitment to reason; rather, he saw the logical positivists as unduly limiting reason to the realms of mathematics and geometry. The criticism was meant to justify the expansion of reason into the realms of values and action. âDe lâarbitraireâ constitutes the 1933 launch of Perelmanâs attempt to yoke reason to the social world, an effort that unfolds into a complete vision in Perelman and Olbrechts-Tyteca 1958 Traité.
At the time of publishing âDe lâarbitraireâ, Perelman had already published three articles: âEsquisse dâune logistique des valeursâ [Outline of a Logistics of Values] (1931); âà propos de la philosophie de M. Dupréelâ [On Dupréelâs Philosophy] (1932); and a brief article âLe statut social de jugements de véritéâ [The Social Status of Truth Judgments] (1933).16 In his 1931 article âEsquisse dâune logistique des valeursâ, Perelman establishes his scholarly agenda, describes his quest for a logic of values, and establishes a research program targeting the fact/value dichotomy, which was accepted as an article of faith by many within the logical positivist movement.17 Two years later, Perelman follows this agenda with two articles. In the very brief âStatut social de jugements de véritéâ, which he explicitly identifies in a note to âDe lâarbitraireâ as the latterâs foundation, Perelman briefly outlines how the social adherence to any particular value judgment functions as an underlying precondition for a judgment to be considered factual.18 In âDe lâarbitraireâ, he expands on this notion, elaborating a framework for a logic of values and offering a vocabulary for dissociating the fact/value paradigm that would later find its way into the architecture of the NRP.
One foundational anchor of Perelmanâs effort to bring reason into contact with the social world can be found in the thinking and writings of his mentor, Eugène Dupréel. Dupréel (1879â1967) taught at ULB from 1906â1949 and advocated value pluralism. In a 1932 article on Dupréelâs philosophy, Perelman uses Dupréelâs theory of consolidation to discuss absolute versus relative values.19 Although Perelman would later criticize Dupréel for failing not only to recognize the need for persuasion and argument but also to secure means for reasoning about values, he readily acknowledged that Dupréel had developed a powerful sociological theory of pluralism and the reasonable.20 Dupréelâs thought was an important presence not only in âDe lâarbitraireâ, but in twentieth-century European philosophy generally. The Biographical Dictionary of Twentieth-Century Philosophers summarizes Dupréelâs philosophy by describing its four major principles:
(1) Values and concepts are defined by and ârelated to their contrariesâ.
(2) Values are âmultiple and in tensionâ.
(3) Philosophy offers reason to help mediate the tension between and among values.
(4) Modern societies sponsor value pluralism in which there is continuous âmoral tension, conflict, debate, and accommodation. â¦â
Unsurprisingly, Dupréel had little sympathy for political systems which propose universally valid value systems.21 Perelmanâs âDe lâarbitraireâ reflects Dupréelâs significant influence. In a footnote to âDe lâarbitraireâ, Perelman describes Dupréel as having made an âindelibleâ imprint on him; he also declares that his ideas were tightly âinterwovenâ with Dupréelâs philosophy.
While Perelman acknowledges both of his mentors in âDe lâarbitraireâ (Barzin would serve as Perelmanâs dissertation advisor; see also Perelmanâs 1961 tribute to Barzin22), in this article he celebrates Dupréel, citing from Dupréelâs De la nécessité [On Necessity] (1928), âConvention et Raisonâ [Agreement and Reason] (1925), Traité de morale [Treatise on Morality] (1932), and Le Renoncement [Renunciation] (1929â1930), as well as Dupréelâs unpublished course on metaphysics, for which as a student he had taken careful notes.23 In âDe lâarbitraireâ, Perelman engages directly with Dupréelâs thinking on values and pluralism, which he adopts and transforms. Perelman explicitly quotes in âDe lâarbitraireâ from Dupréelâs âConvention et raisonâ (âIn order to debate, we must be in agreement about somethingâ). He also observes in a footnote to this article that the âidea of reality [as developed in the article] as a social construction comes from Dupréelâs unpublished âCourse on Metaphysics.ââ His ideas on the psychological nature of necessity, Perelman points out, derive from Dupréelâs De la nécessité.
Careful attention to Perelmanâs language also reveals important implicit allusions to Dupréel. His description of how adopting one particular metaphysicâs definition in the search of a definition of reality would limit the generalityâthe généralité nécessaireâof the definition originally sought, for example, acknowledges the influence of Dupréelâs De la nécessité. Perelman will later state in his 1968 article âà propos dâEugène Dupréelâ that he had developed his notion of the accord des espritsâthe meeting of mindsâfrom Dupréel.24 Perelman draws from Dupréelâs thought in order to position the human community, rather than an abstract conceptualization of reason and logic, as the foundation of judgment. Unlike those within the logical positivist movement who sought to reduce knowledge to mathematical and deductive principles, placing questions of ontology and epistemology outside the reach of the human audience, Perelman asserts in âDe lâarbitraireâ that âan object exists only through the agreement of several people and through the possibility of such an agreementâ.
Perelmanâs elaboration of Dupréelâs thought in âDe lâarbitraireâ functions as a common thread underlying the major topics Perelman introduces here: the interrogation of what comes to be known as the fact/value dichotomy; the development of a nonformal expression of reason; the use of dissociation; and the implications of these for social tolerance. And yet these topics are not mere embellishments of Dupréelâs thought. Perelman will later criticize Dupréel for simply settling for the conclusion that values are multiple without setting forth a means to deal with value disagreement, despite his many discussions of persuasion, common sense, a meeting of minds, and the reasonable.25 To prepare the reader for the translation that follows this commentary, we highlight four themes developed in âDe lâarbitraireâ:
1 The Interrogation and Collapse of the Fact/Value Dichotomy
Perelman uses âDe lâarbitraireâ to question what is best known as the fact/value dichotomy. Hilary Putnam, in the definitive book on the subject, writes that âthe fact/value dichotomy (âisâ versus âoughtâ) and the analytic/synthetic dichotomy (âmatters of factâ versus ârelations of ideasâ), was foundational for classical empiricism as well as for its twentieth-century heir, logical positivismâ.26 Putnamâs book chronicles the history of this dichotomy, which he suggests âcollapsed in the face of criticisms by Quine and others early in the second half of the twentieth centuryâ.27 Quine offered his critique in 1951; Perelman, influenced by Dupréel, produced a dissociation (a creative and new pairing of values) of this pair in the early 1930s. However, it is important to note that in this early article, Perelman does not use the language of âfact/valueâ, as readers of our translation will note. Although Perelman continues to maintain a careful distinction between the real and truth,28 he will later engage explicitly with the lexicon of facts versus values as described by Putnam. Perelman and Olbrechts-Tyteca devote significant attention in the Traité in sections 17, 18, 19 to the relationship between and among facts, values, and truth.
In âDe lâarbitraireâ Perelman puts into play three types of judgments: reality judgments, truth judgments, and value judgments. Perelman follows Dupréel and the sociological impulse (and even anticipates Dupréelâs work on values) in order to define value judgments, but he quickly jettisons reality judgments for truth judgments as a means of defining them, declaring, âJudgments of reality could be considered as those that bear on reality, but what is reality? Every metaphysics will respond differently. ⦠Yet for us a broad agreement is essentialâ. He then asserts that truth judgments are a better means of defining value judgmentsâtruth judgments are, he says, âeasy ⦠to defineââbut carefully points out that âtruth judgments can ⦠be just as false as they are true, but it is necessary that they be demonstrated as suchâ. In fact, we might say that Perelman ties truth (and truth judgments) to their verification, which is core to this article, and which will lead to Perelmanâs concern with the justification of value judgments.
To understand Perelmanâs use of truth judgments and reality judgments, and his oscillation from the latter to the former, we must recall the intellectual context in which he is writing and to which he responds in âDe lâarbitraireâ. His definition of âreality judgmentâ derives from the Dupréelian sociological context (itself originating in Durkheim) in which this work originates; however, with âtruth judgmentâ, Perelman explicitly addresses logicians and philosophers. That is, unlike Dupréel, Perelman here addresses philosophers rather than sociologists;29 his use of examples from the hard sciences (chemistry) and his insistence on verification move the subject of this article away from sociology into the field of analytical philosophy.
Our translation here maintains Perelmanâs original formulation, distinguishing these three types of judgment as a means of insisting on the important nuances of his thought. Our decision to preserve the specific characteristics of Perelmanâs types of judgment also drives our endeavor to sustain the distinctions he makes among different kinds of facts in our translation that follows. Perelman separates sense-dataâthe immediate objects of the senses, of sensory awareness as described by William James and later refined by Bertrand Russellâfrom facts by using the synonymous terms le donné and les données in opposition to faits [facts], which are tied to the construction of truth and, by extension, to truth judgments.30 This distinction is important to maintain, for Perelman also characterizes in this 1933 article sense-data as confused notions, which he later develops as an explicit alternative to Descartesâ âclear and distinct ideasâ developed in Perelman and Olbrechts-Tytecaâs âUse and Abuse of Confused Notionsâ.31 The idea of a confused notion will serve as the foundation for Perelmanâs first book on justice.32 Furthermore, Perelman will later describe agreements [accords], which form premises available to the orator, as données upon which future arguments will be based, and thereby highlight that adherence is the aim of argumentation.33
2 The Emergence of a Nonformal Expression of Reason
In âDe lâarbitraireâ, we also see Perelmanâs initial efforts to challenge the restrictions placed on reason by formal logic. From the beginning of his intellectual explorations, Perelman sought to expand the range of reason. He would write in 1979 that âthe claim has even been made that all nonformal reasoning, to the extent that it cannot be formalized, no longer belongs to logic. This conception of reason leads to a genuine impoverishment of logic as well as to a narrow conception of reasonâ.34 In âDe lâarbitraireâ, we see this broadening of logic not only as a general theme but also in Perelmanâs rather idiosyncratic use of such positivist terms as expérience and expérimenter. Here, these terms fall between their usual meanings, so that they mean, respectively, both âexperimentâ and âexperienceâ, and âto conduct an experimentâ and âto experienceâ. In our translation, we try to capture the fluidity of Perelmanâs thought by not choosing one meaning over the other, but by using that which best fits the specific context; âDe lâarbitraireâ is, after all, Perelmanâs initial effort to argue for a more nuanced and supple definition of reason that is tied to the human experience of values and the search for how to reason about them.
Perelmanâs early critique of Descartes and of Kant in this article serves as a springboard to reflections on the arbitrary, which he distinguishes from necessity. âTo posit universality is to assert its necessityâ, Perelman writes. âYet Kant, as well as Descartes, cannot posit it except by means of an arbitrary judgmentâ. In contrast to Descartes and Kant, he contends that logic and reason cannot be absolute, as they must begin somewhere or fall prey to the problem of infinite regress. Quoting Dupréel, Perelman writes, âAll truth presupposes the arbitrary. In order to debate or to seek agreement on truth judgments, something arbitrary must have been accepted beforehandâ. Perelman not only points out the role of the arbitrary in determining truths and knowledge, but also asserts that the method of verification or the rules used to assess facts or values result from an arbitrary choice. Furthermore, he signals that the arbitrary and the necessaryâin the guise of a relative necessity rather than an absolute necessityâare both operative in agreement.
Perelman dissociates the concept of necessity in âDe lâarbitraireâ, suggesting that the arbitrary differs from absolute necessity because the latter is immutable, beyond the reach of time and debate. Truth may depend on certain rules that are logically arbitrary, he argues, but these are not necessary, and we are therefore not required to accept them. Perelman hints in this article at the âregressive philosophyâ he will later develop in opposition to the First Philosophy of Aristotle and the ancient Greeks (which held to timeless truths), which is the focus of his 1949 article âPhilosophies premières et philosophie régressiveâ.35 A regressive philosophy assumes some arbitrary points in any philosophy, but it continues to test these points through a rational dialogue that uncovers and interrogates the underlying claims that serve as justification for them.
There are three additional descriptions and modifications of formal logic offered in this 1933 article that become significant in the NRP: deduction, induction, and analogy. Perelman begins by confronting deduction as formal logicâs preferred means of verification, pointing out that deduction is not simply arbitrary but also sterile as it prevents the development of reasoning. He then posits induction and analogy as more fruitful means of verification. Concerning induction, Perelman writes in âDe lâarbitraireâ, âFundamentally, it constitutes the most fruitful aspect of every syllogism [â¦] it is the basis of a syllogismâs fecundityâ. Here, Perelman identifies the crucial role played by experience in syllogistic reasoning, which then has the capacity to challenge the major premise making up a deduction. It is through induction that a minor premise creates a major premise in a syllogism.
In addition to induction, Perelman underscored the value of analogy and wrote favorably about its use in history and jurisprudence. As Perelman observes here, â[t]he importance of analogy is extraordinary. It is not only used in history and law, but also constantly in social lifeâ. The fact that âsocial lifeâ was the essential component of analogy had led some logical positivists to declare the analogy as âunacceptably metaphysicalâ.36 Perelman acknowledges some of the weaknesses of the analogy, but he still holds that it can offer insight into judgments. Analogies are based on human experience, which Perelman held could inform and help to determine values.
Forty-six years later, Perelman would again assert, as he had intimated in 1933, that â[t]he argument from analogy is extremely important in nonformal reasoningâ.37 Harold Zyskind, an astute critic of the NRP, detected the importance of analogy and juxtaposition in the NRP. âThe form of inferenceâ, Zyskind observes, âof the new rhetoric gives it its specific character. ⦠The form of inference is neither deductive nor inductive, but comparative.â38 As we see in this 1933 article, and which is expressed in stronger terms in Perelmanâs later work, both alone and in collaboration with Olbrechts-Tyteca, the specific inferential character of the NRP is comparative, often expressed in analogous reasoning, which many in the logical positivist movement denied was legitimate.
3 The Notion of Dissociation
âDe lâarbitraireâ presents for the first time Perelmanâs use of dissociation as a technique.39 In âDe lâarbitraireâ, Perelman finds himself faced with two seemingly mutually exclusive systems of judgments, those dealing with reality/truth and the other dealing with values. Rather than accept the âvalue-reality/truthâ pair as an immutable antimony, he discusses a possible âdissociationâ of the two into a third type that would âsurpassâ the original pair. We find in this 1933 article the clear precursors to the Traitéâs âphilosophical pairsâ and to Perelman and Olbrechts-Tytecaâs exposition of dissociation, which Olbrechts-Tyteca declared the most novel contribution of their magnum opus.40 If Perelman only gestures at the principle of dissociation here, he and Olbrechts-Tyteca will thoroughly develop it in the third section of the Traité, describing it as a mode of argumentative reason, rooting their development of the concept in the works of John Locke, Rémy de Gourmont, and Kenneth Burke.41 Dissociation may also derive from Dupréelâs notion of agreement; as Perelman describes, it permits a new order of values to be established within a social order.42
As Perelman understands it, dissociation allows for the creative problem solving needed when those who must make judgments are faced with antimonies. Perelman identifies a judgment that is neither fact nor value, and asserts that this âstrange âthirdâ [tiers] category of judgment would by its presence alone have dissociated the value-reality pairâ. Further, he dissociates truth from a judgment of truth. As he writes in âDe lâarbitraireâ,
any judgment, which asserts that a given proposition is a judgment of truth or a value judgment, can be considered as a value judgment. ⦠As a result, in our conception, truth judgments can indeed become value judgments â¦; conversely, any value judgment can become a judgment of truth the moment that one adopts a means of verification with which its truth or falsity can be demonstrated.
In âDe lâarbitraireâ, Perelman also presents a line of reasoning that anticipates and prepares for his and Olbrechts-Tytecaâs later reframing of the law of noncontradiction with the notions of incompatibility and dissociation:
we must anticipate the possibility of reaching contradictory conclusions demonstrated by different means of verification, all of which we accept. Unwilling to reject the principle of contradiction, we will be obliged either to limit the scope of certain of our means of verification in a way that the contradiction disappears or to hierarchize our means of verification so that we can discern which of the means is preferable in the case of a conflict.
This explanation anticipates and mirrors Perelman and Olbrechts-Tytecaâs notion of incompatibility presented in the Traité, which allows those involved in argument to avoid the perils of contradiction (either A or B) by engaging in dissociative procedures designed to retain both A and B as reasonable.43
4 Social Tolerance: Moving toward the New Rhetoric
In âDe lâarbitraireâ, Perelman describes a rhetorically inflected sense of social truth:
when it is a matter of convincing other people of a propositionâs truth, we must put ourselves in their place, and take their point of view; we must demonstrate what we assert with their own methods of verification. A means of verification that my interlocutors do not accept does not in any way help me to convince them.
We note here how Perelman affirms social truths, that those who seek audiences to accept propositions of truth must adapt to and use the methods of verifications endorsed and accepted by these audiences. Perelman identifies empathy, persuasion, and the need to verify claims (the mark of reason) as essential. This paragraph reveals, in an undeveloped form, a sensitivity to argumentative truths, persuasion, audiences, and what would become Perelman and Olbrechts-Tytecaâs take on epideictic rhetoric.
One can see in âDe lâarbitraireâ (and in other writings of the 1930s, including his 1939 reworking of Nietzscheâs morality, âLa morale des forts et la morale des faiblesâ [The Morality of the Strong and the Morality of the Weak]),44 Perelmanâs efforts to move beyond logical positivism to a vision of reason that could inculcate and sponsor value pluralism; it is here that we also observe Perelman working to legitimize the status of the European Jew. Perelman fully understood the threat he and other Belgian Jews faced with the rising tides of antisemitism. In his 1931 article in which he formulates a logistics of values, Perelman explicitly mentions the Dreyfus affair, which was at the time a shorthand reference to antisemitism, as an indication of the need for multiple values.45
In âDe lâarbitraireâ, Perelman makes a pointed, implicit reference to the Dreyfus affair via an oblique proverbial saying (âthe truth is not to be found in the depths of a wellâ), which evokes the late nineteenth-century image depicted in the painting La Vérité sortant dâun puits [Truth coming out of the well] by Ãdouard Debat-Ponsan (1898). This painting, held at the Musée dâOrsay in Paris, has as an alternative title âLa Vérité (affaire Dreyfus)â. This painting suggests by means of allegory that the truth of antisemitismâs role in the conviction of Dreyfus would be exposed, and Perelmanâs gesture to it is significant.
As one reads Perelmanâs scholarship in the 1930s and understands that he was active in Zionist affairs, that he emergered as a leader of Belgian Jews, and that his wife was a prominent figure in the Zionist leadership of the era, the motivation for his search for a logic of values, one that would allow for value pluralism, comes into focus. If he could as a scholar help to create a system of reason that would give birth to a logic that allowed for multiple and coexisting values, then the status of the Jew in Europe would be better secured. This aspiration of the âtolerance between groupsâ becomes the polar star of the NRP and of the Traité. The interrogation of reason and the outlines of a system of rhetorical argumentation Perelman sets forth in âDe lâarbitraireâ were designed to create a spirit of tolerance based on ontological and epistemological pluralism, and the concepts he offers in this 1933 article will populate Perelman and Olbrechts-Tytecaâs NRP. Perelman concludes the article by declaring that the âtolerance between groups, all of which are established by means of value judgments, is the most immediate practical consequence of our theoryâ.
Translation and Commentary
Perelman, Chaïm âDe lâarbitraire dans la connaissanceâ. Archives de la Société Belge de Philosophie 5:3 (1933) 5â44
Since the end of the nineteenth century and the success of pragmatism, that philosophy of action, the notion of value has become one of the most common in philosophical vocabulary.
Many new approaches to metaphysics consider the pair âvalue-realityâ as fundamental.1
This distinction was soon introduced into logic, where the notion of value judgment was opposed to reality judgment. But the question of whether all judgments entered into one of these two categories was not asked. Confident in this metaphysicianâs distinction, it was implicitly held that any judgment affirmed either a value or a reality. If not, the entire construction would have undergone an irreversible shock. The existence of a judgment that was neither a judgment of value nor a judgment of reality would have made apparent the imprecise nature, if not the inadequacy, of a theory unable to explain it. This strange âthirdâ [tiers] category of judgment would by its presence alone have dissociated the value-reality pair; it would have demanded as a principle of explanation more fundamental notions that would have allowed us to explain all the terms at hand; the opposition âvalue-realityâ would have been surpassed and reduced to something more general. In this way, the very hypothesis of a third judgment that is neither a judgment of value nor of reality destroys all the relevance of this distinction, which draws its very fecundity from its claim to universality.
Yet logically there are only two correct ways to show that a certain genus [genre] contains only two species [espèce]: by defining the two species and showing that the species, thus defined, depletes the genus completely, or by defining positively only a single species, considering the second as the genus minus the first species.2 It is precisely these two methods that are possible for defining value and reality judgments within the category of judgments.
The first method is obviously more elegant, but it contains a difficulty: the obligation to show that there is no other judgment than those judgments which have just been defined. By doing away with this obligation, one commits a sophism known as the sophism of double definition. If one holds that within a genus A, there are only the two species, M and N, one arbitrarily posits that the positive definition of N is identical to the negative definition âspecies of A that is not Mâ. In the cases with which we are concerned here, it is presupposed that what is defined as a value judgment corresponds to any judgment that is not a judgment of reality.
It may be believed that this troubling conclusion (the importance of which is clearly seen) can be avoided by setting the two species alongside each other without defining either one; it may be thought that readers will thus sidestep [the problem] quite well. But if an author may benefit somewhat from this way of proceeding, its lack of rigor will have the same consequences, since readers, in reading and in trying to understandâand thus in defining the fundamental termsâ, will themselves make the logical error of defining the same term twice. They will comprehend this with greater difficulty than if they had been gently led by an informed philosopher.
If readers are cautious and remain on guard, however, they will make use of the second method in order to define two opposing terms; this method, because of its simplicity, does not assume any implicit postulates. If the readers are limited to âvalue or reality judgmentsâ, they will define one of these terms positively, reserving for the second a negative definition.
Which term must be defined in positive terms? Because this is not important logically, reasons that are practical will serve as a guide, and the term which will be chosen will be that for which it will be the easiest to find a precise definition, and on which the broadest agreement can be reached.
Goblot defined judgments of reality as those that are not value judgments.3 This approach was correct, and by using the second method of definition, he did not commit any errors of logic. However, because he did not base his argument on a positive definition, he unfortunately removed all value from the negative definition. In fact, Goblot speaks of value judgments without defining them. Should value judgments be considered as judgments which bear on values? This only pushes aside the difficulty.4 It is precisely because the term âvalueâ is used so often that this notion of value has been defined in every which way, and choosing between these multiple definitions presupposes a philosophical attitude that is far too determined for someone who wants to construct a general logic of value judgments.
The same type of difficulty confronts the person who wants to define judgments of reality in positive terms. Judgments of reality could be considered as those that bear on reality, but what is reality? Every metaphysics will respond differently to this question. To adopt the definition of any one of them is to limit the generality [la généralité nécessaire] of the definition originally sought. Yet, for us a broad agreement is essential.
We have tried to resolve this difficulty by opposing value judgments not to reality judgments but to truth judgments, which will be easy for us to define. This slight modification will lead us to insist on the special nature of value judgmentsâthat they are arbitraryâas opposed to truth judgments, which are necessary.5
We define truth judgmentsâand this is the fundamental element of the problemâas those whose truth or falsity we can demonstrate. Truth judgments can therefore be just as false as they are true, but they must be demonstrated as such. A simple assertion is not enough: we assert value judgments, which are nevertheless defined as judgments that are not truth judgments.
Every truth judgment must be verified [vérifié] or demonstrated. Yet, to demonstrate a judgment, certain rules called means of verification must be employed. We must be engaged in the investigation and in the analysis of these rules.
The most commonly used means of verification is deduction.6 Verification of a deduction is secured through extension and comprehension. In respect to extension, to deduce is to assert about one member of a class what has been asserted about all the members of the class. All of modern logic assumes the validity of deduction by making use of variables and of the principle of the substitution of variables. Logical construction makes use of this principle in order to assert given propositions that have been postulated based on general propositions.
In respect to comprehension, to deduce is to assert from a subject, a predicate P, when all its predicates, including P, have been asserted from it. We can bring this mode of deduction back to the preceding mode by placing ourselves in the point of view of the predicate: by deducing, we assert of one predicate what we have asserted of all the predicates, that they are inherent to a given subject.
Deduction is an application of the principle âwhat is true for all is true for eachâ; this principle is very generally accepted, but it is not fecund: it can never teach us what we do not know. The assertion that sensory experience is a means of verification has another significance entirely: it is indeed experience that allows us to penetrate the real; deduction cannot suffice.
However, experience alone cannot allow us to verify a judgment. Indeed, we understand by means of judgment the statement [énoncé7] of a relationship between the terms. In this way, every judgment presupposes a symbolism, a language whose fundamental operation consists in establishing a correspondence between the sign and the designated [désigné8]. We establish this correspondence by means of deduction.
Let us take the simplest proposition possible: âPaul is writingâ. In order to verify this proposition, sensory experience is totally inadequate, despite all appearances. Indeed, it is enough to point out that someone who doesnât know the language in which it is written cannot verify it; it will also be extremely difficult to teach a primitive man the distinction between âPaul is writingâ and âPaul is drawingâ. To verify this proposition, the sense of sight is not enough.
The verification presupposes the definition of the two terms âPaulâ and âto writeâ. It seems to me that these definitions are not essential, and we do not know of a method of verification with which we may verify them. We will say that these definitionsâlike all definitions, for that matterâare arbitrary: since they are not truth judgments, they must be value judgments. The verification of the proposition âPaul is writingâ presupposes first two definitions, two value judgments, that will constitute that major premise of a syllogism; the minor premise of this syllogism will be verified by sensory experience (I see that the person designated by the name Paul is drawing letters), and whose conclusion âPaul is writingâ will be obtained from these premises. It is obviously possible to go even further and say that it is not by simple experience alone that we observe that âthe person designated by the name of Paul is drawing lettersâ. And this would be right. But we would have wanted to point out only one direction that may be taken, without hoping to arrive at a conclusion.
For that matter, it is pointless to want to arrive at a conclusion.9 To do so is to admit that we can go no further in reasoning, that we will make no progress. The truth is not to be found in the depths of a well; it descends there with the light, and if the depths are not illuminated, they remain somber and unfathomable.10 The truth is perhaps in that which we already know; it is certainly not in what we do not know, as long as we do not know. Ignorance implies only the arbitrary, since a truth judgment is that which we can verify. In reasoning about the real, we place ourselves straight away within the datum [le donné] that we are trying to progressively clarify; we take the most complicated rather than the simplest route.11 Since the datum [le donné] is a confused notion, the goal of our [faculty of] knowledge [connaissance]12 is to unravel it as much as possible, but knowledge never manages to render it completely clear. We see examples of this in language and in law.
Induction is a much more common means of verification than may be believed. Not only is it the tool characteristic of scientific thought, but it is also indispensable to common sense. Fundamentally, it constitutes the most fruitful aspect of every syllogism. In a sense, it operates in the inverse manner to deduction. It is based on the assertion of the existence, proven by experiment [expérience], of a being [être] E possessing the properties A, B, C, D ⦠K, and on the fact that there has not been found a being possessing the properties A, B, C, D that does not also possess the property K. Induction consists of asserting that any being possessing the properties A, B, C, D is identified with the being E, which possesses additionally the property K. Induction is in this sense the transition from the part to the whole, the assertion that a set of tested [expérimentées] properties is tied to certain other properties.
Let us note in passing just how weak the foundations of induction are. Induction is based on the assertion of the presence of one phenomenon and the absence of another. This assertion, which is thought to be sustained by experience alone, is reinforced, however, by the very attitude of scholars knowledgeable about science [savants] who believe in the permanence either of the beings [êtres] they study or of their evolution. This attitude, quite noteworthy in common sense [sens commun13], becomes in science the belief in the existence of universal laws. And it seems that this may be the only productive attitude for science, whose next goal is to foresee the future by explaining the past, something that it can only do by transferring something permanent. And this belief in the permanence of what exists, that is, of what has been tested, inevitably bears on the belief in the permanence of what does not exist, that is to say, of what has not been tested.
This is the weakness of induction, for it is not clear that what has not yet been tested does not exist. All induction is at the mercy of the greater precision of our tools and methods of research.
We have said that induction is the basis of a syllogismâs fecundity. We will attempt to prove this by means of a few examples, such as the following syllogism: the atomic weight of chlorine is 35.46; this gas is chlorine, and so its atomic weight is 35.46. In order for the major premise to be accepted without any restrictions, it cannot be a simple experimental datum [donnée]; for this, one would have had to weigh all the chlorine in the universe, which is clearly impossible. It would indeed be a matter not only of weighing extraordinary quantities of chlorine wherever it is found as an element, but also of weighing it continually and unceasingly in order to be certain that its weight has not changed over time.
Chemists have resolved this question in a much simpler way. They consider that any pure substance is defined by its physical constants, that is, by its measurable properties that are by definition identical in every part of a homogenous substance. They thus define chlorine as a pure substance possessing a set of physical constants A, B, C, D â¦, including that of having 35.46 as an atomic weight. The major premise derives, then, from the very definition of chlorine by means of deduction.
Let us now consider the minor premise: it asserts that a gaseous sample is chlorine. This observation can be considered as the conclusion of a syllogism whose major premise would be formulated by the definition of chlorine and whose minor premise would be formulated by a judgment verified by experiment and by confirming that the sample at hand possesses all the physical constants of this substance. But in this case the utility of the syllogism for asserting that the sample studied has an atomic weight of 35.46 is not felt; this fact [fait] does not need to be deduced, since it has been verified by experiment.
In reality, for the syllogism to be useful, the minor premise must be verified by induction. It is in having tested the properties A, B, C, D, etc. of the substance in question rather than its atomic weight that one asserts by means of induction that this substance is chlorine. Whereas in deduction, since the major premise reminds us that chlorine has an atomic weight of 35.46, we can say the same of the sample which we determine to be chlorine. The syllogism thus has no purpose [raison dâêtre] if the minor premise has not been obtained by induction. We were thus justified in saying that induction explains the fecundity of this mode of reasoning.14 Before moving to the examination of other syllogisms that would allow us to study the various modes of this operation, we will pause at an interesting point, which is that of scientific definitions.
Let us suppose that someone who has just concluded that the sampleâs atomic weight is 35.46 weighs it, and the scale shows that it weighs 37. This result could be interpreted in various ways: Chemists could say that the method used to calculate the atomic weight of chlorine includes an error rate of 5% or that this error is due to carelessness. They will obviously have to accept this latter possibility if, in reweighing again by the same method, they obtain a more probable result. Chemists could assume that the substance which was weighed is not necessarily pure chlorine, either because of the method used for obtaining it or because of the faulty application of this method. In any case, this means that the substance under study does not correspond to the chemical definition of chlorine. Something more interesting: Chemists could assert that the sample under study is purer than those that have been known up to that point or that the methods used to obtain it are more precise and that the definition of chlorine should be modified by substituting the new atomic weight for the old one. This will mean declaring one definition false. But we had thought that definitions were arbitrary, and now we want to make truth judgments of them!
In fact, a chemical definition is more than a definition. What is arbitrary here is the name that is used to designate a substance possessing certain physical constants; however, this substance exists. By asserting that the atomic weight of chlorine is 37 and not 35.46, one asserts the existence of one substance and denies the existence of another because of the assumption that the difference is due to the person conducting the experiment and not to the samples studied. The last hypothesis that chemists might make is that the difference is not due to the scientist, but to the substance under study, which could have all the proprieties of the previous chlorine but a different atomic weight. In the end, it is a matter of another substance which must be distinguished from the first. If chemists speak of two forms, of two isotopes of chlorine, it is because they believe that common chemical proprieties allow us to see these two substances as only two species of the same genus. But someone like Jean Perrin, who is especially interested in the physical proprieties of matter, will give them a different name; he thus proposes the name âhydrumâ for the isotope of hydrogen that has an atomic weight of 2.15
The classical syllogism that we will take as a second example differs little from the first: all men are mortal; Socrates is a man, and thus Socrates is mortal. The major premise again falls under the definition of the word âmanâ, because a human being [être] who is not mortal will be considered as a god or at least as a demigod, and never as a man. The minor premise has been again verified by induction, for reasons analogous to those noted above. If we replace âSocratesâ by the name of a man who is still alive, we will see that it is quite impossible to verify this proposition in any other way. Indeed, to say of a man that he is mortal is to say that he will die; it is to accord to him a property that is in the end only in potentiality [en puissance], and which we cannot verify before it comes to pass in actuality [en acte].16 It is through the intermediary of such an assertion of a property of potentiality that the induction of coexisting properties shifts to the induction of successive phenomena, which is the [type of] induction that gives rise to most scientific laws.
If we say that the speed of the diffusion of gases is inversely proportional to the [square] root of their density, in the end we only define the gases by a property that is common to all of them. The best proof of this is that it is this very property which has allowed the construction of a kinetic theory applicable to all gases. Let us suppose that a gas does not follow this law and that this divergence cannot be explained in any other way: we will need to limit the significance of the law and say that it applies only to gases that are called âregularâ. This will lead to the modification of the definition of the term âgasâ, and regular gases will be considered as one type [espèce] among others, possessing in addition the property of conforming to the law concerning the speed of diffusion.
This same argument is used in such qualitative sciences as law. The fundamental operation here consists of an induction: in describing a legal case [espèce juridique], other properties are ascribed to itâin accordance with the properties that we are testing, and by using these elements of definition that constitute lawsâfrom which can later be deduced from this description. And it is here that we quickly grasp what is common to all these laws: they are only definitions; their analytical nature is the very foundation of their universality. Their fecundity derives from the fact that they were based on experience [expérience], from which the elements of the definition have been drawn.
If we study something having the properties A, B, C, D and we give it a name, we will be able to assertâwithout any errorâthat anything bearing this name will have the properties A, B, C, D. This is the simplest manner by which to solve the serious problem of the foundation of the universality of laws. We have shown above how the modification of a law is nothing other than the modification of a definition; the term that is defined is said to exist. To conclude, let us note that scientific induction is this very same induction that is used in syllogisms. If logicians were not immediately aware of this, it was because in reasoning about symbols they did not insist upon the manner in which each proposition had been verified. In using examples which are caught up in the real,17 it may be seen how in every syllogism the major premise derives from a definition, whereas the minor premise has been verified by means of induction, which allows, moreover, the syllogism to be an interesting and fertile means of reasoning.
Analogy is a more controversial means of verification than other types, but we have recourse to it much more often than we might think. Indeed, it provides a useful means of conceptualizing something that we cannot test [expérimenter]. It is by analogy that we can reason about a fourth term, which is truly undetermined, when three terms are given as data [trois termes étant donnés]. The fourth term is a state of awareness [état de conscience18] of someone other than us; that is, a phenomenon that is not for us an intuitive, self-evident data [une donnée immédiate]. When we reason about a phenomenon that we can test, we use facts [faits] that we know directly: these are the data [données] derived from our senses and our interior experience. We know our actions and our frames of mind [états dââme]; we also know othersâ actions. If we have understood how our inner state reacts to certain actions and influences others, we reason by analogy by assuming that similar actions will hold similar sway over othersâ state of awareness. We grant this awareness to others by analogy when their behavior is similar to ours.
At first glance, it may be tempting to think that induction is at stake here: we perceive from the point of view of othersâ expressions that we have only ever perceived in ourselves when we were aware of them; these expressions [of other people] must thus be accompanied by [their] awareness [of them] as well. This means of reasoning is set apart from induction on an important point: we believe that we can test any induced property sooner or later; however, it is impossible to test someone elseâs frame of mind [état dââme]. There is no clear path from induction to analogy, as may be believed, since the latter becomes more and more vague, less and less precise. Every time we speak of induction, we are talking about the possibility of experience, which is the case for analogy. As Bergson says, âIt is literally impossible for you to prove, either by experience or by reasoning, that I, who am speaking to you at this moment, am a conscious being. I may be an ingeniously constructed natural automaton, going, coming, discussing; the very words I am speaking to affirm that I am conscious may be pronounced unconsciouslyâ.19 If the foundations of reasoning by analogy are weak, a fact [fait] deduced by analogy can, however, be modified, corrected, amended; it is therefore not exactly arbitrary.
A frame of mind [état dââme] is constructed based on how it is expressed; this construction will be better developed with better knowledge of its expressions. But in order for a truth concerning a specific frame of mind [état dââme] to be established, in order for an agreement to be made, it must be assumed that all the human beings [êtres] that are declared to be analogous have similar feelings when they carry out specific actions; however, this is not demonstrable because we cannot know sentiments other than our own. This similarity [identité], this permanent feature of human psychology, could only be postulated. This is, for that matter, the fundamental postulate of historical criticism.
To derive othersâ frames of mind [états dââme] from their actions requires, in addition to knowledge of these actions, a faculty of imagination that brings the historian closer to the novelist. Indeed, it is not enough to put ourselves in the place of the other person who is under study, under examination, to imagine what we would feel in acting in such a way; we must imagine ourselves living in another time, in another context, educated differently, with a different background. This is much more difficult. In the end, we must create another character based on the data [données] we have at hand. We cannot be ignorant of the influence [valeur personnelle] that the historian20 will have on the reconstruction of the past in making the past come to life in the present.
Reasoning by analogy occurs very frequently in law, and here too it is a matter of reconstructing a frame of mind [état dââme], of seeking what is called the âspirit [esprit] of the lawâ.21 In fact, it is the âspiritâ of lawmakers, their intentions and their goal, that must be reconstructed, according to the [legal] acts they have left us, acts that are equivalent to those which allow us to judge our peersâ sentiments. But the role of the imagination is much less important here than it is in history22 because what matters is not so much the lawmakerâs potentially antiquated intention, but the social utility of a specific law, that is to say, the goal that it would serve if it were promulgated today. In seeking the spirit of the law, we put ourselves in the place of the lawmaker, all the while maintaining current social conditions. We will obviously try to understand the lawmakerâs intentions so as not to fall into too much of the arbitrary, but we will not limit ourselves rigidly to these intentions.23 Judges will modify what they believe to be useful to modify. Whence the important role of jurisprudence whose goal is admittedly to reconstruct based on the past, that is to say, the lawmakerâs intentions, but in modifying it [the spirit of the law] after a confrontation with the present.24 The importance of analogy is extraordinary. It is not only used in history and law, but it is also constantly used in social life.
Whatâs more, without it, it would be difficult to distinguish the exterior from the interior world. We may often distinguish internal from external experience, believing that the latter has greater precision, that only it can become a scientific object. But this distinction, which is believed to be fundamental, rests only upon analogy. When we speak of the reality of the exterior world, we do not always realize that this reality does not depend on us alone. What is a vision, a dream, an illusion, if not a thing that we alone see? A vision that everyone could see would resemble, strangely enough, reality. Our understanding of the real is certainly the belief that others, in putting themselves in our place, would perceive the same things we do.25 And this âsame thingâ: we call it âthingâ and it is thus independent of us, precisely because it makes an impression on another person in the same way that it does on us.
What is exterior can be perceived by others; what is subjective cannot be perceived by others; an object exists only through the agreement of several people and through the possibility of such an agreement.26 Now, and this is the important point, such an agreement presupposes reasoning by analogy; it assumes that others, in putting themselves in our place, will perceive as we do something that we can perceive only indirectly. Knowing only their actions and their words, we must by means of analogy reach a conclusion about individual psychological phenomena [fait de conscience27]. What we have developed above is sufficient indication of the importance of analogy.
The fact remains that this mode of reasoning contains a fundamental difficulty: when can it be said that the acts that we perceive are analogous to our own? The very notion of analogy is obscure and imprecise. In saying that one fact [fait] is analogous to another very often sets forth an arbitrary judgment, which is added to that [judgment] by which analogy has been declared as a valid means of verification. Note too that this judgment can have some significance, for it is based on this that consciousness [une conscience28] or a soul will be granted to other creatures. It would be enough to believe that animals cry in an analogous fashion to ours to reject the Cartesian theory of animal machines. It would be enough to see no analogy between the actions of a man and that of a social group to reject as abstruse any notion of a collective soul.
If we seek to verify a legal judgment, we are confronted with difficulties similar to those encountered when we wish to define a term. Is it true that in Belgium a young man cannot be married until he is fully eighteen years old? Yes, I will be told, for it is affirmed by article 144 of the Napoleonic Code. You must then prove to me that this Civil Code is in effect in Belgium. It will be shown that it came into effect in 1807. By whom, I will ask? Did this person have the legal right to institute such a code in Belgium? I will receive answers, but I will continue to ask questions. There will be a moment when no answers will be forthcoming, but I can always continue to ask questions. In fact, in order for certain legal propositions to be considered as true, and thus as verifiable, we must place ourselves within the law and its conventions. Certain propositions will be accepted, which we will not attempt to demonstrate (and if we were to try, we would be faced with the inextricable difficulties raised by the question of the relationship between law [droit]29 and the State). Every legal system, just as every language, operates from arbitrary data [données], conventions, and value judgments.
Among the methods of verification, we think it useful to admit those that are established by specific conventions. Someone who believes that everything that she or he reads in the Bible or in Aristotle is true accepts the text of the Bible or Aristotle as a means of verification. It could be objected that these means of verification are, in the end, only indirect; that they are based on the exalted idea we have of the authors of the Bibleâs, or Aristotleâs, authority [compétence];30 that we believe them because we consider them more capable than we are of finding the truth.
But it matters little why we grant them our belief. As soon as we accept that everything that they say is true because it is they who say so, we do not need to verify their authority at every moment in the matter at hand (for this would mean proving their assertions by another means); we employ the argument of authority that assumes that we do not exercise the most direct means of verification. By considering the text of the Bible or of Aristotle as relating truth, we grant to them the value of being means of verification.
In this regard, an important question can be posed: are means of verification necessary or arbitrary? Are they judgments of value or judgments of truth? Let us lay out the problem in all its generality. Consider a proposition P: to say that P is a judgment of truth is to assert that it can be verified. This is where it will be necessary to draw a distinction: the preceding proposition can mean that P is a value judgment for Paul who is speaking, or that it is a judgment of truth for everyone. In the first case, in saying that P is a judgment of truth for Paul, a truth judgmentâthat is to say a demonstrable verificationâis still asserted. Indeed, Paul can or cannot verify P; if he can, the proposition is true; if he cannot, the proposition is false, but in each case there is a truth judgment, a judgment that is not arbitrary. The same reasoning could be used for the judgment by which Paul asserts that P constitutes for him a value judgment.
In the second case, by asserting that P is for everyone a truth judgment, one claims that everyone admits a process of verification by which one could demonstrate it: the necessity of a means of verification is basically affirmed. Similarly, when one says that P is for everyone a value judgment, one asserts that no one accepts a means of verification with which we could demonstrate this.
Are these assertions arbitrary, or not? Can they be demonstrated, or not? To answer these questions, let us ask ourselves what it means to assert that means of verification are necessary.31 It means that the judgment âa specific means of verification is validâ is a true judgment for everyone, and thus verifiable by everyone. And yet this can only be verified by another means of verification accepted by everyone, a necessary means of verification that would be again verifiable, by everyone, by a third means of verification, etc., ad infinitum. We see that the assertion of a means of verification as necessary would compel us either to keep climbing a ladder with no end or to stop at a means of verification declared to be arbitrary. If one is tied to the idea of always climbing higher on the ladder of reasoning, the necessity of a means of verification is tied to the necessity of the means that precedes it, etc.; that is, a justified necessity will never be reached.
Let us imagine the other possibility: declaring a given means of verification to be arbitrary; that is, considering that the judgment, which asserts the validity of this means of verification, to be a value judgment. If it is a value judgment, it is logically possible not to admit it; if we do not admit it, the means of verification demonstrated by it becomes in turn a value judgment because it is no longer verifiable. It would thus be permissible to reject it, and to reject all the judgments that follow. By admitting that a single means of verification, the one at which we had stopped, is arbitrary, we remove all absolute necessity, and thus all universality, from an assertion that declares the validity of any one of the means of verification whatsoever. But if we are free to reject an arbitrary judgment, nothing prevents us from admitting it, and it is thus logically possible that everyone may accept a given means of verification. We thus cannot demonstrate the truth of a judgment that asserts the necessity of a specific means of verification, nor can we a priori demonstrate its falsity; we can only do so in making an appeal to experience, that is, to a given means of verification. Yet, as it is not necessary to agree to this means of verification, we can, at the conclusion of this long argument, concede that any judgment that is used to claim the particular universal validity of a given means of verification is arbitraryâan unverifiable judgment, a value judgment.
This leads directly to the notion that any judgment that asserts that a given proposition is a judgment of truth or a value judgment can be considered as a value judgment. Indeed, in order for it to be possible to demonstrate that P is a judgment of truth, it must be shown that everyone accepts the means of verification by which it is demonstrated. Yet, as we have seen, this cannot be demonstrated. In the same way, it cannot be demonstrated in a necessary way that there is someone who does not accept a specific means of verification. To say that P is a judgment of truth is thus an arbitrary judgment, a value judgment. By this it is equally arbitrary to assert that P is not a judgment of truth, that it is a value judgment.
In this way we reach the conclusion that it is just as arbitrary to assert a universal necessity as it is to assert a universal arbitrary. And this is understandable. We mean by arbitrary that which is not necessary. When the limits of the necessary are not definitively defined, we cannot require that the limits of the arbitrary be circumscribed once and for all. It can be concluded from the above that speaking of the universal necessity or arbitrariness of a given means of verification lacks rigor. But if we delve deeper into the mechanism of verification, we see that the arbitrary is at the basis of verification in general. We have in fact seen the dilemma posed by the search for a statute [statut] of judgments asserting the validity of a means of verification when this search continually begins anew or when it stops at a means of verification that is judged to be arbitrary. If the first possibility theoretically forbids us from asserting that a given means of verification cannot be demonstrated, it expressly asserts that there is at least one means of verification (unless there is an infinite number) that we will not demonstrate. It is this means that will be arbitrary. If we call an indemonstrable means of verification the basis of verification, we can say that there is always at least such a basis and, in this sense, verification in general is arbitrary at a foundational level. As a result, in our conception, truth judgments can indeed become value judgments, if a means of verification with which the value judgments had been verified is not accepted; conversely, any value judgment can become a judgment of truth the moment that a means of verification is adopted with which its truth or falsity can be demonstrated. By accepting a means of verification, the number of value judgments is increased. Yet it is not absolutely necessary that a given means of verification be accepted, and thus that a judgment be verified. All truth presupposes the arbitrary. In order to debate or to seek agreement on truth judgments, something arbitrary must have been accepted beforehand. There is nothing more accurate than the well-known expression: âIn order to debate, we must be in agreement about somethingâ.32
The above pages probably contain an assertion that could seem to many people audacious. We consider definitions as value judgments: they are indeed arbitrary. But, someone could answer us, if definitions are not truth judgments, they nevertheless should not constitute value judgments; they may not be judgments at all. In order for this objection to be justified, it must either be proven that a judgment does not constitute the statement of a relationship between terms, or our definition of judgment must be rejected. Now this seems to us a thorny enterprise, and it falls to the person who wishes to take it up to show its utility.
Indeed, our opinion is further strengthened by modern axiomatic data [données]. This, following [David] Hilbert, considers an axiomatic system as a set of a definition of terms called âundefinableâ, that is to say that we cannot define them in an explicit fashion by a nominal definition. Axioms would in the end only constitute definitions, the only way to define the fundamental terms of a deductive system. Definition by postulate is in this way opposed to nominal definition: if the operation occurs in a different manner, it nevertheless serves the same goal, and it may not be obvious why someone should take exception to considering definitions as judgments, when no one contested the claim that judgments are a property of axioms.
Establishing this parallel between axioms and definitions also allows us to respond to those who are astonished that we consider axioms to be value judgments, when we consider the propositions deduced from them to be truth judgments. They will tell us that any proposition can be chosen to serve as an axiom, since in a deductive system axioms and propositions are interchangeable; they will not see why we should consider axioms and the propositions that are deduced from them as having a different status. We will respond to this by saying that the case is the same for nominal definitions and for the analytical judgments that we deduce from them. The proposition âman is mortalâ in the end is only deduced from the definition of the term âmanâ, which is a truth judgment, whereas the definition is a value judgment. It is the same in a deductive system in which an axiom, as unverifiable, is logically arbitrary, and propositions, as demonstrable, are truth judgments.
Even those who consider axioms to be intuitions, which are accepted because they are self-evident, consider these axioms (from the perspective of verification) as different from deduced propositions. If deduced propositions have been demonstrated through deduction from axioms, the axioms have only been demonstrated by intuition. It suffices to reject intuition as a means of verification in order for the arbitrary nature of notions to be obvious, and yet the same is not true of propositions of the deductive system.
To avoid any confusion, it would be useful to insist on the relationship between what is generally considered to be true and what we call truth judgments. All that is ordinarily called true does not constitute a truth judgment; we have just seen this for axioms. It is the same for any truth that is considered indemonstrable. In the end, truths are only truths because we want them to be, and axioms are more akin to definition than to observation. All that is not demonstrable is not logically necessary: being arbitrary, an assertion constitutes a value judgment. On the other hand, truth judgments go beyond the realm of truth: indeed, they can make false statements, that is, statements whose falsity we can demonstrate. Now the assertion of the falsity of a false proposition constitutes a true statement because of the principle of double negation, itself deduced from the principle of the excluded middle [tiers]. This principle constitutes in fact a proposition resulting from a certain definition of the concepts of equality and of negation, which is joined to the Aristotelian logic of the concept. These are all arbitrary things, but we accept them, and it is because we accept them that it is indispensable, in order to remain coherent, to grant the same degree of necessity to false judgments as to true judgments, for the principle of double negation allows us to move easily from one to the other. This is the reason why we call both false and true propositions truth judgments: we grant them the same degree of necessity.
The same reason that prompted us to admit the principle of the excluded middle [tiers] leads us to admit the principle of contradiction. Now that we have admitted it, we must anticipate the possibility of reaching contradictory conclusions demonstrated by different means of verification, all of which we accept. Unwilling to reject the principle of contradiction, we will be obliged either to limit the scope of certain of our means of verification in a way that the contradiction disappears or to hierarchize our means of verification so that we can discern which of the means is preferable in the case of a conflict.
So as not to name the well-known conflicts between experimental data [données expérimentales] and arguments of authority, let us note that this rule [of limiting the range of the methods of verification or placing them in a hierarchy] finds most of its applications in law, where propositions frequently can be demonstrated by means of various procedures of verification. Yet, in the case of conflict [between the various means of verification], it is easy to see that any preference given to one over the other can be only made through arbitrarily accepted value judgments. It will be pointed out that our attitude assumes adherence to certain definitions, to certain conventions, to certain value judgments. This in itself shows that our theory is not at all necessary, since nothing is necessary in an absolute sense. We strive only for a necessity that is termed internal, that which is a necessity following the admission of certain arbitrary rules; it seems to me that this necessity has now been attained.
The rigor of our deductions will not, however, prevent errors in the application of our theory. This is essentially an attempt to classify judgments and, as such, it cannot exclude the shortcomings of every classification. A classification is nothing other than a tool; we must know how to use it; the more that we are in the habit of handling it, the more efficient it will be. I have distinguished between truth judgments and value judgments. Now some may accept that a specific judgment is a truth judgment, whereas they do not accept the means of verification that would allow it to be demonstrated; the contrary is not impossible either. But what matters is that it is possible for someone [une conscience33] to determine the methods of verification that he accepts, since the number of these rules is not infinite. If it were infinite, someone [une conscience] might never know, but rules of which we are unaware are akin to being nonexistent; they are not accepted. Yet a judgment of truth must be demonstrated by accepted means of verification.
II
By solving the problem of truth and the arbitrary, the theory that we have set forth above will allow us to resolve several fundamental problems in the history of philosophy; in particular, the problem of the relationship of subject and object, and the problem of error.
The problem of the relationship of subject and object has provoked inextricable difficulties in every philosophy. And if we are going to run through the various solutions that have been given for these, we do so less in order to provide a history of these solutions than to illuminate several angles from which this problem can be posed. We are thus less concerned with seeking historical fidelity than a logical and precise approach that will allow us to clearly see the flaws and also the advantages of the recommended solutions.
In any empirical philosophy, the subject-object relationship is considered especially from the point of view of the object. According to such a theory, our knowledge is formed exclusively by exterior input. Everything comes from sensation; this strikes our faculty of perception, which resembles a blank sheet whose properties are such that they permit sensations to be imprinted on them. The truth is only a correspondence between the object and the mark it leaves on our mind [esprit]. Of what does this correspondence consist? It is difficult to say. But [the principle of] correspondence requires that the mark left by the perceived object always be the same in order for us to be able to speak of truth. Yet, this is only possible when the subject never changes, and when the matter that has been permeated by the sensation [i.e., the object] undergoes no variation.
In a word, this empiricism presupposes that the knowing subject is comparable to a blank sheet not only at the beginning but also over the course of its existence. It cannot evolve, for in being modified it would influence and even change perception in such a way that the knowledge that would be obtained could no longer be called true. Empiricismâs subject should not get in the way of perception; it should be like soft wax, which is what everyone has already said, and yet even this is too much. The subject would be something malleable, impressionable without making an impression, and, in fact, without any agency. These expectations of the subject can be boldly qualified as contradictory.
Faith placed in the fecundity of the empirical explanation has probably suggested to certain neorealist modernists their particular conception of intuition.34 They have found the empiricist theory of the relationship between subject and object to be so lucid that they have not hesitated to apply it when the object was imperceptible and when it could not be known by means of sensation. In order to explain the manner in which ideas can be known, they have considered them to be the reflection of certain essences, created based on the model of things.
Just as our knowledge of things is only the reflection of the things themselves, our knowledge of ideas is only a reflection of their essence. An essence makes an impression on us as a thing, but instead of impressing us via the senses, it directly strikes a faculty of our mind [esprit]35âintuition. And if we wonder of what this intuition consists, we will see that is nothing other than a tabula rasa even more mysterious than empirical intuition because it is endowed with a special faculty of being able to obtain for us self-evident fact [une évidence], a self-evident fact that goes far beyond sensory evidence and that boasts of being absolute.
These two empiricist conceptions of perception and cognition [intellection] assert the absolute reality of the object. The subject is here reduced to an entirely passive role; it must intervene as little as possible. If it is real, nothing is said about it, however. If it is manifest by means of its own attributes, it only creates confusion and error. Classical rationalism differs less than we would think from the empiricism that I have just described. Before Descartes, the essential difference was in the fact that âuniversal reasonâ was used in place of the tabula rasa. Knowledge consisted of an exterior object imposing itself on an impersonal reason. Here, as in empiricism, the personal subject did not intervene in the development of truth; its influence was only observed in error.
Empiricism and any rationalism of this type hardly differ in their aspirations; they set forth something that is complementary to the object and identical in everyone. For some, it is a tabula rasa without any properties, and it is easy to consider it as universal; for others, it is universal reason, whose properties they do not assert. In this way, truth is explained by the correspondence [identité] of the object and the universality of a same subject. All humans are endowed with reason; it is by this that they are human; it is this impersonal faculty that renders them equal. Each person can assert his or her individual personality only in being mistaken; this error is explained by that which is personalâthe intervention of the senses and of memory.
It is with Descartes that rationalism began to move away from empiricism to draw closer to critique, a doctrine that highlights the place of the subject in the development of knowledge. Descartesâ novelty was to observe that matter and souls are reciprocally inscrutable substances. Our thought cannot be influenced by the corporeal [étendue]; the relationship comprised of exterior object/impersonal reason must thus be overturned. Our knowledge of the noncorporeal [étendu] cannot come from the corporeal [étendue]; the idea of the corporeal [étendue] and of its modifications is an internal criterion of our reason.36 Reason, empty in its prior conception, has been filled with ideas. This is the fundamental divergence between rationalism and empiricism, the essential difference between Descartes and Locke.
Descartesâ internalism continues to be tied to an origin that is exterior to reason; critique, on the other hand, assimilated internalism, making it an integral part of reason. It is by means of these ideas that reason knows; any knowledge implies the necessary application to perceptible data [données] of different categories. Knowledge is thus no longer the simple trace which objects imprint on the subject. The subject now plays an active role in knowledge; one agrees to grant it properties that influence perception. A critique of knowledge will become necessary in order to discover the part played by the object and subject in knowledge, since the object appears altered to our understanding.
This conception leads to a modification of the relation called truth; it is no longer a correspondence between the object and the idea that we have of it. The object as such is unknown to us; we only ever know the object as it appears in the subject, that is, the idea that we have of the object. Truth will be obtained by an internal critic, through the coherence of different ideas of a same object. It could be at first believed that this criterion of coherence, established by idealism, draws its value from the principle of contradiction, but this is far from sufficient. If it allows us to assert that there is an incompatibility between A and non-A, it leaves us mystified when it is a matter of seeing if the ideas A and B can be affirmed simultaneously for the same object without incoherence. In order for this to be possible, we must express B according to A or A according to B, and this requires particular rules as well as arbitrary means of verification and definitions.
In these various conceptions of knowledge, there is an idea whose importance has never diminished and whose role has increased with time: the idea of necessity.37 This idea, as Dupréel has masterfully shown, has only ever been defined as a negative property. In the end it coincides with the sentiment of the impossible. It is thus of a psychological nature; what is necessary is what appears to us to be necessary. Now, any relationship in and of itself has nothing that compels our affirmation to the exclusion of other relationships; to claim so would be to endow it with force.38 And it is useful to insist on the fact that necessity bears an extralogical element. A truth is not absolutely necessary because it has been demonstrated; indeed, its necessity will depend on the necessity of the means of verification, which are a last resort and logically arbitrary.
In fact, the idea of necessity derives directly from reality in itself; necessity is only a tracing of reality. In the same way that any reality is imposed on the sphere of being [être], any necessary truth is imposed on the sphere of knowledge [connaître]. To believe in a reality in and of itself is to set forth at the same time the absolute necessity of a judgment, which asserts such a reality. The necessary judgment is imposed as reality; if reality exists in itself, that is to say independently of the knowing subject, the necessity with which it is asserted is absolute. The subject has only to yield to the feeling of self-evidence that it experiences.
This assertion of the thing in itself, postulated by the idea of necessity, implies at the same time an element common to all perceiving beings; indeed, they must all recognize the compelling presence of this reality that is imposed upon them. They will thus share in the common faculty that can be called universal. The notion of universal reason is nothing other than this faculty. This universality is not tested [expérimentée], nor can it be. A universality never derives from experience [expérience] because we cannot conceive of a particular experience allowing us to assert truths independent of space and time. It is impossible to observe a universality; it can only be posited.39 This is one of the fundamental conditions of Kantianism, and it is so much so an essential characteristic of criticism that it can be easily deduced from it.
To posit universality is to assert its necessity. Yet Kant, as well as Descartes, cannot posit it except by means of an arbitrary judgment; seeking the reasons for any necessity always leads back to the arbitrary.
To arbitrarily assert the existence of an absolute necessity is as much to assert its nonexistence. Whence the need for applying Lequierâs famous alternative.40 He was wrong to use it for the problem of liberty and determinism, but this problem is not so fundamental that we cannot arbitrarily exercise our [faculty of] choice. Scholars will not accept the arbitrary, for they are already located within a system of postulates that they cannot refute. Nor will moralists be able to accept this freedom: they are determined to admit free will through their moral system.41 It is only in placing ourselves on the formal level of necessity and of the arbitrary that we can choose. And here we are in agreement with Lequier. We arbitrarily set forth the arbitrary as the basis of every necessity. As we believe that we have already shown, there is no logical necessity that certain truths must be universally accepted or that particular means of verification must be universally recognized as valid.
Are there extralogical reasons for asserting such a universality? It will be worthwhile to examine this problem. To assume that there are universal rules that every person must accept is to assume that these rules are a part of human ânature;â it is to admit the existence of a nature that imposes such rules. This naturalism places the person who asserts it first and foremost in the obligation of asserting a thing in and of itself. A [human] nature is indeed something that is asserted independently of the knowing subject and that is developed according to necessary rules. As a result, this naturalism is incompatible with our viewpoint, the fundamental consequence of which is the negation of every being in and of itself. Let us move beyond this and place ourselves in the opposing camp. Having set forth a [human] nature, with all that follows, how could our opponents demonstrate the universality of this nature? They could only set it forth as only a universality can be set forth, but they will boast of never having been contradicted by experience. Let us note, however, that this is inevitable. As soon as they observe a rule that is not accepted by someone, they will declare it to be not necessary, and they will declare the rules that have been asserted (whose number they will have progressively decreased) to be all the more universal. It is well known today that it is not experience which will determine a priori truths for supporters and their adversaries. Let us suppose that we manage to find two people who do not affirm the validity of any rules they hold in common. In this case, one will not be convinced by experience; however, one of the two will be declared mad, and excluded from the debate.
We should not believe, then, that experience may rattle the naturalistsâ thesis (and I mean by this those who wish to deduce properties of a [human] nature that they have themselves posited). This thesis also cannot demonstrate the opponentsâ thesis, for if they admit the arbitrariness of all rules and of all means of verification, naturalists will always declare that what matters are rules common to all (there are always some among reasonable people) and which result from their human nature. On the other hand, we have seen that the assertion of a universality of nature, which would demonstrate one personâs thesis and destroy that of another, is arbitrary. We are thus also permitted to contest its necessity, even to place ourselves in the naturalistsâ point of view. No one will blame us then for a certain skepticism, which is essentially the negation of every nature and of every absolute necessity.
The skeptic is essentially someone who believes that a truth is not imposed from without by itself. Skeptics accept some truth like other truths, but accept it for reasons that are not purely logical and that they know to be arbitrary. Perhaps they have sought, more than others, to understand, to grasp, and to explain the nature of truth. Yet one thing can only be explained by means of another, different thing.42 To conceive of knowledge as absolute and final is to consider it inexplicable, and it is a profound delusion to believe that explanatory virtues reside in the fact that they are inexplicable and thus obscure.43
This is the illusion of every absolutist theory, of all those who posit an absolutely final term, of all those who are as one in believing that they see clearly, when they are one only in obscurity and ignorance. Any attempt to explain, if not in vain, debases the absolute as it would debase a concrete substance. An absolute or a substance can only be explained âper seâ. If such an explanation may have meaningâand this we may doubtâto comprehend a substance or an absolute would signify that the term to be explained would be identified with a sort of intuition. But it is the intuition that becomes, then, the absolute, the inexplicable, and the explained term is no longer âper seâ.
But hardly any progress has been made, for it is other inseparable and indivisible termsâintuition, nature, or reasonâthat are posited as the atoms of our knowledge. In the end, these are only words, limits [arrêts]44 to our knowledge, which we would like to render clear by treating them as self-evident. Self-evident fact [évidence] is only the perception of something inexplicable beyond which our knowledge stops and gives in. To posit an atom of knowledge is to say not only that we do not know, but that we will never know. An absolute is such an atom, but science cannot admit anything indivisible; it cannot recognize definitive limits, for it assumes that it is always capable of backtracking. It knows that there will always remain something to explain, but this something will become something else because of the progress of science. To posit an absolute is to say that no further progress can be made, and conversely, as soon as one says this, one posits a term that is absolute because it is irreducible, and that will have the same character no matter the name given to it.
Any stance on the problem of truth is confronted with the following alternative: either it is considered as an irreducible and incomprehensible (and thus irrational) datum [donnée], or it is considered rational, and thus explicable by something other than itself, by something that is logically precarious, by something which by its very nature precedes every logical necessity.
The explanation that we have given concerning truth judgments places us among those who accept this latter position. And the principal consequence that ensues is another conception of the problem of error. For absolutists, knowledge is constituted by a relationship between the subject and the object in which truth would come from the object and the subjectâs error. In defining their position in this way, we recall the analysis that we have given above which excludes from the subject all that is found to be universal in it: reason, intuition, etc. These universal faculties in reality only constitute what is objective in the subject; they are not the cause of error. According to absolutists, error derives only from what is personal in the subject, and most of all, from the senses. Error is explained, then, by what is subjective in perception, memory, and by what is subjective in knowledge and in the association of ideas.
But if a certainâadmittedly harmfulâinfluence in the development of knowledge is attributed to the subject, do we not see that any theory of knowledge based on the subject-object relationship is doomed to fail?45 For isnât it necessary to prove that the subject does not intervene in our understanding of self-evident fact [évidence], which is the sole criterion of truth? And if self-evident fact can mislead us, isnât it necessary to seek another criterion of truth?
The problem of error also destroys a theory that is opposed to truth: that of the pragmatists. For them, if we remain within the strict logic of their reasoning, a proposition is true when it is useful to him who asserts it, when he succeeds in making use of it. We thus understand Jamesâs declaration: âThe truth of an idea is not a stagnant property inherent in it. Truth happens to an idea. It becomes true, is made true by eventsâ.46 For James, the truth of a proposition is dependent on the external. On this point, we are in agreement with him: we cause it to depend on the means that we have to verify it. But unlike the pragmatists, we do not believe that truth is an individual affair. We are able to distinguish the utility of some knowledge, of its truth, for a given individual. There are useful errors and harmful truths. How could we be misled if the truth varied according to our liking? Our position here is intermediate, falling between that of the pragmatists and the substantialists. For substantialists, truth depends on nothing; it is necessary. For pragmatists, truth depends on every individual; it has no degree of necessity. For us, truth depends on certain rules that are logically arbitrary, but which constitute the foundation of all necessity. For some, agreement is necessary; for others, agreement can only be made based on interest: it varies accordingly, and contains nothing that is determinative. For us, the rules of agreement are arbitrary but are necessary for the consequences that ensue.
These rules are not necessary; being indemonstrable, they are arbitrary. We accept them; we are not required to do so. We yield to them for reasons of utility or of efficacy; we admit them by means of value judgments, whose arbitrary nature excludes truth. And it is for this that we are pragmatists, but our pragmatism does not reject all necessity; it limits itself to denying an absolute necessity. A relative necessity is the foundation of every truth. There are necessities relative to arbitrary rules that are accepted. Every scientific collaboration, and every joint effort of thought in general, is based on the admission of the same rules, of the same means of verification. This point of view explains error with the same ease with which immorality is explained with reference to morality. There are moralists who believe in an absolute good, who consider immorality as the fact of transgressing a moral rule. It matters little if we accept it: we are supposed to accept it, because there is an absolute good, and the same morality is applied to all human beings. This is also the point of view of those who believe in the existence of an absolute truth: you are in error because you deny this truth that you are obliged to admit.
We are more tolerant, and we know how to distinguish the immoral from the amoral. The cannibal who follows the rules of his group is not immoral because he transgresses our rules. He is only amoral in respect to the rule that forbids eating human beings, since in order to transgress a rule, one must accept it, at least implicitly. To be immoral is to transgress a rule to which we adhere.
In the same way, to be in error or to be mistaken is to place ourselves in contradiction with a means of verification to which we adhere. I am mistaken if I deny the existence of God while admitting the Bible as a means of verification. If I reject the validity of this means, I am not in error. Any assertion is not either true or false: it can be arbitrary. And it is arbitrary when we assert a proposition that is not a truth judgment, whose truth or falsity we cannot demonstrate by means of accepted methods of verification. The rapprochement that we have just made between error and immorality suggests that profound analogies can exist between moral rules and the rules we set forth as the means of verification. And indeed, what characterizes these two types of rules is that they are both arbitrarily set forth by means of a value judgment. These rules are arbitrary, and it is through this that they both constitute the center of a social groupâs unity [ralliement47].
A group will never be formed in order to assert an uncontested proposition.48 In asserting that two plus two is equal to four, one does not differentiate among those who admit arithmeticâs means of verification. They are not differentiated and separated out from others; they do not form a group apart. To form a group is to agree on an arbitrary proposition or set of propositions. It is the arbitrary that is the soul of a group; it is because of this very arbitrariness that there are multiple groups: the arbitrary leads to pluralism.49
It is the social role of moral rules and of the means of verification that explains all the analogies that can be observed between morality and truth. Morality and truth are presented in three different forms: the belief in a universal good corresponds to the belief in an absolute truth; the ideal of a specific good finds a parallel in the belief in a human truth; the theory of individual good or the morality of a strictly personal inspiration corresponds exactly to pragmatism, to the theory of individual truth. The problems of the theory of knowledge, like those of morality, seem to us to be most easily explained through the consideration of the relationships between and among multiple groups.
However, we should be aware of the fundamental difference between morality and truth: their goals are indeed entirely different. A moral rule has value in itself; by asserting it, we place ourselves on a level that is all the more lofty as it is inaccessible. A moral rule confers distinction upon us: this is what allows us to differentiate ourselves from others. A moral rule is made for us, not for others; our adherence is enough for it. The social role for a moral rule is to allow the constitution of a group of action.
A rule that will allow us to assert true propositions does not have the same social function. A truth is characterized by the fact that we want it to be transmissible. We say a proposition is true when we want others to adhere to it as well. A contemplative group is constituted around a truth; the tendency is for a truth to become universal. And vice versa, for that matter: to pursue the ideal of unification in a single group is to believe in the absolute truth of the assertion upon which the group is based. The members of such a group believe in a universal truth that they are charged with disseminating, but their belief is only an ideal, and what easily proves its arbitrary nature is that a group is formed in order to defend it.
The social goal of the truth explains in practical terms why there are far fewer means of verification than moral rules. Indeed, when it is a matter of convincing other people of a propositionâs truth, we must put ourselves in their place, and take their point of view; we must demonstrate what we assert with their own methods of verification. A means of verification that my interlocutors do not accept does not in any way help me to convince them. The value of a means of verification derives from the fact that it is held in common, and the more that a means of verification is commonly held, the more valuable it is. One will be thus less tempted to invent new means of verification than new forms of a moral ideal, because if the latter are of more value by their quality, the former are of more value by the quantity of their adherents.
What characterizes truth is that in order to communicate it, we must take someone elseâs point of view. Science and religion are essentially different because of this: if science attempts to forever diminish the arbitrariness of its assertions, religion does not moderate arbitrariness at all, but rather demands that you place yourself in a certain position where you may accept its specific value judgments. Despite all its pretensions of universality, religion will serve as the ideology for groups to act only if we accept its means of verification and we are convinced as a consequence. If science comes to us, religion, on the other hand, compels us to come to it.50 This establishes the difference between a group of action and a group of contemplation,51 a difference that even goes beyond the distinction between open and closed groups, since both are here open groups.
To conclude, we summarize once again our position, which is an intermediate position [on the questions of necessity and the absolute]. There is a category of minds [esprits] for whom everything is necessary; every judgment, in their mind, is true or false; there are no value judgments. Value is confused with perfection, which is another name for reality. The universe constitutes a unique order; ordinalism goes hand in hand with monism.52 For others, all is value; there is only a dynamic; everything here is force or the struggle of forces. There is nothing stable, no formalism, no necessary truth. Pluralism is tied to the negation of every ordinalism.
As for us, we assert following Dupréel the existence of arbitrary rules, and thus of a plurality of possible orders. To admit a means of verification is necessarily to admit all that can be deduced from it; every truth depends upon a value judgment, but there are value judgments that give rise to truths. Every necessity depends upon an arbitrary assertion: in denying absolute necessity, we do not, however, abolish all necessity. In the same way that every moral act assumes a rule to which one adheres and that, as such, cannot be described as moral, every true assertion assumes an arbitrary rule, which thus cannot be described as true. Morality and truth are found not in rules, but in the application of these rules. This is why passion can play a role when it is a matter of the foundation of arbitrary rules, but the debate that aims at an agreement on truths, following an agreement on the means of reaching it, will seek to place itself within the field of the necessary. The tolerance between groups, all of which are established by means of value judgments, is the most immediate practical consequence of our theory. For that matter, our theory implies strict necessity within a group, which is expressed by a perfectly legitimate use of notions of truth and error.
This chapter is based on M. K. Bolduc, and D. A. Frankâs âAn Introduction to and Translation of Chaïm Perelmanâs 1933 âDe lâarbitraire dans la connaissanceâ [On the Arbitrary in Knowledge]â. Advances in the History of Rhetoric 22. 3 (2019) 232â275.
C. Perelman, âLe statut social des jugements de véritéâ, Revue de lâInstitut de Sociologie 1 (1933) 17â23.
T. H. Costello, âReview of Perelmanâs âDe lâarbitraire dans la connaissanceââ, Journal of Philosophy 36 (1934) 613.
C. Perelman and L. Olbrechts-Tyteca, La nouvelle rhétorique: Traité de lâargumentation (Paris, 1958); C. Perelman and L. Olbrechts-Tyteca, The New Rhetoric: A Treatise on Argumentation, trans. J. Wilkinson and P. Weaver (Notre Dame, IN, 1969).
All quotations here from âDe lâarbitraireâ may be found in our translation, below.
G. Vannier, Argumentation et droit: Introduction à la nouvelle rhétorique de Perelman (Paris, 2001); L. Gianformaggio, âLa tolérance libérale dans la pensée de Chaïm Perelmanâ, in G. Haarscher (ed.), Chaïm Perelman et la pensée contemporaine (Brussels, 1993), 643â661.
C. Perelman, The New Rhetoric and the Humanities: Essays On Rhetoric and its Applications (Dordrecht, 1979), 8.
Perelman wrote a brief PhD dissertation on Frege in Polish under the supervision of Tadeusz KotarbiÅski, which was published as C. Perelman, âMetafizyka Fregegoâ [Fregeâs metaphysics], Kwartalnik Filozoficzny 14 (1937) 119â142. See J. WoleÅski, âThe reception of Frege in Polandâ, History and Philosophy of Logic 25.1 (2004) 47.
See C. Perelman, âLâantinomie de M. Godelâ, Bulletin de lâAcadémie royale de Belgique (Classe des sciences) 6 (1936) 730â736.
J-P Schreiber, Chaïm Perelman et double fidélité, undated manuscript (Brussels).
R. Creath, âLogical Empiricismâ (2011), https://plato.stanford.edu/entries/logical-empiricism/.
H. Putnam, The Collapse of the Fact/value Dichotomy and Other Essays (Cambridge, MA, 2002).
T. E. Uebel and A. W. Richardson, The Cambridge Companion to Logical Empiricism, Cambridge Companions to Philosophy (Cambridge, 2007), 1.
C. Romano, âRhetorically Speaking: Chaim Perelman Rediscovers Western Philosophyâ, Village Voice, May 17, 1983, 47.
A. G. Gross and R. D. Dearin, Chaïm Perelman, SUNY Series, Rhetoric in the Modern Era (Albany, NY, 2003), 2.
C. Perelman, âEsquisse dâune logistique des valeursâ, Revue de lâUniversité de Bruxelles 3â4 (1931) 486â496; C. Perelman, âà propos de la philosophie de M. Dupréelâ, Revue de lâUniversité de Bruxelles 3 (1932) 385â399; C. Perelman, âLe statut social des jugements de véritéâ, Revue de lâInstitut de Sociologie 1 (1933) 17â23.
Perelmanâs daughter notes rightly that her father outlines the blueprint of his subsequent scholarly career in this article. N. Perelman Mattis, âPersonal Interviewâ, by David Frank (4 April, 1994).
Perelman, âLe statut social des jugements de véritéâ.
Perelman, âà propos de la philosophie de M. Dupréelâ, 391.
See Perelman, âLâoriginalité de la pensée dâEugène Dupréelâ, 70.
S. C. Brown, D. Collinson, and R. Wilkinson, Biographical Dictionary of Twentieth-Century Philosophers, Routledge reference, (London, 1996), 206.
C. Perelman, âLe philosophe Marcel Barzinâ, Le Flambeau 3â4 (1961) 171â176.
E. Dupréel, De la nécessité (Bruxelles, 1928); E. Dupréel, âConvention et raisonâ, Revue de Métaphysique et de Morale 32. 3 (1925) 283â310; E. Dupréel, Traité de morale, 2 vols. (Brussels, 1932); E. Dupréel, âLe renoncementâ, Archives de la Société Belge de Philosophie 2. 2 (1930) 1â36. See C. Perelman, âà propos de la philosophie de M. Dupréelâ, 390â391.
C. Perelman, âà propos dâEugène Dupréelâ, 236.
C. Perelman, âLâoriginalité de la pensée dâEugène Dupréel (à lâoccasion du centenaire de sa naissance)â, Bulletins de lâAcadémie royale des sciences, des lettres et des beaux-arts de Belgique 65.3â4 (1979) 70â71.
Putnam, The Collapse of the Fact/Value Dichotomy 9.
Putnam, The Collapse of the Fact/Value Dichotomy 61.
See, for example, C. Perelman, âLa méthode dialectique et le rôle de lâinterlocuteur dans le dialogueâ, Revue de métaphysique et de morale 60.1 (1955) 28.
C. W. Tindale, âWays of Being Reasonable: Perelman and the Philosophersâ, Philosophy and Rhetoric 43. 4 (2010) 337â361.
See William Jamesâs 1897 âThe Sentiment of Rationalityâ, originally published in Mind, and reprinted in A. Castell (ed.), Essays on Pragmatism (New York, 1948), 3â36; B. Russell, The Problems of Philosophy (Oxford, 1912).
C. Perelman, âThe Use and Abuse of Confused Notionsâ, Et Cetera 36 (1979) 313â24.
C. Perelman, The Idea of Justice and the Problem of Argument (New York, 1963).
Perelman and Olbrechts-Tyteca, Traité §15â43.
Perelman, The New Rhetoric and the Humanities, vii.
See our translation here, as well as in D. A. Frank and M. K. Bolduc, âChaïm Perelmanâs âFirst Philosophies and Regressive Philosophyâ, Commentary and Translationâ, Philosophy and Rhetoric 16 (2003) 177â207.
A. Quinton, âAyerâs Place in the History of Philosophyâ, Royal Institute of Philosophy Supplements 30 (1991) 39.
Perelman, The New Rhetoric and the Humanities 22.
H. Zyskind, âIntroductionâ, in The New Rhetoric and the Humanities xvi.
For a history of dissociation and Perelmanâs use of the notion, see D. A. Frank, âThe Origins of and Possible Futures for Chaïm Perelman and Lucie Olbrechts-Tytecaâs Dissociation of Conceptsâ, Philosophy & Rhetoric 53. 4 (2020) 385â399.
L. Olbrechts-Tyteca, âLes couples philosophiques: Une nouvelle approcheâ, Revue internationale de philosophie 33 (1979) 81â82.
Perelman and Olbrechts-Tyteca, The New Rhetoric, 411.
Perelman, The New Rhetoric and the Humanities 65â66.
Perelman and Olbrechts-Tyteca, Traité §46.
C. Perelman, âLa morale des forts et la morale des faiblesâ, Le Flambeau (1939) 183â192.
Perelman, âEsquisse dâune logistique des valeursâ.
TN: Although Perelman will identify Goblot in this article as his primary source for the distinction of value and reality judgments, he alludes here to Dupréelâs sociological vision of the reality judgment, which itself derives from Emile Durkheimâs definition of reality judgments as existing objects (Jugements 437) that are opposed to value judgments (Sociologie). See E. Durkheim, âJugements de valeur et jugements de réalitéâ, Revue de Métaphysique et de Morale 19 (1911) 437â453; Durkheim, Sociologie et philosophie (Paris, 1967).
TN: Perelman uses here the language of taxonomic classification, which may be associated with his interest in natural law. See F. J. Mootz III, âPerelmanâs theory of argumentation and natural lawâ, Philosophy & Rhetoric 43. 4 (2010) 383â402. Perelman will later co-author with Paul Foriers the entry on âNatural Lawâ in the 1968 Dictionary of the History of Ideas, stressing the necessity to define it in relation to its environment. P. Foriers and C. Perelman, âNatural law and natural rightsâ, in P. P. Wiener (ed.), Dictionary of the History of Ideas (New York, 1973), 13â27.
E. Goblot, La logique des jugements de valeur: Théorie et applications (Paris, 1927), 3.
TN: The answer that Perelman provides hereââThis only pushes aside the difficultyââwill reoccur in his 1938 dissertation on Frege when he describes the contradiction at the heart of idealism (and more specifically, in the idea of the âmoiâ subject as the bearer of representations). That is, Perelman points out that for Frege, if the subject is not a representation, idealism must jettison its founding tenet; however, if the subject is a representation, it must be the representation of a being [être], and as a result, the problem still remains. C. Perelman, âÃtude sur Frege (thèse doctorale non-publiée)â (Université de Bruxelles, 1938), 226.
TN: Perelman is referring here to the demonstrability of truth judgments.
TN: See Tarskiâs 1933 Polish âThe Concept of Truth in Formalized Languagesâ, which can be found in English translation (deriving from a 1935 German translation which contains an added postscript) in A. Tarski, âThe Concept of Truth in Formalized Languagesâ, trans. J. H. Woodger, in J. Corcoran (ed.), Logic, Semantics, Metamathematics, 2nd ed. (Indianapolis, IN, 1983), 152â278.
TN: Perelman use of âénoncéâ recalls the linguistic basis of some of the early work on judgments, especially that of Tarksi (following, of course, Aristotle).
TN: Perelman uses the term âdésignéâ in contrast to Saussureâs sign, âsignfiantâ [signifier] and âsignifiéâ [signified], which may indicate a certain amount of agency and choice.
TN: That is, to go beyond the limits of a deduction.
TN: If the notion that truth is hidden and can only be found at the bottom of a dark well derives from Democritus, this proverbial phrase also refers obliquely to Debat-Ponsanâs well-known painting (La Vérité sortant dâun puits/La Vérité (affaire Dreyfus) [1898]).
This proposition and the spirit [esprit] from which we consider it have been borrowed from Dupréel. If the fundamental idea of this article is due to this need of logic that Professor Barzin has been able to make present to our minds, every informed reader will note how much this idea is tightly interwoven in Dupréelâs philosophy, whose imprint on me is indelible. Although many of his ideas seem personalâso much are they incorporated into my own thoughtâit is my duty to mark out the passages in which I have only followed or elaborated my teacherâs thought.
TN: Perelmanâs use of âconnaissanceâ here points to his conception of it as a kind of specific disciplinary knowledge tied to philosophy.
TN: Although elsewhere Perelman uses Latin, here he is careful to use the vernacular form of the âsensus communisâ, the âsens communâ. Perelman later ties common sense to the reasonable in The New Rhetoric and the Humanities: Essays on Rhetoric and Its Applications (Dordrecht, 1979), 118.
TN: Perelman here smuggles into the syllogism the expectation that experience captured with induction, rather than the formal and abstract structure of the syllogism, is the source of the syllogismâs strength.
J. Perrin, âPossibilité dâémission de particules neutres de masse intrinsèque nulle dans les radioactivités
TN: Perelman gestures here at Aristotleâs distinction between potentiality [dunamis] and actuality [entelecheia or energeia] made in book 9 of the Metaphysics. Aristotle, Metaphysics, trans. L. M. Castelli, 1st ed. Clarendon Aristotle series (Oxford, 2018).
TN: That is, examples which have not been artificially abstracted from the messiness of reality.
TN: Perelmanâs use of âétat de conscienceâ refers to the set of phenomena that exists simultaneously in the consciousness at a particular moment in time, the succession of which represents the subjectâs brain activity; it also reveals how Perelman engages directly with contemporary thinking on psychology that places it in terms of philosophy and logic.
H. Bergson, Lâénergie spirituelle (Paris, 1919), 9.
TN: Perelmanâs description of the historianâs âvaleur personnelleâ recalls Aristotleâs definition of ethos in his Rhetoric. Aristotle, On Rhetoric: A Theory of Civic Discourse, trans. George A. Kennedy (Oxford, 2007), 1356a 2,3.
TN: Perelmanâs gesture here at âlâesprit des loisâ is an invocation of Montesquieuâs 1748 work of the same name, which insists on a system of law adapted to the cultural and natural features of the land in which it is in force. Montesquieu, De lâesprit des lois, 2 vols. (Geneva, 1748).
TN: Perelman is alluding here to the historian who brings history to life by means of his imagination.
TN: That is, âwe will obviously try to understand the lawmakerâs intentions so as not to arbitrarily create interpretations that serve our own interests, but we will not refuse to acknowledge the social utility of the interpretationsâ.
TN: That is, âa proper application of jurisprudence would seek to reconstruct the past in order to understand the intent of those who made the laws, but would modify the spirit of the law when the interpretations of the law makers are confronted with contemporary issues facing the judgeâ.
TN: Cournot explains the human tendency to attribute to animalsâ psychological characteristics as one of analogy. See A. A. Cournot, Essai sur les fondements de nos connaissances et sur les caractères de la critique philosophique, 2 vols. (Paris, 1851), 288.
This idea of reality as a social construction comes to us from Dupréelâs unpublished Cours de métaphysique [Course on Metaphysics].
TN: Perelmanâs evocation of a âfait de conscienceâ here may derive from Cournotâs chapter on the scientific nature of psychology (section 371) from his Essai, where he discusses philosophers who make claim to scientific demonstration, founding their philosophical systems on psychology, and in turn, psychology on observation. These philosophers, relates Cournot, distinguish between the observation via the senses (of phenomena from the outside world, of humans in their corporeal nature) and interior observation (the attentive contemplation of individual contingent psychological phenomenaâthe faits de conscienceâthat occur in consciousness, and which are immediately known to us by means of the consciousness we have of them) (312â313). Perelman will later take notes on Cournotâs work in his carnet 21 which he kept in 1940â1941 (Brussels, Université libre de Bruxelles, Perelman Archives BE.ULB-ARCH/89PP 89 12).
TN: Perelmanâs use of conscience here is in line with Aristotleâs notion of human sentience.
TN: Perelmanâs use of âdroitâ here rather than âloiâ means that he evokes the whole of the judicial system rather than a specific manifestation (i.e., law) of it.
TN: Perelmanâs use of âcompétenceâ here recalls the legal language of authority and jurisdiction.
TN: That is, âlet us ask ourselves what it means to assert that judgments require verificationâ.
For this line of thinking, see Eugène Dupréel, âConvention et raisonâ, Revue de Métaphysique et de Morale 32. 3 (1925) 283â310.
TN: Recall that Perelmanâs notion of conscience means here that the person is endowed with consciousness.
Cf. G. Frege, Blätter für Deutsche Philosophie; Zeitschrift der Deutschen Philosophischen Gesellschaft (Berlin, 1918), 58â77.
TN: That Perelman uses here the term esprit suggests that he is not limiting it to an intellectual faculty.
TN: Perelmanâs use of âétendueâ [corporeal] versus the âétenduâ [noncorporeal], evokes the question of the Cartesian definition of substance, first considered by Descartes as dualâthe separation of the body (which occupies space) [the âétendueâ] âfrom the spirit (and which occupies the space of the intellect) [the ânon-étendueâ]. However, the contrast Perelman makes in this article is not Descartesâ âétendueâ versus the ânon-étendueâ but rather the âétendueâ in opposition to the âétenduâ (the term âétenduâ does not derive from Descartes at all). According to the Trésor de la langue française, âétenduâ is an illegitimate translation of âinétenduâ [or ânon-étendueâ], used by Henri Bergson; Perelmanâs use of âétenduâ may thus reveal Bergsonâs influence.
See E. Dupréel, âDe la nécessitéâ, Archives de la Société Belge de Philosophie 1 (1928) 1â40.
Dupréel, âDe la nécessitéâ 29.
I. Kant, Critique de la raison pure, trans. P. Mesnard (Paris, 1932), 41. âExperience teaches us, to be sure, that something is constituted thus and so, but not that it could not be otherwise. First, then, if a proposition is thought along with its necessity, it is an a priori judgment; if it is, moreover, also not derived from any proposition except one that in turn is valid as a necessary proposition, then it is absolutely a priori. Second: Experience never gives its judgments true or strict but only assumed and comparative universality (through induction), so properly it must be said: as far as we have yet perceived, there is no exception to this or that rule. Thus if a judgment is thought in strict universality, i.e., in such a way that no exception at all is allowed to be possible, then it is not derived from experience, but is rather valid absolutely a prioriâ. If we are in agreement with Kant that experience cannot furnish us with a strict universality, we have tried to show above, contrary to Kant, the analytical character of any universality set forth in such a way. TN: The English translation of Perelmanâs original reference to Kant derives from I. Kant, Critique of Pure Reason, ed. P. Guyer and A. W. Wood, trans. P. Guyer and A. W. Wood, The Cambridge Edition of the Works of Immanuel Kant, (Cambridge, 1998), 137.
âI can assert or deny one or the other (liberty, necessity) only by means of one or the otherâ. J. Lequier, L. Dugas, and C. Renouvier, La recherche dâune première vérité (Paris, 1924), 135.
For this line of thinking, see E. Dupréel, Traité de morale §451â61 (Brussels, 1932), T. II p. 521. TN: Perelman will later describe himself as a philosopher and moralist. See his letter to Friedrich Kambartel 18 August 1970. Brussels, Université libre de Bruxelles, Perelman Archives BE.ULB-ARCH/89PP 037.1.
Cf. E. Dupréelâs Cours de métaphysique [Course on Metaphysics].
Cf. E. Dupréel, De la nécessité (Brussels, 1928), Chapter V, âLes deux inconnusâ.
TN: Perelman uses âarrêtsâ here, which suggests that the acquisition of knowledge ceases in this context.
TN: That is, âdo we not see that any theory of knowledge in which the object is the exclusive source of knowledge is doomed to fail?â
W. James, Le pragmatisme (Paris, 1920), 201.
TN: Perelmanâs use of the term âralliementâ, which we translate here as âunityâ, suggests a bringing together and an accord; it is tied etymologically to the English term ârallyingâ.
Cf. Dupréel, âLe Renoncementâ, Archives de la Société Belge de Philosophie 2: 2 (1930) 35.
TN: Implicit here is the idea that multiple groups necessarily hold different sets of propositions.
TN: That is, âIf science persuades us of its truths, religion, on the other hand, compels us to accept its ideologiesâ.
TN: Perelman here seems to characterize a group of contemplation as one that is dedicated to thinking and knowledge; nevertheless, he stresses the traditional opposition of the vita activa and the vita contemplativa. See D. A. Frank and M. Bolduc, âFrom vita contemplativa to vita activa: Chaïm Perelman and Lucie Olbrechts-Tytecaâs Rhetorical Turnâ, Advances in the History of Rhetoric 7 (2004) 65â86.
TN: Perelman here may be evoking Paretoâs work on choice theory in economics. See his later article, in collaboration with Lucie Olbrechts-Tyteca, on Pareto: L. Olbrechts-Tyteca and C. Perelman, âPareto et lâargumentationâ, Uni-Laisanne 37 (1983) 32â33.