Marcelo Dascal

Ex pluribus unum?*

Patterns in 522+ Texts of Leibniz's Sämtliche Schriften und Briefe VI, 4

by Marcelo Dascal, Tel Aviv University


1.  Introduction

2.  Overview

2.1  At long last!

2.2  Material covered and thematic distribution

2.3  Indices, concordances, and some of what can be learned from them

2.4  The problem of dating and the breakthrough year

2.5  Leibniz's linguistic and publication policies

3.  One picture

4.  Another picture

5.  Two pictures in the same text

5.1  Recommendation

5.2  Purgatory

6.  A three-pronged breakthrough

6.1  General Inquiries

6.2  Discourse on Metaphysics

6.3  Examination of the Christian Religion

6.4  Inclining without necessitating

6.5  Human knowledge

6.6  Persuading without obligating

6.7  Division of labor

7.  Impact


1.  Introduction


“A major scholarly achievement”; “The long awaited for result of painstaking and careful work, which surpasses all expectations and justifies all the effort and funds invested in it”;  “An indispensable tool for understanding Leibniz’s thinking”; “A landmark in Leibniz research”; “A gift that contains hitherto unpublished pearls and reveals hitherto unsuspected patterns in the thinking of one of the greatest and most complex human minds”; “An inexhaustible treasure whence generations of philosophers will draw profound insights and wisdom”—these and many other superlatives have been competing in my mind as I perused again and again the 3500++ pages of Volume 4 of Series VI (henceforth VI.4) of the Academy Edition of Leibniz’s writings, the work I am about to review here.

Needless to say, I fully endorse every single one of the above superlatives: The more than a decade long work of Heinrich Schepers and his collaborators in the Münster Forschungstelle, whose outcome are these massive and dense four books into which VI.4 is divided, certainly deserves all that praise and more. “Bravo!” I would doubtlessly shout, were I present at a well-deserved concert-hall premiere of a book that really turns out to be a masterful performance of a polyphonically and structurally extremely rich creation of one of the greatest masters in the history of human thought. (Here am I, again, adding superlatives to my list.)

Short of further rhetorical resources, I have to get down to business and squarely face the extremely difficult and sui generis task that this gigantic collection of texts of various types and degrees of elaboration—texts stemming from different disciplines and domains, written for different audiences, with different intentions and from different points of view, and the vast majority of which were left unpublished by their author—constitutes for a reviewer. For one cannot avoid the crucial question: Which thread, which filum Ariadnes—as Leibniz’s ubiquitous metaphor would put it—should one follow in sorting out, selecting for special attention and discussion, and making sense of the mass of concepts, theses, projects, arguments, comments on others’ work and discussion thereof, and countless other details stored in this huge magazine of ideas the volume under review in fact consists in—to borrow another Leibnizian metaphor?

The ideal thread, of course, would be a master plan dictated by the intrinsic unity of the leibnizian system. But is there such a thing? Is there a specifiable ultimate goal to which the particular goals of each text or cluster of texts are subordinated, a systematic structure or architecture, a set of fundamental methodological procedures uniformly applied throughout all or most texts, a set of key concepts and/or principles, or any other unifying factor that could indisputably be singled out in Leibniz’s mature philosophical endeavors—which are those stored in the magazine before us? Can one discern in these texts and reconstruct from their careful analysis a unity in terms of which each would acquire its incontestable position, function and meaning? Or perhaps there are several such plans, for each interpreter to pick out at her choice? Perhaps what VI.4 reveals is this "Leibniz pluriel" some interpreters talk about? Perhaps it shows, in spite of the Editors' attempt to organize the mass of material it contains, an ultimately unsystematic or pluri-systematic Leibniz, the plurality being irreducible to any particular unity on pain of missing what it is all about?

Since these are notoriously controversial interpretive questions, we should perhaps settle for a more modest pragmatic thread, perhaps one hinted at by Leibniz himself, a sort of inventory and road map he might had suggested for orienting the reader in the miscellany of materials the book contains. Unfortunately, none of these has been provided by Leibniz, who neither published himself nor collected these writings chronologically and thematically as they are here presented. Perhaps we should simply accept and follow the pragmatic thread fashioned by the editors, clearly visible in the organization of the material and explicitly explained and justified in Schepers’ Introduction. Or else, in case of disagreement with them, the reviewer (as any other reader) should just follow his or her own preferences, picking up for special comment those texts that speak most to her or him, highlighting recurring themes and displaying whatever significant connections between the texts s/he sees as showing their coherence—i.e., following his or her own line of interpretation of the author’s thought, either explicitly or implicitly. The reader of this review will rightly suspect that this is what I will end up doing below—for what else could I or anyone else do in what must be kept within the boundaries (which?) of an unusual review such as the present one?

But I will try to do something else too. I will also discuss both the implicit and explicit criteria employed by the editors in their presentation and suggested interpretation of the texts contained in VI.4. For I believe this is a way to try to shed some light on central issues confronting any serious study of Leibniz’s complex thought and his peculiar forms of writing and (not) publishing—a topic to which the Introduction devotes much attention, and rightly so. Indeed, the considerations underlying the organization of the book, along with the general guidelines for its reading suggested in the Foreword and very substantial Introduction, transcend their usual role of giving the reader some practical orientation. In fact, they purport to be grounded on an intrinsic unity of purpose and method the chief editor, undoubtedly one of the leading Leibniz scholars today, discerns in the texts themselves—as he explicitly argues in the Introduction and lets transpire in many of the short introductions to each particular text.

After an overview exploring the volume's resources and already raising some critical questions (Section 2), an attempt is made to identify the overarching pattern read into the ensemble of the texts by the editors’ strong logically-oriented bias, which is operative both in the selection of texts chosen for particular attention in the Introduction and in their comments on these texts (Section 3). I then argue that this bias leads to overlooking other important Leibnizian concerns and, consequently, to hiding other patterns present in VI.4. In particular, I spell out a cluster of motifs that form an alternative pattern, fundamental in my opinion for understanding Leibniz’s rationalism and the nature of his epistemology—a pattern that becomes visible once one looks at the material through other glasses (Section 4). I try to show how the two patterns often coexist in the same text through an analysis of some examples (Section 5). In particular, I suggest a reading of the relationships between the major texts of Leibniz's mature philosophy now critically edited in VI.4, which differs from—perhaps complementing—the reading favored by the editors (Section 6). I conclude by reconsidering from a broader perspective the issue of the impact the publication of VI.4 is likely to have, both in advancing our understanding of Leibniz’s ideas in the context of his time and—taking seriously the programmatic and open-ended nature of many of his projects—in developing Leibniz’s proposals in the context of our and future generations' philosophical concerns (Section 7).


2.  Overview


2.1  At long last!


In 1901 an inter-academic committee decided to publish a complete critical edition of Leibniz's writings. It was decided to sort out the writings in broad thematic categories and to devote to each of them a "Series" comprising several volumes, as well as a special Series containing his enormous correspondence. Each volume was to contain, in chronological order, the writings or letters of a given period. Soon it became apparent that the total mass of Leibniz's production—especially the unpublished manuscripts, which form its vast majority—was far beyond what had been anticipated. Furthermore, the difficulties in transcribing, comparing different versions, and dating each manuscript were enormous, given Leibniz's chronic lack of able secretaries and copyists and his habit of correcting several times what he had written. In spite of continuous work by editorial teams in three (now four) different locations in Germany, so far only 40 of the (now) envisaged 120 large volumes averaging a thousand pages each have been published. Some of the Series, such as the Political Writings and the Correspondence advanced pretty well and are already reaching the last two decades of Leibniz's life. Others—particularly the Philosophical Writings, whose first volume was published only in 1971—remained somehow stuck in the "young Leibniz" of the first decades, except for the extemporaneous publication of the Nouveaux Essais by A. Robinet (which will unfortunately require a second, thoroughly revised edition).

The fourth volume of the Philosophical Writings (VI.4), published in 1999, nearly a century after the inter-academic decision, is the result of two full decades of work by the Münster editorial team, under the direction of Heinrich Schepers. Fortunately, Leibniz scholarship did not have to wait until 1999 in order to begin to have access to the results of the work going on in Münster. A pre-edition of ten yearly fascicles containing the already prepared texts was circulated from 1982 to 1992 among Leibniz scholars, who could thereby also modestly contribute to improving the final edition. But it is only by feeling in one's arms the full weight of these four massive hard-bound volumes, rather than the much lighter paperback ten fascicles, that one can appreciate the amount of work invested and the magnitude of the result, and get a heavy sensory clue to the 'weight' of its content. Perhaps in the future editions such as these will be delivered only on computer disks, though I am afraid this will never feel the same.


2.2  Material covered and thematic distribution


Volume VI.4 contains 522 items (some of them including several variants of the same writing), based on 612 manuscripts, spanning the period between the beginning of Leibniz's employment  by the House of Hanover as ducal advisor and librarian (1677) and his long voyage to Southern Germany, Austria and Italy (1687-1690). They are arranged in six thematic categories: A. General Science, Characteristica, Universal Calculus; B. Metaphysics; C. Natural Philosophy; D. Theology; E. Moral Philosophy; F. The Science of Natural Law. Each part is, in its turn, subdivided in "Texts" and "Excerpts, To Be Excerpted, Marginal Notes". The selection criterion for inclusion of an item in this volume, reflected in its thematic organization, is based on what is described as a "a broader conception of philosophy" (p. xi), which encompasses also writings on the foundations of natural law (F) [with the exclusion of the lengthy and important comments on Lauterbach's Compendium Juris (1679) which is left for a Supplementary Volume], on theology (D) [with the exclusion of those writings "oriented towards theological controversies" (ibid.)], and "elementary writings" on natural philosophy (C) [the more technical ones being left to the newly created Series VIII which is to contain the Scientific and Technical Writings]. One misses similar extensions of the collection at least regarding the "elementary" mathematical writings and at least the non-polemical political ones—both of philosophical significance. The importance the editors correctly attach to Leibniz's reactions to his readings in the period covered is reflected in the fact that roughly one third of the volume's contents is represented by the "Excerpts and Marginal Notes" subdivisions.

            In terms of the amount of individual texts (247) and number of pages (1345), category A is by far the largest, followed by B (104 texts, 610 pages), D (68 texts, 578 pages), F (36 texts, 198 pages), C (33 texts, 162 pages), and E (34 texts, 46 pages). Leaving aside the Moral Philosophy category, which is surprisingly small and contains many virtually undated short texts, it is worth noticing that, by comparison with their total number of pages, the Metaphysics and Theology categories are those with the largest proportion of pages in their Excerpts and Marginalia subdivisions (more than 40% each), whereas the Scientia Generalis, Natural Philosophy and Law categories have each only slightly over 20% of excerpts and marginalia. What exactly these proportions indicate is hard to figure out. On the one hand they reflect the fact that in metaphysics and in theology Leibniz devoted much time to the critical study of authors whose views he either followed or antagonized—a phenomenon characteristic of his thought. This phenomenon is also non-negligible, though less frequent in the Scientia Generalis category, perhaps because he considered his contributions in this domain as less controversial than, say, in metaphysics.

A question that requires further elucidation is whether the impressive fact that the Scientia Generalis field covers 46% of the material in VI.4 may be indeed taken as a sure indication that it constitutes the core of Leibniz's philosophical work and interests between 1677 and 1690 and irradiates its influence over all the other fields—as the editors seem to take for granted. This impression might rather be the residual result of the original division of the Academy Edition into thematic Series, whose arbitrariness the editors admit (p. lxxxiv) and seek to remedy by broadening the criteria for determining philosophically relevant material—which is of course most welcome. Yet, besides the further extensions I suggested above, I must add that I am not happy with their giving up a separate Epistemology category in the volume's thematic organization. Instead, they included texts that clearly belong to what was traditionally known as the "theory of knowledge" in the Scientia Generalis category. To be sure, the questions of method, of conceptual and linguistic analysis, and of the formal development of special-purpose calculi as well as of a universal calculus, which form the bulk of category A, are closely connected with the epistemological questions about the nature of ideas and their representation, of cognition and of what is cognizable, and of truth. Nevertheless, the latter have been traditionally as well as by Leibniz himself viewed as a distinctive cluster of issues, not necessarily subordinated to the formal approach typical of the former. The fact that one of the rare texts Leibniz published in this period is the well-known Meditationes de Cognitione, Veritate et Ideis (1684), which deals with this epistemological cluster of issues, indicates how important Leibniz considered this domain to be, and would per se justify selecting similar texts as its companions for a separate category. In fact, the list of about twenty texts mentioned on page lviii in connection with the Meditationes would have been a good starting point for this purpose.


2.3  Indices, concordances, and some of what can be learned from them


The editors showed great care in providing 500 pages of indices and concordances of various types (subjects, names, writings referred to, quotations from the Old and New Testament, etc.) and binding these pages handily in a separate sub-volume. Leibniz would be pleased to see that his often repeated recommendation to endow his projected encyclopedia with as many indices as possible was judiciously followed by his literary executors and applied to his own encyclopedic work. These indices and concordances, as well as the Introduction, the Table of Contents, and the editorial introductions to each text, provide valuable systematic information that can be translated into guidelines for the reader interested in where to begin his/her reading, as well as for the researcher interested in discovering hitherto unnoticed aspects of Leibniz's thought. In fact one can draw from these sources a sort of synopsis of the materials contained in the volume—in the spirit of Leibniz who considered synoptic tables to be excellent mnemonic and discovery tools. Here are some examples of what could be learned through a statistical-synoptic exercise of this kind.

From the name, writings, and subject indices, in addition to finding a particular reference or use of a concept, one may gather such important information as who are the authors and writings most often mentioned by Leibniz, what are the concepts most often used by him, etc. One may be in for some surprises. For example, Descartes is by far the most mentioned 17th century author, but the next in line is the Hamburg logician Jungius, exceeding by far such prominent philosophers, logicians and scientists as Spinoza, Malebranche, Arnauld, Gassendi, Hobbes, Galileo, Grotius and Bacon—some of whom are also preceded by churchmen such as Cardinal Bellarmin.

From the concordances linking the pre-edition and the published volume one learns that 26 texts of the pre-edition are not included in the volume, while 9 new ones have been added. The main reason for this is more accurate dating.

Another concordance shows the amount of material here published that was already available in the main earlier editions: Couturat (118 texts), Grua (96), and Gerhardt (48). If we take into account that some of the here published texts were published in previous editions other than these three, that there is some overlapping between these editions (especially between Gerhardt and the other two), and that many of the previously published texts were published only partially and appear here for the first time in complete version, it turns out that VI.4 contains an impressive amount of texts published for the first time—slightly more than the 222 mentioned in the Foreword (p. xli).


2.4  The problem of dating and the breakthrough year


From the Table of Contents we can single out the largest texts, the smaller texts in multiple variants (which indicates how much attention Leibniz devoted to their elaboration), and the one-shot pieces. The introduction to each text explains the grounds for its presumed date. This is no easy or uncontroversial matter, for Leibniz dated only 38 texts (a list of the dated texts would have been helpful). For dating the others, the editors rely on an "external" (paper watermarks) and on an "internal" (thematic and terminological similarity) criterion, neither of which is precise and decisive, thus leaving sometimes an excessively large time interval of several years as the presumed date. Consequently, there is no hope of meeting the ideal of knowing what Leibniz thought/wrote each day, not even with the help of the correspondence. Nevertheless it would be possible to have a chronological concordance (not provided) approaching the ideal in question. This would be particularly useful for those interested in (a) establishing the "correspondence" between apparently unrelated texts written more or less at the same time and (b) suggesting some kind of diachronic evolution in Leibniz's thought.

A more precise dating is of particular interest in view of the fact that the mid-eighties are the years in the course of which the mature thought of Leibniz in several domains crystallizes in the form of thoroughly elaborated relatively large texts, easily identified through the Table of Contents and the headings, apparently 'surrounded' by smaller pieces related to them perhaps as preparatory sketches, auxiliary material, etc. It would be very useful to know the temporal sequence in which all this material was written. The major texts referred to include the justly notorious Discours de Metaphysique, the Generales Inquisitiones, and the Examen Religionis Christianae (also known as "Theological System"). Fortunately, the (broad) dating of these major texts is relatively uncontroversial: in all likelihood they were all written in the crucial year of 1686. Along with the smaller pieces directly related to them and well elaborated texts in neighboring domains (e.g., the Meditationes published by Leibniz and written in the second half of 1684), we can speak of a span of a couple of years of intense philosophical work, in which a cluster of extremely significant texts for the understanding of Leibniz's mature philosophy is produced. Just as Leibniz noted, next to the title of the logical treatise Generales Inquisitiones, "here I have made excellent progress", he could have said the same of its coetaneous counterparts in metaphysics, theology, linguistics, epistemology, and other fields as well. The mandatory question is whether and how the advances he himself was aware of having made in all these fields are correlated—a question to be explored below. The answer, whatever it is, should have fundamental implications for assessing the nature of Leibniz's "system".


2.5 Leibniz's linguistic and publication policies


Another point that emerges clearly from the statistics of VI.4 is Leibniz's peculiar linguistic and publication policies. Concerning the latter, as noted, he published only three pieces out of the whole material here contained. Concerning the former, the data are also overwhelming: very few texts are in German, 12% are in French, and the rest in Latin. If one were to take Russell's dictum that only Leibniz's Latin writings represented his philosophically and scientifically serious views, this would mean that in this period of his life he had very serious things to say, as compared with the later periods, in which he wrote two major books (the Nouveaux Essais and the Théodicée) in French. Pace Russell, however, the fact is that both his publication and choice of language decisions were basically guided by audience- and circumstance-oriented criteria. VI.4 contains, for example, two translations made by Leibniz. In 1685 he translated from German into French a dialogue between a devout person and a confessor that constituted the Preface of the Jesuit Friedrich von Spee's Golden Book of Virtues (1649). This he later (1697) corrected and made available to the Duchess Sophie, with whom he usually corresponded in French. Already in 1677 he had praised Spee's book not only for its "solid piety" but also for its inspiring rhetorical virtues, which makes it fit "to be in the hands of all Christians", Protestant and Catholics alike (p. 2515). "This is the true mystery of secret theology, which I don't know”—says Leibniz—“whether anyone (surely in Germany) except Father Spee has dared to openly divulge to the people, even though nothing is more useful, more efficient, and more necessary" (p. 2516). Also in 1685 he translated from English into German the Foreword of the then published second edition of the Duke of Buckingham's Upon the reasonableness of men's having a religion, a book that had been sent to the Duke of Hanover and whose English Leibniz describes as "very intelligible". In all likelihood, the Duke, who presumably couldn't read English, ordered Leibniz to translate his British counterpart's work. Although Leibniz is known for his appreciation of the concreteness of German, as compared to Latin and French, and his later efforts to raise German to the status of a philosophical and scientific language, he always acknowledged the specific virtues of each language and their consequent appropriateness for different purposes, thus remaining essentially a partisan of linguistic pluralism.

Leibniz's astonishing "non-publishing" record is a more complex issue. Different explanations of this policy for some of the leibnizian writings and projects are considered in the Introduction in some detail. The need for secrecy in order to avoid self-defeating effects is the most often invoked reason. Thus, on the assumption that the mega-project of a Scientia Generalis—which Schepers in the Introduction considers the basis for understanding most of Leibniz's work in this period—is deliberately kept secret by Leibniz, the chief editor explains this decision in terms of the enormous scope of the project, which involved both scientific and practical or political dangers. On the one hand, it would be difficult, he argues (pp. lii, lvii), to find the necessary scientific collaborators if they knew that their work, say, in a new encyclopedia or in compiling definitions, was nothing but a component of the larger aim of providing material for an "art of discovery" and for the analysis of concepts that was to lead to the Characteristica Universalis. They would hardly understand these ulterior aims and might be discouraged by their excessively ambitious scope. (On page lxxxv, a similar attitude is ascribed to Leibniz vis-à-vis the host of collaborators needed for his project of law reform.) On the practical, personal and political level, according to Schepers, Leibniz rejected employment offers (in the university, in the French Academy, in the Vatican library) because he didn't consider these to be appropriate venues for carrying out his mega-project, and accepted the Hanover offer because of its prospects in this respect; he later realized that Hanover's resources were not sufficient and sought support in wealthier and more powerful spheres—the Emperor, the Pope, Louis XIV. Yet, Schepers himself asks (p. lxxxv): what prince or king would in his sane mind espouse and finance a project whose ultimate aim is "the happiness of humankind"? In fact, would any ruler (unfortunately, not only at the time) be able to understand the significance of the project and see in its achievement a mission consonant with his ultimate duties and interests? All indicates that, in spite of his secret contacts and efforts, Leibniz's project was condemned to waiting for the opportunity, the kairós, that never came (p. lxxxvi). Ultimately, he had not only to keep the project secret, but actually to bury it, contenting himself with producing instances (specimina) and general principles (initia) he considered as no more than mere illustrations—themselves kept unpublished—of what the Scientia Generalis might have been had it received the support it deserved.

A similar fate was reserved for his major metaphysical and theological works, for similar reasons—according to the Introduction. The Discours de Metaphysique was not kept in Leibniz's drawers because of the unfavorable reaction of Arnauld, a distinguished and most capable representative of the learned world that Leibniz 'used' as a test ground for his revolutionary metaphysics. The reason, according to the Introduction, was rather its complexity and sophistication (pp. lxx-lxxi). Derived from a few incontestable principles, Leibniz's metaphysics required an "inexorably rational" effort for grasping its radical consequences without falling prey to apparent inconsistencies. Viewed as a whole, it is hyper-complex and requires the sophistication of a mathematical mind—a tinctura matheseos (p. lxxi) alien even to "the Great Arnauld"—for its proper understanding. Its improper understanding, on the other hand, might easily lead (as it in fact did) to accusations of "spinozism" and other politically pregnant allegations of atheism. As for the Examen religionis Christianae, praised in the Introduction for its rigorous demonstrative structure, it was primarily conceived to provide the systematic ground—expressing Leibniz's basic beliefs and principles but also palatable to Rome—for the reunification of Christian confessions. Unless there were previous assurances that the Catholic Church would "tolerate" it for this purpose, there was no point in publishing it. In spite of Leibniz's careful steps to obtain such assurances at least from "moderate" Catholic bishops (pp. lxxvi, lxxviii), he did not obtain them: once more the circumstances prevented the publication of this virtually complete and masterful text (p. lxxx). The four complete dialogues that Baruzi called 'mystical', especially the long and beautiful anti-skeptical Conversation du Marquis de Pianese et du Père Emery Eremite, in their turn, are said not to have been published due to Leibniz's fear of raising the suspicion of  heresy (p. lxxx).

No doubt there is some basis for the restraints in publishing attributed to Leibniz in the above explanations. They correctly highlight the fact that Leibniz was no 'pure' intellectual, but rather a 'political animal' (cf. p. xlviii) in the good sense of the term, for—in the footsteps of Bacon—he considered knowledge, first and foremost, as a tool for the improvement of humankind. They also indicate his awareness of the innovative, sometimes revolutionary, nature of his ideas and, consequently, his doubts about the capacity of his contemporaries to understand them. Nevertheless the explanations adduced are far from compelling, to say the least—and not only because they also present a man of Leibniz's stature as excessively concerned with his reputation (pp. lxxxiii, lxxxv) and exaggeratedly fearful of the personal and public consequences of revealing the truths he was sure to have discovered. Maybe this is true and character was not one of his strengths, although his concerns were no doubt justified: Who would dispute the need for first building a solid reputation for oneself in order to be able to advance one's more ambitious projects (p. xlix) and, in general, "the necessity of a good name and respect" (as he titled one of his little practical morality pieces—A IV 3, pp. 883-887)? Wouldn't it be foolish to overlook the danger of being considered a "chimerical philosopher" if he were to publish his metaphysics (p. lxxi), in the light of Arnauld's allegation that Malebranche's metaphysics was that of a visionnaire—a fear later confirmed when Bayle described Leibniz's own metaphysics as bizarre? Could one forget the grim lesson of Galileo's trial, ever present in Leibniz's mind (p. 2068; p. lxxvi; see also VI.1, p. 265)?.

Still, the explanations offered are, in most cases, not the only possible ones, so that at best they form only part of the cluster of reasons that inclined him not to publish. In the Scientia Generalis case, for example, it might simply be the case that he realized how utopian the mega-project was and settled for whatever specimina and initia he could actually deliver—none of which he considered fit for publication. We should not forget too that he did write quite a few "prefaces" which are in fact something like advertisement tracts depicting the marvels of the mega-project or of some part thereof, which were intended for scholars as well as for possible financial and political supporters. In this way he made his project public, albeit without formally publishing it. Furthermore, in his later, widely distributed memoirs on 'scientific policy', dealing with the creation of academies and scientific societies, he referred to his ideal of a general science and its aims. Similarly, not too long after the Discours de Metaphysique, he published in the Journal des Sçavants (1695) the "New System of the Nature and the Communication of Substances, as well as the Union between the Soul and the Body", where he expounded the essentials of his metaphysical doctrine. True, this publication led to a public debate with Foucher, Bayle and others; but Leibniz engaged in this debate willingly, holding his ground and taking it as an opportunity to clear up misunderstandings and clarify his ideas, as he had done earlier in his private correspondence with Arnauld. Nowhere is perhaps the most relevant sense of the 'political' nature of Leibniz's writing and its consequences for his publication policy more evident than in the Examen and the Conversation. By this I mean the fact that the two writings are explicitly audience-oriented and thus require, for their effectiveness, a shared ground between the author and the intended interlocutors (or between the characters in the dialogue) without which they become communicatively ineffective. In such cases, the realization of the lack of such a shared ground may thus be a decisive, 'rhetorical' reason for non publication.


3.  One picture


It is quite clear that the editors of VI.4—especially the chief editor—see the chronological cross-section of Leibniz's philosophical work collected in this volume in terms of a predominant picture that endows the whole collection with whatever coherence a collection of this kind may reasonably be expected to have.

Unlike previous editions, which were not subject to the constraints of comprehensiveness of the Academy Edition, and therefore were openly selective in their choice of texts, the present edition can be safely assumed to be objective in this respect. Therefore, we can further assume that the systematic pattern or picture the editors discern mainly emerges so to speak "bottom up", i.e., out of their deep familiarity with the texts of VI.4, although one cannot altogether discard the possibility that they have been subject also to the "top down" influence of one or another entrenched conception about "what really matters" for a rationalist like Leibniz.

The image of Leibniz the Introduction espouses from the outset is that of a man animated by a global philosophical project, which requires a "revolutionary work that will ground knowledge—the basis of our rational interaction—upon a new, more secure foundation" (p. xlviii). On this view Leibniz perceived himself as a 'radical rationalist' who, "in all fields of spiritual activity, saw his strength in the radical, i.e., directed to the first roots, use of reason" (p. xlvii).

Not surprisingly, the term 'reason' (Vernunft) is one of the most used in the Introduction—as it is in the texts themselves, as revealed by the Subject Index. To the editors' merit, however, it should be noted that they insist in the fact that this particular rationalist also happened to be deeply concerned with the common good, with justice and with faith, for he was a sincere believer—a fact often forgotten by other interpreters. So that what mattered most for him was to use reason for the improvement of the human condition and to combine or reconcile reason with faith. Still, in the picture they present it is reason rather than his other values that has the upper hand, the latter having to be accommodated in terms of the parameters of the former. Of course, this might be nothing but the effect of the objective fact that 40% of VI.4's texts presumably deal with matters of reason alone, with faith and other leibnizian concerns remaining at best in the background.

What are the parameters of this "radical reason" singled out in the editors' picture? Briefly put, they are essentially logic-related: The analysis of concepts into a few primitive ones yielding precise definitions reflecting conceptual structure; the use of demonstration conceived as a chain of substitutions based on such definitions until one reaches an identity proposition; the formalization of the above through an appropriate notation and a universal calculus specifying all valid inferences, which would ensure a foolproof decision procedure for determining truth and resolving controversies as easily as in arithmetic; the use of the mathematical "art of combinations" as the backbone of the method of discovery; and above all the systematization of all knowledge in a quasi-axiomatic structure leading from the simple to the complex, from the foundations to their ultimate rigorously derived consequences. Along with the project of a new, demonstratively ordered encyclopedia that would be a central tool for the "art of discovery" (for it would make patent the pieces of knowledge still missing as well as those still lacking an appropriate foundation), these parameters are those that characterize the mega-project of a Scientia Generalis—an expression Leibniz began to use in his first years in Hanover.

According to the Introduction, it is this predominantly logical notion of "reason"—where analyze!, demonstrate!, formalize!, calculate!, and axiomatize! are the key injunctions—that underlies and informs the bulk of Leibniz's intellectual activities in the period covered by VI.4. Accordingly, it highlights the presence of these motifs in virtually every single text it comments upon, while downplaying or not noticing at all the elements that do not fit the favored picture. By pointing this out, I am not suggesting that the Introduction's comments on these texts are either "wrong" or useless, but rather that they tend to overlook significant aspects of these texts—aspects which, were they noticed, could lead to an alternative picture.


4.  Another picture


The alternative picture I am going to suggest has also emerged "bottom up". Reading, re-reading, teaching, and discussing the texts of the pre-edition in my seminars and research groups, I had the feeling something was missing in the usual way of viewing them, forcefully spelled out in the 'radical reason' picture. I didn't know what it was, but it seemed to me important enough to keep trying to find out. It was not easy because, as so many researchers, in my previous work on Leibniz I had been working under the spell of this picture, which seemed so natural, so persuasive, so logical. Now I am perhaps under the no weaker spell of the other picture, which leads me, I admit, to detect its traces in virtually every text. Fortunately, the Gestalt shift is not so absolute that, having seen the face, one cannot go back and see the rabbit.

In any case, perhaps under the pressure of too many unexplained things I had to brush under the carpet, another pattern took shape in my mind. It stems from the attempt to re-evaluate the non-negligible amount of texts or parts of texts that are either ignored by the 'radical reason' model or, when noticed, are accommodated by this model only very uncomfortably and partially, thus failing to do justice to what is perhaps hidden in them. The image that comes to my mind is that of a vision that focuses on the easily visible full half of a glass, discarding whatever spots there are in the 'empty' half as worthless dirt or 'noise'. What the alternative vision reveals is that the other half is not empty at all, that the dirt is in fact valuable ore worth mining.

            This ore contains, among many other valuables, the acknowledgment of the essential role of and consequent need to investigate and develop the theory and applications of: presumptions that rationally justify conclusions without actually proving them in a formal sense; a hermeneutics capable of taking account of the specific contextual characteristics in which a text is produced or applied in order to determine its appropriate interpretation; an ethics of communication whose basic principle is a principle of charity that presumes the reasonableness and truth of the interlocutor's or author's utterance or text, an attitude that requires the hearer or interpreter to make all the effort needed in order to understand properly "the other", prior to critique or dismissal; a rhetoric of persuasion whose efficacy is predicated upon the use of situated argumentation, which argues from premises acceptable to the audience and takes into account both the relevant context and co-text; non-quantifiable methods for evaluating reasons pro and con a decision; a 'heuristics' (including, among other things, parts of the Topics and the Rhetoric) as a fundamental component of the "art of discovery", which in principle forms at least half of the General Science project; analogical arguments, whose cognitive import is undeniable and yet have no non-analogical counterparts; systematic structure viewed in terms of horizontal analogical relations rather than vertical, i.e. hierarchic deductive/linear relations; epistemic means that can face the skeptical challenge without relying upon the problematic search of absolute certainty; conflict-resolution negotiations run in the spirit of minimalist rather than maximalist strategies, designed for consensus-building and mutual toleration rather than opposition-enhancing; the actual use of all of these or part thereof in the most diverse domains such as jurisprudence, medicine, theology, politics, science, ethics, epistemology, and in the Scientia Generalis itself.

These phenomena are present, at the practical as well as the theoretical levels, in the texts collected in VI.4. In the following sections I discuss in detail a number of texts, analyzing their relationship with the two pictures. Here, very briefly, I mention a few other examples, belonging to different thematic categories, where the phenomena that configure the second picture can be clearly discerned. The Conversation du Marquis de Pianese et du Père Emery Eremite (pp. 2240-2283), included in the theological category of VI.4, is an epistemologically important dialogue that beautifully illustrates, step after step, the argumentative strategy of arguing from premises accepted by the interlocutor (who presents himself as a skeptic) in order to overcome skepticism. The De Deo trino (pp. 2291-2294) provides a 'proof' of the trinity based on an analogy with the human mind, which can be favorably compared with the analytic treatment of the same topic in the more or less coetaneous De trinitate (p. 2346). In Spongia exprobationum ("A shield against censorship: that no kind of true doctrine should be despised", pp. 729-735), whose value the Introduction sees in its defense of 'interdisciplinarity' (p. lvi), Leibniz in fact argues against the "method of exclusion", on the grounds that every serious doctrine or method, if attentively and charitably examined, contains weaknesses as well as strengths, so that no one can erect itself as the arbiter of all the others—a true eclectic manifesto. Finally, in La place d'autruy ("The place of the other"), which the Introduction (p. lxxii) would have included in the Ethics division had it not been already published (IV.3, pp. 903-904), Leibniz presents the Christian charity principle of taking into account the point of view of the other as much as possible as a fundamental guideline for both ethical and strategic action. Other texts highlighting the second picture can be found in the forthcoming collection of texts G. W. Leibniz: The Art of Controversies and Other Writings in Logic and Dialectics (M. Dascal and Q. Racionero, eds.).

There is, thus, a whole set of phenomena, dealt with in a significant number of texts, which transcend the radical reason picture because, on the face of it, they do not conform to—and perhaps cannot be handled in terms of—its basic injunctions, which call for complete analysis, definition, demonstration, and a special form of systematization. No doubt the procedures underlying these phenomena are systematic in their way, which bears, however, little resemblance to the more geometrico model or any other version of the logical-mathematical ideal favored in the radical reason picture. For, they respond to needs that are best handled in other, 'softer' ways exemplified by the list above, ways that do not privilege the quest for order and certainty above all the rest. By bringing together phenomena such as the ones listed above, a pattern of Leibniz's concern with a kind of 'soft rationality' the radical reason picture does not acknowledge clearly emerges.

Defenders of the radical reason picture might argue that there is nothing in the phenomena mentioned that could not be, in principle, handled by the methods characteristic of the radical reason model. Consider a few possible examples of this objection. Probabilistic inferences—the objection would run—are, to be sure, 'weaker' in the sense that their conclusions have a probability tag attached to them, but as Leibniz himself has shown in his essays on the calculus of probabilities (e.g., in "The evaluation of the uncertain", pp. 92-101), "one can always give demonstrations regarding the probability itself" (p. 707). Legal interpretation and the justification of laws can be systematically handled in terms of precise criteria based on analytic definitions, so as to minimize—if not altogether eliminate—the reliance on dubious hermeneutic and justificatory practices. Even the structure of debates can be formally represented so as to ensure that no contribution to the debate is omitted, thus yielding a "Method of disputing until the exhaustion of an issue"—as the title of a little piece advertises (pp. 576-578).

To these objections one might reply, respectively: (A) The so-called calculus of probabilities in fact provides rules for deducing the probability of a compound outcome given the probabilities of its components, and in this it is perfectly deductive; what 'weakens' this calculus are rather the problems that arise in the quantification of the components' probabilities and their respective weight in the compound outcome; whereas these problems can be solved without principled difficulties in those cases—such as card games or dice (dealt with in "The evaluation of the uncertain")—where the formal structure of the game's constitutive rules in fact determines the analysis, this is not the case in general. (B) A text like the "The interpretation, grounds, and system of laws" (pp. 2782-2791), which the Introduction (p. lxxxiv) considers as a juridical example of the radical reason model, undoubtedly has an analytic and demonstrative flavor; yet, the flavor in this case remains flavor and the pudding is quite another story; for this rigorous text, after saying that interpretation involves both interpreting "what is said" and "what is thought" (by the legislator), and providing approximate guidelines for handling the former question, defines the latter as "the search for what the legislator … had in mind regarding the issue at hand, that is what he would have had to say if the question at stake would have been proposed to him" (p. 2783); I hereby challenge anyone to provide a calculus-like procedure capable of finding out what this deontic counter-factual conditional requires the interpreter to determine in each particular application of a law. (C) As for the formal representation of disputes "until the exhaustion of the issue", it is nothing but a diagram wherein one would see clearly how the disputants' interventions are related to each other, i.e., what is a retort to what, and so on; this formal structure—which it is obviously helpful to have, synoptically, under one's eyes—addresses, however, only the 'syntax' of a dispute, not its semantics and much less its pragmatics, which is where 'softness' is required.

            I am not questioning the fact that formal methods can be of help in handling some aspects of the 'soft' phenomena in question. The texts I brought in to support the imaginary (but not quite so) objections indicate how. What I consider problematic is the suggestion that the latter can be fully accounted for by the former. This reduction amounts to employing a process of purifying the dirty ore into formal shining metal which in fact would suppress its value. For that value lies precisely in the dirt, in the fact that, relatively to neat logical form, soft rationality is essentially 'dirt'—imprecise, uncertain, context-dependent, and therefore incapable of dispensing the exercise of human judgment.

The very expression 'soft rationality' seems anathema when referring to the philosopher whose brand of rationalism has been traditionally seen as solidly grounded in his achievements in mathematics, formal logic and their applications. The idea, however, is not to deny the presence and weight of the formal strand in Leibniz's rationalism, but only to point out another strand, without which his encompassing rationalism would be hardly defensible, and his rational optimism would indeed look like a fool's fantasy. For, Voltaire's challenge can be put this way: If this is what your 'reason' has to say about this crazy world, then the hell with it! To which the reply is that he is right if 'reason' is nothing more than 'radical reason'. If, however, 'soft rationality' is brought into the picture and added to 'reason', thus filling the glass and expanding the leibnizian conception of rationality, then, it is Candide that has no case. After having used the expression 'soft rationality', it was brought to my attention that Leibniz himself, in a text belonging (sic!) to VI.4's thematic category "General Science", speaks of a "softer mode of proceeding (blandior tractandi ratio)", which he claims to be necessary even in mathematics (p. 342).  So I have reasons (soft, of course) to believe that the picture I am here suggesting is less of a figment of my imagination than Voltaire's Leibniz.


5. Two pictures in the same text


While there are undoubtedly texts whose objective is obviously presenting the first of the pictures described above, whence the other picture is virtually absent, in many texts the two pictures—or at least elements thereof—co-exist side by side. In some of them, one of the pictures predominates, the other playing an accessory role. Even in these cases, however, whether one sees the non-dominant picture depends on how strongly one is committed to a single-perspective approach. I believe that the two pictures complement each other and neither can be completely overlooked if one seeks to grasp the full import of any of these texts. I will try to illustrate this by examining in some detail two texts, in each of which one of the pictures is clearly predominant.


5.1 Recommendation


The Introduction acknowledges Leibniz's belief in the "rationality of transmitted knowledge", accompanied by his observation that "much effort is needed in order to bring this rationality to light", and immediately goes on to interpret this observation in terms of the "analyze!" injunction: "Only that which can be understood exists before reason. And one only understands something when its composition out of what is simpler to understand is known" (pp. l-li). Why this jump? Why not to take into account the fact that the effort required for properly understanding a text is, for Leibniz, perhaps mainly due to the difficulty in interpreting a doctrine alien to the interpreter because it is unfamiliar, ancient, or regarding which the reader has little or no relevant contextual information and therefore cannot "put himself in the place" of the author? In other words, the effort required for understanding involves as much semantic as hermeneutic skill, so that the Characteristica Universalis analytic-semantic model is, to say the least, insufficient without an adequate pragmatic-hermeneutic counterpart.

Consider, for example, the Recommandation pour instituer la science générale (pp. 692-713). Leibniz here argues against the esprit de secte that prevails in philosophy and stimulates competition rather than cooperation. He urges philosophers to "imitate the geometers, who are neither Euclidists nor Archimedists: they are all for Euclides and all for Archimedes, because they are all for the shared master which is the divine truth" (p. 695). He recommends, especially in philosophy, the use of "an exact rigor in reasoning" (p. 705), criticizes the "alleged demonstrations" of Descartes and Spinoza (ibid.), defines the practice of "always looking for reasons and expressing them distinctly with all possible precision" as the "true method" (p. 707), and claims that, if followed, these recommendations will "free us from the confusion of the disputes" (p. 704). These claims suggest that this text is perhaps one of the paradigmatic examples of texts expressing nothing but the radical reason picture. Since it is probably one of the texts Schepers had in mind in the remarks quoted in the preceding paragraph, it might also be the one underlying the semantic-analytic jump I pointed out.

Yet, the very same text defends (contra Descartes) the value of the Ancients' contributions to human knowledge, "provided they are equitably interpreted" (p. 703). But what is an "equitable interpretation" if not one that takes into account the constraints not only of the semantic but also of the hermeneutic contributions to understanding? Furthermore, the very same text also stresses those non-conclusive parts of logic that have not yet been developed, such as probabilities, presumptions, clues and conjectures, which would be extremely useful in practice, whenever the certainty of deductive logic cannot be achieved (pp. 706-707).

What is more, I would argue that Leibniz actually employs one of these kinds of non-conclusive inferences in conceptualizing the difference between his and the Cartesian positions and justifying his anti-Cartesian move. For what he proposes is to re-interpret the Cartesian position that categorically demands not to accept any proposition as true unless one has demonstrated it as a "recommendation", i.e., as no more than a presumption. Formulated positively, such a recommendation is equivalent to "always try to base yourself upon reasons" (p. 703), to which Leibniz wholeheartedly subscribes. But this no longer implies the further Cartesian demand to consider all unproven propositions as doubtful and therefore as false, for "doubt has nothing to do with it, since we look daily for proofs of opinions which we do not doubt at all" (ibid.). The final step in Leibniz's argument is to cash out this conclusion in terms of a presumption radically opposed to the Cartesian one, albeit both are now perceived as clearly non-categorical, namely: "the oldest and the most accepted opinions are the best, provided one interprets them equitably"(ibid.). This radical change of attitude towards "transmitted knowledge"—which amounts, in its turn, to taking a definite, albeit 'soft', stand in the then raging "Querelle des Anciens et des Modernes"—is made possible precisely by shifting from narrow to broader criteria of "rationality", from a deductive to a presumptive argument.

In so far as both hermeneutics and non-deductive kinds of inference seem to lie beyond the borders of the narrow radical reason model, we have here a text where the two pictures co-exist side by side. Perhaps it is to their complementary role that the Recommandation's concluding warning also alludes: "I hold that one should distrust reason by itself, and that it is important to have experience or to consult those who have it" (p. 713).


5.2 Purgatory


Whereas the presence of the alternative picture in a text belonging to the Scientia Generalis corpus may surprise some readers, its presence in texts belonging to the other thematic categories of VI.4 is more likely to be expected—pace the editors' tendency to stress even in them the radical reason picture. The theology category is a case in point. No doubt it is part of Leibniz's endeavor to formulate a "rational theology" capable of reconciling reason and faith and reuniting around it all Christians—or at least Western Christendom. Several of the texts in this category are indeed guided by this aim. A particularly interesting example is his attempt to provide an abstract definition of 'Catholic Church'—'catholic' meaning 'correct' or 'true', and hence universal. The concept is thus reduced to two fundamental principles, love of God and dedication to the common good: "Whoever loves God truly and strives for the public good sincerely is truly in the Catholic Church; all the others are schismatic" (p. 2161). These principles are sufficiently general and uncontroversial, so that the new definition would provide the rational basis for reunification and should command the assent of every true Christian believer—or at least define such a believer, since Leibniz believed that there is no "obligation to believe" (see below 6.6).

Yet, Christianity is based not only on reason but also on the "transmitted knowledge" of a revelation—a series of historical events preserved in non-transparent sacred texts, whose interpretation was from the outset quite controversial. The problem for a rational believer like Leibniz, who does not give up the essentials of this revealed knowledge, is, then, to strike a balance between the two components—rationality and revelation—without sacrificing either of them. The solution of this cardinal problem requires, on the one hand, the establishment of reasonable criteria of interpretation of the Scriptures and, on the other, a careful consideration of the adequate model of rationality to employ in this context. In particular, one has to ask whether the radical reason model would not render the task impossible due to its excessive strictures. Furthermore, since the solution must be not only intellectually satisfactory but also politically effective in the sense of satisfying at least the basic vested interests of the conflicting Christian factions, it is not far-fetched to expect that, at least in some of the texts in the theology category, the constraints of strict rationality would give way to a somewhat softer notion of reasonableness, which in turn would render the interpretation task easier to perform.

That Leibniz was deeply involved in the painstaking interpretive work needed in order to deal with controversial theological questions is well known. In VI.4, Leibniz's hermeneutic ability is, for instance, masterfully put to use in an elaborate discussion of the sources adduced in defense of the Catholic conception of the purgatory, which was contested by the Protestants, especially by the Calvinists (Consideratio locorum potissimorum quae pro Purgatorio ex Scriptura Sacra et Sanctis patribus adducuntur—pp. 2123-2147). The Introduction mentions this text (p. lxxiv) without comment, but I believe it deserves to be commented upon. The text is mentioned in the context of Leibniz's efforts to provide a "rationally grounded basis" for the debates on controversial theological issues, relying for this purpose on his newly elaborated metaphysics (p. lxxiii). The difficulty in understanding this rational basis is stressed and a distinction is made between the degree of intelligibility of the proposed solution of the controversial theological issues for which the understanding of the metaphysical "basic elements" is sufficient and that of those issues that require the explanation of the connection between these elements (ibid.). The former are easy issues, the latter, difficult ones. The latter comprise the problems of human freedom and God's grace, of God's power and how he puts it to use, the explanation of evil in his creation, etc. They are difficult because they depend upon the comprehensive understanding of the metaphysics of possible worlds, the complete notion of individual substances, etc. This distinction has of course implications for the communicative and rhetorical strategies to be adopted in persuasively presenting and explaining "the one truth" to different audiences.

The purgatory doctrine is related to the theological doctrines of punishment, grace and resurrection, thus involving both the difficult metaphysical points just mentioned and the not easier one concerning the nature of soul and body and their connection, so that it would presumably be classified as belonging to the difficult side. The amount of work Leibniz devoted to this question supports this conjecture. What is interesting, however, is that his work focuses not on the allegedly difficult metaphysical "rationally grounded basis", but rather on the apparently easy question of how to interpret the passages usually quoted in the purgatory debate—a list of which, following that of Cardinal Bellarmin who took part in this debate, Leibniz provides (pp. 2145-2147). In his discussion, he displays his usual erudition and ingenuity, as well as an extraordinary resourcefulness in making use of a rich and varied arsenal of analytic, hermeneutic, and argumentative moves. It is worth examining how he achieves in the Consideratio a rather peculiar combination of all these tools.

The overall structure of this text is dialectic, being determined by the Protestant opposition to the Catholic doctrine of the purgatory—according to which the soul of a sinner, immediately after death, is subjected to the expiation of his/her sins, so that after being purged from them it can be admitted to beatitude (p. 2124). Leibniz's objective is to verify whether those passages in ancient Christian sources employed to justify this doctrine actually do it. After filtering out the many invoked texts that are clearly irrelevant, he leaves aside also those passages that only defend the practice of praying for the deceased (for this is admitted by the Protestants) on the basis of the dialectical principle that "it is useless to prove that which is conceded" by the opponent (p. 2126). The same principle allows him to omit those passages that can be understood as referring to the "purgatory of resurrection", for "the Protestants concede a great deal" of this doctrine. As we shall see, passages of the latter kind are not omitted at all. In fact, Leibniz's main strategy in the Consideratio turns out to be an attempt to show, for as many of the adduced passages as possible, that they justify this second kind of purgatory but not the first one.

Since the exegesis of the passages will turn on the distinction between the two kinds of purgatory, Leibniz first puts to use his analytic skills in defining precisely the first and then highlighting the second's contrasting features vis-à-vis it. "Purgatory I" is defined as "the place or state of believers' souls in which, immediately after death, they expiate for some time their sins with severe torments and are helped by the living's prayers and sacrifices in order to be admitted, once divine justice is placated, into beatitude even before resurrection" (p. 2126). "Purgatory II", by contrast, takes place after resurrection (rather than immediately after death and before resurrection). It concerns "the whole man after the re-connection of soul and body" (rather than the separate souls alone), and is universal, i.e. applies to everyone (rather than only to particular sinners) albeit with different duration and intensity (p. 2127). Both purgatories are temporary states and in both fire and its cognates figure prominently.

In the light of these contrastive and shared elements, passages that speak only of temporary expiation after death, of fire and of purgation are not specific enough as proof of Purgatory I. Therefore, they can be treated as referring to Purgatory II, being thus irrelevant to the Catholic-Protestant dispute. Consequently, Leibniz declares that they are not going to be considered in the Consideratio, for he does not want to bring up passages "that are so easy for the Protestants to reject" (ibid.). As a neutral and objective analyst and judge of the controversy, he applies thus the dialectical/juridical principle according to which the "opponent" must prove his point by confronting the hardest—not the easiest—pieces of counter-evidence. Another clarification he makes of the notion of Purgatory II might, instead, favor the Protestants. Consider the possible objection that this notion is contrary to divine justice, since its universal application implies that the pious and the impious alike, after having been judged by God, are subjected to the fire of Purgatory II. If admitted, this incongruence would prevent the Protestants from relying on this notion as they do. Leibniz comes to their help by reporting that the tradition distinguishes (a typical disputatio move employed to duck an objection) between two kinds of fire—a purifying and benefic one (for the pious) and a soiling and excruciating one (for the damned). The former is further explained through an analogy: it acts upon the pious like the heat of a high furnace acts upon gold by making it more beautiful (p. 2127).

The framework set, Leibniz gets down to the task of examining in detail the passages that pass the relevance test. In each case, the presumption that the passage proves the existence of Purgatory I is weighed against evidence for interpreting it in terms of Purgatory II. Obviously, such a tactic only makes sense within the dialectical macro-structure of the whole text. But a variety of moves, based on different dialectical and hermeneutic principles, are employed for its implementation.

Consider a few examples. Of the many passages from Origen adduced, all but one are dismissed as interpretable in terms of Purgatory II. In the passage making the exception, however, Origen is shown to reduce Hell to Purgatory, which makes this passage unreliable, for it does not preserve an established distinction (p. 2128). Tertullian's expression 'the jail of hell' is taken to mean "purgatory", but it might mean "hell" tout court (p. 2130). However, the fact that this expression appears in a sentence explaining that sinners are there jailed until "later resurrection" (mora resurrectionis) dispels this ambiguity for it points to temporary punishment rather than to eternal damnation. Nevertheless, mora resurrectionis is itself a "quite obscure" expression (p. 2131). A very detailed syntactic, semantic and contextual analysis of this expression follows, leading to different possible meanings, all bearing on whether 'the jail of hell' means Purgatory I or not. The chosen meaning is the one supporting a positive answer; this is the meaning also favored by Cyprian, a disciple of Tertullian who "usually respects and follows his teacher" (p. 2132). Nevertheless, this reliance on Cyprian's endorsement does not prevent Leibniz from criticizing Cyprian's own passage—another of the sources adduced—for the "many doubts" that plague it, which preoccupied Roman Catholic authors. After analyzing such doubts, Leibniz concludes by attributing to Cyprian a distinction his passage must presuppose, "unless we want to admit that he wrote confusedly" (ibid.)—a nicely formulated use of the "principle of communicative charity". And so on and so forth.

In order to get the feel of the subtle details of Leibniz's deliberative argumentation, let us pause for a slightly longer look at his considerations about what can be learned from Basilius. To begin with, his reference to Isaiah 4 is claimed to refer to Purgatory II, for he explicitly says that the Lord's "examination" (mentioned by the prophet) of the sins of the daughters of Sion in view of their eventual absolution is actually performed through "the fire of the judgment" and "the heat or boiling of the judgment" (ibid.). This not only means that the examination takes place at a stage of afterlife—presumably the "final judgment"—which is well beyond the immediacy of death, but also strongly suggests a literal reading of 'fire', which would be inconsistent with the metaphorical reading implicit in Purgatory I, since souls without bodies cannot be literally affected by "fire". Leibniz admits that he was first inclined to accept the objection that 'judgment' need not be understood here as "final judgment", for it could refer to any examination, including the one that no doubt takes place in Purgatory I (p. 2129). Against this, however, Basilius' further claim, still in the context of Isaiah 4, that the judgment in question occurs "in a future century"—a phrase that usually refers to the final judgment—is decisive (p. 2130). Leibniz singles out another passage of Basilius, this time referring to Isaiah 6. Part of this passage interprets the "burning coal", which an angel picks up from the altar with fire-tongs in order to touch Isaiah's lips thereby purifying him, as acting upon his soul, thus supporting Purgatory I; another part, however, might be seen as stressing the corporeal effect of the fire that "inflames the heart" of sinners, thus apparently inclining towards Purgatory II. The authority of Paul's passage on the straw and salvation "as if through fire", whose 'as if' fits the Purgatory I reading, is brought into the discussion and perhaps is the decisive factor for Leibniz's conclusion that this passage ultimately supports Purgatory I. Commenting on Procopius Gazaeus, who "seems to have had Basilius's words in front of his eyes" (p. 2129) and interprets him as favoring Purgatory I, Leibniz argues that this author inconsistently mixed up the two passages, so that his interpretation is useless (p. 2130). The conclusion of all this is that only one of Basilius' passages, namely, the one in Isaiah 6, can be plausibly understood as referring to Purgatory I. Yet, how much plausibility there is to it? Well, certainly not much, given Leibniz's own doubts and hesitations, the unreliability of earlier interpretations of Basilius, the veiled appeal to authority, and the shaky grounds for deciding between literal and metaphorical readings.

However weak the grounds for the conclusion reached turn out to be, the elaborate process of careful and detailed weighing of the reasons pro and con a purported interpretation of a passage is not otiose. First, because it leads, after all, to a conclusion—albeit a merely plausible one (remember what the inability of deciding in favor of one or the other heap of straw did to Buridan's poor ass!). Second, because it seems to be the only possible way of performing the exegetical task. And finally, because whoever has followed the argumentation cannot but be fully aware of the strengths and weaknesses of the conclusion reached, i.e., of the approximate relative weight to be assigned to it in the final step of deciding about the main issue discussed in the text, namely, whether the current Catholic doctrine of Purgatory I is supported by good ancient textual evidence.

            It is not surprising, therefore, that at one point, after summarizing the conclusions that can be drawn from the passages discussed so far, and pointing out that the scales tend to support Purgatory I, especially due to the clear statements of Augustin and others in this direction, he adds that "the issue is not completely clear", and goes on to express his wonderment as to how far it has been left hanging [by the Ancients] who often kept silent about the purgatory, when the occasion demanded that they talk most and clearly about it. His conjecture is that this shows that the doctrine was not sufficiently established at that time (p. 2139), presumably because it developed later. In any case, "the Ancients did not talk about the things of the other world we today assert—and this is certain" (p. 2140). What is surprising is that, in spite of all the careful weighing, the cautiously hedged formulations he had been using all along, and the general difficulty in understanding the Ancients' views on the other world, Leibniz concludes categorically: "The Ancients would unanimously condemn anyone who claimed that the prayers for the dead are useless and that all souls are either immediately saved or immediately damned" (p. 2140). This is a wonderfully crafted statement that deserves, per se, exegesis, but the reader by now might be weary of this. Let me just say that this syntactically double conditional, which is semantically a hypothetical statement—perhaps even a counterfactual, seems to have been pragmatically very precisely worded in order to fulfill a specific dialectical role: to have the full authority of the Fathers of the Church anachronistically condemning … the Calvinists! (Foucher de Careil saw in the earlier Confessio Philosophi of 1673 the expression of "Leibniz's victory over Calvinism", a view Yvon Belaval rejects on the grounds that Leibniz did not want to defeat Calvinism, but rather to reconcile it with Lutheranism; see Belaval's introduction to the Confessio, pp. 14-15.)

            The punch line aside, it is clear that, of the two pictures, the one that stars in this text is certainly not the "radical reason" model.

To be sure, there is plenty of linguistic and conceptual analysis, necessary for rendering the passages considered sufficiently intelligible for the argumentation to proceed. But obviously the analysis falls short of reaching the level of primitive concepts. Definitions are reached, to be sure, but they are at best of the type Leibniz will call in the Meditationes 'nominal', which provide only distinctive features of the definiendum. No 'real' definitions are within reach, so that none of the key concepts can be even proven to be contradiction-free and thus possible. Consequently, no formalization is sought that would permit a calculus-like procedure to be applied so as to compellingly and finally decide the issues.

To be sure, the laws of logic are respected: invalid inferences and inconsistencies are to be avoided. Yet, since logic alone is insufficient to yield the interpretations sought, other principles of inference—pragmatic, hermeneutic, dialectic—are also used. They include, as we have seen, the logic-related (but not reducible to logic) requirement of intra- and inter-textual coherence (eliminating inconsistencies and "accommodating" one passage with the others as well as with authoritative traditions—p. 2130), as well as the rhetorically effective ad hominem and ad verecundiam arguments, the dialectic distribution of the charge of proof, and the pragmatic maxims underlying the assignment to a speaker of metaphorical or implicit meanings. In fact, a few lines before the end, Leibniz both makes a pragmatic inference and warns against the uncertainty inbuilt in such inferences. Having argued that in one passage the Archbishop of Caesarea, Arethas, tacitly avoids an implication (that all martyrs are immediately saved upon their death), Leibniz points out the impossibility of generalizing this conclusion on the basis of a single obscure and controversial passage: "I don't dare to affirm anything certain about this, especially since, if one takes into account the connection between words, it appears that another sense can also emerge" (p. 2144). In current pragmatic jargon, he would state this as the principle according to which attributions of implicit intentions to a speaker or author are essentially co-text and context sensitive—which makes them reliable only in a case by case basis. Involving as they do all these non-logical principles, the core of the Consideratio's arguments cannot be subsumed under the label 'demonstrative' without further ado.

The text is no doubt conscientiously systematic. But its systematicity does not follow the axiomatic model. It is rather that typical of disputation, i.e. dialectical in nature. Its basic procedure is that of a deliberation in which reasons favorable and contrary to a purported conclusion are weighed against each other until a decision is (or is not) reached—a decision that remains, however, tentative and contestable in the light of further considerations.

Nevertheless, there can be no doubt that a text such as this (as well as similar ones in law, politics, medicine and other domains where interpretive work is essential) instantiates what—if anything—can be called "rational discourse". But in order to grasp its rationality one has to transcend the narrow domain of 'radical reason' and see that what, from the point of view of this model, appears as weaknesses are in fact, from the point of view of a 'softer' model of rationality, strengths that permit reason to apply to a much broader range of human endeavors. It would seem that, in theology, softer rational procedures like the ones displayed in the Consideratio—and it is not the only example thereof—cannot be replaced by stricter ones, which can at most complement them. In so far as this text of the beginning of 1677 must be seen as part of the "Catholic Demonstrations" ambitious project Leibniz elaborated already in Mainz—which according to the Introduction informs his work also in Hanover—it seems necessary to conclude that the word 'demonstration' in such a project cannot be reduced to its radical reason model reading.


5.  A three-pronged breakthrough


As already pointed out, in 1686 Leibniz writes three of the most elaborate and significant texts of the period covered by VI.4—one in logic, one in metaphysics, one in theology. He expressly manifests his feeling of significant progress in one of them, but he certainly felt similarly regarding the other two as well. It is beyond the scope of the present paper to spell out in detail the precise nature of the breakthrough in each of these fields. After an overview of each, I will rather discuss three topics they deal with—which will permit me to address the question, raised above (2.4), of whether and how the progress made in these fields is related. My suggestion is that the answer cannot be given solely in terms of any one of the pictures we have been discussing, but requires taking into account the interplay between them.


5.1  General Inquiries


The "General inquiries about the analysis of concepts and of truths" (pp. 739-788) is a complex and extremely rich text about which much has been written, though no consensus about why Leibniz was so fond of it emerges from the commentaries. The philosophical cornerstone of the text is Leibniz's 'analytic' theory of truth, according to which in every true proposition the predicate is contained in the subject. The purpose is to implement this theory by formal means. The strictly formal approach reaches a high level of generality, for it seeks to develop an abstract calculus whose variables can be interpreted as concepts as well as propositions and whose axioms can serve to analyze (i.e., to 'prove' or 'demonstrate') both and to account for the widest possible range of inferences. In this respect, it goes beyond the Characteristica Universalis, whose 'characters' are not only formal but 'material' (in the sense of traditional logic), i.e., semantically interpreted symbols. One of the sets of axioms in the Generales Inquisitiones, to which the Introduction calls attention (p. lxi)—and which may have particularly pleased Leibniz—boldly purports to capture entirely the notion of 'formal validity' (vis formae).  Its seventh and last axiom is a completeness clause stating that "whatever cannot be demonstrated from these principles, does not follow by virtue of [logical] form (vi formae)" (p. 785). In spite of being incomplete and consisting in a not quite systematic sequence of attempts to develop complete axiomatic systems, the Generales Inquisitiones can be rightly seen as the flagship of the radical reason picture, where traces of the softer picture are difficult—but not impossible—to find.


5.2  Discourse on Metaphysics


The "Discourse on metaphysics" (pp. 1529-1588), conceived as a concise summary to be submitted to Arnauld's examination, is in fact the first comprehensive formulation of Leibniz's mature metaphysics. The Introduction describes it as a "grandiose" and "masterfully" organized work, containing "unheard of" new ideas (p. lxviii). It introduces indeed the notion of the complete concept of an individual substance—the basis of what Leibniz later called 'monad'. This notion, in conjunction with the principle of maximum perfection of God (hence, of the created universe) and of the analytic conception of truth, yields most of his well-known metaphysical theses. Leibniz believes his system solves several crucial metaphysical and theological problems—e.g., the body-mind relation, the creation, causality, evil, divine grace. He is particularly proud to have found a way out of the "labyrinth of freedom", i.e., the notoriously difficult centuries-old problem of overcoming the apparent antinomy between human freedom and divine providence (or determinism), as well as between God's freedom and his rationality. The key to this solution is the new way he proposes of accounting for the fundamental distinction between necessary and contingent truths, which will be discussed below.


5.3  Examination of the Christian Religion


The "Examination of the Christian religion", originally untitled, is "the most comprehensive text of this volume" (introductory note, p. 2356). It is also the longest (pp. 2356-2455). It undertakes to examine the ensemble of the Catholic dogmas, but is abruptly interrupted in the midst of a sentence and therefore does not fulfill its plan in its entirety. Among other things, it does not discuss the extremely controversial issue of papal primacy and authority, about which it at one point promises to "say much later".

The text was presumably intended to carry out the earlier plan, submitted in 1683 to Count von Hessen-Rheinfels, of writing an "Exposition of the Faith" dealing with some controversial points, a text which would be approved by the new Pope Innocence XI. Prior to further diffusion and publication, the text would be sent to "moderate bishops" in order to check its "tolerability" by the Church and, upon their approval, to Rome and then eventually made public. In order to ensure its success, the identity of the author would not be revealed (p. lxxvi). The Introduction conjectures that what prevented the carefully worked out and nearly complete Examination to be published by Leibniz was the lack of positive Catholic response he had expected (p. lxxx).

About a century after its composition Catholic theologians realized the value of this text for them, and hastened to publish the manuscript, which received several editions in the course of the 19th century. The editor of one of these editions, which contains other texts as well (Exposition de la Doctrine de Leibniz sur la Religion, Paris, 1819), M. Emery, General Superior of St. Sulpice, understood "what an interest for the Catholics" this "religious testament" of a man that all his life was concerned with "the points of doctrine separating Catholics and Protestants" held (p. i). Another Catholic scholar, who apparently was in charge of preparing the Examination for publication (under the title "Theological System") in the same volume, spells this interest out, by saying that in this work, which "would make more sensation than all other writings of Leibnitz", he "defends with so much zeal the Catholic religion, even on those points that have been most vividly debated between Catholics and Protestants, that one could hardly believe he was its author" (id., p. xix). Alas! The Catholic acknowledgment Leibniz needed for his reunification endeavors came too late.

I mention these details because they are important for understanding the dialectical context to which the Examination belongs and, thereby, its mode of argumentation, structure, contents, and even its political failure. Even a hundred years later, the interest Catholic readers attached to this "Catholic" text resides in its having been written by a Lutheran, who had no intention whatsoever to convert. Looking at the text alone, one might perhaps be fooled by its professed neutrality and objectivity, regardless of the author's identity. But for the active participants in the ongoing negotiations in Leibniz's time, it was impossible not to see this text as what it really was: a move by one of the parties involved. To be sure, the Introduction describes the circumstances and explains, in terms of them, why the move failed. But it goes on to comment on the text's contents, setting aside these circumstances, which it views as "external". This purely semantic approach, which is part and parcel of the 'radical reason' picture, is what leads it to emphasize—as in other cases—those virtues of the text that fit such a picture, overlooking its not less significant pragmatic-dialectic virtues which the other picture, naturally, would highlight.

Accordingly, the Examination is praised for its use of the "method of exclusion" that mandates the uncompromising rejection of undisputable errors of various sects, for the terminological rigor and clarity that avoids verbal disputes, for the systematic organization of the material in view of the aim it pursues, and above all for choosing "right reason" (recta ratio) as the supreme examining authority (p. lxxviii). The Introduction acknowledges that this text purports to provide a basic platform for the reunification negotiations (ibid.), but points out that it does not follow the usual dialectical position vs. position and move vs. counter-move pattern of writings prepared for particular reunification "colloquia" (p. lxxviii). Instead, it displays a quasi-axiomatic demonstrative order, which begins with a consensual general definition of God and unfolds its consequences up to the justification of the sacraments and mysteries of faith (ibid.). It is this distancing from the circumstantial in search for rational foundations that grants a text like this a place in a volume of "Philosophical" rather than "Political" writings (p. lx). The question is how large is the distance, i.e., how far can one decontextualize the Examination, ignoring its dialectical underpinnings, and still understand it properly.

The text is no doubt systematic. It starts with the notion of God, the sources of knowledge, the nature of theological discourse and truth; and then proceeds stepwise to deal with specific doctrinal points such as trinity, Christology, incarnation, sanctification, cult, idolatry, the appeal to saints, the use of relics, the sacraments, the mass; with organizational matters such as priesthood, celibacy and Church hierarchy; with eschatology, e.g., the purgatory, the nature of beatitude, resurrection—and this is not an exhaustive list. A concise list of the theses Leibniz argues for (which he himself partially provided in the coetaneous text "Positions"; pp. 2351-2354) shows that neither the list of topics nor their order is casual. It contains, in particular, the principal points of contention raised by the Protestants. The "positions", thus, are in fact answers to well-known Protestant arguments, but they cannot transgress Rome's equally well-known "red lines". Leibniz's proposals and arguments have willy-nilly two attentive audiences to satisfy and if they are to be of any use, they must be "tolerable" and appear "reasonable" to both.

The dialectical nature of this situation imposes upon Leibniz's recta ratio a situatedness that severely constrains its freedom of operation. This self-appointed mediator, who in fact is an active participant by intervening as he does in the debate, knows very well that neither he nor anyone else can pretend he is landing from an outer world wherefrom he brings to bear upon issues that have been hotly debated and fought about for so long the irresistible wisdom of a "view from nowhere". The author of the Examination is well aware that the ethos he projects in the opening paragraph is problematic, since it comprises conflicting elements: the 'natural', uncorrupted reason of a "neophyte" as if just arrived from the new world, the neutrality of "as much as humanly possible non-partisanship", and the experience with religious controversies of someone who is "well-versed" in them. It seems to me that the main philosophical interest of this text lies not in opting for one of these components, to the exclusion or minimization of the others, but rather in attempting to strike a balance between them.


6.4  Inclining without necessitating


Focusing now on a few specific points briefly mentioned in the preceding sub-sections, let us consider first the necessary/contingent distinction. Both necessary and contingent truths must of course be analytic, according to Leibniz's theory of truth. The distinction between them is made, in the Generales Inquisitiones, in terms of the quasi-mathematical distinction between "perfect" and "imperfect" or "finite" and "infinite" analysis. "A true necessary proposition can be proved by reduction to identical propositions", whereas a true contingent proposition cannot be thus reduced and "is proved by showing that if the analysis is continued further and further, it constantly approaches identical propositions, but never reaches them". Epistemologically, the latter kind of proof "transcends every finite intellect" and "God alone, who grasps the entire infinite in his mind, knows all contingent truths with certainty". Logically, such a proof yields only 'certainty' but not 'necessity', because the proposition thus proved "can never be reduced to an identical proposition or its opposite to a contradictory proposition" (pp. 775-776; Parkinson's transl.).

In the Discours de Metaphysique the distinction necessary/contingent acquires, in addition, an ontological dimension. Paragraph 13 of the Discours, the one Arnauld reacted to most negatively, concludes by explaining the difference in terms of the opposition between two realms—that of possibilities and that of existence. The former concerns "the essences themselves"; the latter, "what is or seems to be the best among different equally possible things". The former is ruled by the principle of contradiction, the latter by "the principle of contingency or of the existence of things" (p. 1549); while the former governs the realm of the possible, the latter does not affect it (p. 1548), since its violation does not imply impossibility. The dichotomy finite vs. infinite analysis is neither the definitional criterion nor the reason for the different modal operators ('necessary' or 'certain') attached to necessary and contingent truths. What is decisive are rather the grounds for their respective proofs. The proof that the individual substance 'Caesar' has the predicate 'crosses the Rubicon at t' "is not as absolute as" a geometric demonstration because it presupposes certain "free divine decrees" (such as "I will always do that which is most perfect" and "humans will always freely do that which seems the best"). Decrees such as these are certain because, although they might be violated in some possible world, in the actually created maximally perfect world they are not. If their opposites do not materialize, "it is not [due to] their impossibility but [to] their imperfection". Consequently, "every truth that is grounded on this kind of decree is contingent, though certain" (p. 1548). A contingent truth is thus 'weaker' than a necessary one because it relies upon premises or principles that have the status of presumptions, i.e., that can be in principle denied without contradiction. This is why the reasons they provide only "incline without necessitating" (p. 1548). Like any presumption, they and their consequences can be overruled, provided reasons for this are produced, the burden of proof being always upon he who attempts to overrule them. They are, however, very well-grounded presumptions that provide "strong reasons" (rationes fortes; p. 806), so that it is very hard to find overruling reasons—which is why these principles, and the conclusions they ground, are 'certain'. Nevertheless, they remain presumptive or 'hypothetical' truths—as Leibniz dubs them.

As far as I know, the locution incliner sans necessiter is a terminological innovation of the Discours de Metaphysique and coetaneous texts (e.g., "The principle of human science", pp. 670-672; "On the primary principles of contradiction and of sufficient reason", pp. 803-806), but I leave it to the reader to check more attentively than I perhaps did the Subject Index. In any case, Leibniz's later recurrent use of this locution indicates the undisputable pivotal role the status of contingent truths it expresses acquires henceforth in his metaphysics and theology. In the Discours itself (paragraph 30), for example, God's alleged responsibility for human sins is explained away on the grounds that he "inclines"—as strongly as "determining"—our will towards the choice of what appears as the best, without however "necessitating" it to actually choose in this way, so that we remain free to choose and the responsibility for our choices and actions is, ultimately, not his but ours. And in the Théodicée, what is perhaps the most important argumentative strategy employed against Bayle consists in showing that the reasons he adduces against such a presumption are not powerful enough.

The recourse to the notion of inclination immediately—and deliberately, I believe—brings to mind the image of the scales used at the time—the model for what Leibniz elsewhere called "scales of reason" (trutina rationis; cf. A VI 1, pp. 548ff.), i.e., scales where reasons for and against a conclusion or decision are weighed. This procedure, in contrast to the decision procedure provided by a formal calculus or demonstration, rarely if ever leads to a necessary and incontestable result. Nevertheless, it can and, if properly conducted does, lead to a reasonable one, i.e., sufficient to opt rationally for one of the scales—the one toward which the combined weight of the reasons incline, albeit this option may be turned upside down if other reasons are later brought up. This metaphor is a central topos of the 'soft reason' picture. Its implicit presence in the Discours, and the key role the notion it helps to conceptualize plays in this text, can therefore be a sure indicator not only of the presence but also of the importance the picture in question has in Leibniz's mature metaphysics. Furthermore, the fact that he unabashedly makes use of a metaphorical locution to express the distinctive feature of a fundamental notion such as contingent truth, reveals the presence of another typically 'soft reason' trait in the Discours, namely the need—perhaps unavoidable—to employ metaphors and analogies in order to convey novel concepts that transcend the dichotomies embedded in traditional philosophical discourse—which is what Leibniz's account of contingency undertakes to do. But—one may suggest—cannot this metaphor be, in fact, easily cashed out in formal terms?

This is precisely what the Introduction, in the wake of Leibniz's own pronouncements, suggests. The Introduction repeatedly argues (e.g., on pages lxii, lxv, lxx) that Leibniz, the creator of a calculus that successfully deals operationally with infinitesimal magnitudes, surely possesses the means for providing a non-metaphorical account of contingent truths, namely, the notion of an "infinite analysis" of a proposition. In fact, this is precisely what he seems to claim in the Generales Inquisitiones, as we have seen. The Introduction's claim, however, overlooks the fact that, after introducing this notion and stating that "the distinction between necessary and contingent truths is the same (idem) as that between lines which meet and asymptotes, or between commensurable and incommensurable numbers" (p. 776), Leibniz warns against taking this statement too literally. For, if so taken it would imply more than what it actually yields: "We can prove that some line—namely, an asymptote—constantly approaches another, and (also in the case of asymptotes) we can prove that two quantities are equal, by showing what will be the case if the progression is continued as far as one pleases; so human beings will be able to comprehend contingent truths with certainty. But it must be replied that there is indeed a likeness (similitude) here, but there is not a complete agreement (convenientia)" (p. 776; Parkinson translation). So, what the mathematics of the infinite provides for understanding contingent truths is a handy and suggestive analogy, viz. a metaphor, which breaks down precisely because it cannot offer the tools for actually computing or proving such truths.

Leibniz adds what seems to be a possible explanation for this breakdown of the analogy, namely, the existence of "relations (respectus) which, however far an analysis is continued, will never reveal themselves sufficiently for certainty, and are seen perfectly only by him whose intellect is infinite" (ibid.). If we recall that the complete analysis of a contingent truth would require referring to an individual substance's relations with all other individual substances and to the result of the global comparison between all other possible worlds in order to determine that the created world is the most perfect one, we have a glimpse at why such an analysis involves a complexity that only God can actually grasp and deal with. Even Leibniz was aware that, however suggestive his analogy, there was a leap—perhaps unbridgeable—between the proven powers of his mathematics of the infinite and the tools necessary to fully unfold, analytically, the grounds of contingency. The trust he often expresses that posterity will be able to do what he and his contemporaries couldn't do is of no avail to him here, for at least to this day the mathematics of complexity is very far from providing the tools the "infinite analysis" of contingent truths need. Perhaps in this domain we face the limits of the capabilities of formal thought. In any case, in his time this metaphor couldn't be cashed out formally, and therefore to appeal to it in support of the 'radical reason' picture is rather like appealing to a Trojan horse.


6.5 Human knowledge


The Discours de Metaphysique locates human beings at the top of creation (paragraph 35; pp. 1584-1586), as the creatures closest to the creator. The "principle of Arlequin" establishes that "everywhere and at all times things are like they are at our place and now" (GP, III, p. 340). The principle of continuity precludes gaps of any sort, for their existence would imply imperfection. Nevertheless, a profound ontological gap and a deep difference separate God from us, humans—the gap between absolute perfection and whatever is less than that. If proof is needed of its existence, recall Leibniz's speculations about the existence of a host of non-corporeal species of spirits filling the gap between the human species and God (GP, VI, p. 548). This ontological gap has crucial epistemological consequences. The impossibility for humans to perform the infinite analysis that would formally prove contingent truths is one of its manifestations. Lacking God's epistemological perfections, humans must endeavor to reach whatever knowledge they are capable to achieve by employing less than perfect means. They need to develop 'helps'—to use Bacon's expression—to overcome as much as possible their limitations and to make the knowledge they achieve as sure and reliable as possible.

            In fact, the search for methods of proof that culminates in the idea of formalization is a search for helps God does not need but men do, even in order to secure knowledge of those truths—the necessary truths—which are not in principle beyond human reach. For an analysis and a calculus that yield the knowledge of these truths to work, an adequate notation is essential—a notation obeying what Leibniz calls the 'law of expression': "in the same way that the idea of a thing to be expressed is composed of the ideas of the things that compose it, the symbol (character) expressing the thing must be composed of the symbols of its components" (p. 916). Without such a notation we couldn't go far, due to the limited capacity of what psychologists call our 'short term memory'—namely, the amount of information we can keep simultaneously under our focus of attention. With its help, however, we are able to compress and code the information into symbols, which relieve our mind from the need to be aware, each time we think of a thing or concept, of the myriad of its components. This is what Leibniz calls, in the Meditationes, 'blind' or 'symbolic' thought (pp. 587-588). This is not to be understood as a pejorative denomination, for it is what increases significantly our cognitive capacity, even though if compared to God's capacity of grasping "at one glance" (uno obtutu) the whole complexity of the universe, it remains insignificant.

            If notational helps are essential ingredients in our formal, demonstrative methods, semiotic and other means of organizing our quest for knowledge are even more needed in those parts of the Scientia Generalis, belonging to the "art of discovery", which are of an informal, heuristic nature: a well-ordered (not necessarily 'demonstrative') encyclopedia; synopses in the form of summaries, lists, tables, or diagrams, etc. Such tools as the encyclopedia, cooperatively and painstakingly developed bottom-up from the knowledge available at a given time, not only suggest to us what is missing and ought to be discovered or elaborated henceforth; they also allow us to rise a few inches above the ground in our efforts to see 'the big picture' that God, at the top, has always so easily in front of his eyes. Even the imperfect—because historically evolved—grammatical and semantic structure of natural languages can—with the help of some systematization—provide a useful thread for helping us to see, uno obtutu, "the entire variety of what is thinkable" (pp. 594-604).

            Leibniz incorporated in the text of the Discours the contents of two of the three papers he published in this period which are included in VI.4: the "Brief demonstration of a memorable error of Descartes" (1686; pp. 2027-2030), in paragraph 17 (pp. 1556-1558) and the "Meditations on cognition, truth and ideas" (1684; pp. 585-592), in paragraphs 24 and 25 (pp. 1567-1570). While in the former he employs a mathematical demonstration in order to prove the mistake in question, thereby purporting to show the need to introduce "final causes" in physics, in the latter he elaborates a theory of knowledge that acknowledges the value even of obscure and non-distinct ideas as forms of 'knowledge', as well as the indispensability of 'blind thought'. Except for a few cases in which we are able to know something through direct access to the relevant concept's complete analysis—a type of knowledge he calls 'intuitive' –, in the vast majority of cases, we have to rely upon what he in the Discours he calls 'suppositive' knowledge, since for even mildly complex notions we must content ourselves with presupposing their analysis to be given. From the point of view of an epistemology dealing with a limited mind like ours, no source of knowledge and no kind of help can be neglected. All fall short of an unachievable divine ideal, and all—demonstrative and inclining alike—are mentioned in the same tree of types of knowledge of the Meditationes, and appear next to each other in the same text, as it happens in the Discours.

            If mild complexity already poses problems for human cognition, the huge complexity of the entire world, which underlies every contingent truth, only magnifies it. As we have seen, we are unable to unveil all the details of this extremely complex structure and thereby give a full analytic explanation of each proposition. Were we to take this as a reason for giving up the attempt to discover at least some of the regularities and order that govern the world, we would be like credulous creatures that take the shortcut of seeing miracles everywhere as the only explanation of what happens in a fundamentally disordered and irregular world. What helps us out is, first, to realize that regularity and order, even in geometry, is not always simple: if a straight line does not connect a series of points, it may simply be due to the fact that they are connected by a more complex line rather than arbitrarily juxtaposed, for "when a rule is very complex what fits it may pass for irregular" (Discours, paragraph 6; p. 1538). Epistemologically, this teaches us to adopt a presumption of regularity rather than of irregularity and search harder for the former before giving up to the mere appearance of the latter. For Leibniz this practical maxim, which allows us to pursue our scientific endeavors, is grounded on the general presumption of the order and rationality of a world created by a perfect being, to which even real miracles—whose existence he, as a believer, cannot deny—are subordinated. At most, he contends, miracles violate "subaltern" laws, but fit the "general order" (Discours, paragraph 7; p. 1538). Nevertheless, the fact that not even admitted miracles—whose underlying "rule" we cannot find—do not violate the general presumption of rationality, only shows how strong and well justified this presumption is, and how useful it is as such in our search for knowledge. But, however strong, it remains a presumption. No doubt it yields a powerful epistemological guideline or 'help', which, in spite of its force, only inclines without necessitating. Still, it is its generality and power that ranks it—or the principle upon which it is grounded, namely "that there is nothing without a reason" (quod nihil est sine ratione)—"among the highest and most fecund [axioms] of all human cognition, upon which most of metaphysics, of physics and of moral science are erected" (p. 806).

            Within the full range of epistemological 'helps' at our disposal, the formal ones favored by the 'radical reason' model cover only a small portion of the whole realm of knowledge, most of which requires 'softer' forms of rationality. Although these formal helps are not free of imperfection, the neatness and certainty they provide entice one to see in them the ideal to be pursued everywhere. This ideal did indeed entice Leibniz, and he ceaselessly worked toward its realization—as the texts in VI.4 abundantly testify. But he never overlooked the need to develop also the 'softer' helps, even at the height of his formalistic drive. As his system crystallized in the breakthrough year of 1686, the formal and the informal, the hard and the softer, the necessitating and the inclining epistemological tools found their respective places and raison d'être, side by side, within the global perspective of a 'system'. Had VI.4 'epistemology' as a thematic category—under which perhaps the Scientia Generalis category should be subsumed rather than vice-versa—this global picture would have become synoptically visible for the reader.


6.6 Persuading without obligating


Referring to "On the obligation to believe" (pp. 2149-2155), one of the earliest texts of the period covered by VI.4, the Introduction comments that it "provides a paradigm of what Leibniz understands by a demonstration" (p. lxxvi). No doubt this text—elaborated through various versions, whose conclusion is that "there is no obligation to believe, but only to investigate with utmost application" (p. 2154)—displays a paradigmatic demonstrative form, with its definitions, lemmas, theorems and even the result of an "experiment" (p. 2153). But there is a catch.

The text has a peculiar subtitle that suggests that Leibniz was not so eager to consider it as exemplifying a particularly unproblematic or praiseworthy demonstration, namely Ratiocinatio quae pro demonstratione venditatur. This can be interpreted either as commending the "reasoning" the text presents as amounting to a demonstration, or else as mocking its appearance of a demonstration—two possibilities the standard figurative readings of the frequentative verb vendito ('to offer again and again for sale') permit. The editors seem to have opted for the former reading; I am rather inclined towards the latter. After all, the virtue of a good demonstration is that its formal force shows itself and need not be advertised as such, so that the subtitle in fact hedges the reasoning's pretense to be a demonstration. The erased phrase mihi communicata ('communicated to me') next to ratiocinatio indeed indicates that Leibniz distances himself, for some reason, from this 'demonstration'. The reason cannot be the conclusion, which expresses a view Leibniz holds and whose first part at least is in full consonance with the Lutheran thesis that faith cannot be imposed being rather a gift of God. Nor is it some formal defect.

A hint at what the reason for his qualms may be is given by Leibniz himself, in the following erased passage: "Nothing can be objected to this demonstration if the definitions I have given are followed. If, however, someone does not admit them, the dispute ends, for [the opponent] thereby declares he employs the words obligation, to believe etc. in a sense different from the one I attribute to them. As for myself, I declare that I am satisfied with demonstrating the proposition according to the sense I have assumed the words have—which I think is not different from their ordinary usage" (p. 2154). His definitions, however, are rather unusual, as Grua (p. 181n) remarked about the definition of 'obligation' as "a necessity imposed under the fear of a just punishment". The same could be said of the definition of 'believing' as "being conscious of [the] reasons that persuade us" (p. 2152). The usefulness of a demonstration—especially for the 'analytic' Leibniz—depends upon the definitions it relies upon. If they are not acceptable or not accepted by those the demonstration is intended to persuade, it is useless. If the intended 'audience' of the demonstration and of the definitions is the author himself, he can of course declare his satisfaction. But how much weight does a demonstration that satisfies only its author add to the reasons he must be conscious of in order to believe its conclusion? In such a case, the suspicion may rightly arise that, having first adopted (by virtue of other reasons) the belief in the conclusion, the author crafts definitions that 'demonstrate' it. But if this is the case, the 'demonstration' can hardly function as an independent reason for the author's belief and its only viable function is to persuade someone else—in which case, if the addressee does not accept the definitions, it cannot fulfill its purpose.

Demonstrations in this respect are "situated", especially when used as arguments in a dialectical context where an explicit or implicit opponent is expected to fulfill the critical duties of his or her dialectical role by raising objections. In the case of the thesis 'proved' in "On the obligation to believe", Leibniz's allusion to possible opponents is not gratuitous. He knows quite well who they are: the Catholics who, since they do not accept the Lutheran conclusion, are likely to use their scholastic abilities to undermine the definitions Leibniz uses, and thus the weight of the 'demonstration' allegedly proving it. The proponent of a demonstration, as of any other argument in such a situation, cannot forget that his primary task is persuading an audience, by providing reasons acceptable to the audience and in this way leading it to espouse the beliefs the proponent is interested in promoting. Formal validity alone won't do. Nor the appeal to recta ratio per se will be sufficient, unless the concepts, the principles, the authoritative sources, the procedures of interpretation, and even the set of valid arguments are those the audience is able and willing to accept.

In the complex dialectic situation underlying the demonstration-like arguments of the Examination, where the implicitly self-appointed, participant-observer mediator has to propose positions tolerable to the two disputing parties—with special attention to one of them (the Church of Rome) which is the primary intended addressee—Leibniz has to perform a skillful tightrope act. And his performance is indeed admirable, beginning with the opening announcement of an impartial use of uncorrupted reason. Nevertheless, the traces of what Albert de Broglie, a 19th century translator of the Examination, has called Leibniz's "subtle dialectics" (in P. Lacroix, ed., Système Religieux de Leibniz, Paris, 1846, p. xi), are not difficult to detect.

To begin with, the workings of the recta ratio must be "confirmed"—a word aptly chosen by the Introduction (p. lxxviii)—by ecclesiastic tradition, i.e., by the faithful transmission of the opinions and decisions of the "pious antiquity". This vague expression refers to a pool of 'ancient' Christian authorities, naturally pre-Reformation and therefore usually accepted by both Protestants and Catholics.  Given the enormous weight of the reliance on "authorities" in theological argumentation, the choice of the most favorable set of them is extremely important, since bringing up the appropriate authoritative opinion is often the decisive move—unless, of course, there are even weightier arguments. For instance, regarding the hotly disputed question of transubstantiation, "if it were possible to demonstrate by irrefutable arguments of metaphysical necessity that the essence of bodies consists in extension", then "one would have to admit that, even through divine intervention, a body cannot be in several places [at once], just as the diagonal cannot be commensurable with the sides of a square" (p. 2420). Were such a demonstration available, one would have to interpret Christ's words "This is my body" allegorically (ibid.). Yet those philosophers that purported to provide such a demonstration failed clamorously (a clear allusion to Descartes' "memorable error") and it is unlikely any demonstration of this sort is forthcoming. So, the literal interpretation prevails on the strength of the fact that "pious antiquity has openly declared many times that the bread is transmuted into the body of Christ and the wine in his blood", i.e., that transubstantiation actually does takes place (ibid.). If the voice of "pious antiquity" is not unanimous or is liable to different interpretations—and we have seen (5.2) that this is not unusual—then each party, as well as the "mediator", can pick out those 'ancient' opinions that most fit its needs at each step of the argumentation. Leibniz's incredible erudition in Church history and canonical law—he was able to refer to precedents or councils nobody seems to remember—is of great help in allowing his recta ratio to operate efficiently under the vigilant eye of a "pious antiquity" not quite innocently interpreted by his partisan contemporaries, among whom he himself must be ranged. To be sure, the best rule in all these matters, as he writes after discussing the issue of divorce, is "to follow the Church's judgment and to acknowledge its power" (p. 2448), but there are different ways of interpreting what precisely that judgment is and how to follow it.

The more recent and therefore more controversial sources, especially the Council of Trent, cannot be neglected, but the rope in this case is so tight that one wonders how Leibniz nevertheless managed to walk it. The Protestants' denial of the ecumenical legitimacy of the council because they were not admitted to it, and Rome's intransigence in considering the acceptance of its decisions as a sine qua non for their re-admission into the Church seem to preclude any possible compromise. Under these conditions the best argumentative strategy would be to refer to the contentious Council of Trent as little as possible. Leibniz, however, does not avoid altogether such references.

He concludes his discussion on the issue of divorce, for instance, by quoting a passage of the council's decision endorsing the most rigorous prohibition of divorce under any circumstances and condemning as anathema any criticism of this decision as being a mistake. But, in a subtle pragmatic move, he goes on to discern in this very strong statement signs of "some moderation", since it does not condemn those who have thought (and acted) otherwise, but only those who have said the Church was wrong in adopting the doctrine it did. The latter, he admits, is indeed an example of obstinacy (which amounts to contesting the doctrine of the infallibility of the Church) that deserves to be anathematized. Furthermore—he adds—these obstinate critics forget that, in practice, when it comes to the enforcement of the decision, it is in the power of the Church to be indulgent and to mitigate its doctrinal stance whenever there are "mighty reasons" (maximas rationes) against its implementation—a power the Church has actually exercised in certain cases (p. 2147). The innuendo is obvious: were the Catholic moderation matched by a similar Protestant moderation, this and similar contentious issues might be resolved. Needless to say, this typical mediating move, which the 'moderate bishops' he intended to first submit the Examination to might perhaps accept as 'tolerable', was far from being viewed as such by defenders of a strict adherence to orthodoxy, on the Catholic side, and partisans of direct confrontation with such an orthodoxy, on the Protestant side. The already mentioned translator of the Examination, for example, adds a note to this passage, in this spirit, adverting that "the Church in no way arrogates for itself the power that Leibniz so generously attributes to it of permitting polygamy and divorce in cases of absolute necessity" (in Lacroix, op. cit., p. 383).

Another interesting example of Leibniz's subtle 'mitigated acceptance' strategy vis-à-vis Trent authority is his use of extensive extracts of the council's resolution on the cult of images, on the grounds that their verbatim quotation is "the surest way to correctly learn what this cult consists in" (p. 2395)—notwithstanding the fact that the council's description of this practice may be seen by those who criticize this cult as rather tendentious, since the decision is to preserve, rather than to eliminate it. Leibniz goes a step further, by not only suggesting (through the approval of the council's description) that the decision is reasonable, but also by asserting that it is: "I don't see what can be reprehensible in these words of the council", especially since the council explicitly adds that "if abuses have crept into this practice the council vehemently wishes them to be extirpated" (ibid.). Rome must certainly be happy with Leibniz's endorsement of a resolution of the Council of Trent, thereby implicitly acknowledging its legitimacy. And the Protestants too—he suggests—since the idolatry they are concerned with occurs in the "abuses", not in the ancient and reasonable practice of granting the images of Christ, the Virgin and the Saints "the honor and veneration they deserve, not because one attributes to them any divine virtue … [but] because honoring them one honors the original they represent" (council's words).

The topics of images and their cult occupy about twenty of the hundred pages of the Examination. A wide variety of pro and con arguments—historical, anthropological, psychological, sociological, theological, and semiotic—are employed. I want to single out the latter, due to its relevance to Leibniz's conclusion. From a semiotic point of view, Leibniz stresses, nothing is clearer than the distinction the council uses between a sign and what it signifies, whose application to the case at hand he nevertheless undertakes to further clarify (p. 2396). He also relies upon his semiotic views and experience (manifest in both the Discourse on Metaphysics and the General Inquiries) for supporting the council's claim that images are efficient to induce in believers states of mind that contribute to their piety. In fact, according to Leibniz signs are necessary for our mental and spiritual life, since we are not pure spirits. "If it is possible for us to form internal ideas of things that do not fall under the senses, we cannot however either fixate upon them our attention or engrave them in our mind without adjoining to them certain external signs—words, characters, representations, analogies, examples […]; furthermore, the more efficient signs are those that are more significant (significantiores), representing more properties of the thing represented, especially if the signs are in themselves striking and attractive" (p. 2387)—an interesting extension of the "law of expression" (see 6.5). To this brief summary of his semiotics he might have added what is perhaps its central point, namely that signs, which are indispensable for the fixation and recall of abstract ideas, are also essential for thought and reasoning—that is, that they are in fact constitutively involved in mental processes.

Why is all this relevant to the conclusion, in which Leibniz reiterates that, "everything considered", the council's decision to preserve "the veneration of the original through the image (which is what the cult of images consists in) is wise and pious, provided it is kept, with extreme precautions, within its limits" (pp. 2399-2400)? If indeed the real point of contention, as suggested above, consists in the abuses of the cult (one of them illustrated in a story Leibniz tells immediately after his conclusion), then the additional precautions in Leibniz's proviso should definitely satisfy the Protestants. It is in fact Leibniz's emphasis on these precautions that embodies his "mitigated acceptance" strategy in this particular case. If the Protestants still resisted, maybe it was not only because they saw in the cult of images as such a flagrant violation of the Law of God as delivered to Moses, but also because they were perhaps aware of Leibniz's semiotics and suspected that, if sign and signified are so close and enmeshed in the workings of a logician's or mathematician's mind, in the mind of ordinary persons—the practitioners of the cult of images—it might well be impossible to disentangle them.

Let me try now to disentangle the conclusions that follow from what has been said in this section. That there is an underlying dialectical structure governing the argumentation in the Examination and that it is clearly and heavily situated should by now be apparent. That the specifics of this structure—two contenders asymmetrically addressed by an observer-judge-participant-mediator—impose specific requirements and constraints should also be apparent. In particular it imposes a deliberative mode of proceeding, manifested in a constant flip flop marked by concessive conjunctions (but, although, etc.) and locutions such as on the one hand …on the other hand. This mode evokes the scales image, and implies that the most one can expect from the conclusions is that the sum of the arguments supporting them inclines but does not necessitate, so that their force and eventual contribution to the ultimate objective—reunification—lies in their ability to persuade the contenders, not to compel them by the sheer force of 'radical reason'. The mediator's presence and active participation in re-interpreting the positions of the adversaries and formulating reconciling alternatives—as the  'judge of controversies' envisaged by Leibniz (e.g., in "On controversies", a text of 1680 not included in VI.4, but in IV.3) is expected to do—modifies the rules of the dialectical game. Instead of a dispute where the objective is to win, it becomes a negotiation, where the objective is to win over the adversaries to a common ground 'tolerable' to both, whence a reasonable resolution of the conflict can be found.

The specificity of the Examination can be seen if we compare it with the Consideratio on the one hand and the Positiones, on the other. While the former can be viewed as a discussion of one of the many issues the Examination deals with and as employing a similar flip flop, dialectically-based mode of deliberative argumentation, as a model for resolving controversies it lacks: (a) the comprehensive coverage and systematic structure that would permit to see where a specific issue it deals with fits in and what is its contribution to the whole controversy, and (b) the freedom to muster potentially relevant arguments of all kinds, rather than only those narrowly defined as relevant (in this case, those based on the authority of authoritative sources) by the 'localized' nature of the debated issue. The Positiones, on the other hand, strips the arguments and conclusions from their dialectical clothing and presents them in the minimalist nudity of a strict definition-demonstration sequence; by thereby suppressing their situated nature, they may gain in conciseness and clarity but are likely to loose in the ability to persuade the real participants in the drama. It seems to me that what is peculiar in the Examination is that it somehow succeeds in combining the virtues of the two texts mentioned while avoiding their vices. Maybe its main achievement lies in providing, in this way, an intermediary and perhaps more effective model for handling complex controversies.


6.7 Division of labor


If one acknowledges and puts together the 'radical' and 'softer' reason pictures, rather than seeing them as competing, a compound 'big picture' emerges, where apparently disparate endeavors of Leibniz can be seen as occupying their proper place in the puzzle. If we look at the three 1686 texts we have examined in this way, we can conceptualize their respective contributions to the big picture, as well as their inter-relations, in terms of a 'division of labor'.

            Indeed, in 1686 Leibniz's rationalism crystallizes into a system where "two great principles" divide between themselves ontology as well as epistemology into separate areas of jurisdiction—the principle of contradiction ruling over necessary truths, the principle of perfection or sufficient reason over contingent truths. Although the notion of truth itself is the same in both domains, knowing these truths requires different epistemic means. Since the "inclining" means needed for the knowledge of contingent truths cannot be reduced to the "formal" means needed for the knowledge of necessary truths, a split in 'reason' seems thus to be implied. But the split is only apparent if one foregoes the picture that identifies reason with the latter and views both as different but complementary components, necessary for reason to fulfill its encompassing job of accounting for rationality in every domain: the possible as well as the actual, the theoretical as well as the practical, the divine as well as the human. Given the complexity of this job for us humans, the split it seems to imply is nothing but a division of labor.

            From this perspective, it is clear that the General Inquiries is concerned with the realm of the possible, i.e., with all that can be accounted for in terms of one of the two great principles alone. It "inquires" about the foundations of this realm, develops the tools for discovering its truths, and explores the limits of such tools. The Discourse on Metaphysics, in its turn, is mainly concerned with the realm of the actual, i.e., with that whose rationality depends, in addition, from the other great principle. It focuses on the foundations of this domain, on the general characteristics of the epistemic tools it requires, and gives some examples of them, drawn mainly from the field of physics. But the Discourse does not pause to develop such tools in detail and systematically. This task is left for the 'new logic' that will comprise, in addition to the calculative means developed by the General Inquiries, the softer means characteristic of the knowledge of the contingent. In his past work as well as in the work of others, Leibniz is able to pick out "examples" (specimina) of these tools. But, although the task is clearly delineated by Leibniz in the following decades, its magnitude is such that its fulfillment remains—as in most of his other projects—only fragmentarily dealt with.

            But life does not stop. Contingency presses. Praxis cannot wait for researchers to develop the theory and design the tools. In particular, religious controversies must be rationally handled before they lead to further devastating wars. The 'new logic' must be applied even before it exists as a corpus of doctrine and practical methods. The Examination comes to fill this gap. In so doing, it reveals how little one should rely on the dichotomy theory vs. practice. For, once involved in participating in an actual controversy, one has to find out by experience, as it were 'intuitively', the most efficient strategy to adopt as well as the most efficient tactic to follow at each step. One must learn, for instance, that "there is one truth, but not everything is easy and intelligible by everybody". This is a quote from the Introduction (p. lxxiii; see also 5.2), which I interpret—as might be expected by now—in my own way, as meaning that, in the practice of controversy one soon learns that the persuading argument is not necessarily the one that expresses luminously or demonstrates the "one truth", but rather the one that skillfully argues "from the point of view of the adversary", taking into account not only what is intelligible to him but also what is possible for him to accept, given the circumstances. One learns that for an appeal to the recta ratio to be effective, it must take into account not only universal principles of rationality, but how the opponent concretely interprets such principles in a given context (i.e., one must ask why is it that what I consider to be contradictory is not necessarily so for my opponent). These lessons were not forgotten by the Leibniz of the Nouveaux Essais and the Théodicée. They should also be kept in mind by the theoreticians and developers to come of the 'new logic', because they suggest it should include significant chunks of rhetorical principles and methods—something the 'old logic' would certainly abhor.

            If we were to order the three texts considered in the present section on a scale of context-bound or 'situated' vs. context-free or 'ethereal', the Examination is certainly the most situated of them, the General Inquiries the most ethereal, and the Discourse on Metaphysics should be located fairly close to the latter. I submit that the breakthrough of 1686 does not consist in each of these texts' achievements per se, significant as they certainly are. It rather consists in the fact that, mostly implicitly but sometimes explicitly, they interact, both by borrowing from each other whenever appropriate and—even more significantly—by revealing each other's limitations. What more could one expect from a successful division of labor?


7. Impact


It is time now to wind up with some speculation. Every edition of Leibniz's works which increased the public's access to the unpublished trove well kept in the Leibniz Archive in Hanover has had an impact on the understanding of his philosophy. That the present edition, with its impressive 222 hitherto unpublished pieces, will also have such an impact is certain. Perhaps, however, its main impact will not be due to the novel information any single new piece may contain. As I have been arguing throughout this review, it may rather come from the possibility of having under one's eye the ensemble of Leibniz's philosophical production in a crucial period of its development, whether previously published or not.

In this respect, the present collection differs significantly from the two previous 20th century editions of unpublished manuscripts that had such a major impact—Couturat's Opuscules et fragments inédits (1903) and Grua's Textes inédits (1948). Whereas both of these editions did not purport to be in any way comprehensive, chronologically or thematically, volume VI.4 purports in fact to be thematically exhaustive for a given chronological period. Therefore, whereas both Couturat and Grua could be unabashedly selective, VI.4's inevitable selectivity can only marginally affect the picture or pictures that emerge from the volume. Both Couturat and Grua indeed selected the texts they were to publish in the light of their comprehensive interpretations of some aspect of Leibniz's philosophy, which they had developed on the basis of those texts—interpretations made public in their books on, respectively, logic (1901) and jurisprudence (1953). One can say that an edition informed by an interpretation naturally bears its heavy imprint and highlights the pattern or picture the editor has discerned in the chosen collection of texts. The editors of VI.4, and especially Heinrich Schepers, also had in their minds an interpretation and pattern, clearly expressed in the Introduction, and marginally in the thematic organization and a few decisions about the inclusion of certain groups of texts. But the imprint of Schepers' interpretation upon the ensemble of texts offered to the reader is much weaker than in the case of Grua and Couturat. This is extremely important, as this review tried to substantiate, because it makes it easier for the reader to discern in the collection patterns other than those the editors have seen.

            Let me now turn to further aspects of VI.4's eventual impact. First, from the point of view of an internal-structural historiography, the appearance under the same cover of such an ensemble of texts necessarily yields a better understanding of the connections, often subtle or altogether hidden, between the various strands of Leibniz's mature thought—the commercium between its plural aspects and parts he often advertises as characteristic of his 'system'. In the tracing of these connections, the Subject Index is of extraordinary help. We should remember, regarding this point, that for Leibniz words have a constitutive role in thought, so that they are a precious guide if we wish "inside information" on thinking in general (see 6.5) and on the workings of his thinking in particular. The many erased and modified variants of a text that appear at the bottom of each page, as well as the appearance, spreading and dominance of new terms and locutions in his vocabulary (which can be followed in the index), are examples of how VI.4 can provide this sort of insight.

            From the point of view of a broader, context-oriented historiography, VI.4 in fact also provides a detailed panorama of Leibniz's philosophy in the context of his time. This philosophy emerges as a multi-faceted alternative to the towering thought of Descartes and its offshoot, the Cartesianism Leibniz considered to be an unfit sequel to the master's genius. The many pieces in which he criticizes Descartes and the Cartesians reveal how much his own thought was shaped in a contrastive way. An equally formative role is discernible in his careful study and discussion of Spinoza, Malebranche, Arnauld, Hobbes (and, later, Locke and Bayle), as well as of lesser figures, with most of whom he held actual debates. In fact, the decision to include in each of the six thematic divisions of the volume abundant evidence of his attentive critical readings is a blessing in this respect.Readers are encouraged not to overlook these pieces, for they are essential for understanding the positions this deeply dialogical-dialectical thinker adopts. Had more of his controversy writings—especially in theology—been included, this characteristic would have become even more salient. But in what is included there is enough to highlight not only Leibniz's modus operandi, but also to what extent intellectual production in the République des Lettres was at the time substantially based on swift communication, debate and criticism.

The impact of a volume such as this must also be considered from the point of view of a 'historiography of the future'. Leibniz's legacy contains endless new ideas, some of which are half-baked or difficult to understand, as well as insights, hints, suggestions and plans—all of which amount to as many tasks for us to perform: interpreting, comparing, reconstructing, criticizing, and eventually developing and applying. Virtually all his projects remained uncompleted, part of them due to their intrinsic open-ended nature, which mirrors his historical-dynamic conception of human knowledge. I happen to believe in a "user's" rather than an "observer's" approach to the history of ideas. For me the history of ideas is not only a repository of past thoughts to be preserved, displayed and admired as if in a museum.  It is also a source of insights that can be put to use and developed with a view to solving the problems—theoretical as well as practical—one's own era faces. This is especially important in the case of an imaginative thinker such as Leibniz, who preserved so much of his thought "for us", his posterity.

In previous generations, a philosopher of the stature of Frege attempted to realize with his Begriffschrift what he understood to be the gist of one of Leibniz's most cherished projects, the Universal Characteristic. A century later, a different picture of Leibniz's rationalism—wherein Frege's interpretation (shared by many other 20th century philosophers and Leibniz scholars) is but a part of a bigger puzzle—begins to emerge. The 'new logic' this broader and softer conception of rationality implies was sketched out by Leibniz but remains to be developed. Other patterns that the careful reading of this volume may reveal to other eyes surely will also imply unfinished tasks to be performed. Thanks to the editors of volume VI.4, we have now in our hands an invaluable additional part of a still incompletely edited philosophical legacy that permits us—indeed, requires us, as Leibniz's heirs—to continue to explore the patterns it contains and make use of the insights they reveal. It is up to us, and to future generations, to pick up this challenge!

* Forthcoming in The Leibniz Review, Volume 13, December 2003.