Gabriel Falkenberg
The author, Englishman by birth and continental philosopher by choice, is primarily known for his careful work on formal ontology and semantics in the tradition of early phenomenology, early Russell and Frege. In the present collection of articles from 1982 to 1988, he presents and interprets some important aspects of the exciting philosophical tradition in which he feels at home, the formative era of analytic philosophy. This is a tradition neglected or distorted in standard treatments of that period, and therefore to a large extent reconstructed here.[*1]
The book has sixteen chapters and falls into roughly two parts. The first part deals, in single chapters, with Bolzano, Brentano and Tarski on logic, Husserl's mereology, Frege's number theory, the Meinong-Russell dispute on what there isn't, and Meinong's influence on Lukasiewicz. The second part treats, each in more than three chapters, Lesniewski's "ontology" and the earlier Wittgenstein (each in three and a half chapters). Brentano's logical ideas, used as a tool for comparison, also plays a prominent role in the second part of the book.
The articles, collected on the proposal of Witold Marciszewski, present a unique blend of historical exegesis and systematic reconstruction which makes the well-designed volume a pleasure to read. Due to works as Dummett's on Frege or Bennett's on Kant, the style in which the history of philosophy is now pursued has certainly changed for the better: masters of the past are not seen as items in a dusty hagiography, but as thinkers making contributions to a still ongoing debate; philosophers from which we can learn not by reciting their dogmas, but by confronting their writings with the most advanced reasoning available today. Thanks to books such as the one under review (or the recent by Hylton on the early Russell) the genre is presently both more rewarding and more demanding, requiring logical skill, detailed and wide-ranging knowledge of the actual texts as well as of the state of the art. For the reader, it is also more fun.
The Introduction considers the role of Mitteleuropa in the History of Modern Philosophy. Analytic philosophy - or perhaps better, that part of the analytic movement which we now are able to perceive and study as past - can be said to begin in 1837 with the appearance of Bolzano's Wissenschaftslehre, and to end, for Central Europe, with the fall of Warsaw in 1939. "Central Europe" means the former Habsburg Empire plus imperial Germany, and their respective successor-states. (To be fair, there is nothing on Hungarian philosophy here, and Brentano and Frege are the only German-born philosophers treated.) The essays concentrate on Austrian and Polish philosophers.
Brentano (phase two), who had corresponded with Mill and Spencer and opened the eyes of continental philosophy to British empiricism (8), cultivated the following virtues: strictly empirical analysis (mainly via psychology) [*2], anti-idealism, clarity and objectivity, piecemeal analysis rather than grand system-building, and almost an obsession with truth and representation. One might add: nominalism or "reism" (after the "Abkehr vom Nichtrealen").
It is evident that, with this perspective on analytic philosophy, no linguistic turn of any dramatic kind was involved; rather, close attention to matters of language formed an integral part of the whole enterprise all along. What proved particularly important was Brentano's strong Aristotelian background, inherited from his teacher Trendelenburg (comparable to kd the classical erudition of Oxford philosophy from Cook Wilson to Austin).
As a whole, Kantianism was and is more popular in protestant than in catholic cultures. Indeed, Bolzano, Brentano, Stumpf, Marty, the Vienna Circle were anti-Kantian. So was the Warsaw School.
Lukasiewicz at one place expresses his profound vision of logic in the following memorable words (cited with approval by Geach as well as by Father Bocheński):
...whenever I work even on the tiniest logistic problem, e.g. trying to find the shortest axiom of the implicational calculus, I always have the impression that I am confronted with a mighty construction, of indescribable complexity and immeasurable rigidity. I sense that structure as if it were a concrete, tangible object, made of the hardest of materials, a hundred times stronger than steel and concrete. I cannot change anything in it; by intense labour, though, I discover in it ever new details, and attain unshakable and eternal truths. Where and what is this ideal structure? A Catholic philosopher would say: it is in God and His thought.[*3]I am inclined to think that all this is more than sheer coincidence. It actually suggests to me the bold conjecture that the formation of analytic philosophy is part and parcel of the great process of secularization. After the Catholic enlightenment in the Donaumonarchie had already helped to loosen the grip of the Church, analytical methods were greeted by many as a kind of Ersatz religion. This may also help to explain the almost ecclesiastical zeal with which the "exact method" was preached at the beginnings of the movement (as well as some of its sociological features: priestlike behavior of leading figures, orthodoxies and heresies, forms of excommunication from the inner circle, etc.). It was the training in Aristotle and the Schoolmen that seems to have predisposed them so well to logic and formal reasoning. And Platonism and a World 3 realm are of course still well in accord with some sufficiently wide understanding of catholic philosophy; the next step, nominalism, less so. [*4]
Of course, with the growing establishment of analytic philosophy, especially in its third and fourth phases, and the development of British analytic philosophy, some of the original driving forces grew weaker or were overridden by other concerns. What is said here is therefore to be taken as referring only to Central European analytic philosophy in its inception. In particular, I do not wish to assert that Polish logic developed out of Polish catholicism; that would be a silly claim. It is known that Polish analytic philosophers who were catholics in no way took sides with catholic philosophy, and that Polish thomists were at times more influenced by logic than their logic by religion. (The Vienna and Berlin Circles are of course on record for their high percentage of agnostics - Carnap, Hempel, Reichenbach, Neurath - most of them with socialistic leanings.) And we should never forget those brilliant Polish logicians who were Jews. But the strong connection with Catholicism seems to me a remarkable fact about the rise of analytic philosophy in Central Europe.
This strong "spiritual" undercurrent of Central European analytic philosophy, especially in its first two phases, is set aside in Simons' treatment which, quite legitimately and in accord with its orientation, concentrates on what can be spelled out clearly and formally. But if the Roman Catholic Church had not closed up and petrified in a lost cause against the scientific orientation of its day, Central European culture would surely have taken a different path. Remember that the officially commissioned author of the 1869 memoir of the "modernist" German episcopate against the impending dogma of Papal infallibility had been no one other than Brentano. Perhaps it may not be very popular to point this out in Poland today, but all in all, without Vaticanum I and its logically outrageous decision, there might not have been what is now known as analytic philosophy. So much for my little expansion on this theme.
The logically and systematically most sophisticated group - the "purest", in a way - proved to be the Polish "connection" under the aegis of Brentano's grandsons (so to speak); perhaps the fact that it had no binding material doctrine helped to make it so successful in logical and methodological matters. The active center of philosophy in Mitteleuropa switched for some time from Vienna to Warsaw, which was, as a result of the First World War, the capital of an independent and energetic state again. Here we reach phase four.
The Warsaw School managed to establish the earliest and best contact with English philosophy and logic and, hence, to the growing international scene. (When the Vienna and Berlin Circles caught up at the beginning of the thirties, things of course began to take a different turn. But Goedel was never a real member of the Vienna Circle.) Russell was to write later that, of the six people known by him to have studied Principia Mathematica in full, three had been Poles (My Philosophical Development, p.86; presumably he meant Chwistek, Leśniewski and Zawirski; the last name was suggested to me by J.Jadacki).
The invasion of Poland by Hitler and Stalin brought the shortlived Golden Age of Polish philosophy to a ruthless end, while its methods and results became common property of the Anglo-American scientific community. But nearly a whole generation of young Polish philosophers was wiped out, in the ghettos, in the Warsaw Uprising, or in the concentration camps. Many beautiful thoughts were lost forever.
Let me remark that, with the Wehrmacht's crushing the Polish army, Nazi-Germany effectively wiped out the German-based philosophical culture in Central Europe. For example, when the German Stuka bombers set Warsaw to flames, what they also destroyed beyond repair were the printing plates for the planned international journal to be published in German, English and other languages, Collectanea Logica. One cannot but feel, at this blind self-destructivity, a sense of tragic irony. It is a fact that, around 1939, philosophy became predominantly Anglo-American, and the German language lost its position as a lingua franca for philosophy and science that it had held in Central Europe for so long. The emigrants who brought the ideas of the Berlin, Warsaw and Vienna Circles overseas (Reichenbach, Hempel, Carnap, Goedel, Lukasiewicz, Tarski; Wittgenstein had left before them) switched to English as their second language, and in any event did not return to the Continent after the Second World War.
Truth values as understood by Frege never gained popularity in the Warsaw School. The reason was a different approach to logic: in Poland it was Peirce's concept of logical value that was predominant. Whereas Peirce's "Truth" or "1" can be substituted for a sentential variable, this would of course make no sense for a Fregean truth value.[*6]
Since Simons does not think that the logical constants can be satisfactorily explained in this way, he sketches an alternative: a hierarchy of logical objects starting with truth values. The true and the false themselves can of course not be allowed to exchange roles, as this would destroy our notions of truth and validity; one must restrict permutations here to those by which each truth-value is mapped into itself (25). Tarski's invariant idea about what constitutes a logical object is then given its place. Next, a Bolzanian "language", in fact a simple type theory, of intensional An-sich-propositions and An-sich-concepts is introduced which allows, via a notion of constituency, to define the logical constants. (Simons seems to disallow self-constituency.)
The price of this alternative is pointed out as well: whereas Tarski cautiously eschews World 3-entities in favour of the more tangible sentences of a language, full use of Bolzanian intensional entities has strong epistemological drawbacks. First, the realm of the An-sich is, because of its universality, threatened with inconsistency by a diagonalization argument. And second, Simons thinks that propositions in which the same concepts, but in different order, occur, cannot thus be represented adequately, because "the type, or its abstract meaning, is something unique" (39) and cannot as such occur twice in World 3 (39). (The second argument strikes me as too undeveloped to be convincing: one might as well hold that there are only 26 letter-like elements in a Bolzanian heaven.)
CHAPTER 3 resurrects Brentano's Reform of Logic, according to which any simple judgement is an assertion or denial of existence. This was a break with the then (1870/71) predominant subject-predicate tradition, predating Frege's far superior one by nearly a decade. Given Brentano's dislike of mathematical logic, it is remarkable how much can be achieved without quantifiers and a clear grasp of relations.
Brentano's logic is actually a term logic which employs both extensional and intensional notions. Acceptance and rejection are the basic qualities attached to a judgement: There are men comes out as something like Men are accepted. No existential import is assumed, so inferences remain valid when interpreted in the empty domain. Interestingly, propositions are introducible, under the guise of terms. Two "shuttle" principles, Nominalization (sometimes also called "Abstraction" in the literature) and Denominalization, allow us to move back and forth between propositions and their respective that-terms (59). All this even lends itself to an algebraic treatment. But on the whole, Brentano's logic turns out to be much too inelegant and cumbersome by modern standards and is best regarded as a historical "curiosity" (63). Nevertheless, as the remaining chapters show, Simons somehow falls for it. Earlier treatments of the topic by Terrell and Chisholm (Midwest Studies in Philosophy 1, 1976, 81-95) are not mentioned.
CHAPTER 4 is a long and careful formalization of Husserl's Theory of Wholes and Parts in the 3rd Logische Untersuchung which promises to be of permanent value. Several notions of ontological dependence and independence are distinguished that have applications in different areas of science, from physics to linguistics. I refrain from paraphrasing or commenting on this important contribution to formal ontology (first published in 1982), because its contents are known to those interested in the matter, and because it overlaps with the more comprehensive treatment in Simons' book on Parts (Oxford: Clarendon 1987) which would merit a separate review.
Frege's Theory of the Reals is rescued from oblivion in the self-contained CHAPTER 5. That part of Grundgesetze - largely neglected in the literature as compared to its treatment of the natural numbers - has come to us only as a fragment, with extended polemics in between. Simons draws attention to Frege's unerring critical instinct of always finding the weakest spot in the theories he attacks, but also to his related neglect to do justice to the positive contributions of his adversaries (e.g., Cantor and Dedekind). Frege's objection to a modern, set-theoretic treatment of the reals - as propounded in the standard text book by Suppes - would have been that it "gives us no information at all about the application of the real numbers in the physical sciences" (124). By contrast, Frege held that the applicability of the reals in measurement must be built into their very essence. He consequently conceives of real numbers as ratios of magnitudes, i.e. the extensions of a relation between magnitudes, making ingenious use, in obtaining the relation, of the availability of natural numbers as previously defined by him. Hence, separateness of the naturals from the reals is of strategic importance for Frege. The reals are planned to be obtained not stepwise, but in one big leap.
Simons also sketches the "likely continuation" of Frege's treatment beyond that point at which it actually broke off at the receipt of the Russell letter. It throws some light on Frege's logicism "which is in fact a gigantic existence proof" (135): he set himself the task to show that there are logical objects satisfying the requirements of natural and real numbers. But, as already with the naturals, the notorious multiple representation problem rises its ugly head here as well: any attempt to tie the reals to certain applications is faced with logically distinct codifications between which it has to choose, on some grounds. Today, the logistic thesis is said not to be upheld globally anymore, only locally. (There has however been something of a revival of global logicism in recent years which Simons does not mention, e.g. in works of Bealer, D.Bostock, C.Wright, and others.[*7]
The Anglo-Austrian Analytical Axis is the topic of CHAPTER 6 which is closely related to the introduction and does not quite seem to fit anywhere. On the whole, I find it somewhat too broad in coverage compared to the other pieces, and as containing little that is not elsewhere available in the book. It deals with the third subperiod in the history of analytic philosophy in which "the best philosophy coming out of England and ... out of the German-speaking world were running more or less parallel" (144). But at that happy time, few people were able to perceive this fact clearly and to make good use of it (compare, e.g. Husserl's Philosophie als strenge Wissenschaft of 1911 with Russell's On Scientific Method in Philosophy of 1914); anyway, the First World War effectively put an end to this interlude, closed up the possibilities of to any international discussion for some time and saw quite a number of respected philosophers on both sides of the channel turning into raging nationalists (Russell and Husserl were the exception, McTaggart and Scheler the rule).
CHAPTER 7 is a richly documented and closely argued exegesis, step by step, of the Meinong-Russell Controversy over Being and Nonbeing. Weighing the respective merits of the different proposals and the theoretical price to be paid for each of them, both sides are given their due. Russell was right - as he was against Frege - that there is an indispensable indexical element to language, and that immediate acquaintance is required for knowledge; whereas Meinong's ontology, with its doctrine of Aussersein and its incomplete objects, proves flexible enough to devise answers to all other objections of Russell's which tended to look so overwhelming when first propounded. In fact, recent developments in free, relevance and paraconsistent logic allow one to formulate some of Meinong's ideas in even more precise dress, showing him to be both eminently reasonable and far ahead of his time. Simons, following the footsteps of Findlay and Chisholm, definitively sets matters right for Meinong and his keen "sense of reality".
Particularly helpful is Simons' clarification of Meinong's strong semantic resources, e.g., incomplete objects take on the part of Fregean senses. The exposition is marvellously clear and leads straight into current debate. If I would have to pick out my favourite paper in the volume, it is this. There is only one tiny point where I disagree (161): I find a return to some of Meinong's opinions (not particularly those that were at stake in the earlier debate) in Russell's later works from 1940 onwards, most clearly in Human Knowledge (1948); I have not seen this remarked anywhere in the literature.
Lukasiewicz, Meinong, and Many-valued Logic, CHAPTER 8, is a convincing argumentation to the effect that Meinong, by his work in probability theory, and perhaps also by personal communication, was influential in Lukasiewicz' creation of many-valued logic around 1917. Meinong can therefore justly be counted among its founding fathers.
As an appendix, there are two interesting letters to Meinong one of which (dated Lemberg 12.IV.1912) mentions a "kleine Schrift von Frege" that is sent together with the letter but not mentioned by title. Lukasiewicz writes: "Die Fregesche Ausfuehrungen ueber Annahme und Urteil sind auch von B. Russell akzeptiert worden" (221).
Simons (222, note 111) speculates about which of Frege's works may be meant here, but without success. I presume the wording strongly suggests a reference to the booklet Function und Begriff (1891, p.21 bottom), where both Annahme and Urteil occur. That Russell is said, by Lukasiewicz, to have accepted Frege's statements, may perhaps allude to Principles of Mathematics where Russell refers to the above-mentioned place in Funktion und Begriff, adding the confused footnote: "Frege, like Meinong, calls this an Annahme", i.e. the truth-value of a Gedanke called by Russell assumption in the text (cf. Anscombe, An Introduction to Wittgenstein's Tractatus, 2nd ed.,1963:105). Given that this is the place in Russell's writings that Łukasiewicz had in mind (if in fact he had seen it), one may wonder how he could interpret it as Russell accepting what he concocts from Frege. Apart from the doubtful correctness of Lukasiewicz's claim about Russell, the supposed Fregean distinction between Annahme and Urteil (to which Lukasiewicz intended to direct Meinong's attention in the above-mentioned letter) was of course no news to Meinong, his attention having been drawn to it by Russell himself, in the latter's 1904 Mind article (cp. Meinong, Ueber Annahmen, 2nd ed., 1910, p. 6). To what extent Meinong seriously studied Frege, I do not know. He is quite enthusiastic about him, but did he have more than an earnest intention?
Chapters 9-12 all deal, in one way or another, with Lesniewski's "ontology". The first, On Understanding Lesniewski (Ch.9), is a basic introduction to this part of his logic for outsiders, which does not proceed by interpreting Lesniewski in terms of type or set theory, but starts from zero level, i.e. by indicating what features of a natural language it can be said to capture formally. Simons argues that not common nouns but definite terms (empty, singular and plural) best fill the bill. On this interpretation, a [epsilon] b comes out as something like a is one of b (or one of the b's).
The definite article is introduced as an operator on common nouns to form singular terms, which allows to translate formulas from the Goodman-Leonard Calculus of Individuals smoothly into Mereology. It turns out that Simons belongs to the growing number of sceptics about the general feasibility and naturalness of set theory, and prefers starting with plural terms instead, an account surely nearer to the actual workings of language; but classes can be introduced into the framework if we wish.
A Brentanian Basis for Lesniewskian Logic, CHAPTER 10, shows that something akin to a Brentanian logic, in the manner reformulated above, can be used as an alternative underpinning for "ontology", requiring as primitives only existence and nominal conjunction. This sort of exercise in possible history is a notable feature of the chapters under review, and turns up at several places in the book, always with some unexpected cognitive gain.
CHAPTER 11 considers the Relation of "Ontology" to Classical and Free Logics. Successively richer logical systems can be embedded in extensions of "ontology" which proves the latter's unique combination of maximal power with supreme elegance. The chapter closes with the question how an intensional logic formulated in Lesniewskian fashion would look like. (I presume that part of the answer can be gained by glancing at Bealer's 1982 logic with copula and intensional abstraction.) Critical remark: the first parts of this chapter overlap with the beginning of Ch.9 to a rather annoying degree. Some tightening up of the text would surely have been in order here.
In CHAPTER 12, Lesniewski meets Wittgenstein: A combinatorial Semantics for Ontology is developed that eschews, in conforming to Lesniewski's nominalism, commitment to abstract entities. The basic and, I think, highly suggestive idea is that of a way of meaning which is a generalization of Wittgenstein's truth-table method in the Tractatus. But ways of meaning are not be identified with objects named. If we allow for semantic categories, some important principles about designation and truth-values fall nicely into place.
But as there are more ways of meaning (presumably infinitely many) than objects named, adverbial talk about ways apparently turns out to be both Platonistic and intensional, contrary to the nominalistic intentions of the outset. Further work on whether we can have it "both ways" is needed, i.e. a semantics that presupposes neither referential objects nor World 3 entities.
CHAPTER 13, The Old Problem of Complex and Fact, turns directly to the Tractatus. What is that problem? Wittgenstein refused to assimilate sentences to names, or facts to objects. Russell, in speaking of complexes in regard to such "objects" as a in the relation to b, had in fact confused the two. Now, either complexes are further analyzable into simples and facts, or recognized as existents, henceforth nameable and, for Wittgenstein, simple.
Simons holds that this problem is of Wittgenstein's own making. Under the name of analysis, he ran together two different activities: resolution and clarification. Resolution cannot preserve truth-value, but clarification must. (The further connection between the two, elaborated in the last paragraph of section 4, p.331, eludes me.) I take it that "the old problem" is really one of the hangovers from Hume in logical atomism. Expelling complexes from his official ontology, Wittgenstein introduced, besides facts, Sachlagen (Sachverhalte, for the atomic case), but these got somehow squeezed between complexes and facts. In order to achieve a consistent reconstruction of the Tractatus, Simons urges a reading of Sachverhalte as atomic complexes.
CHAPTER 14, Tractatus Mereologico-Philosophicus, looks at the early Wittgenstein through Brentanian eyes, a type of move to which we have grown accustomed by now. Viewing states of affairs (Sachverhalte) as thing-like instead of fact-like, a certain amount of part-whole theory can be discerned in - or squeezed out of - the Tractatus. However, purely negative situations (Sachlagen) cannot be rescued in this way.
CHAPTER 15, Wittgenstein, Schlick and the A Priori, shows Schlick's denial, against phenomenology, of the synthetic a priori as beside the mark: Husserl had in fact a narrower concept of analyticity at his disposal by which sentences about colour incompatibility (which were the critical examples) do qualify as analytic. Poor Schlick also got his master's voice wrong (well, nearly everybody did) which was ever more subtle than he was able to perceive. (Simons, by the way, takes no notice of the book by Delius on the matter.)
The final CHAPTER 16, Categories and Ways of Being, is a short piece rambling from Aristotle to Ingarden. Compared to the high standards set elsewhere in the volume, it is rather sloppy, achieves disappointingly little and does - to my mind - next to nothing to make a study of either Ingarden or the categories look more attractive. (In fact, it falls short of recent work on the categories by A.Code, Tugendhat and Furth.)
On the whole, this is an excellent and inspiring volume, full of interesting ideas. The notion of Central European philosophy is established beyond doubt and should prove indispensable to any serious historiography of modern intellectual culture.
Moreover, Simons' strategy of confronting and integrating theories that did not have the chance to meet in real life is developed into a sophisticated methodological tool for historical and systematic understanding. Subjunctive history writing (as it may be called) opens up our sense of what was contingent and coincidental in the actual history of philosophy, what might have happened instead and it stimulates us to think about other possible and enlightening encounters. How about Bolzano meeting Goedel? Or Frege arguing with Lesniewski?? Or Stumpf criticising Freud?! There are no limits to such ideal hermeneutic undertakings in which thinkers can be put to argue with one another in constructive dialogue, notwithstanding barriers of space and time.
F o o t n o t e s
[***] Peter Simons, Philosophy and Logic in Central Europe from Bolzano to Tarski. Selected Essays. Dordrecht, Kluwer Academic Publishers, 1992, 441 pp. Nijhoff International Philosophy Series 45.
[*1] I thank Henry Hiz, Peter Simons and Jan Wolenski for helpful comments on an earlier version.
[*2] The influence of Brentano, through lectures and through personal contact with Breuer, on Freud's psychoanalysis (in particular his fascinating 1895 Outline of Psychology) may merit closer study.
[*3] Jan Lukasiewicz, W obronie logistyki. Mysl katolicka wobec logiki wspolczesnej, 1937 (In defence of logistic, last paragraph); reprinted in Z zagadnien logiki i filozofii (Ed. J.Slupecki), Warszawa 1961. Compare next footnote.
[MU Editor's Note. A literal translation of the beginning of last sentence in the quoted passage would run "A believing philosopher" where the adjective means "a believer in God" (in some Polish contexts the couterpart of "belief", viz. "wiara", means religious faith). The identification of a believer in God with a Catholic, as suggested by Dr. Falkenberg's translation, could not have been intended by Lukasiewicz.][*4] "[...] most analytic philosophers were in open opposition to Catholicism, some to any Christianity, many to all religion. Warsaw philosophers had not only Brentano as predecessor, also Warsaw Positivism, a philosophy popular in Warsaw 1865-1910, inspired mainly by J.S.Mill. Lukasiewicz [...] was different. He advocated the philosophical neutrality of mathematical logic, and defended this neutrality against people from right to left. In the thirties he had long public disputes with conservative Father Jakubisiak." (Henry Hiz in a letter to the author, quoted with permission).
[*5] Alfred Tarski, `What are Logical Notions?' In: History and Philosophy of Logic 7, 1986, 143-54.
[*6] See Henry Hiz, `Peirce's influence in logic in Poland', in: Proceedings of the Peirce Sesquicentennial Congress (to appear).
[*7] David Bostock, Logic and Arithmetic: Rational and Irrational Numbers, Oxford 1979; Crispin Wright, Frege's Conception of Numbers as Objects, Aberdeen 1983.
Submitted March 5th, 1993. Revised July 1997.