REVIEWS
239
B. F. McGuinness, T. Nyberg, and G. H. von Wright (eds.), Wittgenstein's Prototractatus, with an historical introduction by G. H. yon Wright, Cornell University Press, 256 pages, $ 18.00. The motivation for printing the work under review began in 1965 when G. H. yon Wright, who co-edited the Prototractatus and provided its introduction, discovered a 121-page manuscript (with no page 89), in Wittgenstein's own hand, which contained an early version (circa 1918) of the Tractatus along with an addendum and foreword. Subsequent investigation led yon Wright to conjecture that at least one proto-Prototractatus existed, supposedly extracted from 7 to 9 of Wittgenstein's notebooks dating from 1914 to 1918; and that the l'rototractatus immediately preceded the final version of the Tractatus, which was completed in August of 1918. By way of comparison, essentially all of the passages in the Prototractatus are shared with the Tractatus. While von Wright has uncovered approximately 400 places where the wording between the two texts differs, he found most of these to be of minor significance. It is interesting to note, however, that 106 actual non-correspondences of passages occur within the two volumes; in particular, there are 67 passages which occur in the Tractatus but not in the Prototractatus. Section 5 of the Tractatus seems to have received the most attention from Wittgenstein during the last days before the completion of the work, revealing the greatest number of alterations. In this section occur 33 of the 67 nonrepeated passages, the remainder occurring in sections 4, 6, 3 and 2 (in descending order of occurrence). Of the subsections in the Tractatus revealing the greatest dissimilarity with the Prototractatus, perhaps the passages concerning expressions and variables (3.311-3.314), probability (5.152-5.155), operations and internal relations of propositions (5.2-5.231), operations versus functions (5.24-5.25), and truth functions and numbers (6.-6.021) are the most notable. The second part of the manuscript, consisting of an addendum to the first part, contains remarks corresponding to those of the same number in the Tractatus; and the foreword is, with the exception of one sentence, indentical to the foreword which appears in the Tractatus. While only the first two sections (9 of 34 pages) of von Wright's introduction concern themselves directly with the research associated with his Theory and Decision 6 (1975) 239-240. All Rights Reserved Copyright 9 1975 by D. Reidel Publishing Company, Dordrecht-Holland
240
REVIEWS
discovery, the remaining six sections are no less interesting. They offer the reader a splendid account of the history behind the publication of the Tractatus after the work was completed in the summer of 1918. They tell of Wittgenstein's attempt, while on leave from the service in 1918, to get the Tractatus published in Vienna; of the humiliation he suffered when the Viennese publisher Braumuller demanded that Wittgenstein assume complete financial risk as a condition for publication; of Frege's effort to get the Tractatus published; of Russell's assistance and Wittgenstein's rejection of his introduction; and of Wittgenstein's gift of the manuscript to Russell in 1920, asking only that Russell credit himself with any changes made in the text, should he ever get it published, to name just a few instances. As yon Wright shows, the labors involved in the final publication of the Tractatus were monumental, and only after Wittgenstein had given up hope of ever publishing it did the Ostwald edition come to pass - and only then because of the weight carried by Russell's introduction. Von Wright describes this entire history, in appreciable detail, with a good bit of Wittgenstein's correspondence during the period reproduced in toto. The Prototractatus naturally falls into 4 parts: (1) yon Wright's historical introduction, (2) facsimile of the entire 1965 manuscript, (3) a translation of the first part, the Prototractatus proper, by Pears and McGuinness, and (4) tables of collations and non-correspondences within similar passages in the Tractatus and Prototractatus. University of Nebraska
H . L . BERGHEL
REVIEWS
241
Anna-Teresa Tymieniecka (ed.), Analecta Husserliana, Vol. II, The Later Husserl and the Idea of Phenomenology. Idealism-Realism, Historicity and Nature. Papers and debate of the International Phenomenological Conference, held at the University of Waterloo, Canada, April 9--14, 1969. 1972, vii +374 pp., Dfl. 85,-. In the center of the papers and debates during the International Phenomenological Conference lies Husserl's late work 'The Crisis of European Science and Transcendental Phenomenology', in which the author takes an extremely idealistic view concerning transcendental phenomenology. On the one hand, the conference's study of the brilliant master might lead us to the supposition that the discussion does not go beyond the scope of Husserl's work - in other words, that it does not transcend phenomenological self-introspection. On the other hand, however, the subtitle excites the expectation of a controversy with other philosophical standpoints, independent from Husserl's text. The editor gives expression to this expectation in her paper 'Phenomenology Reflects upon Itself', in saying that the 'philosophia prima', the 'mathesis universalis" which phenomenology pretends to be, has to meet the standards for validity and excellence not only as basic inventory of cognition but also as an 'ars inveniendi'. Thus it is quite legitimate to judge both Husserl and the contributions of the symposium by the self-chosen criteria. As to Husserl himself, Tymieniecka points out that she has left the criterion of intersubjective validity in transferring the constitution of the philosophia universalis into the life-world (p. 7). Life-world contains such a variety of aspects and associations that an anchorage of cognition in it transposes the ideal of validity into the realm of the unattainable. Husserrs transcendental turn in which not only the function of cognition but also the constitution of the existence of environment are based on the 'Ego of the transcendental reduction' signifies a depravation and devaluation of reality. The intentional acts are separated from the empirical ones. This means that the world is created in thought and that thereby the fundamental ontology of rational structures is transposed into the field of transcendental consciousness. The first part of the volume deals with this problem of idealism-realism. But the claim by Tymieniecka is not satisfied, a claim which proposes to tackle the problem in an undogmatical way and to integrate natural Theory and Decision 6 (1975) 241-244. All Rights Reserved Copyright 9 1975 by D. ReidelPublishing Company, Dordrecht-Holland
242
REVIEWS
sciences, the origin of evidence and of getting to the bottom of things without prejudice, into the discussion and to open the way for an overdue contest with the arguments of analytical philosophy. The participants of the symposium mainly confine themselves to the immanent phenomenological controversy between Husserl and Ingarden about that question. On the whole, most lectures are of a very high level. A series of particular questions of which Husserl had merely scratched the surface in his Crisis are outlined systematically in order to complete the master's work. Thus this book is an advantageous and indispensible work for already convinced phenomenologists, though the epistemologically interesting questions are not dealt with in a presence-related way. Positivism and logical empirism are presented only as shapeless dummies, not in their modern variants. The multitude of lectures only permits a selective mode of procedure in order to illustrate this judgement. In his lecture on 'Hyletic Data' A. Lingis deals with the phenomenological concept of experience and sensation. Because Husserl rejects empirical psychology (even explicitly Gestaltism) as a legitimate approach to phenomenology, he used the Aristotelean term 'hyletic data' instead of the term 'sense-data' in order to avoid misunderstandings about the empiristic position on that concept. The sense data (hyletic data) reflect intentional objects to consciousness (p. 97). They only establish contact to the things but do not influence the act of cognition in any determined way. Consciousness is 'encumbered" with hyletic data, it 'lives them', but they do not entail the unquestionable spatial and temporal zero-point of cognition; this zero-point is freely displaceable to intentionality (19. 98). Consequently, the decisive element is the 'Sinngebung' by consciousness. Here it becomes evident that empirical sciences have - according to the phenomenological view - to stop in front of the sanctuary of intentionality. The ornate and complex phenomenological language and the respectful orientation to the great master block a serious discussion about the concept of sensation. This concept is - in spite of the phenomenological supposition - in the revised perspective of modern analytical philosophy not a crude sensualistic one but a theory-laden one. Lingis' lecture is an example of the phenomenological tendency to overlook progression in the scientific environment. In his lecture about 'The Material Apriori and the Foundation for its Analysis in Husserl', G. Brand focuses on the life-world as the fundamen-
REVlnWS
243
tal element of phenomenology. According to this author, Husserrs imperfections are rooted in his solipsistie narrowing which obstructed the constitution of intersubjectivity. As one might expect, Brand then joins the ordinary-language approach of the later Wittgenstein. However, the emphasizing of the linguistic and historical prestructuring of life-world does not even lead him to the concept of ' Verstehen', as it is laid down by Weber or Schtitz. His analysis is purely intentional, it remains in a phenomenological-reductionistic sphere and thus cannot contribute to an empirical controlled theory of intersubjectivity. The only contribution in this book which attempts a direct relation to non-phenomenological philosophy is G. Kiing's lecture on 'Ingarden on Language and Ontology (A Comparison with some Trends in Analytic Philosophy)'. The author avoids the repetition of his colleagues' stereotypes about the so-called 'objectivation' of natural sciences which in their mind pretend to be able to recognize the 'an-sich" of things. Kfing states the developments of analytic philosophy, passing from a strict realism to a more Kantian position. As to Carnap and Frege, they consider the metaphysical question about the existence of the world as a pragmatic one. The world can be described in various systems. Every 'functioning' ontology is allowed. Kting conceives this process as partial convergency of the phenomenological and analytic position - in spite of fundamental differences between both. According to him, the primary interest of phenomenology concentrates on the richness of content and subtle nuances, whereas analytic philosophy is interested in the construction of formal systems according to the principle of economy (p. 213f). Each line of thought can complete the other: the phenomenological descriptions enrich the logical empiricist's understanding of the ontologies which he is committed to in his systems; phenomenology then serves as an inspiration to the process of system-building. On the other hand, phenomenology can learn from analytic philosophy that formal systems are tools to strengthen its intuitions (p. 214). The so-defined division of labor between the different approaches results in a determination of the function of phenomenology as heuristics for hypothetic-deductive sciences. This - in my eyes realistic -judgement hardly differs from that made by analytical philosophers about this question. After reading this volume, edited by Tymieniecka, a person who is interested in solving the problems of the philosophy of science gains the
244
REVIEWS
following impression: Phenomenology is not an empirical science and offers no methodology for the empirical sciences. Therefore, it cannot compete with the explanatory models of analytic philosophy, though it is an appropriate heuristics for the inspiration of scientific theories. Its scientific-theoretical function exhausts itself in this task. Apart from this it offers a possible approach to the problems of relation between theory and practice. Its basic thesis reads as follows: feed back scientific acts and the results of the special disciplines to the 'natural' life-world! The idea of a free subject seems to be, at least as a regulative idea, quite meaningful in terms of challenging science in front of practical problems of everydaylife. The human being in his inescapable pressure to make continuously practical decisions can be regarded as a legitimation for the subjectivism of phenomenology. In this line of interpreting phenomenology one would have expected the urgent ethical and political problems, with which even the natural sciences are confronted today, to be treated by a less scholastic but rather original practical philosophy, based upon the results of the exact sciences. But this was not the ease with this phenomenological symposium because of its restrictive introversion.
Institut fiir Soziologie Albert-Ludwigs- Universitiit Freiburg
HERMANN KOTTHOFF
RBVI~WS
245
J. Leach, R. Butts, and G. Pearce (eds.), Science, Decision and Value, D. Reidel Publishing Co., Dordrecht and Boston, 1973, vii+213 pp., Dfl. 27.50 (paper). Science, Decision and Value is the first volume of the new University of Western Ontario Series in Philosophy of Science. It contains the four papers read at the 1969 philosophy colloquium held at that university as well as three papers previously published elsewhere, and comments on the invited papers. There are no contributions f r o m the philosophers in the host institution. The piece de resistance is a new approach to deontic logic proposed by Patrick Suppes in his paper 'The Concept o f Obligation in the Context of Decision Theory'. Faithful to his slogan a b o u t elucidating all philosophic and scientific concepts in set-theoretic terms, Suppes formulates in this language an axiomatic theory o f obligation. The primitive concepts of this theory are those o f state of nature, decision, obligatory decision or act, and a preference relation. The theory presupposes the controversial construal of events as sets (as suggested by Kolmogoroff) and that of decision or act as a function on all states of nature. The central intuition is supposed to be conveyed by the following example. "Let A be the event of a two-year-old child's starting to run across a busy street in Jones' presence. Let f be the act or decision that for every x in A is the act of stopping the child, and let f, for x CA, be the (passive) act of watching the child play. Then intuitively, not f, but f restricted to A, in symbols:f[ A, is the obligatory act. So tP is the set of such partial functions, i.e., O is just the set of obligatory acts" (p. 3). The obligations thus characterized are then conditional: ' f ] A ~ 0' is read 'actf is obligatory given event A'. The basic concepts are defined axiomatically. One o f the axioms reads: I f decisions f and fl are obligatory given events A and B respectively, then their union too is obligatory provided events A and B are mutually exclusive. On the other hand, if the domains o f two obligatory acts overlap, then the two acts are identical on the c o m m o n domain. The strongest axiom o f the theory bears on the notion of m a x i m u m obligatory domain ( M O D ) of a decision, defined as follows: Event A is the M O D o f action f i f f f I A is obligatory and i f A ~ _ B , but (if) A ~ B thenfl B is not obligatory. The axiom in question is this: I f A is the M O D of f and B is the M O D of #, and either o f the events is possible, t h e n f i s preferable to # Theory and Decision 6 (1975) 245-247. All Rights Reserved Copyright 9 1975 by D. Reidel Publishing Company, Dordrecht-Holland
246
REVIEWS
just in case A is more likely than B. Therefore "one act will be preferred to another or chosen in place of another just on the basis of obligation and without any consideration of utility". This, which may be regarded as a formalization of Kant's categorical imperative, is regarded by Suppes himself as being probably too strong. Suppes' theory of obligation has some strange ingredients. One of them is this: Theorem 6 states that i f f is obligatory both in the face of A and in that of B, t h e n f i s obligatory given A and B. What if B happens to be the complement of A ? Why should anyone be under any obligation in the nonface of the null event? Another suspect is Theorem 16: I f p is an obligatory act and q is permitted then q is obligatory. Example: if offering water to the thirsty is obligatory, and smoking is permitted, then smoking is compulsory. Henry Kyburg takes issue with this consequence in his comments on the paper (though does not use my example). But he also notes, and justly so, that Suppes' paper is "a gem of sparkling clarity" and one more piece of evidence that "set theory has rendered many of the other techniques of philosophical inquiry obsolete". Quite so. What can you infer about a person's beliefs on the strength of your observations of his choices from a set of courses of action? In other words, given a body of information concerning somebody's behavior, what can you conclude about his beliefs ? This is the problem Richard Bevin Braithwaite addresses himself to in his paper 'Behind Decision and Games Theory: Acting with a Co-agent versus Acting Along with Nature'. Braithwaite proceeds on the assumption that the relation between a person's choices and his beliefs is a logical relation not an empirical one. But this very presupposition should have been examined thoroughly for, if true, then the philosopher is qualified to tackle the problem, otherwise he must leave it to the psychologist. Unless of course what is intended is a normative not a descriptive theory of the relations between belief and action. But this is not what Braithwaite is after: he wants to implement "the pragmatist's programme of determining what a man is (in the way of believing and preferring) by finding out what he does". We shall not discuss here Braithwaite's theory, commented on by Isaac Levi, Ronald Giere and I. J. Good, but will wait until the appearance of his book Pure Theory of Applied Belief, announced by the Cambridge University Press. In 'Measurement: A Systems Approach', C. West Churchman reviews an earlier work of his, Theory of Experimental Inference, in the light of the
REVIEWS
247
scientific and technological developments that have taken place since its publication over two decades ago. Despite its title the paper does not deal with measurement but deals instead with a large assortment of problems, among them the exaggerate claims made in the 1950's for the capabilities of the information sciences, the equally exaggerate claims made nowadays by the systems approach to social problems (claims subscribed to by Churchman himself), what Kant really meant by 'critical', the serious-joyous opposition, and the problem of teleology. Isaac Levi and Ronald Giere manage to offer some coherent comments on that incoherent paper. J. Sayer Minas' 'Emergent Utilities' makes the important though usually neglected point that "preferences per se, and their induced utilities are not the whole of the story concerning decision-making in any normative sense", if only because the preferences themselves may be wrong to the point of aberration. "One expects the human subject to exbJbit both consistency and regularity on the one hand and change and development on the other. Consequently, we require a concept of preference in terms of which changes in the preference structure of the individual may be seen in some cases to be part of his 'natural' (or 'appropriate' or 'proper') development while others may be seen as evidence of inconsistency. That is, preferences and their associated utilities must not be seen as ultimates or as brute facts of the decision-making world if they are to play a significant role in that world" (p. 160). It is interesting to note that David Braybrooke champions independently a similar revision of standard preference theory (or rather the static model of consumer's behavior), emphasizing that a more adequate model must make room for detecting and rectifying mistakes in preferences. See his recent 'From Economics to Aesthetics: The Rectification of Preferences", Nods 8 (1974), 13-24. The volume includes papers by Peter C. Fishburn, I. J. Good, H6ctorNeri Castafieda, and Alex C. Michalos, which according to the editors have appeared elsewhere. It also includes a useful Bibliography compiled by Alan Ross and Danny Steinberg. Foundations and Philosophy of Science Unit, McGill University, Montreal
MARIO BUNGE