quarta-feira, 25 de janeiro de 2017


Drafts of summaries:


This book proposes ways to solve some main problems of present-day theoretic philosophy by returning to something more akin to ordinary language strategies. The stalemate in contemporary philosophy of language is due to a continuing, implicit positivistic effort to mimic the methods of natural science. This scientistic philosophy of the ‘latest novelty’ made for ‘immediate consumption by specialists’ has entrapped itself in self-created pseudo-problems, fragmenting analytic philosophy in almost unrecognizable ways and sacrificing the indispensable element of comprehensiveness. By reviving forgotten methodological procedures, this book tries to unify in a systematic way some of the best insights of not only exponential philosophers like Wittgenstein, Frege, Russell, Husserl, but also of influential theorists like Dummett, Tugendhat, Searle, Donald Williams, and even of some classic authors. Furthermore, it exposes some of the main fallacies popularized by the clever but scientistic expansionist reductionism of formalist philosophers from Quine to Kripke and Putnam. In this way it shows how some central questions of philosophy could regain the proper integrity of their true epistemic place. The book is simply and clearly written and can be understood by any reader with a basic training in analytic philosophy.

This book is extemporaneous, exemplifyig a new way of treating a subject. According to it, our present philosophy of language is like a headless mule that has lose its ways through a continuous, unconfessed positivist effort to mimic the procedures of science. This scientism has fragmented analytic philosophy in such a way that it has lose a fundamental treat of any philosophy, which is comprehensiveness. This book finds and unifies creatively the best insights of philosophers like Wittgenstein, Frege, Russell, Husserl, Dummett, Tugendhat, Searle, Williams, and even some classics, in a systematic way. Then again it criticizes the challenging but scientist reductionism of formalist philosophers from Quine to Kripke and Putnam, which are references to the mainstream. In this way it shows how central questions of philosophy could regain the integrity proper of their epistemic place, with swiping results for the contemporary philosophy. The book is exceptionally readable and made to be understood by any person with a basic training in analytic philosophy.

This book is a sustained attempt to develop systematic philosophy based on materials presented by the contemporary analytic philosophy. Cognitive meanings are given by semantic-cognitive rules and their combinations in building thought-contents or verifiability rules. Existence is the effective applicability of such rules. Truth is the correspondence between only conceived verifiability rules and the effectively applied ones, what includes coherence as intermediator. The criteria for the applicability of these rules are tropes and constructions out of them. The world is constituted by tropes. And as far as these criteria satisfy the conditions of external reality they may be considered tropes and constructions out of them belonging to the external world. The arguments in defence of these views cut deep in the inherited wisdom of much of the present mainstream philosophy.

This book was inspired by my growing awareness that the philosophy of language developed before the Second World War often seems more powerful and nearer to the truth than the kind of compartmentalized scientistic philosophy of language that the community of ideas has increasingly adopted since the War and that is almost hegemonic today. After W. V-O. Quine and particularly after Saul Kripke, this later Anglo-American philosophy of language has accumulated many very insightful scientistic-formalistic objections against all kinds of truisms accepted by the old analytic philosophy, with the result that descriptive metaphysics – which is a systematic enterprise relying on commonsensical truisms and aiming at comprehensiveness – turned out to be almost impossible to realize. As this book will show, the stalemate of much of the present philosophy of language results from the authority of these new views, which often corrupt philosophy through a positivistic-scientist abusively formalistic deformation of procedures that deprives central philosophical questions of their proper epistemic place. This book aims to show the power of the old views, on the one hand by systematically reconfiguring them under new models, and on the other by offering evidence of the ultimately equivocal character of many new ideas. Its guiding thread is a critical blend of ideas mostly developed by Wittgenstein, Frege and Husserl. If the views defended in this book are correct, they should have sweeping implications for the present stage of philosophy of language.


The book Philosophical Semantics will be published by Cambridge Scholars Publishing (2017/1)


- 1 -

In such theories, it seems to me, there is a failure of that feeling for reality which ought to be preserved even in the most abstract studies. Logic, I should maintain, must no more admit a unicorn than zoology can; for logic is concerned with the real world just as truly as zoology, though with its more abstract and general features.
Bertrand Russell

Would the old orthodoxy of the philosophy of language that prevailed in the first half of the Twentieth Century, with its insistence in the centrality of meaning, its eroded principle of verifiability, its correspondentialism, its sharp distinction between analytic and synthetic, its crude descriptivist-internalist theories of proper names and general terms, its naive monolithic dichotomy between the necessary a priori and the contingent a posteriori, be nearer to the truth than the now dominant causal-externalist orthodoxy?
   I wrote this book in the conviction that this question should be answered in the affirmative. In my view, the philosophy of language of the first half of the Twentieth Century was in some ways more profound, wider and nearer to the truth than those of the current orthodoxy, and its insights were more powerful. The reason resides in the socio-cultural background. All culture is born of great conflict; and the time between the end of the Ninetieth Century and the Second World War was a period of increasing social turmoil that made all established cultural values dubious, opening the gates for sweepingly original innovation in the sciences, arts, and also in philosophy.
   However, in saying this I am not attempting to dismiss the ‘normal philosophy’ that came later. I don’t reject, for instance, the interest of anti-verificationist arguments like those of W. V-O. Quine; nor do I reject the deep philosophical relevance of the new causal-externalist mainstream founded by Saul Kripke and Keith Donnellan in the early Seventies, which was later expanded by Hilary Putnam, David Kaplan and others. They are relevant and in a sense even indispensable to what I am trying to do.
   However, the importance of these achievements is in my judgment predominantly negative, since I think that their results are intrinsically less than the truth. In other words, in my view the relevance of these achievements consists mostly in being dialectically useful challenges, which if adequately answered would bring us to an enriching reformulation of old descriptivist-internalist-cognitivist views of meaning and reference, although more complex in an almost unrecognizable way.
   The aim of the present book is to contribute in the proposed direction, though often in an indirect methodological way. My approach to the issue here consists in gradually developing – aided by a critical examination of some central views of traditional analytic philosophy, particularly those of Wittgenstein and Frege, but also of others – the defense of a purely internalist, cognitivist and neo-descriptivist view of the meaning of our expressions and the mechanisms of reference. This task will be complemented by a renewed reading of the concept of existence as a higher order property, by a reconsideration of the verificationist view of meaning, in which the main objections to it are answered and, finally, by a thoroughgoing reassessment of the correspondence theory of truth, seen as complementary to the proposed form of verificationism.
   The obvious assumption that makes my project prima facie plausible is the idea that language is a system of rules, some of them directly accountable for meanings. The most central meaning-rules are those responsible for what Aristotle called apophantic discourse – the representational discourse, whose meaning-rules could be generally called semantic-cognitive rules. Indeed, it is prima facie highly plausible to think that the cognitive meaning (i.e., informative content and not mere linguistic meaning) of our representational language cannot be given by anything other than semantic-cognitive rules or combinations of such rules. This is a plausible idea, particularly because our knowledge of these conventions is usually tacit, implicit, non-reflexive, that is, we are usually unable to expose them by verbal means.
   My ultimate aim should be to investigate the structure of the semantic-cognitive rules by means of a careful examination of basic referential expressions, which are singular terms, general terms and even declarative sentences, in order to furnish an appropriate explanation of their reference mechanisms. In the present book this will be done only very partially, often in the appendices, summarizing ideas already presented in my previous work (Costa, 2014, ch. 2, 3 and 4), though still demanding development. In the body of the present book, my central tenet is rather to clarify my own assumptions on the philosophy of meaning and reference.
   By trying to develop these ideas, I realized, in retrospect, that what I wanted to do could in its core be understood as reviving a program already speculatively developed by Ernst Tugendhat in his classical work Vorlesungen zur Einführung in die Sprachanalytische Philosophie.[1] This book, published in 1976, can be seen as the swansong of the old orthodoxy, defending an anti-externalist program that was gradually abandoned during the next decade under the ever-growing influence of the new causal-externalist orthodoxy. Tugendhat’s strategy in developing this program can be conceived of in its core as a semantic analysis of the fundamental singular predicative statement, since it is not only epistemically fundamental, but also the most indispensable unit for building our first order truth-functional language.[2] In summary, giving a statement of the form Fa, he suggested that:

1)      the meaning of the singular term a should be given by its identification rule (Identifikationsregel),
2)      the meaning of the general term should be given by its application rule (Verwendungsregel), which I will also call characterization or, preferably, ascription rule,
3)      the meaning of the complete singular predicative statement Fa should be given by its verifiability rule (Verifikationsregel), which results from the combined application of the first two rules (Tugendhat & Wolf 1983, pp. 235-6, Tugendhat 1976, pp. 259, 484, 487-8).[3]

The verifiability rule is obtained by jointly applying the first two rules in such a way that the application of the identification rule of the singular term must be made first, in order to make possible the application of the general term’s ascription rule. Thus, for instance, Yuri Gagarin, the first person to orbit the earth from outside its atmosphere, gazed out of the small window of his space capsule and claimed: ‘The earth is blue!’ In order to make this into a true statement, he should first have identified the earth by applying the identification rule of the proper name ‘Earth’; then, based on the result of this application, he was able to apply the ascription rule of the predicative expression ‘…is blue’. In this combined application, these two rules work as a kind of verifiability rule for the statement ‘The earth is blue’. If these rules can be conjunctively applied, then the statement is true, otherwise, it is false, which Tugendhat saw not only as a form of verificationism, but also as a kind of correspondence theory of truth (a conclusion contested by some readers).
   One can critically ask if it is not possible that one first applies the ascription rule of the predicative expression. For example, one sees a fire at a distance in the night, without identifying what is on fire. Only after approaching it does one see that it is an old building. It also seems that the person in this example first applied the ascription rule and later the identification rule. However, by suggesting this we forget that in order to see the fire one first needs to direct one’s eyes at a certain spatio-temporal region, thereby localizing it: the region where something is on fire. Hence, a primitive localizing identification rule is first applied. Thus, initially the statement will not be; ‘That old building is on fire’, but simply ‘There is a fire over there’ or ‘There is a fire in that place’. Later, closer to the building, one can make a more precise statement. In this same way Gagarin could think ‘There is blue color down below me’ before saying ‘The earth is blue’, while looking out the window of his space capsule. But in these cases the ascription rule cannot be applied without the earlier application of some identification rule, even if only to be able to identify a vague spatio-temporal region.
   Tugendhat came to his conclusions as result of purely speculative considerations, without analyzing the structure of these rules and without answering external criticism of the program, like the numerous objections that others have already made against verificationism. But what is extraordinary is that he was arguably right, since I believe I can make his view more plausible. I can do this by investigating in some detail the structure of the semantic-cognitive rules, that is, first by analyzing the identification and ascription rules, then by refuting the main arguments against verificationism, and finally by explaining some fundamental attributes related to them, particularly existence and truth, in a more satisfactory way.
   My methodological approach, as will be seen, is also different from that used in the more formally oriented views opposed in this book, which are mostly inherited from the philosophy of ideal language in its positivist developments. It is primarily oriented by the communicational and social roles of language, which are seen as our fundamental units of analysis, being in this way more influenced by the so-called ordinary language tradition than by the ideal language tradition.[4]
   Since I consider that a comprehensive understanding of language must situate it in its unavoidable involvement in our overall societal life, I will always begin with commonsense and ordinary language, often seeking support in a more careful examination of concrete examples of how our linguistic expressions are really used.
   Finally, my approach is intended to be systematic. The chapters shall be interconnected so that the plausibility of each one is better supported when regarded in its relation to arguments developed in the preceding chapters and their often critical appendices. Even if complementary, these appendices are placed as a counterpoint to the chapters, aiming to justify the expressed views, if not to add something complementary to them.

[1] English translation under the title Traditional and Analytical Philosophy: Lectures on the Philosophy of Language (2016).
[2] In this book I will use the word ‘statement’ in most cases as referring to the speech act of making an assertion.
[3] According to J. L. Austin’s correspondence view, an indexical statement (e.g. ‘This rose is red’) is said to be true when the historical fact correlated to its demonstrative convention (linguistically expressed by the demonstrative pronoun ‘the’) is of the type established by the sentence’s descriptive convention (the given red rose type) (Austin 1950, p. 122). This is to me the first approximation of similar conventionalist strategies later employed by Dummett in his interpretation of Frege (see 1981, pp. 194, 229) and still later more cogently explored by Tugendhat.
[4] The ideal language tradition (inspired by the logical analysis of language) and the ordinary language tradition (inspired in the workings of natural language) represent opposite views. The first was founded by Frege, Russell and the early Wittgenstein. It was also strongly associated with philosophers of logical positivism, mainly Rudolf Carnap. With the rise of Nazism, most philosophers associated with logical positivism fled to the USA, deeply influencing American analytical philosophy. The philosophies of W. V-O. Quine, Donald Davidson, and later Kripke, Putnam and David Kaplan, along with the present mainstream philosophy of language with its metaphysics of reference, are in direct or indirect ways a later product of ideal language philosophy. The ordinary language tradition, on its side, was represented after the Second World War by the Oxford School. It was inspired by the analysis of what Austin called ‘the whole speech act in the total speech situation’. Its main theorists were J. L. Austin, Gilbert Ryle and P. F. Strawson, although it had an antecedent in the later philosophy of Wittgenstein and still much earlier in G. E. Moore’s commonsensical approach. Ordinary language philosophy also affected American philosophy through relatively isolated figures like Paul Grice and John Searle, whose academic influence has been foreseeable less. For history, see J. O. Urmson (1956).

terça-feira, 24 de janeiro de 2017


The Philosophical Inquiry: Towards a Global Account, was published by the University Press of America, Langham, 2002)

             CLAUDIO F. COSTA


                                          Nun scheint mir, gibt es ausser der Arbeit  des
                                          Kunstlers  noch  eine  andere,  die   Welt   sub
                                          specie aeterni einzufangen. Es ist – glaube ich,
                                          der Weg des Gedankens,  der  gleichsam  über
                                          die Welt hinfliege und sie so lässt,  wie sie  ist
                                          – sie von oben von Fluge betrachtend.**
                                           Science is what we know; philosophy is what
                                           we don’t know. (…) Science is what we  can
                                           prove to be true; philosophy is what we can’t
                                           prove to be false.
                                           (B. Russell)

*   The sibyl with raving mouth uttering her unlaughing, unadorned, unincensed words reaches out over a thousand years with her voice through the god. (tr. K. Freeman)               
**  Now it seems to me that besides the work of the artist there is still another, the world sub specie aeterni to capture. It is – I believe, the way of thought that, so to speak, flies over the world and leaves it as it is – from above, from flight regarded.


1. Methodological remarks
1. The shortcuts of the critique of language
2. Philosophy as analysis of language
3. The objectual fallacy in analytical philosophy
4. Concluding remarks
1. The conjectural character of all philosophical inquiry
2. The idea of philosophy as protoscience
3. Origins and divisions of science
4. Some examples of protoscientific philosophical insights
5. Fission
6. The remaining core of philosophical problems: Two hypotheses
7. Our general idea of science
8. Towards a nonrestrictive view of science
9. Why should philosophy be viewed as a protoscientific endeavor?
10. Consequences of the proposed view
1. Philosophy and religion: The genetic approach
2. Comte’s law of three stages
3. A brief appraisal of Comte’s law
4. Philosophy as a transitory inquiry between religion and science
5. Conclusions
1. The artistic flavor of some philosophical writings: External similarities
2. Internal similarities between philosophy and art
1. Philosophy as a derivative cultural activity
2. An integrative account of the philosophical activity
1. Forms of philosophy
2. Three historical phases in the evolution of philosophy
3. Linguistic-analytic philosophy in the wheels of history
4. The future of philosophy


This book was written during the time I spent as a visiting scholar in the Department of Philosophy of the University of California at Berkeley. It is an attempt to put together in a systematic way some metaphilosophical ideas I’ve published in Brazil in the last five years. As a whole, it contains the outline of a global account of the nature of philosophical inquiry, developed from a broad historic-cultural perspective.
     I’m very grateful to professor John R. Searle, who accepted me to do post-doctoral research at Berkeley. I also thank Ms. Sharon Hudson for her efforts to improve a text written in a foreign language. This book would not be possible without the support of CAPES.
Konstanz, 2001


                                           AIMS AND METHODOLOGY

Among the many philosophical problems, the problem of the nature of philosophy is certainly not the most important or exciting. Nevertheless, it is one of the most troublesome for a philosopher. For how can he intend to do philosophy, or to do it rightly, if he can’t even tell what kind of thing he is trying to do? This book is an attempt to give a general explanation of the nature of philosophical inquiry, not from a particular philosophical perspective, but from a comprehensive examination of philosophy in its historical development and in its connections to other cultural activities.
     One objection often made to the attempt to give a unified account of the nature of philosophy is that it is so multifarious and changeable a subject that any effort to enclose it in an appropriate theoretical framework is destined to failure. One can’t classify clouds by their forms, as Wittgenstein once said. However, why would it not be possible to approach philosophy theoretically, if we were able to encapsulate it under criteria originated from a sufficiently general and flexible perspective? After all, in a general way, at least, meteorology has long classified kinds of clouds by their forms. In the following chapters, I will try to show that a general theoretical approach to philosophy is possible; a succession of arguments will be advanced in order to provide us with a comprehensive theoretical framework capable of giving us explicit means to identify and map the philosophical territory. But before doing so, some previous methodological observations must be made.

There are two methodological points to be considered. The first concerns a distinction between two different kinds of approach of the nature of philosophy: the prescriptive approach and the descriptive approach.
     The prescriptive approach intends to say what philosophy should be; it is a proposal for what should be called by this name. Carnap’s definition of philosophy as an investigation of the logical syntax of scientific language(1), Heidegger’s view of philosophy as the science of the beingness of Being (of Being as such)(2), Wittgenstein’s view of philosophy as a therapy against the bewitchment of our understanding by means of language(3)… all of these were proposals for what these philosophers believed that philosophy should be. A prescriptive approach can’t be shown to be true or false simply by being compared with the real historical praxis of philosophy, since this approach is not made in order to represent this praxis. Relative to the philosophical praxis, a prescriptive account can only be successful if adopted, or unsuccessful if not. And in fact, some prescriptive accounts have been successful. The epistemological turn, unwittingly impressed on modern philosophy by Descartes, was a successful prescription, at least for some time. And the same can be said about the linguistic turn that Frege, Russell and Wittgenstein have impressed on the philosophy of the twentieth century. Saying what philosophy should be, the prescriptive approach remains unconcerned with the past practice of philosophy. To say it figuratively, it “looks to the future”.
     The descriptive approach, on the other hand, doesn’t intend to say what philosophy should be, but what philosophy has been. It “looks to the past”, trying to make explicit the criterial conditions that the philosophical community has implicitly admitted for the identification of philosophy, in its technical or academic sense, during the whole history of this subject or at least a regional segment of it. Descriptive accounts are the kind of explanation more likely to be found in dictionaries and philosophy textbooks, than in the doctrine of particular philosophers, since the latter are more committed to their own particular perspectives. However, when C. D. Broad understood philosophy as the search for a general view of the world and of man’s place in it(4), when G. E. Moore suggested that philosophy, under other things, is an attempt to give a general description of the largest classes of things in the universe and the way they are related to one another(5), and when E. Tugendhat wrote that philosophy is the elucidation of the network formed by the concepts constitutive of our understanding as a whole(6), they were trying to meet a descriptivist paradigm, by covering as far as possible the extension of what has always been called ‘philosophy’.
     Time seems to work for descriptive approaches since it seems that, as time passes, room for prescriptive approaches grows smaller, while room for descriptive approaches grows larger. If someday philosophy comes to an end, there will remain no room for proposals. Today, as some are speaking about the decline or even the end of philosophy, the descriptive approach seems to be the most reasonable. The metaphilosophical account pursued in this book will also be a descriptive one.
     It is important to make clear in what sense I will speak of ‘philosophy’ under the descriptivist perspective. It is not in the vernacular sense of the sum of unexamined beliefs necessarily maintained by everyone in order to direct life, and it is also not in any popular sense, like speaking of philosophy as a general guide to the conduct of life. The investigation will be restricted here to the privileged scholarly, academic or proper sense of the word ‘philosophy’, the sense in which the occidental philosophical tradition has used to refer to itself, and which is paradigmatically exemplified by the work of the most outstanding philosophers of this tradition. By making this scholarly sense explicit, we can expect to make explicit the criteria by which we can use the word ‘philosophy’ indicatively, in order to point out what belongs and what doesn’t belong to it; and more than this, we can hope to find a rationale for the existence of such criteria.
     Now, what entitles us to hope that it is possible to give such a unified account of the nature of philosophy? The task seems to be prima facie plausible because we not only have the (perhaps misleading) feeling that the word ‘philosophy’ has some kind of unified scholarly or academic sense, but also because we are generally able to decide with great assurance what counts (or not) as philosophy in this sense. From this it seems to follow that, by sufficiently careful examination of the applications of the term ‘philosophy’ we should also be able to make explicit the conditions that have guided our decisions, explaining them theoretically and then organizing them in the form of a general characterization. Though it seems also possible that the term ‘philosophy’ doesn’t have any kind of unified technical meaning, I will adopt from now the thesis that it has such a meaning as a working hypothesis to be evaluated by its results.
     One could also object that the assumption that we are unaware of the criteria we apply to identify the designata of general names belongs to ordinary language philosophy, being as such applied to some general names of our language, names like ‘knowledge’ and ‘truth’, which express timeless categories of thought; ‘philosophy’ however, is not such a term. That this objection is insufficient can be shown when we consider that we are also unaware of the criteria we apply to more recent technical terms, like ‘theory’, ‘explanation’, and ‘observation’, terms that are used in the sciences. If we ask a scientist what ‘scientific explanation’ means, he will tend to answer by means of examples. It is the task of the specialized philosopher of science to make explicit precisely what those terms mean. Now, why couldn’t this idea also apply to the word ‘philosophy’ itself? Indeed, the concept of philosophy was introduced in our academic culture a long time ago, subsequently undergoing a natural growth on its own, a growth apparently sustained by the nature of the subject.  If we could make explicit the criteria for the identification of philosophy in a way that also provides a rationale for the use of the word, we would not only understand better what the philosopher is trying to do, but also prevent a deceptive practice of philosophy.
     The second point I wish to address, concerns two opposite dangers usually faced in metaphilosophical investigations. The first may be called poverty. The definition of philosophy as an account of the whole world and of the place of man in it, though very inclusive, is plainly too vague and uninformative and, if we consider it more closely, we see that it does not succeed in giving a necessary condition, since there are philosophies that don’t do this, nor a sufficient condition, since religion can also do it. Poverty handicaps most of the descriptivist accounts. Escaping from such vacuity, then, one often succeeds in saying something more definite at the expense of generality. This second kind of inadequacy may be called reductivism, and is the unavoidable limitation of the prescriptivist approach. Carnap’s definition of philosophy as the investigation of the logical syntax of language is a prime example of reductivism, paying for the advantage of precision an exorbitant price in exclusion.
     Intending to preserve the extent of the subject matter without falling into the constraint of an excessively uninformative characterization, I will pursue here the descriptive approach in an integrative way, trying to conjoin what seems to be descriptively true of some well-known views of the nature of philosophy in order to uncover and justify the most distinctive features of philosophical inquiry.


                           PHILOSOPHY AS CONCEPTUAL ANALYSIS:
                                A CASE OF REDUCTIVE DEFINITION

When we look at the history of philosophy as descriptivist metaphilosophers, there are some accounts of its nature that we are tempted to reject without further consideration. This is the case of any account based on the proper subject matter or on the proper method of philosophy. For there are as many distinct philosophical objects or methods as philosophies or philosophical movements. Only the prescriptivist metaphilosopher can still have the hope (or fancy) of finding the subject matter or method proper of philosophy; the descriptivist will tend to see these kinds of explanations as inherently reductive, narrowing unnecessarily the frontiers of philosophy.
     Since my intention is constructive rather than critical, I will examine only one view of the nature of philosophy that relates it to a proper method and often also to a proper subject matter, for this view underlined the main developments of philosophy in the twentieth century. It is the extremely influential and still widely accepted view that the proper method of philosophy is that of conceptual analysis and that the proper subject matter of philosophy is what might be called the logical-grammatical structure of our most central concepts. This view was sustained in different ways by philosophers like Wittgenstein, F. Waismann, A. J. Ayer, P. F. Strawson, M. Dummett, E. Tugendhat and many others.
     The view of philosophy as conceptual analysis was seriously challenged through the so-called “naturalist turn”, promoted especially by Willard V. O. Quine(7). For him, philosophy is more then merely a matter of linguistic-conceptual investigation, since it is not something essentially distinct from empirical science; in fact, there is no real distinction of subject matter here: philosophy forms a continuum with science, and the distinctions that might be traced are merely artificial, somewhat like state frontiers(8). Though this point of view has some advantages of its own, the fact is that no advocate of this naturalist or scientist turn has been able to explain why the rest of us remain not very disposed to see the frontiers between science and philosophy as the result of arbitrary conventional agreements. The thesis that the distinction between philosophy and science results from an artificial decision does not explain why we feel such a strong resistance to the idea of altering the present frontiers, calling science what has been called philosophy or vice versa. Moreover – and what seems to me decisive – Quine’s thesis doesn’t explain why we don’t need to appeal to any new conventional agreement in order to identify a new idea as philosophical instead of scientific or vice versa. The view of philosophy as conceptual analysis at least has the merit of trying to answer these questions by spelling out what should be characteristic of philosophy.
     Though there are many different versions of the view of philosophy as linguistic-conceptual analysis, I will reduce them somewhat artificially into two general forms, in order to show better the view’s intrinsic limitations. I call these forms critique of language and analysis of language. By doing critique of language we seek to analyze or elucidate concepts in order to dissolve philosophical confusions; by doing analysis of language, we seek to analyze concepts for the sake of a better understanding of our own conceptual framework. I will explain what I understand by each of these views, showing that in spite of their own merits, they fail to give an adequate explanation of the nature of philosophy.

The critique of language seeks to uncover flaws in philosophical arguments, many of them belonging to traditional philosophy. This has historically been done in two ways. First, as an analysis of the logical structures of sentences – what I call a syntactically oriented form of analysis. The second kind of analysis is a close examination of the meanings or uses of the expressions of our ordinary language in their actual interpersonal contexts – what I call the pragmatically oriented form of analysis. I use the expressions ‘syntactically’ and ‘pragmatically oriented form of analysis’ to replace the old and misleading dichotomy between ideal language philosophy (guided by logic) and ordinary language philosophy (guided by natural language). This dichotomy is misleading because the history of analytical philosophy has shown that nothing prevents an inquiry into ordinary language being conducted from a logical perspective, as in cases like J. R. Searle’s formalized treatment of the theory of speech acts by means of illocutionary logic, and P. F. Strawson’s truth-functional account of the concept of presupposition; on the other hand, nothing prevents us from conducting ideal language inquiries from the perspective of their realization in ordinary language, as the investigation of the uses of logical particles in natural language shows.
     The syntactically oriented form of the critique of language can be exemplified by the observation of analytical philosophers (like Russell and Ryle)(9) that one reason underlining Plato’s creation of his doctrine of ideas might be a confusion generated by the superficial similarity between the linguistic grammar of sentences like “The beauty is pleasing” and “Socrates is bald”. Led by such similarities, Plato concluded that, since the subject of sentences like the latter refers to something, the subject of sentences like the former must also refer to something. However, since there is no “the beauty” in the visible world, “the beauty” must be an inhabitant of a world that is only intelligible, which Plato calls the super-celestial world of ideas. Against this conclusion, the critique of language, based on modern predicative logic, shows that the logical structures of both kinds of sentences are only apparently identical, since the former has a logical structure that is very different from its superficial linguistic-grammatical structure. A sentence like “The beauty is pleasing” is merely an abbreviation of the sentence “For all x, if x is beautiful, then x is pleasing”, where ‘beautiful’ is shown not to be a name at all, but a predicative expression. The suggestion is that the sameness of the superficial subject-predicate form of both kinds of sentences has mislead Plato, leading him to construct a metaphysical castle of cards.
     The second example – now of the pragmatically oriented form of the critique of language – concerns the critical exposition of linguistic distortions that underline the argument of illusion, an argument posed by representationalist (and phenomenalist) epistemologies opposing direct realism. In this argument, cases are considered in which objects seem different from what they really are, like a straw that looks bent when in a glass of water. The consideration of these cases leads us to the conclusion that what we directly perceive are not material objects, but only our representations (or sense-impressions) of them. Against this, linguistic critics such as John L. Austin have shown that in such cases we don’t say that we don’t directly perceive the objects, but only their representations; it would be more appropriate to say that we still see the objects directly, though not how they really are. So when I look (with both eyes) at my own nose, I don’t say that I really see two noses, but rather that I see my own nose duplicated; and when I see a coin that looks elliptical, I don’t say that I see an elliptical object, but that I see a round object that looks elliptical (10).
     Examples like these serve to show not only the qualities, but also the limits of the critique of language. For it is clear that Plato’s doctrine of ideas, as an attempt to explain our grasp of generality, and the representationalist objections to direct realism, mainly by means of the argument of science, remain beyond the reach of a purely linguistic criticism. A reason to think so is that the arguments for the admission of ideas in order to explain generality and predication, like the arguments for the admission of representations as the most immediate objects of experience, mediating unavoidably our access to the external world, have both a substantive content, which seems to be refutable only through material considerations in the correspondent domain of questioning. A further reason that such theories remain resistant to a purely linguistic criticism is that, based on this substantive content, one could always claim to be innovating language, using analogical speech as a means to convey new and original intuitions for which our natural language does not currently provide enough resources.
     Generally, the critique of language is not seen as a view of the nature of philosophy, but only as a critical way of doing it. Nevertheless, the critique of language became a view of the nature of philosophy in the hands of Wittgenstein, who seems to have viewed philosophy as a kind of linguistic therapy without any positive content of its own(11). Even if it is questionable to what extent Wittgenstein would endorse this view, since he also made pronouncements that run against it, this view can be easily picked out from his texts, and I will expose it for what it can teach us(12).
     The therapeutic view of philosophy maintains that much of philosophy (especially traditional philosophy) is a result of linguistic confusion. Philosophers are people possessed by a metaphysical craving for generality(13), which disposes them in an irresistible way to be misled by the superficial structures of our language, leading them to the construction of  theoretical “castles of cards”, when not turning them into hopeless prisoners of their own “knots of thought”. In face of this, a sound philosophy must be therapeutic: the aim of the therapeutic philosopher is to undo the theoretical castles of cards of the speculative metaphysician and to unknot the knots of thought into which more austere thinkers have knitted themselves. And the way of doing this is not by constructing any theory, nor by explaining anything, but only by describing the real ways we use our words – by bringing these words, as Wittgenstein says, from their metaphysical uses back to their ordinary ones. Hence, philosophy must be a purely destructive undertaking, only succeeding when the philosopher is liberated from his metaphysical preoccupations and, like a patient liberated from his neurotic fixations, becomes able to forget philosophy.
     The problem with the therapeutic view of philosophy is that it cuts the branches too short. No criticism of language has succeeded in being entirely nontheoretical and nonexplanatory. Wittgenstein’s own work is a good example of this failure, though this fact is usually hidden by the fragmentary and elusive character of his writings(14). Consider, for example, Wittgenstein’s remarks on proper names in the Philosophical Investigations(15). These remarks are generally intended as a criticism of the “label-theory” of proper names, by which the meaning of a proper name is the object presented by it, in a way similar to a bottle’s label presenting its content. However, by refuting this theory Wittgenstein is, intentionally or not, framing a more sophisticated version of the bundle-theory of proper names, which explains the meaning of names like ‘Moses’ by the different descriptions we bind to it, like “the man who led the Israelites through the wilderness”, “the man who lived at that time and place and was then called ‘Moses’”, or “the man who as a child was taken out of the Nile by Pharaoh’s daughter”. (Using Wittgenstein’s own vocabulary, we could add that these descriptions are expressions of rules for the identification of the named object, rules that jointly may constitute what we mean by the proper name.) Now, Wittgenstein’s suggestions are theoretical, since their therapeutic efficacy depends on a suggested generalization for all proper names; and his remarks are also explanatory, since they aim to explain how individuals can be identified by proper names. These two features are clearly shown later, by the explicitly theoretical suggestion of a bundle-theory of proper names along the same line by Searle(16). Examples like this show that a philosophical therapy, to be effective, to cure the disease and not only to relieve this or that occasional symptom, must be based on generalizations with explanatory power. These explanations, when put forward, force us to abandon the terrain of ordinary-language descriptions in the direction of more and more elaborated theoretical constructions. Criticism and theory, we conclude, can’t be completely set apart from each other; they are the two opposite sides of the same philosophical coin, and it seems to be merely a circumstantial matter if one prefers to stress one side more than the other.

The failure of the purely therapeutic view of philosophy leads us to consider what I call the view of philosophy as analysis of language. The analysis of language is the constructive, theoretical side of the philosophical coin, capable of furnishing support for the critique of language and possibly of including it as an extension of itself. Analysis of language can also be done in a syntactically or in a pragmatically oriented way. An example of the syntactically oriented form is Rudolf Carnap’s outline of a general structure required by the syntax of every language, by means of his distinction between formation rules (specifying symbols and sentences) and transformation rules (determining the possible relations between sentences)(17). On the other side, an example of the analysis of language in its pragmatically oriented form is Searle’s theory of speech acts, which maintains that the structure of our communicative actions is generally reducible to the form F(p), where p is a propositional content and F the illocutionary force, namely, the kind of interpersonal commitment to be associated with this content(18).
     Analytical constructions like these are also very general theories, having interest on their own, since they are cognitive enterprises able to bring us near the horizons of science. In fact, Carnap’s characterization of the general features of the syntactic structure of any language has long since been incorporated into the formal science of logic (and not into a philosophical discipline, despite his misleading claim that philosophy is logical syntax) and the speech acts theory belongs today to the pragmatic field of linguistics, more than to philosophy. Although such theoretical constructions can also be used as critical tools, this is not the chief reason for their development, which is to widen the frontiers of our knowledge.
     In what follows, I will expose a full-blooded version of the view of philosophy as analysis of language. This version belongs to the pragmatically oriented form, but extends it to its limits of tolerance and defensibility, incorporating syntactic forms of analysis when necessary. Something near to this version can be found, with individual differences, in the views of the later and more well-advised practitioners of analytical methods, like Peter Strawson and Ernst Tugendhat.
     A basic assumption of the full-blooded view of philosophy as analysis of language is the idea that we are unaware of the extraordinarily complex structure of the most central concepts of our natural language, concepts which are intrinsically related to one another, like the concepts of truth, knowledge, belief, perception, cause, time, good, justice, beauty, etc. This lack of awareness has an explanation: we don’t learn these concepts by means of explicit definitions, but from childhood through a non-cognitive praxis of positive and negative exemplifications, where our learning is repeatedly submitted to interpersonal correction. Consequently, though it seems clear that we know the meanings (or concepts) expressed by words like ‘truth’, ‘time’, and ‘beauty’, since we know how to use them, we remain unable to describe how we use these words, to make the rules constitutive of their meanings (concepts) explicit. This is the reason why, though we can use such words adequately, we are all seriously embarrassed when we are asked to explain what we mean by them. Due to this lack of awareness of the rules governing the use of words, philosophical confusions can easily arise: philosophers, particularly those doing speculative metaphysics, have systematically misunderstood the uses of our expressions; and we already saw how the critique of language works, analyzing the logic-syntactical structures of the relevant concepts or making a pragmatic analysis of the uses of the words expressing them, in order to demonstrate the implausibility of these attempts. In itself, analytical philosophy of language is not a critical undertaking; its main concern is to construct theories intending to make explicit our most central conceptual structures. Containing generalizations, these theories also have explanatory value. And their most distinctive aim is to furnish what we, with Wittgenstein, could call a perspicuous representation (übersichtliche Darstellung): an overview of the grammatical structure of the most fundamental concepts of our language(19). Since these concepts are generally interrelated, the perspicuous representation might also make explicit the systematic relationships among them, aiming to elucidate what Tugendhat called the conceptual network (begriffliches Netzwerk) constitutive of our understanding as a whole(20).
     To complete our picture, it is important to say something about the most pervasive feature of analytical philosophy. It is what Quine called the semantic ascent(21) and what I – not being afraid of intensions – prefer to call the linguistic-conceptual accent, meaning by this the discursive emphasis on the linguistic and conceptual elements. By means of the semantic accent, linguistic-conceptual aspects of our expressions are focused in order to make fine distinctions explicit and to prevent linguistic-conceptual confusion. To give examples: the question “What are numbers?” was rephrased by Frege as the question, “What is the meaning of sentences containing number-words?” and the assertion, “The world is made of facts, not of things” was rephrased by Carnap as, “The concept-word ‘world’ is so understood, that by means of it only a system of facts, not of things, can be referred to”. The notion of semantic accent is reminiscent of Carnap’s concept of the formal way of speaking, for him the way of speaking that is adequate to philosophical issues, namely, the linguistic-conceptual ones. However, as Quine insightfully noted, semantic accent differs from the formal way of speaking by being viewed as applicable not only to philosophical sentences, but to every conceivable sentence: “Semantic ascent”, writes Quine, “applies anywhere. ‘There are wombats in Tasmania’ might be paraphrased as ‘‘Wombat’ is true of some creatures in Tasmania’, if there were any point in it. But it does happen that semantic accent is more useful in philosophical connections”(22).
    The notion of linguistic-conceptual accent can be explained more clearly when we consider that for technical reasons, by doing analytic philosophy, we present our arguments – in a more or less informal manner – in a semantic metalanguage, which allows us to center our discourse in our words and the concepts conveyed by them. But it is important to stress that this is usually done by means of a semantic and not a merely syntactic metalanguage, since with this we will be able to answer the objection that analytical philosophy, being a linguistic endeavor, unavoidably takes the world away (see note 25). To make this suggestion clear, compare the two following sentences:

(a)  “’Kraków’ is a name-word with six letters.”
(b)  “’Kraków’ is a word naming a city located 50° north of the
             equator and 20° east of the meridian of Greenwich.”

     In sentence (a) we use a syntactic metalanguage to speak about a word. In sentence (b) we use a semantic metalanguage to speak not only about a word, but also about what the word means. Using Fregean terms, we can say that using a semantic metalanguage we are making explicit the senses of our words and that by doing so, we are also speaking about their references, at least insofar as these references are available to us by conceptual means (Frege called the sense of a name the Art des Gegebenseins eines Gegenstandes). In sum: by means of a syntactic metalanguage, we speak only about the signs in abstraction of their meanings; this is the way of dry formalism. By means of a semantic metalanguage, we preserve the meanings as well as the signs, speaking about both; this is the philosophical way, by which the analysis of language can be extended from words to what is meant by them. The linguistic-conceptual accent is a way of centering our attention on language without excluding anything valuable that might be represented by language.
     Though the syntactically oriented form of the analysis of language, practiced by philosophers like Carnap, Quine, D. Davidson and S. Kripke, also employs the linguistic-conceptual accent, it differs importantly from the full-blooded view of analysis in its attitudes towards commonsensical constraints. Syntactically oriented philosophers give much more weight to the internal consistency of their formally-oriented theories than to their eventual agreement with our ordinary-language, commonsensical intuitions, usually being prepared to sacrifice the latter for the former. In fact, many of their ideas are in flagrant contradiction to these intuitions. What is the reason for this? The answer is not difficult to find. One is perfectly able to learn the syntax of a language – the rules for the combination of its signs – in a state of ignorance, without knowing the references of these signs and their combinations, without knowing their meanings and how to use them in concrete situations. But the opposite is not conceivable: one can’t adequately grasp the meanings of our combinations of signs and the ways these signs are used without knowing their syntactic functions, the ways they can be combined in the construction of well-formulated sentences. This means that, though the understanding of the syntactic dimension of our language does not presuppose the understanding of the pragmatic dimension, the pragmatic dimension, to be fully understood, does presuppose the understanding of the syntactic (and semantic) dimension (see chapter VII, section 3). This also means that this pragmatic dimension carries with it, as least as assumption, the whole set of meaning-rules of the language, a set that constitutes our linguistic-conceptual intuitions, our commonsensical intuitions about the meanings of our expressions, the ways we use them. This means that syntactically oriented analytic philosophy, being independent of the pragmatic dimension, might be developed in abstraction, and consequently also, in disagreement with the pragmatic dimension, without loss of intelligibility. A consequence of this is that the syntactically oriented conceptual analyst feels himself free to confront the usual constraints of reasonability and to propose views apt to challenge our common visions of the world, even if eventually in an illusory way, since this is done reductively, by arbitrarily abstracting assumptions that constitute this vision. (This explains, for example, why Quine’s and Kripke’s arguments can so easily run against linguistic common sense, while Searle’s and Strawson’s arguments cannot).
     In the next section, theoretical consequences that philosophers have drawn from the described views will be critically considered in order to show that the view of philosophy as analysis of language (and, consequently, also as a critique of language) is incapable of showing us what philosophy is.

Many defenders of philosophy as conceptual analysis think that their views lead to the conclusion that, since the philosopher is exposing the conceptual structure of our language, he is

(a) in no way advancing any speculative hypothesis about the world, like traditional speculative philosophers have done,

      and he is

(b)  in no way advancing any empirical hypothesis about the world, like the natural scientists (even if the undertaking of describing the way language actually works might be seen as empirical(23)).

     My aim in this section is to show that neither claim (a) nor claim (b) can be fulfilled by the real praxis of philosophy as analysis of language, and that the claim that they are fulfilled is an insidious objectual fallacy. Moreover, by showing that these claims are fallacious, I hope also to prove wrong the assumption that from the point of view of its subject matter, analytical philosophy distinguishes itself from traditional speculative philosophy and from empirical science, since it deals with the structure of our concepts and consequently, not with the world. The lengthy argument I use to show this is not a paragon of linearity and transparency, but here it goes:
     To show that the conceptual analysts haven’t succeeded in establishing that conceptual analysis possesses a different subject matter from traditional philosophy and from science in general, we need to begin with a consideration of its actual praxis. The theses (a) and (b) could indeed be consistently maintained if conceptual analysts had limited themselves to a logical analysis of the structure of sentences or to a tedious, quasi-lexicographic description of the meanings of the philosophically relevant concept-words of our natural language. But this is not what they actually do. In order to achieve any kind of philosophical relevance, conceptual analysts must take a step beyond: they must inquire into their actual praxis of thought about things, discovering in this praxis concepts for which there aren’t any words in our language yet, such concepts being chosen by virtue of their explanatory richness. Since these newly discovered concepts can be expressed only through new concatenations of words, conceptual analysts are often led to replace these concatenations of words with new terms of art, invented for discursive economy. Some examples illustrate this procedure: the proponent of a theory of communicative actions can make an analysis of our “speech acts” under the perspective of their “illocutionary forces”; someone engaged in philosophy of content can try to analyze the representational function of our statements, their “factual meaning” in terms of “rules of verifiability”; an epistemologist can suggest an analysis of the concept of our “propositional knowledge” (“knowing that”) in terms of “ultimately undefeated justified true belief”.
     When we reflect about these procedures, the first point to be considered is that the supposedly purely analytical procedure contains a hypothetic-synthetic moment. Deep conceptual structures are first hypothetically uncovered and only then analyzed (see note 57). But by doing this, the philosopher is already doing a work of generalization or synthesis – or, we may also say, he is trying to uncover a kind of synthetic unity that (at least to the pragmatically oriented analyst) would be already present in the uses of our language. However, the adequateness of this uncovered conceptual unity is highly hypothetical. This is shown by the fact that the meanings of the highly abstract words used to explain this new conceptual unity are themselves controversial; indeed, the philosopher is trying to establish a newly discovered concept embedded in the whole network of beliefs consciously or unconsciously assumed by him as the most consistent and natural, what makes his endeavor inevitably tentative. In so doing, he poses eventually fruitful hypotheses. These hypotheses are about the empirical structure of language in the case of the speech act theory, about the representational function of our statements in the more speculative case of the verifiability principle, and about our mental form of evaluation of our “knowing that” in the case of the definition of propositional knowledge. The whole endeavor might be analogous to the work of discovering a “law of nature” in the natural sciences, namely, something able to explain a variety of individual cases and able to be later confirmed or disconfirmed.
     I think that a liberal conceptual analyst would have no great difficulty in accepting these objections. But he will usually maintain that, even if his actual analytical procedure is preceded by a hypothetic-synthetic moment, he is always trying to make explicit what already belongs to our conceptual system and never, like the empirical scientist or the speculative philosopher, going beyond this system by making hypotheses about the actual empirical world. However, when we examine the given examples, we see that they can also be interpreted as dealing with empirical facts, even if very general ones, related to our relationships to our mental or linguistic representations of the world rather than to the world itself. Moreover, when we examine other examples of analysis, like those coming from the field of analytical metaphysics, we see that these empirical facts may also concern the physical world itself. Consider the analysis of scientific laws as relations of “nomological necessitation holding between universals” sustained by D. M. Armstrong(24). By posing (synthetically) the hypothesis that natural laws can be analyzed in this way, the philosopher is suggesting in an Aristotelian manner the existence of some kind of universal in re, able to establish the ways by which certain empirical events can be related to one another. However, this is a hypothesis that concerns the natural world too, even if in a non-straightforward way.
    Though this seems to be a clearly unsatisfactory conclusion, the conceptual analyst still has an answer to it. He would say that the conclusion is acceptable, since, as the world is reflected in the structure of our concepts, by analyzing concepts we are also saying something about the world. As A. J. Ayer said, “The distinction between ‘about the language’ and ‘about the world’ isn’t at all sharp, because the world is the world as we describe it, the world as it figures in our system of concepts. In exploring our system of concepts you are, at the same time, exploring the world”(25).
     Though this is true, and confirmed by our previous considerations of the linguistic-conceptual accent, this answer points to the fact that we can’t distinguish the proper subject matter of philosophy by referring to the analysis of our conceptual structures. For in a similar sense we can suggest that, like the conceptual analyst, the empirical scientist and the speculative metaphysician are doing a work of “conceptual analysis”, the only difference being that they are not aware of it, and do not have to worry about focusing on the linguistic-conceptual aspects of their inquiry by means of a semantic metalanguage. I will try to make this last point clear by raising separate objections against (a) and (b).
     Consider thesis (a): analytic philosophers are not making assertions about the world, like speculative philosophers.
     Against this thesis, it is important to consider that the recent history of analytical philosophy has shown that all realms and positions of traditional philosophy can be reached by the work of analytical philosophers. This history has made clear that somewhat corresponding distinctions, which philosophers hold between

    critic philosophy               vs.    speculative philosophy
     (occupied with  the  de-            (aiming to reach general conclusions a-
     finition and critical ana-            bout the nature of the universe and our
     lysis of the concepts of              position and prospects in it)
     our daily life and scien-              (C. D. Broad),

     descriptive  metaphysics    vs.   revisionary metaphysics
     (occupied with the des-            (attempting to create a new structure
     cription  of  our  actual             of thought (P. F. Strawson),
     structures  of  thought)

     immanent metaphysics       vs.    transcendent metaphysics
     (limiting itself to the                  (aiming to pass beyond the world of
     word of  the  senses)                  senses,  relating itself to  the  super-
                                                     sensible, considered as the truly real
                                                     world) (W. H. Walsh),

find a certain parallel in the domain of analytical philosophy in the distinction between

      the results of the pragma-    vs.   the results of the  syntactically
      tically oriented analysis               oriented analysis of language.
      of language

     In fact, there is a deep reason for the existence of this parallel: the dependence on commonsensical and linguistic intuitions maintained by critic philosophy, as well as by immanent and descriptive metaphysics, corresponds to the dependence on ordinary-language intuitions maintained by the pragmatically oriented conceptual analyst. On the other hand, as we saw, the syntactically oriented conceptual analyst usually does not have the same regard for ordinary intuitions, resembling in this way the speculative philosopher.
     These considerations suggest that the distinction between analytical and traditional philosophy doesn’t afford to be a distinction of subject matter. Indeed, if we are sufficiently imaginative, all speculative metaphysics can be translated into a linguistic-conceptually accented way of speech, namely, put in a way that legitimates the claim that the speculative philosopher is doing philosophical analysis in the same extended way as the conceptual analyst. To take a radical example, consider the concept of pure subject in Fichte’s transcendental metaphysics. The pure subject is something only intellectually accessible, that posits (setzt) the outer world in order to posit itself (by Selbstsetzung) simultaneously as a necessary opposition to it. Now, I would not be surprised if a contemporary supporter of Fichte’s ideas decided to translate such claims into an analysis of the concept of an “elusive I” as constituting and being constituted by the social reality under antirealist assumptions. Even if such neo-Fichtean antirealism suffered from the same weak intelligibility and low plausibility of the original model, it would be no less defensible than some social-constructivist contemporary ideas(26). Though this kind of move could easily be made by the syntactically oriented conceptual analyst, it would give some trouble to a pragmatically oriented one, since it seems to clash clearly against ordinary-language and commonsensical intuitions. Nonetheless, even here such a move is possible: the pragmatically oriented analyst can maintain that the disagreement with our intuitions is only apparent, and try to show that there is a way of harmonizing what he is saying with the background of our ordinary beliefs (Berkeley anticipated such a strategy when he claimed that his immaterialism was merely reflecting the real commonsensical expectations of plain men).
     To summarize: Because the work of pragmatically oriented analytic philosophers includes hypothetic-synthetic moments in which new concepts are thought to be discovered, this work is able to contain (even if only indirectly), unexpected metaphysical speculations, which may also have an empirical import. The syntactically oriented conceptual analyst can make such speculations with lighter conscience, since he is able to sacrifice agreement with our intuitive expectations without losing the intelligibility of his arguments, since for him this intelligibility is heavily sustained by their formal coherence. But even the pragmatically oriented analyst can make metaphysical speculations by claiming that the new concepts are required to fit with our common views of the world in its more adequate ways. It seems also that all domains of traditional metaphysics can be in some way reached by philosophy as linguistic-conceptual analysis. Hence, to maintain that there is really a distinction in subject matter between philosophy as conceptual analysis and traditional philosophy, even in its more speculative forms, is to “hypostasize” the role of the linguistic-conceptual accent.
     A similar argument applies to thesis (b), which claims that philosophy differs from the empirical sciences by restricting itself to a conceptual investigation.
     That this thesis is a false claim should already be clear, since our last example of conceptual analysis concerned the physical world too, even if in a non-straightforward way. But the point can be posed in a more dramatic way. Suppose, first, that there is an entirely consistent conceptual analyst, who, assuming the broad view of analysis we have described, believes that concepts are the distinctive subject matter of philosophy. Then how would he consider science? It would not be difficult for him to perceive that Einstein was in fact analyzing the concept of simultaneity, when applied by observers moving at great relative speeds, for it is certain that he was not analyzing the actual empirical objects moving in space or even the actual experience of them. As for the work of the famous contemporary cosmologist Stephen Hawking, our conceptual analyst would easily see that Hawking was not involved in any division of black holes in themselves, but that he made an important astrophysical analysis of what must be meant by the concept of black hole, if we want to have a coherent understanding of the phenomena. The concept of natural evolution, he would see, was first analyzed in the right form by Charles Darwin, as a consequence of reflections based on his zoological and botanical observations; Gregor J. Mendel analyzed his concept of gene and Watson and Cricke their concept of DNA; Carl Jung has found and analyzed the concept of the collective unconscious, and T. B. Veblen, that of the leisure class… Were all those people doing philosophy? Accepting, as our conceptual analyst does, that our conceptual world is the object of philosophy, he could not avoid a positive answer: they were all doing philosophy as analysis of language, though without a linguistic-conceptual accent. Indeed, all theoretical work of thought would become, in one way or another, a work of conceptual analysis, and thereby a work of philosophy.
     Now, the opposite situation can also be imagined: Suppose that there is a hard-minded empiricist, who decides to start with the premise that empirical scientific knowledge is not essentially conceptual, even if it is only accessible conceptually, since these concepts deal with empirical facts, their applicability depending ultimately on them. How would he consider the majority of questions asked in philosophy? Since the speech acts theory is about human communicative actions in a real context, since the verificationist analysis of the factual cognitive sense of our statements concerns the ways our minds can achieve knowledge of the world, since realism about scientific laws is a thesis about the possible constitution of reality, he would be led to see much of philosophy as dealing with empirical phenomena, even if here they are pervasive and all-embracing.
     The case of the consistent conceptual analyst shows that an inquiry which is not about concepts, like science, can always be interpreted in a way that makes it about conceptual content; the case of the hard-minded empiricist shows that an inquiry usually seen as being about concepts, like that of analytical philosophy, can generally be interpreted in a way that makes it an inquiry that goes beyond concepts. What conclusions might be drawn from this? The first conclusion is that, the objects of analytical philosophy don’t need to differ essentially from those of traditional speculative philosophy or from those of science, since analytical philosophy can’t claim to differ from these endeavors because it deals with our conceptual structures. Moreover, the cases show that the claim that the proper subject matter of philosophy is the structure of our concepts, if taken seriously, ends up by blurring the real distinction between analytical philosophy and any other theoretical undertaking. A further conclusion is that, even the method of conceptual analysis can’t be seen as the exclusive proper method of philosophy, since, if the analytical philosopher works with concepts in the liberal way described above, then scientists are also allowed to do the same if they consider it appropriate.

What is then the real difference between philosophy as conceptual analysis on one side, and speculative philosophy and science on the other, if it is not a difference in their proper subject matter? The obvious answer seems to be that it is a contingent methodological difference, a difference in the ways the subject matter is questioned. Analytical philosophers submit their inquiries to a much stronger methodological control, by presenting their views in a semantic metalanguage, by submitting them to new logical and linguistic devices, and by setting them against the context of our scientifically informed contemporary world-view. This being so, we are led to conclude that analytical philosophy is only the name for a more refined way of doing philosophy developed in the twentieth century, which requires the accent of the linguistic-conceptual medium mainly for reasons of methodological rigor. Since philosophy is a kind of heuristic game with argumentative moves, which are based on linguistic symbols, it is easy to understand why the use of analytical tools became fashionable in contemporary philosophy.
     A revealing parallel can be traced between the historic assimilation of the “propaedeutic” doctrines of the Aristotelian Organon and the assimilation of the analytical procedures in the current philosophy. Aristotle considered the new logical and methodological doctrines contained in his Organon a necessary instrument for the adequate exercise of philosophical (and scientific) reasoning. His Organon contained a theory of proposition and their constituents, a theory of deductive reasoning (syllogistic), remarks about the nature of definition, a sketch of a theory of inductive reasoning and of scientific explanation, and a classification of fallacies and their solutions. Indeed, the assimilation of the doctrines contained in the Organon changed, slowly but definitively, our ways of doing philosophy: the Aristotelian tools of inquiry were assimilated and improved during the Middle Ages, mostly under the name of dialectics, setting up new and irreversible procedural standards in philosophical reasoning. Now the so-called analytical philosophy can be explained as the outcome of a similar methodological revolution. Since the end of the eighteenth century, extremely important developments in the same domains covered by the Aristotelian Organon have arisen. Some concern the nature and structure of propositions (as in the case of Frege’s semantics), others concern the deductive logic (the predicative logics of the first and second order, modal logic, epistemic logic, etc.), the inductive reasoning (theories of probability and decision, etc.), the pragmatics (the theories of verification, contextualist views of meaning, speech acts theory, etc.), and the domain of the philosophy of science (like the theories of explanation and confirmation). It would be a wonder if philosophy were not definitively changed by such developments, able to set up better standards of clarity and rigor, increasing impressively its heuristic potential. The assimilation of all these new procedures has allowed and will allow us to see more and more things in more clear and distinct ways, in an upheaval that in the long run will be comparable with the discovery of the telescope in astronomy.
     Recapitulating our main results: The chief reason that analytical philosophy seems to have only concepts and language as its object is its general “propaedeutic” concern with the linguistic-conceptual element, made perceptible mainly through its linguistic-conceptual accent. This fact has misled analytic philosophers, leading them to mistake new procedural devices, which can also be used elsewhere, for the peculiar philosophical method, and leading them further to mistake the object of the application of these devices for the object peculiar to philosophy. The fact that, by stressing language in philosophy, we generally appeal to a semantic metalanguage, which forces us to a rigorous treatment of linguistic-conceptual structures without ignoring either their senses, or their references (as conceived through their senses), therefore, without closing the way to the world, is the main constitutive element of what in a rather misleading way has been called analytical philosophy. Indeed, if ‘conceptual analysis’ is the name of anything, then it is the name of philosophical ways of inquiry that incorporate within themselves a certain linguistic-conceptual accent, along with related methodological and heuristic procedures that became common in the twentieth-century philosophy. In sum: ‘conceptual analysis’ is the name applied to the most salient procedural features of a historically contingent state of the art – of a style, rather than of a thing. Later (VII, 3) we will see that the emergence of analytical philosophy can be much better understood as a contingent historical event, generated by the rise of what might be called the “semiotic sciences”, than as the discovery of the true object of philosophy or of its own, inalienable method.


                                       PHILOSOPHY AS CONJECTURAL
                                         ANTICIPATION OF SCIENCE

In this chapter, we begin the descriptivist search for the criteria used to identify philosophical discourse and thought. My first claim is that, even if we can’t find anything like the essential subject matter of philosophy, or anything methodologically relevant that is exclusive to it, we are still able to find something very peculiar to philosophy, if we pay attention to its form.

Even if the descriptivist metaphilosopher can’t find any proper identifying feature of philosophy in the material aspects of its investigations, he can always find a striking formal feature common to all philosophical inquiry, namely, its conjectural character. Philosophy is essentially a conjectural or speculative undertaking, in the sense that philosophers are not able to produce sufficient consensual agreement concerning their ideas, doctrines, and even their most general views. There is no philosophy whose results are taken for certain or which is indisputable. The reason for this is not difficult to find. To achieve consensual agreement about the results of our intellectual endeavors, we need to share some basic assumptions. But philosophy lacks such shared assumptions in almost every step of its inquiry. Particularly important in this respect is the absence of common assumptions capable of producing consensual agreement about the right problems and their formulations, and about what counts as truth-evaluative procedures, namely, the verificational and/or falsificational argumentative procedures, able to prove or disprove, to confirm or disconfirm, a proposed solution. Without such assumptions, it seems impossible to expect anything like agreement about results.
     To give an example, consider once again Plato’s doctrine of ideas. This doctrine was suggested as a solution to the problem of generality or predication, and constructed under the assumption that for something to be an object of knowledge, it must be unchangeable. Now, since the visible world is always subject to change, the only possible objects of knowledge must be what Plato called ideas or forms, objects existing timelessly in a purely intelligible world. This doctrine allows Plato to explain generality and predication: We can predicate beauty of many visible things because they exemplify the abstract idea of beautifulness. However, this kind of explanation leads to difficulties. One is this: How can an idea preserve its unity when it is shared by the many individuals to which it applies? To solve this problem, Plato appeals to the concepts of participation and copy, used in a strange metaphorical way: Many things can participate in one and the same idea, but without dividing the idea into parts; being a copy of an idea, a thing is similar to the idea, although the idea can’t be similar to the thing…
     What do the critics of Plato’s doctrine have to say about this? First, they are free to reject the general assumption that knowledge must be of something unchangeable, casting doubt on the necessity of Plato’s recourse to a non-empirical object of knowledge. Further, they are also free to consider Plato’s new concept of participation and copy as ultimately incoherent, since they are metaphors that resist any further explanation, what casts doubt on the intelligibility of Plato’s own concept of a timeless idea or form. Are these objections justified? Yes, I suspect. But to be sure, we don’t know for certain. Indeed, the doubt is expected, since philosophy is the creation of uncertain theories, founded on uncertain grounds. This is a quite depressing fallibilist conclusion which past philosophers have tried to deny, but which contemporary philosophers have long since learned to accept as unavoidable. Indeed, there is no exception. Even the supposedly purely descriptive therapeutic philosophy attempted by the late Wittgenstein has shown itself incapable of producing agreement: Where Wittgenstein saw a remedy, others could see only a placebo or even a poison.
     This impossibility of agreement provides also the most striking contrast between philosophy and science: Unlike in philosophy, prior sufficient agreement about general assumptions (which allows for the existence of common problems) and procedures of truth-evaluation (which allows for the possibility of common solutions) are present in all sciences, empirical and formal; and this is what enables further agreement about experimental results in physics, as much as about the proof of theorems in mathematics. Because scientists, unlike philosophers, have found such common assumptions, they can arrive at agreement about the results of their investigations and hope for progressive development.
     Paying attention to the conjectural nature of the philosophical endeavor helps us to explain another of its features: The fact that philosophical discourse and thought is characteristically argumentative and aporetic, with few (and often questionable) exceptions. Philosophers are always posing or arriving at some uncertain principles and trying to validate them by showing how much follows from them. This procedure is related to the conjectural character of the philosophical inquiry, since by the very fact of working with conjectures, philosophers proceed by a constant critical comparison of the argumentative consequences of the assumptions they believe to be correct, and by a critical comparison of the quality of the arguments used to arrive to these consequences, in an apparently unlimited task. The conjectural character of philosophy generates its argumentative, dialogical and aporetic praxis.
     Could philosophy be defined in terms of its conjectural or speculative character alone? Not without qualifications, since not all conjectures are philosophical. We can, for example, make conjectures about the climatic conditions on earth in the next fifty years, but this does not amount to a philosophical quest. The reason why this question is not philosophical seems to be that it lacks a theoretical point: It is only a conjecture about empirical events. On the other hand, Chomsky’s theoretical conjecture about the existence of universal grammar innate to all men can’t be easily proven, being in a certain sense philosophical. But it is philosophical in a less relevant way, since it is a specific problem for which linguists hope soon to find the path to an empirical answer. Could we then define as ‘philosophical’ all argumentative conjectural endeavors with a theoretical point and a broad concern? This seems sound, though still brief and uninformative.

A deeper answer to the question, “Why is philosophy a conjectural form of inquiry?” can be given when we accept the thesis that philosophy is a protoscience, namely, a conjectural undertaking anticipating science, and that the enduring wit of philosophical theories come from the scientific truths anticipated in them.
    That at least part of philosophy is (or was) an anticipation of science is not a speculative thesis at all, but a statement of fact. At the time of the Greeks, when all fundamental empirical sciences were still to be formed, the word ‘philosophy’ was indistinctively applied to the whole domain of human inquiry. Only much later, with the emergence of those sciences, the application of the word ‘philosophy’ has been gradually restricted. By ceding parts of its domain to science, philosophy has been, as Antony Kenny wrote, the womb from which the particular sciences have been born(27). The concept of philosophy as a protoscience was impressively stated in a well-known metaphor by John L. Austin: “Philosophy is the initial central sun, seminal and tumultuous, that from time to time throws off some portion of itself to take station as a science, a planet, cool and well regulated, progressing steadily towards a distant final state. This happened long ago at the birth of mathematics, and again at the birth of physics: Only in the last century we have witnessed the same process once again, slow and at the time almost imperceptible, in the birth of the science of mathematical logic, through the joint labors of philosophers and mathematicians”.(28)
     Now, insofar as philosophy is conceived of as a speculative inquiry dealing with a material of thought potentially yielding its place to science, we have a deeper reason to explain its conjectural, argumentative, aporetic nature. If philosophy is what can be done before science is possible, then we can expect that the most diverse hypotheses can be suggested and the most diverse lines of argument can be developed in the attempt to justify them, and that, since there is no decisive way of truth-evaluation at hand, the dispute over the right hypothesis and the best argument is likely to continue indefinitely. This state of affairs only ends when the path of the scientific inquiry is found, namely, when researchers finally achieve conditions to find sufficient agreement about the basic assumptions underlying a certain field of investigation, which sets limits on the allowed questions and to the kinds of answers that can be given to them. When this agreement is achieved, the researchers no longer call they field of research ‘philosophy’: They simply redefine it as a field of science.

Before we discuss in more detail the derivation of science from philosophy, it is advisable to say something about the classification and emergence of the most important sciences.
     Sciences are obviously of two kinds: formal and empirical. These two kinds of science have been always in some measure interdependent in their development. The fundamental formal sciences are logic and mathematics. Mathematics had its origins in antiquity. Elementary arithmetic and geometry had been detached from philosophy already under the Greeks, when these subjects were, for the first time, considered separate from the practical problems they were called on to solve. Logic also began early, with Aristotle’s syllogistic, although this was a very limited form of logic.
     Could we speak of philosophical protomathematics or protologic? Certainly. Parmenides’ poem, for example, contains a sketched affirmation of the logical law of non-contradition, and Plato had a rudimentary theory of predication. Moreover, the Pythagorean philosophers were so impressed with the first achievements of abstract mathematics that they believed that numbers were the principles sustaining all reality, confusing the formal with the empirical. Nevertheless, the real questions about the nature of numbers were in their time still hidden in the most complete darkness.
     Turning to the empirical sciences, I will adopt here a corrected and actualized version of the old classification of sciences proposed by Auguste Comte, since it seems to be still the most reasonable, apart from being able to give us a rationale for the historical order of their appearance. Its principle of classification maintains that the fundamental sciences are related to one another in an order that ranges from (a) greater to smaller generality in their scope, which corresponds to an inverse order ranging from (b) smaller to greater complexity relative to the phenomena investigated by them. Applying the principle, we can distinguish five fundamental sciences: physics, chemistry, biology, psychology, and social science(29). The following schema summarizes this classification:

   PARTICULARITY                                                  COMPLEXITY
                                5. social science       human                  
                                4. psychology           sciences
       (a)                                                                                     (b)
                                3. biology                 natural
                                2. chemistry              sciences              
                                1. physics
   GENERALITY                                                         SIMPLICITY

     Physics is the first fundamental science, having in its scope all empirical things without exception. Chemistry has a more restricted scope, applying to empirical things formed by combinations of atomic elements; biology applies only to chemical compounds constituting living organisms; psychology applies only to living organisms possessing consciousness; and social science restricts its scope to conscious organisms socially structured. The progressive loss of generality corresponds to a gain in the complexity of the investigated phenomena, since complexity is inconceivable for the domain of the most general.
    The relations of generality and complexity also help us to explain the order of the cognitive apprehension of the fundamental sciences by us and, related to this, the order of their historical development. Indeed, to learn physics we don’t need to learn any chemistry, but chemistry usually presupposes some understanding of physics; the learning of psychology presupposes some understanding of biology, though not the opposite, etc. The development of the more specific and complex fundamental sciences in a certain way depends on the development of the more general and simple ones. This dependence also involves the developments in the applications of the more general sciences: How could, for example, biology be developed without the microscope, whose construction depends on physical knowledge? This explains why, after the Renaissance, the first empirical science to emerge was physics. Though there were rudiments of physics even in antiquity (ex: Archimedes’ discovery and measurement of specific density), only after Galileo was experimental physics able to emerge as a whole corpus of knowledge. After physics, the other fundamental sciences, chemistry, biology, psychology, and social science, have subsequently detached themselves from philosophy – the last two, it seems, until now only partially, in an echeloned, gradual, and convoluted process.
     Furthermore, these dependencies help us to explain why the establishment of psychology and social science has been much slower, laborious and echeloned. We find a clear epistemological rupture(30) in the birth of physics as a corpus of scientific knowledge with Galileo and Newton in the seventeenth and eighteenth century; in the birth of chemistry with Lavoisier, Cavendish and others at the end of the eighteenth century; and even in the more scattered organization of biology as a corpus of scientific knowledge during the nineteenth century, by scientists like Pasteur, Claude Bernard, Mendel and Darwin. Ruptures occurred in these sciences when appropriate experimental methods were found, that provided the possibility of a consensual agreement about the predictive and explanative power of their theories. However, we don’t find the same clear rupture in the more complex domains of psychology and social science. A reason for this might be the irreducible element of internal (or introspective) evidence, which always plays a role in the human sciences. This element of internal evidence is not open to interpersonal observation, and for this reason, it is not so easy to be objectively considered – though in no way is it helplessly subjective, as some philosophers suppose(31). But another reason for the more gradual constitution of the human sciences might be the fact that in the domains of higher complexity and diversity of studied phenomena, the evaluative procedures require a massive background knowledge provided by the more fundamental sciences: The human sciences require their maturity and depend also on their technical application (one might wonder, for example, how much more scientific psychology will look after an adequate causal explanation of its phenomena through a fully developed biological neuroscience).
     There is a reason why the sciences we have considered deserve to be called ‘fundamental’. The remaining empirical sciences we already have either are particularized subdivisions of the fundamental sciences – as linguistics and economics as parts of social science – or they combine the results of fundamental sciences, applying them locally to certain specific kinds of objects or regions in space or time. Examples of the last type are history, which applies psychological and sociological insights to the understanding of historical change; geology, which applies physics and chemistry to the study of the earth; and neurophysiology, which applies biochemistry and biophysics to the investigation of the brain’s work.
     Finally, it might be noted that the emergence of fundamental sciences has always replaced philosophical speculation: The emergence of physics as an experimental science put an end to the reign of the speculative Aristotelian physics; a similar fate befell the doctrine of the four elements after chemistry appeared, likewise befell the vitalism after biology developed further.
     In this and the following chapters, I will assume this modified Comtian classification of the fundamental sciences, since it remains the most intuitive and indisputable, at least as long as we don’t subordinate it to further issues about theoretical reduction or to a metaphysical problem concerning the unity of the sciences.

In this section, we will consider three examples of philosophical ideas anticipating scientific ones in the fields of physics, biology, and psychology. These examples can be misleading since, as will be seen later, they concern only with one kind of philosophical anticipation. And they can also be misleading by implying that the present philosophical inquiries should be related to future science in the same way as past philosophy has been related to our fundamental empirical sciences. Nevertheless, it is instructive to consider them.
     The first example is the idea, sustained by Anaximander in the sixth century B.C., that the earth is held up by nothing, being stationarily suspended because it is equally distant from all things and it is impossible for it to move simultaneously in opposite directions(32). K. R. Popper holds truly enough that this is one of the boldest and most portentous ideas in the whole history of human thought, making possible the theories of Aristarchus, Copernicus, and even others, because “to envisage the earth as freely posed in mid-space, and to say ‘that it remains motionless because of its equidistance or equilibrium’ is to anticipate to some extent even Newton’s idea of immaterial and invisible gravitational forces.”(33) Though anticipative of science, Anaximander’s hypothesis couldn’t be viewed as a scientific one, since by the time it was formulated, no procedure of truth-evaluation that could lead to general agreement was conceivable. By contrast, Copernicus’ and Newton’s ideas were able to be evaluated and to attain consensual agreement about their truth or falsity, since this condition of scientificity was already satisfied by them by the time of their formulation.
     The second example concerns the first hypothesis about biological evolution, also suggested by Anaximander; he held that life began in the water, that living creatures can be spontaneously generated from moisture, and that human beings evolved from lower species (maybe fish), since in early ages they would have died if they were as defenseless as they are today after birth.(34) Though Anaximander’s ideas were strictly speaking erroneous, since he believed in spontaneous generation and that men had been gestated in fishes, emerging fully-formed from them, instead of evolving gradually(35), there was already a hint of evolutionism in them, pointing to ways of thought that could only be adequately settled within a scientific framework more than two thousand years later, when an adequate background of scientific knowledge made possible the pursuit of such inquires in non-speculative ways.
     Does this mean that the ideas expressed by the sentences “The earth is held up by nothing” and “Man has evolved from inferior forms of life” were once philosophical and have now turned out to be scientific? In a certain sense, the answer is affirmative. These ideas turned out to be considered scientific for us. Nevertheless, this does not mean that they were not to be philosophical for men in other times: precisely because we are considering them as ideas of past thinkers, they must be viewed in their contexts, where they must be seen as philosophical ideas, since in the context in which they arose they could be only speculatively addressed. Under our perspective, the concept of being philosophical only makes full sense when related to the historical context in which the ideas are considered.
     The last example, related to psychology – a field of investigation that has not yet been completely developed as a science – concerns Plato’s doctrine of the tripartite soul (Republic, IV, 446 A f.). According to this doctrine, the most primitive part of the soul is formed by its bodily appetites, desires, and needs; the next part is the spirited element, formed by emotional drives such as courage, anger, ambition, pride, protectiveness, honor, loyalty, etc.; the third part of the soul is formed by reason, acting as an inhibiting principle that commands the others. In the dialogue Phaedrus (246 f.) Plato compares reason with a charioteer driving a pair of winged horses, one good, representing the spirited element, who strives upward towards the realm of ideas, and the other bad, representing the bodily appetites, endeavoring to pull the team into the realm of the earthly and causing much trouble for the driver.
     Now Plato’s doctrine of the tripartition of the soul can be seen as a forerunner of Freud’s structural theory of mind(36). Freud also divides the mind into three central institutions: the id (Es), which is unconscious and represent our instincts, the super-ego (über-Ich), mostly unconscious, representing introjected ideals and moral constraints, and the ego (Ich), mostly conscious, being immediately linked to conscious will, perception and motor control. The two theories correspond only roughly: Freud’s id corresponds to the bodily appetites, but includes also volitional elements attributed by Plato to the spirited part of the soul, like anger; the super-ego corresponds approximately to the spirited element, the good horse in Plato’s analogy; and the ego might correspond to the rational element, the charioteer, trying to satisfy the opposite demands of the id and the super-ego.
     When we confront these two theories, we find a kind of difficulty similar to that found when we try to compare two philosophical theories. Indeed, psychoanalysis satisfies too bad our conditions of scientificity, since there is not enough agreement about psychoanalytic ideas, even by its practitioners. Nevertheless, while Plato’s suggestion was based only on his personal knowledge of himself and human behavior, Freud’s theory is also based on the observation of repeating patterns of behavior in the psychotherapeutic context. Moreover, Freud’s theory introduces a new element, absent in Plato’s schema: the unconscious; and Freud’s theory is much less metaphoric, much more articulated, detailed and precise: it endeavors to tell us more; and although Freud’s theory in its fine details is apt to be even more uncertain than Plato’s, it suggests a psychologically more workable general framework.
     We conclude this section with a terminological remark about the concept of anticipation of science. The given examples might be considered good anticipations: Anaximander’s ideas about the earth and biological evolution show in a very rough way the direction to be followed by science, and Plato’s theory of the tripartite soul anticipates the form of Freud’s later theory, which apparently approaches science. However, many philosophical undertakings are bad anticipations in the sense that they show us the wrong direction. A famous case of this is the hypothesis of the phlogiston, suggesting the existence of an element liberated by flames and responsible for it, which was completely wrong and retarded the advancement of chemistry for nearly a century. The most notorious example of bad anticipation is the aprioristic Aristotelian physics, which, accepted by the church as a matter of dogma, retarded the development of physics during the entire Middle Ages, until the experimental discoveries of Galileo made it unsustainable. Finally, the concepts of good and bad anticipation are relative to the extent of deviation from truth that we are willing to tolerate, which varies according to the context: Anaximander’s evolutionism, for example, might be seen as a bad anticipation in a context in which we wish to exclude non-Darwinian accounts of evolution as fundamentally misleading.

Antony Kenny, considering the way that philosophical thought gives way to science, has noted that this occurs by a process of parturition, which he calls fission(37). He makes this process clear by means of an example concerning one of the central problems of philosophy in the seventeenth century: The question about innate ideas. Initially, the problem was this: Which of our ideas are innate and which are acquired? After Kant, this confused question broke up into two others: The question of the roles of inheritance and environment in the constitution of our ideas, and the question of how much of our knowledge is a priori and how much is a posteriori. The first question, says Kenny, was earlier on handed over to psychology, while the second, addressing the justification of our knowledge, has remained philosophical. Some time later, the remaining philosophical question about the a priori split again into non-philosophical and philosophical questions, branching into a number of questions, one of them being: Which propositions are analytic and which are synthetic? For Kenny, the notion of analyticity found a precise formulation, through the works of Frege and Russell, in terms of mathematical logic, and the question “Is arithmetic analytic?” found a precise mathematical answer through Kurt Gödel’s incompleteness theorem; nevertheless, residual problems concerning the nature and justification of mathematical truth were left behind as questions of philosophical dispute. The following schema sums up  Kenny’s version of the process:

                                 philosophical problem
                                      of innate ideas

psychological questions about        philosophical problem of how
the role of heredity  and  envi-       much of our  knowledge  is  a
ronment in the constitution of        priori
our ideas
                         logical-mathematical           remaining  philosophical
                         questions  about  the           problems about the natu-
                         definition and extent           re  and  extension  of   a
                         of apriority in mathe-          priori knowledge  in  ge-
                         matics                                 neral

     The suggested pattern of development is that of a broad, confused philosophical problem that splits itself in two parts; one part of it condenses itself into scientific questions, capable of achieving consensual answers, while the other part remains philosophical. And this same process tends to be repeated again and again within the remaining philosophical questions, perhaps until their final disappearance.
     But the most interesting point is that, the loss of part of philosophy to science produces changes that might affect the whole organization of the remaining field of philosophical inquiry. As the example shows, after the fission, the part of the problem that remains philosophical must be reformulated, generating new answers. But the changes don’t remain circumscribed. All related problems belonging to the remaining philosophical field must be accommodated to the new state of affairs, together with their speculative answers; this is done by means of a more or less extensive reformulation of the problems and their answers, and by a relocation of their places relative to other problems and answers. To give an example: Kant’s reformulation of the remaining philosophical problem of innate ideas in terms of his doctrine of a priori knowledge and concepts, led to a further reformulation of the questions about the concepts of world, soul, and God. Kant no longer saw these three concepts as actually referring to their objects, but as ideas of reason: “As if” a priori concepts, generated by the nature of reason, whose function is only that of orienting our processes of inference; so we must proceed intellectually as if there were an external world that is a closed causal totality, in order to seek further causal chains; we must proceed as if there were a simple permanent subject (the soul), in order to seek a unified understanding of our own psychic phenomena; and we must proceed as if there were an intelligent creator (God) of all of nature as an intelligible system in order to investigate nature. As a consequence of this reformulation of the concepts of nature, soul, and God as directive a priori concepts, a relocation of their places within the whole conceptual framework follows; the concept of God, for example, can’t be viewed as able to have the same function as that of the concept of the veracious God still maintained in the “pre-critical” philosophy of Descartes.

As a result of the described process, philosophy has contracted itself into a resistant set of philosophical issues. These issues certainly include the philosophies of the fundamental sciences, which take the already existing particular sciences as their objects. Since these philosophies are dependent on the development of these sciences, they must develop later. Consequently, it is not unreasonable to expect that these philosophies will someday find consensual agreement as metasciences (sciences of sciences).
     However, the resistant core of the present set of philosophical issues essentially consists of the more central traditional disciplines of philosophy, like epistemology, metaphysics, philosophy of content, philosophy of mind, and ethics. These central domains have until now resisted any conversion into science, and it is important to realize how peculiar they are. They are not on the same level as the fundamental sciences or even of the philosophies of these sciences. Indeed, what strikes us in disciplines like metaphysics and epistemology is their extraordinary comprehensiveness. In the case of metaphysics, ultimate problems such as those of the universals, substance, the nature of causality, space, and time, concern the world in the most general way, involving as much objective as subjective objects of experience and crossing the subject matters of all the fundamental sciences. In the case of epistemology, the questions are not less comprehensive, since they are not concerned with this or that form of knowledge, like the epistemologies of sciences, but with knowledge in general, crossing all these local epistemological endeavors. Our question about what is the general nature of philosophy could be here replaced by another, not less important question: What is the proper nature of the core disciplines of philosophy?
     Now, the most serious question concerning the idea of philosophy as an anticipation of science is not about the fact of the derivation of science from philosophy, but about the extension of this derivation. Does the remaining set of philosophical issues, or some part of it, belong essentially to philosophy, in a way that resists any transformation into science? Did the domains of philosophy already handed over to science indeed ever belong essentially to philosophy, or were they only mistakenly confused with it?
     Philosophers are divided about this. Some, like Keith Lehrer, have suggested the progressive hypothesis that philosophy is “only a collective name for the pot of resistant problems still untouched by science”(38); and the fact that some philosophical questions have waited two thousand years for a consensual answer does not mean that no answer will ever be found. Nevertheless, most philosophers remain more reserved. Antony Kenny, for example, holds in his book on Aquinas’ philosophy of mind, the more conservative hypothesis that even if philosophy has in its past handed over parts of itself to science, the central remaining philosophical problems are the only genuinely philosophical; at least those problems comprehending epistemology, metaphysics, ethics, and theory of meaning will remain forever philosophical(39).
     Trying to justify this claim, Kenny, influenced by Wittgenstein’s idea of perspicuous representation (see note 19), suggests that philosophy, unlike the particular sciences, deals with our whole knowledge, since it aims to organize what we already know, in order to provide us an overview of our own knowledge, rather than to acquire new truths. This aim gives philosophy a kind of comprehensiveness that can’t be found in any particular science. This comprehensiveness, argues Kenny, is the reason the philosophy of mind of Aquinas remains in many ways relevant: “Philosophy is so all-embracing in its subject matter, so wide in its field of operation, that the achievement of a systematic philosophical overview of human knowledge is something so difficult that only genius can do it. So vast is philosophy that only a wholly exceptional mind can see the consequences of even the simplest philosophical argument or conclusion.”(40)
     In what follows, I will argument for the first and more progressive hypothesis, though not in the way the reader might suppose.

My argument that all philosophical questions will eventually be able to be absorbed into science is not of a constructive kind; I will not try to prove it, nor do I believe such proof is possible. But I intend to show that the progressive hypothesis can be made plausible, insofar as we can remove the main reasons that make the philosophers reject it.
     There seems to be two deep reasons why many philosophers have come to reject the idea that the whole of philosophy is anticipatory of science(41). The first is that, when they think about science, they have in mind, if not the formal sciences, at least the already well-established experimental sciences of nature. Considering not only the limitation of the scope of most of those sciences, but also their more direct empirical character, to accept the progressive thesis about the nature of philosophy seems to commit ourselves to an impoverishing and reductive view of the remaining core of philosophical problems, a view that seems to take away all of philosophy’s comprehensiveness and relevance, by putting its problems on the same level as those of particular sciences. To agree with the progressive hypothesis seems then to leave us with nothing, except some kind of pedestrian philosophical scientism, which is intrinsically narrow and inimical to the comprehensiveness and abstraction that belong to genuine philosophizing.
     The other reason to dismiss the progressive hypothesis is the implicit adoption of views of science conveyed by our contemporary philosophy of science. Philosophers of science are only able to build interesting and detailed theories of science insofar as they take developed science as a model. But since not all scientific fields have yet arisen or are well developed, philosophers of science usually take the natural sciences – physics in particular – as the basic models of science, since these are the most advanced available forms of science. This procedure can lead to fruitful results concerning the philosophy of these well-established sciences. However, when the results are interpreted as building a general criterion of demarcation between science and non-science, valid for all future candidates for the role of “science”, the consequences are liable to show that this criterion yields a narrow and obstructive view of the limits of science. This is true even for a natural science distinct from the science used as the model, which can be shown by Karl Popper’s criterion of scientificity as the falseability of our theories through decisive experiments. His criterion seems to apply well enough to theories of physical science, like the theory of relativity, which Popper often uses as an example. But his criterion leads to the rejection of the scientific character of many psychological and socio-historical theories, even including the theory of evolution – a biological theory whose scientificity none would dare to deny. What kind of experiment could enable us to falsify a theory that explains a process extending over a period of millions of years in the past? And even if the theory can be in some indirect way tested, failing to pass the test would certainly not be seen as a decisive falsification(42). For reasons like this, I think that Popper was right when he claimed that his methodology is not descriptive of what people (including the scientists) think of as belonging to science, but rather a proposal: A rationally argued suggestion as to what kind of investigation should be included in science(43). Now, when a philosopher conceives of science having such models in mind, he will probably not admit that philosophy is anticipation of science, for it will be clear to him that the central areas of contemporary philosophical investigation, by their own nature, could never accommodate the exigencies of those views of science.
      Nonetheless, I think that the two reasons given above to mistrust the progressive hypothesis when clearly false: when we hold that philosophy is anticipatory of science, we don’t need to limit the word ‘science’ to the already existing particular sciences; and we are by no means forced to accept what philosophers of science have told us science is. Indeed, what naturally comes to mind, when we contrast philosophy with science, is the contrast between a conjectural endeavor, in which there is no possibility of agreement about results, and a non-speculative undertaking, in which agreement about what is true is likely to be found, along with the progress resulting from this agreement. It seems further that the idea of science as a non-conjectural truth-evaluative undertaking matches quite well what we – scientists, educated people, with the eventual exception of philosophers of science – naturally mean by the word ‘science’. Indeed, to judge if a theory belongs to science, we don’t immediately ask, for example, if it can be submitted to an empirical test. What we necessarily ask is whether an interpersonal agreement about its truth or falsity in the scientific community can be achieved, what is not beforehand attainable only by means of empirical test. The possibility of an adequate consensual agreement seems to be the most general and decisive requirement, diversely from the many varied ways through which this agreement might become possible. Though never adequately explored, this idea has not been completely ignored in the philosophical inquiry about the nature of science: John Ziman, at least, has made a case for it, maintaining that the unifying principle of science in all its aspects lies “in the recognition that scientific knowledge must be public and consensible.”(44)
     In what follows, I will approach this general view of science in a way that differs from most philosophers of science. I’m not making a proposal. My whole approach is descriptivist: I will try to rescue the most general scholarly meaning of the world ‘science’ by making explicit the criteria scientists and educated men usually use to identify science. This is an endeavor parallel to the descriptive metaphilosophical procedure; in fact, if the descriptivist procedure brings us to the view of philosophy as a protoscience, it seems clear that the “science” to which philosophy is “proto” must also be dealt with within a descriptivist account. Indeed, a descriptivist account of science seems to be the only coherent way of imagining the contrast between philosophy and science under the descriptivist metaphilosophical approach. Only after we have explored this idea of science in more detail will we be able to judge if the concept of philosophy as anticipatory of science is really restrictive.

My aim is not to develop a complete descriptivist characterization of science based on the analysis of the criteria of demarcation actually used by scientists. It is rather to make sufficiently explicit – for the purpose of contrasting science and philosophy – what might be called a consensualist view of the nature of science: The intuitive view that the unifying principle of science is that it is a truth-evaluative inquiry enabling progress through consensual agreement among the members of the community of ideas. To achieve an analytical grasp of this idea and its ramifications, three conditions of scientificity will be identified; so general are these conditions that they can be understood as applicable to all sciences, the empirical as much as the formal ones.
    The first condition is that in its period of formation, a science must behave as a progressive undertaking, in the sense that the theories once suggested in a scientific field are capable of being refined or replaced by new ones, with greater explanatory power. Moreover, this condition says that in the process of its constitution, a science must be knowledge-cumulative, in the sense that it allows the community of ideas to recognize the truth of an increasing number of propositions. This condition of progressiveness can be stated as

     S1: The scientific inquiry must be a potentially progressive
           and knowledge-cumulative undertaking.

     Condition S1 applies primarily to a whole science, formal or empirical, understood as composed of a corpus of interrelated scientific theories, and it only derivatively applies to the acceptance of the scientificity of the theories within this corpus.
     The satisfaction of condition S1 presupposes the satisfaction of condition S2. S2 is the prevailing condition, applicable primarily to the theories (hypotheses and systems of hypotheses) aspiring to scientificity and only derivatively to a scientific corpus. This is the central condition of consensuality, which can be stated as

     S2:  A proposed theory is scientific when consensual agreement
           about its truth or falsity can be rationally achieved by a criti-
           cal community  of  ideas.

     Necessary for the adequate understanding of condition S2 is an adequate grasp of the concept of a critical community of ideas, which will tell us who is entitled to evaluate supposedly scientific ideas and how. If there are people who don’t believe that the theory of natural evolution has received sufficient empirical confirmation, we will not conclude that this makes false our belief that there can be scientific agreement about the truth of this theory, since there is such agreement. If an authoritarian government decides to call some spurious ideology a science, imposing agreement on its scientific community (as occurred in the Soviet Union and in the Nazi-Germany), we will not conclude that this ideology is indeed a science. And we also don’t think that a community of ideas that bases its truths on the authority of sacred scriptures or on the visions of prophets is acting like a scientific community.
     To eschew such inadequate outcomes, which would make our characterization of science inevitably defective, I will make use of an idea originally suggested by Jürgen Habermas in his consensual theory of truth, requiring that the decision about what counts as truth must rely on a discussion (Diskurs), under the presupposition of an ideal situation of speech (ideale Sprachsituation)(45). It seems clear that a community of ideas, to be able to evaluate scientific hypotheses, must do it under certain assumptions, like those of its own rationality and freedom – assumptions that depend on something very similar to what Habermas called an ideal situation of speech. Therefore, a critical community of ideas can be characterized as a community which, so much as it is practically possible, satisfies a set of criteria, which can be roughly described as follows:

(a) A critical community of ideas must be composed of equally well-trained and well-informed members (the scientists).
(b) The members of a critical community of ideas must be engaged in submitting their ideas to a critical rational scrutiny.
(c) The members of a critical community of ideas must have full access to information, equal chances of evaluating ideas, and similar rights of intellectual exchange.
(d) The members of a critical community of ideas can’t be subjected to any constraints on the views they hold, except the constraints of the best justification.

     It is important to realize that such criteria form an ideal cluster that can never be completely satisfied by any actual scientific community. However, these criteria must be at least in a satisfactory measure fulfilled, since no community of science could reach any reliability without their minimal realization. And clearly, not only must we all presuppose that such criteria are satisfactorily fulfilled, when we accept a scientific discovery purported to be true, but in addition the scientist working on research must do it under the constant assumption that his results will be accepted by a community of ideas satisfying the conditions such as (a) through (d), using this assumption to guide his own evaluation of what he is doing, even in cases in which this assumption is clearly contrafactual. So understood, condition S2 turns out to be the most relevant requirement for our acceptance of theories as belonging to science.
     Agreement on the truth or falsity of theories within a critical community of ideas requires a third condition of scientificity. As we have discussed, consensual agreement on truth among the members of a community of ideas is only possible if there is a previous agreement about general assumptions and methods of truth-evaluation. So the possibility of the satisfaction of condition S2 presupposes the satisfaction of S3, a second condition that a theory must satisfy to be called scientific. This condition of objectivity can be stated as

S3:  Scientific agreement about the truth or falsity of a  propo-
      sed theory can only be achieved when the critical commu-
      nity of ideas has found consensual agreement about what
      counts as basic assumptions necessary for the  evaluation
      of this theory.

     This agreement requires the satisfaction of several basic assumptions that can be roughly categorized as follows:

(i) Assumptions about what can be counted as the (empirical or formal) elementary data, constituting the epistemic domain to which the theory belongs;
(ii) Assumptions about what can be accepted as adequately formulated questions to be asked in this domain (the theory must answer meaningful questions);
(iii) Assumptions about what can be accepted as an adequately constructed theory in the epistemic domain (in its internal consistency as well as in its external coherence within a wider conceptual framework); and
(iv) Assumptions about what counts as the procedures of truth-evaluation of a theory in its epistemic domain (which involves some kind of agreement with the facts the theory tries to explain).

     That basic assumptions like these must be satisfied for the achievement of consensual agreement is an experiential truth that the critical community of ideas has learnt about its own ways of working.
     So understood, the conditions of progressiveness, consensuality, and objectivity seem to constitute a sufficiently reliable descriptivist criterion for demarcation between science (empirical or formal) and non-science.

The point to be underlined is that our consensualist view of science is in direct contrast to philosophy. In philosophy, as in science, a critical community of ideas must be presupposed, even if sometimes counterfactually. It is expected that philosophers have competence in their activity, that they are disposed (even if somewhat grudgingly) to submit their philosophical theories to free critical scrutiny by other competent thinkers, that they have equal information and possibilities of interaction (a complaint against dogmatic philosophy is that it fails to satisfy this), and that their ideas are not submitted to any ideological constraint (in fact, the main complaint against medieval philosophy is that this condition couldn’t be fully satisfied). Nevertheless, philosophers are not able to build a work that satisfies any of the three conditions of scientificity we have considered. This allows us to characterize philosophy negatively, as the heuristic undertaking of a supposed critical community of ideas, in which those conditions are not satisfied. The negative conditions are, first

       NS1: Philosophy fails to satisfy the condition of progressiveness,
               since it is not a progressive and  knowledge-cumulative en-

     Philosophy can be cumulative only in the sense of a content-cumulative undertaken, namely, in the sense that our philosophical views can increase in number, and can be subdivided and combined, building an always greater amount of possible truths, which makes the speculative net of possibilities formed by such subdivisions and combinations increasingly dense. The content-cumulative but non-knowledge-cumulative character of philosophy can be easily grasped when we compare different philosophical theories about the same thing. Consider, for example, the doctrines of kinds of knowledge in Locke and Spinoza: Each seems to illuminate different aspects of the problem, each seems to have some truth, and both together seem to have more truth than each one of them; the trouble is that we are not in a position to tell with sufficient certainty where the truths are, or even to eschew any skeptical doubt about their existence.
     Condition S1 is not satisfied by philosophy because philosophy don’t satisfy its precondition

      NS2:  Philosophy fails to satisfy the condition  of  consensuality,
                since no agreement about the truth or falsity of its theories
                can be achieved by its critical community of ideas.
     And this occurs because in different ways

     NS3    Philosophy fails to satisfy the condition of objectivity S3,
               since a philosopher is not  able, before the critic commu-
               nity of ideas, to satisfy basic assumptions.

     The philosopher is not able

(i) to achieve general acceptance about what might count as the elementary data constituting his epistemic domains;
(ii) to make other philosophers sure that his questions are not basically misleading (pseudo-problems);
(iii) to achieve general acceptance about the adequateness of his theories (internal and external coherence); and
(iv) to develop procedures of truth-evaluation (arguments) that will be generally accepted by his neighbor philosophers (showing that his theory agrees with the facts it tries to explain).

     The conditions of progressiveness, consensuality, and objectivity are probably not the kind of thing that philosophers of science would find interesting; they are too unexciting. But they seem to match nicely the criteria that we intuitively use when we are called to distinguish what belongs to science from what belongs only to philosophy.

When we consider the whole of philosophy as a science-anticipative undertaking, the adoption of the foregoing view of science leads to some interesting consequences.
     First, considering that our criteria for what might count as science leaves open the concrete ways by which an investigation can come to be considered scientific, the proper identity of this investigation also remains open. In other words, the suggested criteria don’t presume the proper character of a new scientific field; particularly, they don’t presume that the sciences eventually destined to take the place of the present domains of philosophical inquiry will have any similarity to the experimental sciences already known by us. Given the proposed view of science, even comprehensive speculative theories, like Freud’s metapsychoanalysis, Comte’s law of three stages, or even Fichte’s doctrine of the pure I, could in principle become scientific if consensual agreement about their truth or falsity could be achieved in a critical community of ideas, or (more plausibly) if they could be restated in order to make such agreement possible. Even a view of philosophy similar to that which I’m suggesting here could eventually come to be scientific. To see this, suppose that a view of philosophy as protoscience similar to our consensualist view is more adequately and completely developed, and that this view comes to receive in the future more and more confirmation by the emergence, from our philosophical discussions, of new fields of scientific knowledge. As a consequence, the critical community of ideas of the future would eventually come to accept the truth of the view I’m proposing as a matter of consensual agreement, coming to see it as a plain scientific truth. The view would self-satisfy the condition of scientificity built into it.
     Second, by accepting the suggested criteria of scientificity, we would not necessarily eliminate the comprehensiveness of our philosophical visions by seeing them replaced by science. In fact, there are reasons to expect something different. Speaking about the interdependence of philosophical problems belonging to the residual core – like those of metaphysics and epistemology – philosophers have claimed, with some exaggeration, that such problems are so deeply intermingled with one another that each of them will only be solved when all the others have already been solved. This claim hints at how our central philosophical problems could give way to science: Not so much by means of constructing theories that can more clearly be shown to match or not match the facts they have to explain, but by means of the heuristic support that theories are able to give one another. Some kind of intertheoretical support can easily be found, even in the natural sciences: Darwin’s theory of evolution, for example, received very important heuristic support when, many years later, the scientific community became aware of Mendel’s achievement. Something similar can occur with interrelated philosophical problems: consensual agreement can arise in these domains, not so much as a result of what counts as objective confirmation (though some kind of objective confirmation may be necessary), but through the inter-theoretical support that a solution of a problem can give to the solutions of others and vice versa.
    There are three conclusions that might be drawn from this prevalence of inter-theoretical support in philosophical inquiry. The first is that, we are not devoid of reasons to maintain the optimist belief that even in the most resistant domains of philosophy, we will be able at some future moment, to find a path of consensual agreement. Second, we also have reason to expect that the subject matter of these agreements will not be a set of independent theories of narrow scope, but will be comprehensive clusters of scientific theories; in this case only the conjectural form of our problems will necessarily be lost, not their comprehensiveness. A third conclusion, at least suggested by the heuristic interdependence of theories, is that we can’t dismiss philosophical attempts in fields like epistemology, metaphysics, and ethics, by comparing them to what has happened to the philosophical theories anticipative of sciences like physics, chemistry, or biology, where the corresponding philosophical theories were shown to be simply too wrong or too rough to have more than a merely historical importance. Indeed, in the case of the natural sciences, there were deep epistemological ruptures distinguishing science from prescientific (non-consensual) philosophical inquires, which often came to play the role of bad anticipations, delaying the emergence of science. However, in higher forms of knowledge, where inter- theoretical support might be the prevalent sign of truth, it seems that the transition from philosophy to science tends to be more gradual, since it involves correction of the interrelated theories, maybe deep corrections, but does not leap to something totally new. This means that philosophical speculation in its central fields can be heuristically more relevant, since this speculation accumulates more truth (though we don’t know where) before consensual agreement becomes strong enough to produce, in more urbane and discrete ways, a real qualitative change. Attention to this might rescue much of the importance of our central philosophical disciplines from positivist and scientist dismissing.


                                   RELIGION AND THE MYSTIC REMAINDER
                                                        OF PHILOSOPHY

We can understand why philosophy is a conjectural form of inquiry by viewing it as enticipation of science. But not all features of the philosophical inquiry can be explained in this way. The features indicated in the historic definitions of philosophy as a “search for wisdom”, the wonder about the existence, the frequent appeal to transcendent principles of explanation, the drive towards a comprehensive understanding, aiming to integrate our experiences in comprehensive worldviews enabling us to explain the world as a whole and our place in it, the production of philosophical systems trying to put forward and to justify such worldviews – all these aspects can hardly be understood if we continue to think about philosophy only as a cognitive enterprise anticipating science. In this chapter, I will try to show that an answer to these questions can be found when, instead of investigating the way philosophy gives rise to science, we investigate the way philosophy originated. This endeavor leads us to compare philosophy with another of its proximal relations, namely, religion.

There are two main features usually shared between philosophy and the religious thought, which might be called transcendence and comprehensiveness. Monotheist religions, at least, obtain the feature of transcendence by their appeal to a transcendent God, generally viewed as a personal, efficient, and sustaining cause of the world. In this way, these religions also achieve their comprehensiveness: Usually the concept of God stays at the center of a doctrine that aims to integrate our views in an explanation of the world as a whole, together with the human place in it, from which is derived a set of directives for human conduct in life. Philosophy often preserved similar aspirations of transcendence and comprehensiveness, though realizing them without appeal to a personal God. Traditional philosophers were moved by a need for comprehensiveness, which has lead the best of them to the construction of all-embracing philosophical systems, explaining reality as a whole and often deriving from this explanation general directives for human conduct. Though the aspirations of today’s philosophy are not so high, broadness of purpose still remains an important element in the evaluation of the pertinence and importance of a philosophical inquiry. As for transcendence, though philosophy does not appeal to the supernatural in the same way as religion, it traditionally appeals to metaphysical principles of explanation, which lie beyond our actual possibilities of experience or understanding. Though these principles are not spiritual beings, like the Gods of religion, they might not be completely different from them: They often can’t be adequately grasped through human cognitive powers, they often have some mental attribute, they might relate to the experienced world in an obscure and mysterious way. To see the importance of such metaphysical principles, we need only consider their central place in the history of philosophy. Here is a sampling, from Thales to Wittgenstein:

- water (Thales); unbounded (Anaximander); air (Anaximenes); earth   (Xenophanes); feuer (Heraclitus); Being (Parmenides); the atoms (Democritus);
- the ideas, especially the idea of the good (Plato); the being qua being or substance or God (Aristoteles); the Uno (Plotino); nature (J. Scotus); the omni-God (Aquinas and many others);
- the infinite thinking substance (Descartes); substance-nature-God (Spinoza); monads (Leibniz); minds (Berkeley); the noumenic ocean  with its thing in itself and the transcendental I (Kant); the pure ego (Fichte); the absolute spirit (Hegel); the will (Shopenhauer); the will to power (Nietzsche); the beingness of Being (Heidegger); the inexpressible (Wittgenstein).

     The relationship between philosophy and religion can be historically and genetically approached by means of the consideration of such principles or entity-principles, since they are entities acting as principles. It is a well-known historical fact that occidental philosophy was born from the soil of Greek mythology and religion. At some point, the Greek thinkers came to be dissatisfied with the explanations of the events of nature and human life furnished by mythology, and began to replace these explanations with philosophical ones. Historians of philosophy have suggested that the contact with other cultures, which their different gods and values, might have contributed to weaken their belief in their mythological explanations(46). But this fact would never in itself be enough to trigger philosophical speculation, since many other cultures were similarly exposed without developing any kind of argumentative philosophy. Also the use of generalization, which the first scientific reasoning suggested to the human mind(47), would not be enough to cause the emergence of philosophical thought, since commonsensical generalizations about ordinary phenomena have always existed.
     In my view, a more compelling reason for the birth of philosophical speculation is the following. The Greeks, partially as consequence of their exposure to other cultures, made scientific developments in arithmetic, geometry, physics, and astronomy.  While other people saw the results of science only as an aid for the achievement of practical aims, the Greeks for the first time considered them in abstraction from these aims, namely, as scientific generalizations. This abstraction enabled them to become aware of the intrinsic characteristics of these generalizations. They could see that scientific generalizations have an explanatory power, which enables them not only to explain what is openly available, as in the case of commonsensical generalizations, but also the “hidden nature of things”. They also saw that the scientific form of explanation is based on the assumption of the existence of regularities, found in empirical nature as well as in mathematics, regularities not only able to be reflected in generalizations, but also, when empirical, to allow projections to the future, as astronomical predictions proved. In other words, they grasped what we could call the idea of scientific generalizations, constituting rules, principles, and laws. This idea amounted to a new kind of explanation, very different from the anthropomorphic explanation provided by religion. Indeed, it seems that it was the discovery of the possibility of replacing religious or mythological explanations with explanation by means of rule or law, applicable even to what was unobservable or hidden in nature, that was the spark that lit the fire of philosophical speculation in the minds of the first Greek thinkers. The underlying idea that must have come to the mind of the first philosophers was simply that the whole world could be explained, not by appealing to the wishes of deities, but by means of regularities similar to those discovered by science. For this reason, it is not surprising that the first philosopher of the occidental tradition – Thales of Miletus – was also an astronomer and a competent mathematician who once predicted a solar eclipse. His hypothesis that water could be the principle (arché), namely, the efficient and sustaining cause of all things, was the first attempt to replace the explanation by appealing to gods with the non-anthropomorphic kind of explanation provided by science. Certainly, the explanations he was able to give couldn’t be adequately constructed as scientific ones: They couldn’t be constructed in a way that enables the kind of consensual agreement we have seen to be distinctive of science. Neither he nor his successors could achieve any suitable scientific explanations in such broad matters as the ultimate constituents of the physical world, since consensual agreement about these explanations depends on the realization of sophisticated scientific experiments, which only today is possible. Nevertheless, the pre-Socratic thinkers were at least able to philosophize about these matters; they were able to have conjectural glimpses of the nature of things: vague, incomplete, inevitably defective suggestions, but still able to order, direct, and even deepen our understanding of reality. What philosophers like Thales and, with more refinement, Heraclitus and Parmenides, were able to produce, were may-be ideas, forms of theories, explanatory sketches working as protoscientific products of speculative imagination. Under the pre-Socratics, these principles took the form of efficient or sustaining causes of the world experienced by us, being at first sensible things, like water and earth, but quickly becoming more evanescent things, like fire and the invisible air, being eventually more consistently replaced by completely non-empirical things, ranging from the ‘unbounded’ of Anaximander to the ‘inexpressible’ of Wittgenstein. They sought principles to ground the natural world and, as later philosophers have done, to ground practical life as well. I will go deeper into the analysis of these principles, but first I need to consider some related ideas of Auguste Comte.

The historic consideration of the fact that philosophy was born as a substitute for the explanations of mythology and religion brings to memory the so-called “law of three stages”, developed by Comte as an ordering of the mind’s journey from superstition to science(48). I will make some use of this law in section 4. But now, since I believe that Comte’s law is important and has been misunderstood and unjustly depreciated, I will explore it in some detail, answering in the next section the most influent objections raised against it(49).
     The law of three stages is applicable on three levels: (a) the level of the development of human culture in its distinct branches; (b) the level of the development of the individual mind; and (c) the level of the development of human society.
     It is at level (s), as a general law governing the development of human culture, that the law of three stages is particularly important. For Comte, underlying the emergence of each fundamental science (see III, 4), there is an evolutionary process in which the human mind passes through three successive stages: the religious or fictive, the metaphysical or abstract, and the scientific or positive.
     The religious or fictive stage is the necessary point of departure of our cultural evolution. This stage is dominated by anthropomorphism: The human mind attempts to explain the anomalies of nature by projecting its own features into the external world. The natural phenomena, particularly the deviant ones, are explained as caused by the will of personal beings with superhuman powers: the gods or God. Knowledge about these supernatural entities, which is supposedly acquired in this stage, is considered as absolute. Nonetheless, this supposed knowledge is not a product of reason, but of the imagination alone.
     The religious stage assumes subsequently three forms, each of them undergoing a high level of abstraction. In the first, the animist sub-stage, physical objects like trees, animals, and heavenly bodies are vaguely conceived as possessing life, passions, and will. In the second or polytheist sub-stage, such objects are replaced by gods, living beings of supernatural nature, normally invisible, intervening arbitrarily in the course of nature. Finally, in the monotheist sub-stage, the divinities of polytheism are melded to form the kind of non-cognizable omni-God, typical of the Judeo-Christian religion. Comte sees in this movement a progress of the cultural mind within the theological order: tending to abstract and reduce the number of causes for the explanation of the phenomena, the mind begins the process of replacing imagination with reason.
     The second, metaphysical (philosophical) stage is in Comte’s view only a transitional one. This stage represents notable progress, for the principles of explanation cease to be searched for in supernatural deities and come to be searched for in nature itself. Though these principles might belong to nature, they are there in a hidden way. They are called “natural powers”, “essential properties”, or “abstract entities”. Examples of such principles are for Comte the phlogiston preceding modern chemistry and the ether in the early stages of physics. Such principles, Comte claims, have a fundamentally equivocal character: They should give a natural explanation of the phenomena, as scientific principles, namely, as law-like regularities holding between phenomena, though they fail to do this; on the other hand, they can’t be viewed as personal agents without regress to the theological stage. They are what Comte suggestively called “personified abstractions”, pointing out their internal inconsistency.
     Comte has a negative view of the intrinsic value of the first two stages. For him, they are basically dependent on imagination, and the explanations given through their conceptual constructions are not genuine. The value of these stages belongs only to their practical consequences. But only by means of these conceptual constructions can the path for the scientific stage be prepared. The human mind, says Comte, can’t investigate without being guided by some kind of theory; the theological and metaphysical stages furnish theories through which the human mind can pursue investigation and, motivated by an illusion of knowledge, persevere in the cumulative observation of facts that eventually leads to science. A good example of this process is provided by the transition from astrology to astronomy: The continuous observation of celestial bodies, aiming to foretell human fate, led to the development of mathematical measurements, which created the conditions for the emergence of astronomy as a science.
     For Comte, the metaphysical stage is intermediary and provisory, being nothing more than a long and laborious preparation for the emergence of the positive stage. Only in this last stage is science established as the unique suitable form of inquiry, and the old theological and metaphysical questions are abandoned and anathematized as unanswerable and sterile.
     In the positive or scientific stage, what is sought ceases to be an absolute and infallible kind of knowledge, becoming a relative kind of knowledge; relative to the unavoidable uncertainty of human inquiry (indeed, by what means could we recognize absolute knowledge, in case we find it?). The intention to explain the world as a whole is also recognized as an illusion: We can’t do more than explain its constituents, which is done by the particular sciences (indeed, how could concepts intended to classify the constituents of the world be applied to the world as a whole?). Moreover, in this stage, the phenomena cease to be explained by imagination and come to be explained by reason alone, this reason being identified with scientific thinking, which doesn’t seek the hidden essential causes of phenomena, but the discovery of laws, namely, the discovery of verifiable regularities holding between phenomena. The knowledge of these regularities allows us to infer the occurrence of some phenomena based on the occurrence of other phenomena, and in this way to make predictions. And this power of making predictions leads to a real – and not just imaginary – dominion over nature.
     For Comte, the law of three stages also manifests itself in the development of the individual mind, evincing its biological root. As he noted, we are all theologians in childhood, since we live in an imaginary world, believing in mythical beings; we are metaphysical in adolescence, when we become able to apply reason though lacking true knowledge of facts; and when we achieve adulthood (insofar as we achieve it), we become “physicists”, admitting only positive knowledge, firmed and confirmed by scientific means.
     Finally, the law of three stages also manifests itself at the level of social organization and practices. But this manifestation is dependent on the previous realization of the stages in the domain of culture. Considering that the fundamental sciences necessarily were formed in different times, since the development of one fundamental science presupposes the development of another, and considering also that technological development is mostly a result of theoretical development in science, it is to be expected that the social effect of the formation of the fundamental sciences in the “positivation” of the social and economic organization of society is a rather late phenomenon. Comte’s suggestion is that at the level of social organization, the theological stage lasted until the end of the Middle Ages – this organization being characterized by an authoritative and militarist society dominated by priests and kings. After the Protestant Reformation, metaphysical ideas began to direct society, establishing a reign of law and abstract rights. Only after the French Revolution, in a period where all the fundamental sciences had achieved or were in the process of achieving their “positivation”, the positive or scientific stage at the level of social organization became possible. This last period is characterized by the emergence of a peaceful society, in which the economic life of men becomes the center of attention; in this society, science is destined to exercise a determining role, leading to a society organized and regulated by an elite group of scientists.

Comte’s law was always subject to criticism. Some criticisms, such as the accusation of rigidity and of an excessive discrediting of nonpositive forms of thought are in my view well justified. But the central objections seem to me misleading and I will answer the most important of them.
     The first objection is that the law of three stages is itself metaphysical, since it is attained a priori without recourse to observational facts(50). This is certainly false. Comte says explicitly and shows through his writings that his law originates from an attentive examination of the facts concerning the evolution of our culture and the emergence of the fundamental sciences, along with a thoughtful consideration of human nature. Against the further objection that the law itself can’t be adequately inferred, since there is only one, unfinished, historical instance, namely, that of our own civilization, I suggest that the law of three stages could be better justified as an inference for the best explanation, the only capable of putting under only one hat a myriad of socio-cultural facts in their historic progression. Indeed, it is because the explanation provided by this law gives some coherence to the historical progression of human culture, and because this coherence is confirmed by our understanding of human nature, that the law tends to impress itself in our minds as a natural and compelling explanation. Furthermore, because the law can be gradually confirmed or refuted by a careful investigation of past socio-cultural facts and also future ones, it is not in the end much less confirmable or refutable than, for example, the theory of biological evolution.
     The second objection is that, when applied to the explanation of the three stages at the social level, Comte’s law can’t adequately account for the order of emergence of the sciences: Mathematics, for example, had already emerged among the Greeks in the theological stage, and astronomy and physics had already emerged when the society still was in its metaphysical stage. Like the first one, this objection is also explicitly answered by Comte. For him, each fundamental science should arise after the metaphysical and theological stages have occurred in its own field, and since there is an order of presupposition among these sciences, they can’t arrive at their positivation simultaneously. So the stages at the social level are a result of a sum of partial changes. In a similar way, a child can anticipate some traces of the adult’s mind and an adult’s mind can preserve some adolescent or even childish features.
     A third and more serious objection is that Comte’s use of the word ‘law’ is abusive and misleading: The uniqueness of the considered events, the vagueness and uncertainty of the considered process, give us no right to use this venerable word; the best we can do is to speak about a socio-cultural tendency or trend(51). One answer to this objection consists of accepting it. In fact, what Comte discovered is certainly only a tendency, valid in vague probabilistic terms; consequently, what he discovered was not a real law. Although it appears sound, I disagree with this reasoning. I prefer to think that the proper form of a historic-cultural law is that of a general tendency. We can’t expect that a historic-cultural law maintain the same precision and lack of exception as the laws of physics or chemistry. A historic-cultural law works somewhat like a statistical law: It would be unreasonable to expect from its statement more than the suggestion of a general tendency, since the diversity of variables that can intervene in the process is undeterminable. However, it is false to think that the vagueness and uncertainty of a law compromise its status, except when we misleadingly assimilate our general concept of law to concepts like that of a physical law, as the philosophy of science invites us to do. What most distinctively characterizes the statement of a law is not its universality and precision (for in this case, no statistical law, even those of quantum-mechanics, would satisfy the condition), but our assumption that the generalization made in this statement is non-accidental. Indeed, the assumed non-accidental character of the regularity asserted by a generalization might be admitted as the only characteristic that must be common to all kinds of law. The fact is that science needs a term to cover all kinds of generalizations that we assume to be non-accidental, and ‘law’ seems to be the most suitable term for this job. The law of three stages fulfills this requirement. It seems reasonable, for example, to predict that a society constituted by human beings biologically identical to us, under similar circumstances, in the process of becoming a scientific civilization, would probably follow a similar order of stages in the development of its branches of knowledge instead of, for example, jumping directly to the scientific stage. If we accept this thought as plausible, this is because we are already assuming that the sequence of stages is a non-accidental generalization, namely, that it is a law in the liberal but justified sense of a historic-cultural tendency.
     The conclusion is that, under a sufficiently tolerant and flexible interpretation, the idea that the progression of human culture tends to follow the three described stages is defensible. My next step will be to consider the traditional philosophy through this view and see how much further it can take us.

We can summarize Comte’s view of the place of philosophy between religion and science through the following schema:
         RELIGION       >       PHILOSOPHY     >       SCIENCE
         (explanation               (explanation                    (explanation
           by gods)                   by principles)                   by laws)

    Despite the obvious metaphilosophical appeal of this idea, Comte didn’t apply it sufficiently to the central domains of philosophy, presumably due to lack of a more detailed acquaintance with its history; usually, his examples are of metaphysical principles belonging to the prehistory of the positive sciences, such as the phlogiston before chemistry and the ether in the infancy of physics.
     To put Comte’s evolutionary outlook to work in the analysis of traditional metaphysical principles, the first thing we need to do is to make explicit the most distinctive features of the entities claimed by religion to be supernatural or divine. These features, which I call theomorphic, can be reduced to four:

(i) Physical transcendence: Mental entities are made of a stuff which is different from and superior to that of physical bodies (for example, the Cartesian God is an infinite thinking substance);
(ii) Hypermentality: The mental powers of the mental entities are altered and extended, maybe infinitely (they can forecast the future, some of them are omniscient, etc.);
(iii) Hyperphysicality: The physical powers of the mental entities are altered and extended, maybe infinitely (they can change human fate, contravene physical laws, etc.); and
(iv) Mind-body idiosyncrasy: The mental entities, when eventually bound to physical bodies, are not necessarily bound to bodies ordinarily known by us as possessing mind, nor they are bound to physical bodies in the usual dependent ways known by us (they might have no physical body, they can inhabit non-living beings, they can change freely the bodies they chose to inhabit, etc.).

     These features can be seen as supposed criteria of identification, which would enable us to describe and eventually recognize the supernatural and the divine. Not all of them need to be present: In Epicurean materialism, for example, since gods are made of utterly fine material atoms, they must be physical, failing to satisfy adequately the criterion (i). Typical of theomorphic features is that they aren’t objects of our daily mental or physical experience; but in fact, we can conceive them in a secondarily, by altering and extending what we already know from our ordinary experience.
     Now, if following Comte we wish to consider the metaphysical entity-principles as something subsisting between supernatural godness and the regularity of scientific law, we must understand them as consisting of something between

     A. what is theomorphic, namely, what possesses at least one theo-
          morphic feature, and
     B. what is natural,  namely, what possesses only the ordinary phy-
         sical and psychological features recognized by  common  sense
         and possibly also by science, since science can be easily unders-
         tood as a critical extension of common sense.

Assuming this, we are prepared to distinguish some basic kinds of metaphysical entity-principles. The first is,
(a) +A+B: Hybrid (or inflationed) metaphysical principle.

 The constitution of a metaphysical concept intended to refer to a principle of this kind is semantically dependent (even if in an elusive way) on both the theomorphic features (which are constitutive of the supernatural) on the one side, and, on the other, the normal physical and mental features, accessible to our ordinary commonsensical and scientific experience (which could give us scientific laws).
     Spinoza’s Deus sive Natura could serve as an example of a hybrid metaphysical concept. At first view, his God seems to be only a natural entity-principle (+B), since it is the whole world, accessible to us under its essential attributes of extension (our experiences of the physical) and thought (our experiences of the mental) as our natural world. Nevertheless, Spinoza’s view is not so absolutely naturalist and free from anthropomorphism as he intends: It implies, for example, that all physical things are also mental, possessing some kind of sentience, since each modus of extension must be accompanied by a correspondent mental modus, and this means mind-body idiosyncrasy (+A). Furthermore, nature as ‘God’ is “hypostasized” as having the capacity of loving itself with infinite love (Ethics, Book V, prop. 35), which means that Spinoza’s God also has some kind of theomorphic feature of hypermentality (+A).
     Perhaps, the best example of a rich and colorful hybridous first principle is John Scotus’ concept of nature: Nature must be the biblical God (as nature creating and uncreated), possessing all theomorphic features (+A); but nature also consists of all empirical things (as nature created and uncreating) (+B). Consequently, on one side nature is the same personal God of Christian religion, possessing something like consciousness, intentionality, and freedom of the will; but on the other side, nature is our own world, evolving in a law-like way, unavoidably, in the direction of its ultimate fate as the uncreated and uncreating nature – this hybridism is so flagrant that it has always struck the critics as an insurmountable inconsistency. Another mixed principle, which in some way reminds us of Scotus’ nature, is Hegel’s spirit (Geist), which on one side is hypermental (since it is the origin of all reality) (+A) and mind-body idiosyncratic (since all reality belongs to it) (+A), while on the other side, it must unfold itself following impersonal dialectic laws (+B).
     Still another example of a hybridous entity-principle is Leibniz’s monads. For him, the real world is constituted by an infinite number of mental points called monads. On the one hand, a monad has its own impersonal laws, relating to all other monads through the appearances of space-temporal nature (+B). On the other hand, each monad is also a lively force, possessing some grade of perception and consciousness, which extends itself in some measure to the whole universe of monads! Consequently, monads also have theomorphic features, like mental idiosyncrasy (because material things are phenomenal appearances of aggregates of monads) and hypermentality (because monads are to some degree omniscient) (+A).
     Certainly, the relative amount of +A and +B can vary: Spinoza’s Deus sive Natura is nearly naturalist (we could represent it as +A++B), while Leibniz’s monads and Hegel’s spirit distinguish themselves by their theomorphic features (we could represent them as ++A+B). J. Scotus’s nature is to be placed somewhat in the middle (+A+B). The majority of entity-principles of speculative metaphysics are of an inflationed kind, alluding to both theomorphic and naturalist features in order to be cognitively grasped.
     The next kind of metaphysical principle has the form

(b) –A–B: Elusive (or deflationed) metaphysical principle.

 The constitution of a metaphysical concept intended to refer to a principle of this kind is explicitly conceived as devoid of any semantic dependence on either theomorphic features or the usual mental and physical features, as they are ordinarily experienced and known through common sense and science.
     The consequence of this explanatory strategy is that the elusive entity-principle becomes in itself unknowable. In fact, if the elimination of references were carried out in a consistent way, the concept-word used to refer to such a metaphysical principle should be senseless.
     Historically, the first example of an elusive metaphysical principle seems to be Plotino’s Uno, which was conceived as something completely unattainable through our cognitive powers (the Uno can be approximated only through what it is not, since it is nothing that we can know). But the most notorious example of the elusive principle is Kant’s noumenical world, including as its inhabitants the thing in itself and the transcendental self. Contemporary examples of elusive principles are Wittgenstein’s concept of the inexpressible (Unausprechlich), pointing to what can only be shown but not said, and Heidegger’s concept of Being, understood as the beingness of beings, which, as he claims, can be only metaphorically approached by means of literary language. The deflationary kind of metaphysical principle has the advantage of not running the risk of being internally inconsistent; but the price for this advantage is that of not being a concept at all. And this semantic vacuity is eventually able to contaminate the rest of the related philosophical discourse with rhetorical vacuity, as the development of Heidegger’s work has made clear enough.
     There are ways by which inflationary and deflationary strategies can be combined in the process of the constitution of metaphysical concepts. Consider the case of Shopenhauer’s concept of will. Originally, the will is only a name for the Kantian thing in itself, which is unknowable. The best one can say of this will is that it is an x, its concept having the form –A–B. But then Shopenhauer begins to attach properties to this x in the form of its immediate phenomenal manifestations: The will is a blind drive that manifests itself as force in inanimate nature, being most directly objectified in our internal experience of the will to live, which is able to show its presence in the whole world, organic and inorganic. In this way, the harmless thing in itself ends up manifesting itself as a wicked cosmic will, which pervades all of nature and which is the real source of all the endless suffering of mankind. We see that the will, which begins by being conceived as –A–B, receives features that transform it into a principle which, taken as a blind natural force, might possess something of the character of a universal natural law (+B), though simultaneously involving, in its manifestations as a universal will to live, theomorphic features: namely, mind-body idiosyncrasy and some kind of hypermentality (+A). This is so, even if Shopenhauer applies the old philosopher’s trick of denying what he has done after having done it. His concept of will can be seen as resulting from a conceptual composition of the form +A(–A–B)+B (the brackets enclose what comes first in the process of conceptual constitution).
     Looking for alternatives between +A+B and –A–B, the hybrid and the elusive principles, we still find two more basic possibilities:

(c) +A–B: The constitution of a concept intended to refer to a principle of this kind is semantically dependent on theomorphic features unaccompanied by natural ones.

     This combination is obviously unattainable, since it brings us back to religion: Entities that are physically transcendent and/or hypermental and/or mind-body idiosyncratic without any appeal to naturalist explanations are just spiritual entities like gods, totems, etc.

(d)  –A+B: Naturalist philosophical principle.

 The constitution of a philosophical concept intended to refer to a principle of this kind is semantically dependent on natural features, admitted by common sense and eventually science. (However, the more adequate and precise explanation of these principles as objects of scientific or consensual knowledge remains practically unattainable).
      The main difference between a naturalist principle and a scientific law lies in the absence of a possible consensual agreement about the truth-values of the often too vague statements of naturalist philosophical principles.
      Pre-socratic speculation is rich in examples of this kind, like Anaximander’s theses that the earth is suspended in the void and that human beings have evolved from animals, already discussed in chapter III (sec. 4). But the standard example of a naturalist principle is the atomistic theory of materialist philosophers like Leucipus and Democritus, maintaining that concrete things are constituted of eternal and invisible portions of matter. For Democritus, the atoms can have different forms, responsible for the different properties of matter; they can attract one another in order to hold together chunks of matter, etc. Though the atoms might be “theoretically” divisible, since they have forms and sizes and weights, they are at least physically indivisible(52). Certainly, since the atomist hypothesis results from reflexion based on our ordinary experience of physical things, being devoid of any appeal to theomorphic elements, the philosophical concept of the atom, like the scientific one, has the form –A+B.
    Naturalist principles are those which most easily show their protoscientific character because they occur most often in ancient anticipations of today’s well-developed natural sciences. Concerning atomism, the pattern of this development is the same as that discussed in the examples given in chapter III: The ancient atomists could not identify the properties of their atoms, measure them, or observe their traces in consensually attainable ways, as physicists do today with electrons or quarks; but they could speculate about their existence and their theory assumes a form, which we may view as common to our atomic theories too, since the claim that matter is not divisible in a continuous way, but into discrete components, is common to both. (It seems that the more remote from scientific realization is the idea that the philosopher is searching to grasp, the more theomorphic the explanation will tend to be.)
     Another example of the naturalist principle is Parmenides’ Being, since it is devoid of theomorphic features. For Parmenides, the “path of truth” is of that which is. Substantivating that which is by the Being (to on), he attributes to it the predicates of unity, uniqueness, eternity, unchangeability, indivisibility, homogeneity, and limitation, as if Being were a thing, though only capable of being grasped by thought, not by the senses; further, since to think what is not is completely impossible, Being is the only object of thought and “the same thing is there both to be thought of and to be” (to gar auto noein estin te kai einai).
     Parmenides’ strategy exemplifies the non-determining semantic suggestivity (see V, 1) so often found in philosophical discourse: The vagueness and incompleteness of the argument, together with the apparent inconsistency among the different properties attributed to Being, suggests an indefinite number of interpretative keys, none of them entirely satisfactory. My own hunch is that Parmenides’ Being would be better identified with what we could call today the totality of conceivable propositional contents. This interpretation satisfies the principle of charity, saving most of Parmenides’ claims. Indeed, the totality of conceivable (true and false) propositions is all that can be thought of (it is “what is there to be thought”), being eternal, changeless, unperceivable by the senses, and not divisible and heterogeneous as are empirical entities. Moreover, to be excluded from the totality of conceivable propositions are the inconceivable ones, like the contradictions, which justifies Parmenides’ dictum that one can’t think what is not. Finally, according to this interpretation, the “path of truth” admits false propositions, so that the Parmenidian Being becomes immune to Plato’s objection that it is impossible for Parmenides to say what is false. If my paraphrase is correct, Being is an anticipation of what Plato tried to approach with his hypothesis of a world of ideas, the stoics with their doctrine of the lekton (as the incorporeal matter of discourse conveyed by linguistic signs), Frege with his reign of thoughts, Popper with his world 3 (of cultural creations), Peirce with the category of thirdness... If this is correct, then here we have an impressive example of speculative anticipation of something that later philosophers arrived at in more advanced ways, though always with a very limited success. Even though all these doctrines differ deeply, we are not entitled to dismiss the hypothesis that there is something relevant to be found here, which in principle could itself become a matter of consensual (scientific) agreement.
     Examples of the form –A+B are interesting because they can, in certain cases, be shown to be speculative anticipations of science that do not hide any deceptive anthropomorphic intention – they are constructed only to satisfy our speculative curiosity about questions that lie beyond our present possibilities of consensual evaluation. These cases prove that Comte’s depreciative claim that the metaphysical inquiry is a mere product of imagination, without any consequence except that of preserving, through hope and illusion, the disposition to investigate, was too pessimistic.
     Finally, it is to be noted here that also different strategies can be combined in the process of conceptual constitution. This seems to be the case with Plato’s concept of idea or form. To make this concept conceivable, Plato must appeal to analogies taken from our ordinary experience, beginning with the psychological meaning of the word ‘idea’ and the special meaning of the word ‘form’, which means the addition of +B. Besides, a platonic idea should be viewed as a non-theomorphic entity (–A). Consequently, the Platonic idea should have the form –A+B. However, since Plato maintains that ideas belong to a world of purely intelligible things existing absolutely apart from the sensible world, the concept of idea should also have the form –B. As a consequence, it seems that the Platonic idea should be referred to by means of a conceptual composition of the form (–A+B)–B.

The first conclusion of our analysis of metaphysical principles under the perspective opened up by the law of three stages is that it shows a certain limitation in Comte’s own view. At least when we consider the cases –A–B and –A+B, we see that Comte’s thesis that metaphysical principles are inconsistent personified abstractions falls too short. Further, the case –A+B shows that philosophical speculations are also able to be a purely heuristic undertaken, motivated by intellectual curiosity alone, without a theomorphic orientation. These speculations then appear as conjectural explanatory sketches, constituting not a provisional stage of inherently equivocal ideas, but the speculative beginnings of science, which eventually are capable, at least in their contours, of being admitted later as obvious part of our scientific achievements.
     The recognition of such possibilities also shows, particularly in the examined case of Parmenides’ Being, that vagueness and obscurity in philosophy can be justified if a philosopher is trying to point out (like Parmenides, Heraclitus, Kant, Wittgenstein…) something that lies beyond the conceptual resources at his disposal. As H. H. Price once pointed out in a very suggestive passage, “There may very well be some things which in the terminology available at the time can only be said obscurely; either in a metaphor, or (still more disturbing) in an oxymoron or a paradox, that is, in a sentence which breaks the existing terminological rules and is in its literal meaning absurd. The man who says them may, of course, be confused. But it is possible that he is saying something important. Nevertheless, his successors may be able to divine what he is trying to convey. The terminological rules may eventually be changed. And the wild metaphor or outrageous paradox of today may become the platitude of the day after tomorrow.”(53)
     Though I don’t believe that philosophers can think something precise or adequate that they can’t put in sufficiently precise or adequate language (language is always plastic enough), it seems clear to me that philosophers often have important but imprecise and inadequate intuitions, which they are able to spell out only in correspondently imprecise and inadequate terms. If so, we come to the conclusion that even if inherently contradictory and misconceived, as the strategies of inflationary or deflationary entity-principles certainly are, these unsupported intuitions can always be pointing to something important lurking behind the stage.
     Finally, one more word about the question of comprehensiveness. We saw that the wish for comprehensiveness found in philosophy is inherited from the wish found in religion to find an integrated explanation of the whole world and of the place and prospects of man within it. However, this is not necessarily the unhappy heritage of an impossible quest. When we consider that the present central philosophical questions are always in some way linked one with the other, it seems that comprehensiveness, when preserved under more reasonable limits, may be a well-justified aspiration of philosophy as a science-anticipatory endeavor. If this is true, then even the religious search for comprehensiveness wasn’t wholly misplaced.


                                          THE RELATIONSHIP BETWEEN
                                               PHILOSOPHY AND ART

We have compared philosophy with two other fundamental cultural activities, science and religion, showing how philosophy stands between them. Philosophy is not only a science-anticipative endeavor, but it retains some traces from the religious thought, not only in the broadness of its theoretical and practical aims, but also by appealing to principles of explanation that, like God, remain in some way beyond our cognitive powers. Now it is time to compare philosophy with a third fundamental cultural activity: Art.
     Based on the fact that there is a certain similarity between philosophy and art, some philosophers have advanced the thesis that philosophy is essentially a form of art. As J. H. Gill, a supporter of this idea, has suggested, philosophy is “not like a lens, through which we penetrate and scrutinize reality, nor like a lamp, by which we explore previously undiscovered dimensions and horizons of human existence, but like a prism with which fascinating and provocative conceptual patterns and thought sculptures are created”.(54) In what follows I will consider the interface between philosophy and art, in order to show that the properly artistic aspects of philosophy, far from constituting a sufficient condition of philosophy, are not even necessary. To make this thesis plausible, we need to begin by distinguishing between two kinds of similarities between philosophy and art: (a) external similarities, namely, those that are due to the utilization of artistic resources in philosophy, which are not always present, and (b) internal similarities, namely, similarities in nature between the two cultural practices, which are always and necessarily present. We will begin with the former.

Similarities between philosophy and art are external when the philosopher uses literary means. There are various reasons for a more literary approach. One is to communicate ideas in a more effective and impressive manner, or in order to create a suggestive discourse. But the most serious reason is that philosophers often find no alternative, having to choose between setting forth straightforward but faulty arguments, and expressing themselves in more elusive ways, which are open to different interpretations and are less incorrect, but also less informative. This is a legitimate reason to produce what might be called a metaphorical or, as I prefer to call it, a semantically suggestive discourse in philosophy: A discourse in which words and their combinations evoke things that are not literally meant by them. Consider the recourse to similes and myths by Plato, to poetic imagery and allegories by Nietzsche, to aphorisms by Wittgenstein, and we see how important and powerful the use of figurative language in philosophy can be.
     These multiple aesthetic resources are art: They are art in the philosophy. However, they are not to be confounded with philosophy itself. The use of literary resources in philosophy is external to the philosophical endeavor. To see why the external use of artistic resources in philosophy don’t turn philosophy into art, we need only to consider the case of religion. Religion has always made external use of artistic resources in order to perform its exhortative functions. Not only mythological stories, like Hesiod’s Theogony, but also the Bible, are also literary works of the highest quality. Nevertheless, nobody would say that the Theogony and the Bible were intended only as works of fiction, or that religion can be reduced to a form of art. If this is so with religion, why should it be different with philosophy? And if religion can conceivably exist even without being colored by artistic means, why not philosophy?

There are also internal similarities, namely, similarities in nature between philosophy and art. If philosophy could be considered a kind of art, this should be due to internal similarities. Nevertheless, we will see that the properties that are similar, though possibly necessary for both philosophy and art, are sufficient for neither of them, what will lead us to reject any essential identity.
     The first internal similarity between philosophy and art, one could suggest, is that philosophy is a cultural activity without a further aim: Like art, philosophy is an end in itself. To a certain extent, at least, this is true: Philosophy justifies itself as something highly pleasurable in its own right, much more than for some external advantage that it can bring to us. However, the importance of this similarity can’t be exaggerated, for in the case of philosophy we can find more direct external ends intrinsically bound with it: The philosophical views we adopt have an indirect influence on the ways we judge and act. However, one can’t adopt the views expressed in artistic works as such, since there are no such views. The best one can do is to adopt some views that one has achieved under the influence of some aesthetic experience.
     A second element in common concerns what we can call the integrative function of art. Art envisages the integration of our sensible and emotional lives, enabling us to bring inner harmony to our feelings and enrichment to our existential experience. Something similar can be said for philosophy. It has an integrative function too, not so much for our sensible and emotional lives, but for what has been called the life of understanding and reason. It seems that philosophy does with the material of abstract concepts what art does with the material of sensible intuition. In the production and appreciation of art, sensible imagination is at work, while in the case of philosophy, it is the “intellectual imagination” that is laboring. So it seems that philosophy could be classified as a kind of “art of reason”, in contrast with the customary “art of emotions”. However, that the word ‘art’ is used here in a purely analogical sense is revealed by the fact that something similar can be said about religion. Religion also has an integrative function, relative to our vision of the world and our place in it. Is religion then something like the “art of spirituality”? And what about science? Doesn’t science have some integrative function too, concerning our knowledge of the world and even of ourselves? On this basis are religion and science also forms of art?
     Another similarity between philosophy and art concerns creation. Like art, philosophy is to some extent a work of imagination. Creation in art is directed not only to the production of customary beauty and harmony, but also to the production of unexpected contrast, able to suggest to each of us a reorganization of the emotional values we attach to things. Philosophical creation, for its part, might produce such contrasts with the cognitive material of abstract concepts. This is an aspect of philosophy that is strikingly similar to certain artworks, namely, its capacity to provide unexpected contrast in the form of tauma, the Greek word meaning astonishment. Here again we see philosophy at work as the “art of reason”, endeavoring to show the most unexpected possibilities of rearrangement of our intellectual world. This can be seen in transcendental metaphysical systems, like Plotino’s theological construction of the world and Fichte’s subjective idealism. Such systems don’t show how the world certainly is (despite the explicit intention of those philosophers), but how the world could be or possibly (but very improbably) is. This is an interesting point, but again, doesn’t show that philosophy is art or depends on art. It shows only that philosophy is a creative activity, more than science, though less than art.
     The thesis that philosophy is a form of art is more definitely disqualified when we consider that there are also essential differences between the two practices. Unlike art, philosophy has explicit heuristic purposes: It aims at discovering the truth. Even philosophers of the skeptic variety usually aim at establishing the truth of their skeptical claims. Though I don’t wish to deny that good art also aims at truth, it aims at truth in an uncommitted, indirect way: It makes us open to look at ourselves and the world around us in more correct ways. Philosophy, however, aims at truth in a more direct way: It intends to say what is true or at least to indicate it. And, though philosophy is not progressive and knowledge-cumulative in the same sense as science, it is, as we have noted (III, 8), content-cumulative, fulfilling more and more a spectrum of truth-possibilities. Indeed, if philosophy occupies the epistemic places of unknown scientific domains, as we have suggested, then we may expect that the subdivision within a philosophical domain have a limit in number, while the same might not occur with art.
     However, philosophy, like religion, remains somewhat close to art, closer than science. How can this be explained? I think that the psychoanalytical theory can give us some help here. Under this theory, philosophy and art have in common the fact that both are in some measure a result of what Freud called the primary process (primäre Vorgang) of thought, a form of thought based in the principle of pleasure rather than on the principle of reality(55). For Freud, this form of thought occurs in dreams, in the work of the neurotic and psychotic imagination, in the creation and appreciation of works of art, and also in religious and philosophical reasoning. In the primary process, the emotions or cathexes (Besetzungen) are not firmly linked with their correspondent representations. So cathexes related to unconscious or pre-conscious representations can be ceded to other representations related to the original ones, so that the last ones come to consciousness carrying these affective cathexes, producing pleasure by decreasing the levels of endopsychic tension. It is important to note that the mechanisms by which the charges of non-conscious representations are ceded to representations able to emerge into consciousness are essentially two: displacement (Verschiebung), by which the cathexis of a representation R is ceded to a representation R1, which comes to consciousness, and condensation (Kondensation), by which the cathexes of multiple representations R, R1, R2… come to consciousness by being ceded to a representation R. A consequence of this process is that representations will be combined in the consciousness in much more flexible ways than are found in the secondary process (sekundäre Vorgang), which is more characteristic of our practical and scientific reasoning, which are based on the principle of reality. What I called semantic suggestivity is something obviously dependent on the primary process, since it involves condensation and/or displacement.
     Now, the fact that from the psychoanalytic point of view the philosophical thought can be understood as being in some measure dependent on the primary process seems to corroborate the idea that the similarity to art isn’t essential. For, if the primary process were sufficient to characterize art, then it seems that we would need to assimilate other products of the primary process into art, for example dreams. However, it is clear that we would not say that dreams are artistic manifestations simply because the manifest content of a dream is related to its latent content through displacement and condensation. Such considerations don’t prove, but do reinforce our conclusion that neither the internal nor the external similarities are sufficient to characterize philosophy as necessarily requiring art, even if philosophy, like religion, might be greatly enriched by aesthetic means.


                           TOWARDS A GLOBAL ACCOUNT: INTEGRATING
                                                THE CRITERIAL CLUSTERS

In this chapter, I will conjoin the achieved results in an attempt to build an integrated descriptivist account of the nature of philosophy. This account might furnish a more perspicuous representation of the criterial clusters expected in the identification of philosophical discourse and thought.

Understanding as a cultural practice a set of predominantly mental activities, sustained by a social background and not immediately related to the satisfaction of the practical necessities of life, we have seen that philosophy can share similarities with three fundamental cultural practices. They are:

                                            a)   SCIENCE,
                                            b)   RELIGION,
                                            c)   ART.

     I call these three cultural practices ‘fundamental’ because of their importance to human life in general; if there are other cultural practices (ideological activities, cultural games, etc.), they seem to be derivative, combining the former practices with one another, or associating them with non-cultural activities.
     Admitting the fundamental character of these three cultural practices, the following question arises: Is philosophy a fourth fundamental cultural activity, on the same level as science, religion, and art, though different from them? Past philosophers have tried to confer to philosophy a status of its own, independent of these activities, maybe above them, but the attempts have been rather unconvincing. Our previous consideration of the protoscientific character of philosophy, about the religious heritage of its concerns, and about the literary aspects of its discourse, leads us to the conclusion that we should be more modest. We should recognize that there are in fact only three fundamental kinds of cultural activities, philosophy being a derived kind, as much in its motivations as in the material it uses and in its methodological procedures. The place of philosophy among the more fundamental cultural activities can be roughly compared with the place of the opera among more fundamental forms of art. Opera is a combination of music, drama, and poetry. Philosophy can similarly be seen as a kind of composite, resulting from a combination of scientific, religious, and artistic elements. And in the same way as poetry is not a strictly necessary element of the opera (unlike music and dramatic action), so the artistic element seems not to be strictly indispensable to philosophy.
     This analogy to the opera, like all analogies, has its limits. Though put together in a suitable way, in order to produce a more impressive result, music, plot, and poetry can easily be separated in the case of opera: We can hear the melody alone of some arias presented in a purely instrumentalized form, or independently read some poetic strophes or a summary of the plot. The same can’t be so easily said of philosophy. Philosophy is not just a collage of elements originating from science, common sense, knowledge, and religion, maybe put together through artistic means. Philosophy is also not a perfect combination, resulting in a completely new product, like a new chemical compound made from other chemical compounds. The metaphor of an amalgam seems to be somewhat more helpful here. In an amalgam, different chemical elements are not simply mixed by chance, nor they are combined to form a new chemical compound, but they are joint in a way that changes their macrophysical properties. A similar idea pertains to philosophical inquiries: They join alien elements in order to provide something new and in some sense unique, yet remain examples of a derived cultural practice, since the resulting unification of elements don’t constitute something intrinsically new.

Trying to replace the foregoing analogies with something more literal, I suggest that philosophy is a derived kind of cultural activity in its motivations, semantic material, and procedures.
     Concerning motivations, it seems that philosophy derives them (a) from inquisitive curiosity bounded by scientific forms of investigation, the wish to acquire a consensual (scientific, concrete, manipulative) knowledge of the nature of things; (b) from religious motivations, wich include the speculative drive, aiming to integrate our experiences and to provide a comprehensive worldview, often by appealing to something beyond the world of ordinary experience, in a way capable of organizing and guiding our intellectual grasp of the world; and (c) from art, in its search for “cathartic effects”.
     Concerning the semantic or conceptual material, the primary data manipulated by philosophy also don’t properly belong to it. (a) Part of this material is the same as the data of our ordinary experience and the data of science. As we saw in chapter IV, in the case of the naturalist concepts of entity-principles (–A+B), these data might be all the data in need of consideration. (b) We also saw that in the case of hybrid metaphysical concepts (+A+B), philosophy can take recourse to theomorphic features (of transcendence, hypermentality, etc.), which were originally present in the constituition of the spiritual beings of religious worship, treating these features as elementary data or indicative of these data. This theomorphic semantic material, as we saw, is already a modification of the material taken from our ordinary (commonsense even scientific) experience, which is then secondarily organized into the semantic constitution of the hybrid concepts of metaphysical principles. (c) Semantic material loaded with emotional suggestivity usually takes part in aesthetic devices.
     Heuristic procedures are also not originally philosophical: (a) The methodological procedures of philosophy are not essentially different from ordinary pre-philosophical reflexive procedures or from the procedures of the formal or empirical sciences. The geometric method of rationalist philosophers (such as Descartes and Spinoza), reflected in the aprioristic way they ground their arguments, mimic the axiomatic procedures of mathematics; the historical method of empiricist philosophers (such as Locke and Hume), based on introspection and empirical information about the world and human behavior, has the same origins as the more accurate procedures attained by the natural and human sciences (also the contemporary analytical devices are not the exclusive property of philosophy). (b) As we saw (IV, 4), philosophical reasoning commonly rests on the assumption of metaphysical principles, which might be represented by the incoherent (+A+B), or the senseless (–A–B), or the merely vague (–A+B) metaphysical concepts (the first two kinds of principles at least, retaining something from the often incoherent or unknowable supernatural beings of religion). While concepts of the form +A+B and –A–B occur mostly in transcendental metaphysics and rationalism, concepts of the form –A+B are more suited to naturalism and empiricism. (c) The works of imagination in the use of rhetorical devices, the construction of similes, etc. are all suggestive resources, able to produce aesthetic effects.
   The columns in the following diagram summarize the main derived features that may belong to philosophical discourse and thought.


A) From    Scientific curiosity     Formal (grasped) or       Use of hypothesis
SCIEN-     about   the   world     empirical       (expe-       and  use  of  argu-
CE            (wish to attain con-    rienced) data                  mentative reason-
                  crete scientific                                                ing   in  order   to
                  knowledge)                                                     justifty them
B) From    Speculative  curio-    Theomorphic  featu-       Recourse to meta-
 RELI-      sity, aspirations of     res,  (transcendence,       physical  intuition
GION        transcendence, se-    hypermentality,  hy-        (of principles, etc.),
                  arch for a  world-     perphysicality, mind-      mystic experience,
                 view ordering the       body idiosyncrasy)        exhortative means
                 intellectual world
C) From    Wish for cathartic     Symbolic-sensory           Application of li-
ARTS       experience,    har-     data, loaded with se-       terary resources
                 monizing the sen-     mantic  suggestivity        of semantic sug-
                 sory-emotional                                                 gestivity

     The diagram makes it clear that, far from being a self-contained cultural activity, philosophy only combines what it appropriates from other fields of human culture. We can interpret the three horizontal levels of the diagram as representing the three possible dimensions of the philosophical inquiry: A) a heuristically oriented dimension of science-anticipative conjectures; B) a mystically oriented dimension, containing ungrounded metaphysical principles, cognitively problematic and generally assumed as a matter of belief; C) an aesthetically oriented dimension, manipulating the medium of the philosophical discourse in order to suggest possibilities and to increase its efficacy.(56) Consideration of these dimensions makes explicit the criterial clusters actually involved in identifying philosophical discourse and thought, from our descriptivist metaphilosophical perspective. Let us review the three levels once more:

     A)  The heuristically oriented dimension. This first dimension is motivated by the operative scientific curiosity. It is constitutively science-anticipative or protoscientific, hence essentially cognitive and heuristic.. This dimension is based mainly on hypothetical generalizations(57), followed by arguments aiming to show what might result from them, and trying to reinforce their plausibility through the consistency of these arguments and the range of the achieved results – a task always accomplished under the assumption of a (real or imaginary) critical community of ideas. In this heuristically oriented dimension, philosophy distinguishes itself from science negatively, by leaving unfulfilled the conditions of shared assumptions, of consensual agreement in truth-evaluation, and of progress as an accumulation of beliefs held true by a critical community of ideas (see III, 8).
     This first dimension is characteristically argumentative and truth-seeking, relying on constative statements. But the following two dimensions are not essentially cognitive, depending essentially on performative utterances: B is primarily exhortative, relying more on truthfulness than on truth, while C is primarily expressive.

     B) The mystically oriented dimension. Speculative curiosity and the drive for comprehensiveness, along with the wish for transcendence, is the motivational element of this dimension of the philosophical inquiry. This dimension contains essentially ultimately non-rational and non-cognitive elements, which inevitably affect all philosophical speculation, particularly those appealing to metaphysical entity-principles of hybrid or of elusive kind. (Using Wittgenstein’s metaphors, the mystic dimension is not of what can be said, but of what can only be shown; being cognitively elusive, certain metaphysical principles can – through the help of what can be said – only be shown.)

     C) The aesthetically oriented dimension. This dimension contains the properly artistic elements, acting expressively and suggesting cognitive possibilities.

     My arguments for the criterial cluster constituting the heuristically oriented dimension were presented in chapter III, and those for the criterial clusters constituting the other two dimensions were presented in chapters IV and V. The question now is: how we can organize these criterial clusters in a way that helps us identify what counts as philosophy in the scholarly sense of the word? My tentative proposal is the following. The presence of the criterial cluster constituting the heuristically-oriented dimension might be considered the primary criterion, namely, a necessary condition for something to be called ‘philosophy’ in the broadest scholarly sense of the word. But is this criterial cluster also a sufficient condition? I suppose not, for scientific curiosity is not the same as speculative curiosity, and the former alone would not lead to the kind of comprehensive conjectural undertaking independent of consensual results, that is usually expected of philosophy. It seems that it is the same drive that in ancestral times led to the formation of religious explanations that now leads us to philosophical speculation. If this is true, then the criterial elements constitutive of the mystically oriented dimension are also necessary for an appropriate philosophical inquiry. About the aesthetically oriented dimension, our considerations about the role of art in philosophy lead us to the suggestion that the artistic element is a secondary criterion, the kind of thing that Wittgenstein also used to call a symptom. Being devoid of content of its own, the artistic element is not a necessary factor, though it is relevant and enriching, like the color added to a picture (a confusion between criteria and symptom seems in this case to have fostered a whole tradition of cultural pathology).
     The variations in the importance of each dimension can be illustrated, if we draw a triangle joining the three fundamental cultural activities, so that philosophy occupies its internal space:

     RELIGION                                                                 SCIENCE



     To philosophy belongs everything inside the triangle. The arrows show that the relationships between the dimensions are actually and historically dynamic. Through time, religious explanations have gradually given way to philosophical ones. And the religious remainders of philosophy have been gradually replaced by forms of inquiry nearer to the patterns of science. As the figure also shows, religious activity and discourse are usually deeply intermingled with artistic expression, which explains why philosophy, particularly in its origins, often remains intermingled with art. Nevertheless, when philosophical inquiry approaches the edge of consensual discourse of science, artistic expression tends to vanish, being replaced by more direct and precise forms of presentation. (Will philosophy be completely replaced by science? The answer to this question depends on the answer to other, presently unanswerable questions, like those about the finitude of possible knowledge.)


                                             COROLLARIES AND PROSPECTS

This chapter suggests some applications of the proposed account of the nature of philosophy. They consist of a more intelligible differentiation between different forms of philosophy and of a new account of the historical succession of the distinctive ways of doing philosophy, including the linguistic-analytic way.

Moved only by a practical cartographic intention, we can classify philosophies according to the place occupied by them in the interior of the triangle drawn at the end of the last chapter. Indeed, philosophical inquiries can be comparatively situated in that space, in accordance with the relative weight of their protoscientific, mystic and aesthetically oriented dimensions. Consider the case of Wittgenstein’s Tractatus Logico-Philosophicus: For its (protoscientific) attempt to build a theory of representation, for its mystical doctrine of the inexpressible, and for its aesthetic structural and rhetoric resources, this work could be placed more or less in the middle of the triangle. However, the most impressive example of a philosophical system to be situated in the middle of the triangle is that of Plato’s writings. His philosophy has a proper protoscientific, cognitive, theoretical side, to be found in the essentially argumentative nature of his writings, at whose center – the doctrine of ideas – ontological truth is pursued in connection with a theory about our cognitive capacities, along with moral and social concerns. But Plato’s philosophy also has a mystical dimension, recognizable in his attempt to create a speculative worldview and, more obviously, in his appeal to the Orphic myths, in his doctrine of the world-soul and in his fast religious worship of the form of good. And there is also the aesthetical element, turning his dialogues into literary works of great beauty and permanent appeal. Indeed, achieving an ideal balance between the three considered dimensions, Plato’s work remains a paradigmatic example of a classical philosophical endeavor on the verge of perfection (Plato is the Mozart of the philosophical opera).
     Nevertheless, the role of these different dimensions is seldom so equitably distributed. There are borderline philosophies, to be placed near a border or a corner of the triangle. Aristotle’s philosophy was, by his motivations and achievements, nearer to the scientific corner than Plato’s philosophy, and many important specialized philosophers of our scientific times – I think of Frege, Carnap, and Quine, not to mention Russell and the empiricist tradition – do a kind of work that could be placed near the scientific corner of the triangle (which is to a certain extent expected, since philosophy seems gradually to approach the scientific edge). On the other side, philosophies like those of Augustine and Plotin are, mainly because of their motivations, to be placed nearer to the mystical/religious corner (Augustine’s philosophy also possesses an appealing literary dimension). A pre-Socratic philosopher like Heraclitus, and also writers like Montaigne and Nietzsche, are examples of “artist philosophers”, whose work would be placed near to the artistic corner of the triangle. And Kierkegaard’s philosophy would be placed near the artistic/mystic side of the triangle. There are also cases, which are really borderline: Novalis and Cioran are as much artists as philosophers. These borderline cases are to be distinguished from those cases of artists working in proximity to philosophical frontiers, like Hölderlin and Goethe. Borderline cases on the edge of religious thought are, for example, the mystic doctrines of J. Böhme or Meister Eckhart (Eckhart’s religious Sermons are embedded in deep philosophical insights into human psychology). And there are, of course, borderline cases to be placed between philosophy and science. Consider, for example, Freud’s psychoanalysis: While still dependent on non-consensual subjective interpretations, psychoanalysis allows for insights unattainable through unaided introspective methods. Yet another example of borderline work is given, I think, by the imaginative anthropological writings of Claude Lévi-Strauss: They satisfy a philosophical, an aesthetic, and also a modest scientific ambition.
     One can also classify whole philosophical movements and even traditions according to their relative places in the triangle. Contemporary North American philosophy, for example, is typically influenced by science; it is often a naturalist undertaking, seeking eagerly, even if deceptively, to reproduce the standards of clarity, rigor, and objectivity exhibited by science. German philosophy is in its way of operation nearer to the mystic/religious corner: Historically it possesses a heavy mystical accent, which underlines its elusive discourse and the supposed depth of its metaphysical obscurities. French philosophy has been more and more influenced by an ideal of artistic expression, being centered in the (unfortunately inessential) aesthetic dimension. As a consequence of this, it turns to be mostly a rhetorical game without a serious heuristic commitment, where argument is only vaguely suggested. The insistence on doing this results in a literary persiflage of the real work of philosophy: Like a child playing with a toy as if it were the real thing, the French philosopher feigns doing philosophy.
     As theoretically worthless as the present cartographical exercise might be, it seems to bring some order to the botch of philosophical forms. Moreover, it shows the universal applicability of the proposed integrative account, even if it is still in need of a more complete development.

As would be expected, the relationships between post-religious and protoscientific elements change with the emergence of science. The consequence of this is that the whole historical development of philosophy can be viewed from the perspective of changes in the dynamic relationship between philosophy and science. This view leads us to divide the history of philosophy into three major periods, according to philosophy’s relationship to science.
     In the beginning, there was only religion, art, no idea of science, and, consequently, no room for philosophy. The first period in the development of philosophy, which might be called preformational, since it precedes the formation of the fundamental sciences as systematic bodies of knowledge, began with the early Greek philosophers. As we already noted (IV, 1), the emergence of philosophy became possible, not only because of dissatisfaction with mythological explanations, but essentially as an aftermath of the idea of science. It was the idea of science, the spark that lit the flame of philosophical speculation. The fragmentary rise of the first scientific theorizations brought within it the idea that problems (in arithmetic, geometry, physics, astronomy), whose solutions were hidden to us, could be answered through generalizations in the form of scientific hypotheses (theorems, laws), made in abstraction of their practical application. Now, this awareness must have suggested to the human mind the splendid idea that possibly the whole world, whose hidden nature was previously explained by religion, could be explained through similar kinds of generalizations. Though this undertaking was completely impossible as a matter of fact, it was always possible as a matter of conjecture or speculation, and this was precisely what the first philosophers, who were also scientists, tried to do. In doing this, these philosophers often mixed their speculation with the old anthropomorphic explanations, but we have seen that even this was not an indispensable element of the philosophical endeavor.
     This first period of philosophical inquiry endured until the end of Middle Ages. During all this time, though developing new dialectical procedures and always being guided by the idea, originally suggested by science, of explaining the hidden nature of things by means of conceptual generalizations, philosophy was not in need of maintaining any dialogue with science, since the latter still remained in too rough and fragmentary a form to be able to challenge our ordinary views of the world.
     The second, paraformational phase of philosophy, was marked by the emergence of the fundamental empirical sciences – physics, chemistry, biology, psychology, and social science – as systematic bodies of knowledge, along with parallel developments in the formal sciences (see III, 3). This phase arose with Descartes and flourished at least until Hegel. After Descartes, philosophy was in considerable measure a reaction to the growth of science: It was mostly a response to this growth in the form of a conjectural accommodation to the rest of our philosophical worldview. The philosophical task was not so much to disclose new scientific fields, but mainly to produce a reformulation and relocation of the material of ideas belonging to philosophy’s remaining fields in conformity with new scientific ideas, formal and empirical. Taking Descarte’s metaphysics as an example, it is difficult to see how he could believe in the fruitfulness of the geometric method in philosophy, without witnessing its heuristic power in science, and it is difficult to see how he would feel the need to answer the skeptic in the way he did, if he were not conscious of the argument of illusion in its scientific form, or that our souls are not in our chests.
     Finally, we come to what might be called the postformational phase of the development of philosophy, after the emergence of the fundamental sciences. These sciences required certain order of development, from physics to social science, since it is difficult to imagine a more complex and particular fundamental science being developed before a more general one. These days many localized scientific developments emerge, which require the prior existence of the fundamental sciences, since they apply them. Contemporary philosophy is, more than ever, a participation in the origination of these developments, and it is also a response to them, as we philosophically adjust our remaining related philosophical views to the new perspectives created by these developments.

The consideration of the last phase of the development of philosophy under the proposed perspective helps us to understand why philosophy in the twentieth century has been mistakenly considered solely matter of linguistic-conceptual analysis. One reason for this seems to be that, as the emergence of fundamental empirical sciences took the places once occupied by philosophy as anticipatory empirical speculation, philosophy was mostly reduced to second-order kind of inquiry, more reflexive and detached from direct empirical concerns. However, the central reason for the consolidation of analytical philosophy was the development of new methodological devices to control argumentative procedures, requiring metalinguistic consideration of the linguistic-conceptual element. Under such circumstances, it was easy to mistake philosophy for a purely linguistic-conceptual endeavor. However, the detachment of philosophical speculation from more direct empirical concerns and the linguistic-conceptual accent are historically contingent facts. To say that philosophy in the twentieth century has been mostly a linguistic-conceptual inquiry is to describe the form philosophy has taken in a certain historical period, rather than to diagnose its proper nature.
     If we adopt this perspective, we can arrive at a better understanding of the internal development of the linguistic-analytic philosophy in this century. My suggestion is that we can understand the main achievements of linguistic-analytic philosophy as intrinsically bound to the development of a general scientific constellation that might be called (in the broadest conceivable sense) semiotics. Linguistic-analytic philosophy is bound to semiotics on one side by the proper conjectural way by which the semiotic field has been gradually disclosed, and on the other by a reformulating and relocating of our views of traditional philosophical problems, resulting from the achievements in the semiotic field.
     In order to argue for this suggestion, I need first to explain briefly what I mean by the word ‘semiotics’. Calling ‘signs’ all things that are used to represent other things, semiotics is the name we can give to the vague and all-encompassing idea of a general science of signs. The projected scientific field of semiotics is usually divided in three great domains(58): syntactics, semantics, and pragmatics. The first, syntactics, consists of the investigation of the rules combining signs and combinations of signs, in a way that essentially includes the logical syntax. The second domain of semiotics is semantics, understood as the investigation of the rules relating the signs (and their combinations) to their designata. The third domain is pragmatics, understood as the investigation of the rules relating signs (and their combinations, along with the relations they have to their designata) to their speakers and circumstances through the concrete use of language. It is easy to see that there is a certain order of presuppositions here: In a general way, semantics presupposes syntactics and pragmatics presupposes both syntactics and semantics. This becomes clear when we consider that one can investigate the syntactics of a language in abstraction of its semantics, but one can hardly investigate the referential relations of a language without being able to identify its syntactic unities; and one can understand  syntactics and semantics without considering the use of language in the concrete circumstances of human communication, but one can’t make full sense of the use of signs by speakers when one doesn’t know their possible syntactic articulation or their references (see II, 2). The consequence of this is that it is natural to expect that the historical development of semiotic knowledge follows the same order, beginning with syntactics and going further with semantics and pragmatics.
     Now, these considerations help us understand the historical development of analytic philosophy in the twentieth century. Indeed, analytic philosophy appeared in three successive waves of investigation. The first wave was mainly a syntactic one. At the end of the last century, Gottlob Frege developed for the first time a complete symbolic formulation of the predicative calculus. This was an essentially syntactic contribution of unprecedented importance in the development of logic, so valuable that it would be no exaggeration to say that logic as a science began with Frege. We can consider the logical atomism of Bertrand Russell and of the early Wittgenstein in the Tractatus as the most important philosophical responses to this, trying to accommodate our remaining views in philosophy of content and theory of knowledge to this fundamental development of logical science. Though there were already semantic developments – like Frege’s distinction between sense and reference, the elusive ontology of the Tractatus, and Russell’s speculations about the designata of logically proper names – they remained a complementary and mostly schematic endeavor.
     The second wave was mainly a semantic one: Wittgenstein, in the intermediary phase of his philosophical development, followed by logical positivists like Moritz Schlick and Rudolf Carnap, suggested a full-blooded philosophical semantic, mainly in form of the verifiability principle, whose consequences for the reformulation of our remaining philosophical worldview were paradigmatically developed in A. J. Ayer’s book, Language, Truth and Logic.
     The third wave focused on the efforts directed to creating a science of pragmatics, and accommodating other philosophical problems to its results. It began with the scattered reflections of the late Wittgenstein about the meaning of expressions as equivalent to their use in the context of language-games. But the emergence of pragmatics as a systematic reflection on communicative actions was due to the efforts of J. L. Austin and was later carried further by J. R. Searle in the theory of speech-acts. Pragmatic investigations also led to attempts to accommodate old philosophical problems. An early example of this was the restructuring and relocation of the mind-body problem – even if in an ultimately misguided form – as a result of pragmatic reflections about a necessarily interpersonal moment in the learning of language. Part of this attempt can be observed in the behaviorist analysis of mental concepts made by Gilbert Ryle in The Concept of Mind, and also in the work of the late Wittgenstein, for example, in his argument against the possibility of a private language and in his attempt to escape its paradoxical consequence – the rejection of our mentalist language – through a doctrine of criterial expression. Another attempt to reformulate philosophical problems emerging as an application of pragmatic developments, especially though not entirely of the theory of speech acts, is found in Jürgen Habermas’ attempt to read basic social structures (and their possible distortions) in our forms of communication. Once more, we see the role of the linguistic-conceptual accent as a historically contingent and non-essential feature of philosophy.

What can be expected in the future? Certainly, we may expect that someday the present philosophies of science will be transformed into metasciences through the achievement of adequate consensus about their explanations of what constitutes the more basic sciences. But our greater expectations are directed to the central core of philosophical problems, which seems to remain as distant as ever from scientific agreement. Domains of speculation like metaphysics and epistemology are all-embracing, since metaphysics must explain, independently of the science to which they belong, the ultimate categories of reality, which are constitutive of all objects of knowledge, and epistemology must take into account all forms of scientific and non-scientific knowledge. Though not so comprehensive, ethics seems to be integrated in a very complex form into the center of human social activity, therefore also requiring the same kind of conjectural argumentative approach.
     The global account of the nature of philosophy outlined in this book gives us some clues in thinking – though very cautiously – about the future of philosophy as a whole. If philosophy is an intermediary or derived cultural activity, stability is not to be expected. To see this, take again the example of the opera: It grew up with the development of music, after the Renaissance, arriving at its highest development in the eighteenth and nineteenth centuries, only to lose almost completely its importance in our century, yet still surviving in its minor forms, like the operetta and the musical. Probably something similar can be said about philosophy, at least in the classic and predominant sense of the word: The big times are gone. They originally belonged to the great systems of Plato and Aristotle, and in modernity, to the period of the configuration of the fundamental sciences, when philosophy was in great measure an accommodation of remaining philosophical views to these revolutionary achievements; this intellectual movement culminated in Kant’s work. Today, philosophical inquiry, mostly restricted to a core of questioning that in its nature remains highly dependent on argument, seems to progress in its minor forms. Nevertheless, it is good to remember that this diagnosis is not necessarily a value judgment, for contemporary philosophy can be at times important and even fascinating (the Drei Groschen Opera is a minor form of opera, but not less interesting than Die Götterdämmerung, at least for those who refuse to be impressed by Wagnerian pathos).
     Indeed, in our times, science has grown so much that it has taken the place of much of philosophy, although, one could remark, very little from its most important domains. Nevertheless, adopting the tolerant view of science proposed in this book, there seems to be no strong reason to deny that sometime in the future, science will take over the whole field of philosophy. This will not occur if there is no limit to the creation of new philosophical questions, if philosophical problems are self-multiplicative, if there are limits on the possibility of consensus. In this case, philosophical speculation will always exist. However, since what we have experienced until now is only a sequence of subdivisions and applied combinations of the fundamental sciences, there is some reason to expect that there is a limit to the acquisition of essential scientific knowledge. If so, then it might be that one day philosophers and scientists will find themselves jobless in an intellectually satiated world, where all things worthy to know will already be known, and no important discovery remains to be made. At this time, there will remain no place for philosophical accommodation of the rest of our philosophical worldview to science, since there will be no “rest” of our worldview at all: The sum of scientific knowledge will be our worldview, nothing more being allowed, since a search for the whole beyond this knowledge would be clearly recognized as an unnecessary and conceptually senseless endeavor.
     Indeed, if the observations made here are correct, if the emergence of new scientific fields is not an indefinitely self-multiplicative possibility, it is not difficult to foresee that, when the dust of the conceptual confusion that has arisen and continues to arise through the formation of all the new ramifications of science has fallen, a day will come in which philosophy, even in its minor forms, will come to an end. However, this does not mean that central philosophical conjectures will be replaced by a multiplicity of narrowly focused, unrelated, unexciting scientific theories – as the positivist-scientist fragmentation of the field of experience suggests – provided that our open concept of science allows very comprehensive scientific realizations to take the place of our central philosophical conjectures, preserving in this way the suspected worth of their questions.


1 R. Carnap, “On the Character of Philosophical Problems” in, R. Rorty (ed.) The Linguistic Turn, p. 54.
2 Cf. M. Inwood, A Heidegger Dictionary, p. 164.
3 L. Wittgenstein, Philosophical Investigations, sec. 109.
4 C. D. Broad, Scientific Thought, p. 20; see also B. Blanchard, On Philosophical Style, p. 6.
5 G. E. Moore, “What is Philosophy?”, p. 23.
6 E. Tugendhat, “Die Philosophie unter sprachanalytischen Sicht” in, Philosophische Aufsätze.
7 W. V. O. Quine, Word and Object, p. 270 f.
8 W. V. O. Quine, “A Letter to Mr. Ostermann”.
9 See, for example, G. Ryle, “Systematic Misleading Expressions”.
10 The classical ordinary language criticism of the argument of illusion is found in Austin’s Sense and Sensibilia. A very sharp though schematic criticism of the same argument can be found in J. R. Searle’s book Language, Mind and Society: Philosophy in the Real World, chap. I, p. 28 f.
11 L. Wittgenstein, Philosophical Investigations, sec. 109, 118, 119... See also A. Kenny (ed.), The Wittgenstein Reader, pp. 263-285.
12 I prefer to think that Wittgenstein was speaking of his personal, minimalist way of working with philosophy, rather than proposing the proper method of philosophizing. That he also held a different and maybe incompatible view has been pointed out by his best interpreters (see A. Kenny, “Wittgenstein and the Nature of Philosophy”).
13 L. Wittgenstein, The Blue Book, pp. 17-18.
14 So writes A. J. Ayer about Wittgenstein’s therapeutic method, “His repeated preference for description over explanation and the avoidance of theory which he claimed to practice and enjoined upon his readers are not characteristic of his actual procedure at any stage of his development, including that of the Philosophical Investigations. That his explanations are runic does not reduce them to descriptions: his theories do not cease to be such by being covertly set out.” (A. J. Ayer, Ludwig Wittgenstein, p. 137.)
15 L. Wittgenstein, Philosophical Investigations, sec. 79.
16 J. R. Searle, “Proper Names”. The same results we find so clearly stated in Searle’s paper can be also inferred from a careful reading of Wittgenstein’s text.
17 R. Carnap, Logische Syntax der Sprache, part I.
18 J. R. Searle, Mind, Language and Society: Philosophy in the Real World, p. 138.
19 So wrote L. Wittgenstein: “a main source of our failure to understand is that we do not command a clear view of the use of our words. – Our grammar is lacking in this sort of perspicuity. A perspicuous representation produces just that understanding which consists in ‘seeing connexions’. Hence the importance of finding and inventing intermediate cases” (Philosophical Investigations, sec. 122). On the concept of perspicuous representation (übersichtliche Darstellung), see the interesting comments of G. P. Baker & P. M. S. Haker in, Wittgenstein: Understanding and Meaning, p. 489.
20 See E. Tugendhat, “Die Philosophie unter den Sprachanalytischen Sicht“, in his Philosophische Aufsätze.
21 W. V. O. Quine: World and Object, p. 270 f.
22  W. V. O. Quine, Word and Object, pp. 271-272. Quine saw correctly that a formal way of speaking can’t be used to identify philosophy, since it can be universally applied. For this reason, he rejects Carnap’s thesis that the possibility of translation into a formal way of speaking can be used as a way of distinguishing the questions belonging to philosophy, choosing the expression ‘semantic accent’ to mark his own naturalist perspective.
23 Kai Nielsen underlines the obvious but remarkable fact that when philosophers describe the uses of our expressions, “they are making empirical remarks about how language works.” (“What is Philosophy?” in, History of Philosophy Quarterly, 10, 1993, pp. 389-390).
24 See D. M. Armstrong, What is a Law of Nature.
25 A. J. Ayer in interview with B. Magee (B. Magee, Men of Ideas, p. 127). Magee’s objection to this remark of Ayer and to similar remarks of J. R. Searle – an objection answered here in a more systematic way – is that the analytical inquiry, like any metalinguistic inquiry, unavoidably takes the real world away (See B. Magee in Confessions of a Philosopher, pp. 74-76).
26 A specimen of this is B. Latour & S. Woolgar, Laboratory Life: The Construction of Scientific Facts.
27 A. Kenny, Aquinas on Mind, cap. 1, p. 4.
28 J. L. Austin, Philosophical Papers, p. 232.
29 See A. Comte, Cours de Philosophie Positive, Oevres, vol. I. I don’t follow his classification in detail, since Comte makes at least two obvious mistakes: the inclusion of astronomy under the fundamental sciences and the exclusion of psychology, which was non-existent as science in his time. The principles of classification, however, remain valid.
30 One could object that the very idea of an epistemological rupture distinguishing science from pre-science is misleading, since the usual criteria of scientificity really don’t allow us to identify such ruptures. I agree with this. But I also claim that we find no difficulty in identifying these ruptures intuitively and that the criteria of scientificity suggested in section eight of the present chapter is able to rescue this intuition, allowing for a clear identification of epistemological ruptures. Indeed, the epistemological rupture occurs when truth in a whole domain of investigation becomes consensually attainable.
31 As J. R. Searle has noted, it is an error to believe that because objects of inner experience have an ontologically subjective mode of existence, they must also be epistemologically subjective, hindering the ways of science (see his Mind, Language and Society: Philosophy in the Real World, pp. 43-45).
32 G. S. Kirk, J. E. Raven and M. Schofield (eds.), The Presocratic Philosophers, pp. 133-134.
33  K. Popper, “Back to the Pre-Socratics” in his Conjectures and Refutations, p. 138.
34 G. S. Kirk, J. E. Raven and M. Schonfield (eds.), The Presocratic Philosophers, pp. 140-142.
35 See discussion in W. K. C. Guthrie, A History of Greek Philosophy, vol. I, p. 103.
36 S. Freud, The Ego and the Id.
37 Aquinas on Mind, pp. 4-5.
38 K. Lehrer, Theory of Knowledge, p. 7. See also W. James, Some Problems of Philosophy, p. 23.
39 Aquinas on Mind, p. 5.
40 Aquinas on Mind, p. 9. I agree with Kenny’s motivation, though not with his conclusion. My aim is to show that to believe that the progressive thesis endangers the comprehensiveness of philosophy is to confuse the nature of the scientific answers (i.e., consensually obtainable answers) eventually destined to replace the core issues of philosophy, which we don’t know, with the current endeavor of the particular sciences, which we already know.
41 See J. Passmore, “Philosophy”, in P. Edwards, The Encyclopedia of Philosophy, vol. VI, pp. 219-20.
42 See K. R. Popper, Conjectures and Refutations, pp. 339-340. The standard example of decisive falsification used by Popper is the deflection of the light of the stars observed in the eclipse of 1919. Ironically, precisely this kind of test would later be considered too unreliable to be significant when taken in isolation.
43 See K. R. Popper, The Logic of Scientific Inquiry, chap. II
44 “What is Science?”, p. 42. Science, as a corpus of knowledge, as what scientists do and as an institution, says J. Ziman, can’t be treated separately, more than a solid can be reconstructed from its projection upon separate Cartesian planes (ibid. p. 42).
45 See J. Habermas, “Wahrheitstheorien”. By adopting this idea and by calling my characterization of science ‘consensualist’, I’m in no way suggesting that science is a matter of some kind of arbitrary consensual decision. Our collective experience has shown that it is only because facts, which we conceive as independent of us, can be matched by our propositions, that we are able to achieve interpersonal agreement about the truth-value of these propositions within a critical community of ideas.
46 Cf. G. Reale, A History of Ancient Philosophy, vol. I, p. 14.
47 Guthrie, W. K. C., A History of Greek Philosophy, vol. I, p. 36 f.
48 The law was first suggested by A. Turgot in his Réflexions sur la Formation et la Distribuition des Richesses (1750), but only Comte developed it in all its implications. From A. Comte, see Cours de Philosophie Positive, Oevres, Paris 1968 (1830-1842), vol. I; see also, Discours sur L’esprit Positif, Oevres, Paris 1968 (1844), vol. XI, p. 2 f.
49 Comte’s development of the so-called law of three stages has often been misunderstood, I think under the influence of prejudice. Its plausibility is defended by W. Schmaus in, “A Reappraisal of Comte’s Three-State Law”. See also C. F. Costa, “Filosofia, Ciência e História”.
50 J. Habermas, Erkenntnis und Interesse, p. 92.
51  See K. R. Popper, The Poverty of Historicism, chap. IV.
52 Though Democritus never said this, the conclusion is difficult to avoid, given the internal special properties he attributes to atoms (for a discussion, see W. K. C. Gutthrie, A History of Greek Philosophy, vol. II, p. 396).
53 “Clarity is not Enough” in, H. D. Lewis (ed.), Clarity is not Enough, p. 40.
54 “Philosophy as Art”, Metaphilosophy 14, n. 2, 1983, p. 141. See also J. Deleuze and F. Guattari in, Qu’est-ce que la Philosophie? J. H. Gill tries to confirm his proposal historically, by showing the central role of aesthetic metaphors in the great philosophical systems, but the meager result suggests rather the opposite conclusion (see J. H. Gill, Metaphilosophy, chap. 6).
55 See S. Freud, Traumdeutung, chap. VII.
56 C. F. Costa, “A Conjectura Filosófica”, p. 29 f.
57 When I say “hypothetical generalization”, I’m not denying that the philosopher usually comes to such a generalization a posteriori, relying on previous arguments and examples. My point is that there is always a “jump” to the generalization, which then needs further argumentative confirmation or disconfirmation, in a way essentially analogous to the hypothetic-deductive procedure in science.
58 See (for example) C. W. Morris, Foundations of a Theory of Signs.


Armstrong, D. M., What is a Law of Nature, Cambridge 1983.
Austin, J. L., Philosophical Papers, Oxford 1961.
- Sense and Sensibilia, reconstructed from manuscript lecture notes by G. J. Warnock, Oxford 1962.
Ayer, J. L., Language, Truth and Logic, London 1936.
- “Metaphysics and Common Sense” in, Metaphysics and Common Sense, London 1969.
- Ludwig Wittgenstein, New York 1985.
Baker, G. P. & Haker, P. M. S., Wittgenstein: Understanding and Meaning, Oxford 1980.
Berlin, I., “The Purpose of Philosophy” in, Concepts and Categories: Philosophical Essays, (ed.) H. Hardy, Princeton 1998.
Blanchard, B., On Philosophical Style, New York 1969.
- “The Philosophic Enterprise” in, C. J. Bontempo and S. J. Odell (eds.), The Owl of Minerva, New York 1975.
Bobik, J. (ed.), The Nature of Philosophical Inquiry, Notre Dame 1969.
Bontempo, C. J. & S. J. Odell (eds.), The Owl of Minerva, New York 1975.
Broad, C. D., “Critical and Speculative Philosophy” in, J. H. Muirhead (ed.), Contemporary British Philosophy, First Series, London 1924.
- “Two Lectures on the Nature of Philosophy” in, Inquiry 1, n. 2, 1958.
- Scientific Thought, Peterson 1959.
Carnap, R., “On the Character of Philosophical Problems” in, R. Rorty (ed.), The Linguistic Turn: Recent Essays in Philosophical Method, Chicago 1967.
- “Überwindung der Metaphysik durch die logische Analyse der Sprache“ in, Erkenntnis 2, 1931-32 (translated by A. Pap as “The Elimination of Metaphysics Through Logical Analysis of Language” in, A. J. Ayer (ed.), Logical Positivism , Illinois 1959).
- Logische Syntax der Sprache, Vienna 1934.
Chappell, V. C. (ed.), Ordinary Language: Essays in Philosophical Method, E. Cliffs, 1964.
Comte, A., Cours de Philosophie Positive in, Oevres, Paris 1968 (1830-1842), vol. I.
- Discours sur L’esprit Positif in, Oevres, Paris 1968 (1844), vol. XI.
Copleston, F. C., Religion and Philosophy, Dublin 1974.
Costa, C. F., Filosofia Analítica, Rio de Janeiro 1992.
- “Wittgenstein: a Gramática do Conceito de Significado” in, C. F. Costa, A Linguagem Factual, Rio de Janeiro 1996.
- “A Conjectura Filosófica” in, C. F. Costa, A Linguagem Factual, Rio de Janeiro 1996.
- “Filosofia, Ciência e História” in, C. F. Costa, Estudos Filosóficos, Rio de Janeiro 1999.
Deleuze, J. & F. Guattari, Qu’est-ce que la Philosophie?, Paris 1991.
Diels, H. & W. Kranz (eds.), Fragmente der Vorsokratiker, Berlin 1961.
Dummett, M., Frege: Philosophy of Language, London 1981 (2nd ed.).
- “Can Analytic Philosophy be Systematic and Ought it to Be?” in, M. Dummett, Truth and Other Enigmas, London 1978.
- The Origins of Analytic Philosophy, London 1993.
Edwards, P., “Heidegger’s Quest for Being” in, Philosophy 64, 1989.
Erickson, G. W., Negative Dialectics and the End of Philosophy, Wolfeboro 1990.
Freud, S., Die Traumdeutung, Frankfurt 1987 (1900).
- The Ego and the Id, (Das Ich und das Es), (ed.) J. Strachey, 1960.
Gill, J. H., “Philosophy as Art” in, Metaphilosophy 14, n. 2, April 1983.
Metaphilosophy, Washington 1982.
Glock, H-J. (ed.), The Rise of Analytical Philosophy, Oxford 1997.
Goodman, N., The Languages of Arts, Indianapolis 1968.
Guthrie, W. K. C., A History of Greek Philosophy, Cambridge 1962 (4 vols.).
Habermas, J., Erkenntnis und Interesse, Frankfurt 1973.
- “Wahrheitstheorien” in, H. Fahrenbach (ed.), Wirklichkeit und Reflexion, Frankfurt 1973.
Hook, S., “Scientific Knowledge and Philosophical ‘Knowledge’” in, The Quest for Being, New York 1991.
Husserl, E., “Philosophy as Rigorous Science” in, Q. Lauer (ed.), Phenomenology and the Crisis of Philosophy, N. York 1965.
Hussey, E., The Presocratics, London 1972.
Inwood, M., A Heidegger Dictionary, Oxford 1999.
James, W., Some Problems in Philosophy, Lincoln 1996 (1911).
Kant, E., Kritik des Urteils, Frankfurt 1976.
Kenny, A., Aquinas on Mind, London 1993.
- The Wittgenstein Reader, Oxford 1994.
- “Wittgenstein and the Nature of Philosophy” in, A. Kenny, The Legacy of Wittgenstein, Oxford 1984.
Latour, B. & S. Woolgar, Laboratory Life: The Construction of Scientific Facts, Princeton 1986.
Lehrer, K., Theory of Knowledge, San Francisco 1990.
Magee, B., Men of Ideas, New York 1978.
- Confessions of a Philosopher, New York 1999.
Moore, G. E., “What is Philosophy?” in, Some Main Problems of Philosophy, London 1953.
Morris, C., Foundations of the Theory of Signs, Chicago 1970 (1938).
Nielsen, K.,  “What is Philosophy?” in, History of Philosophy Quarterly 10, n. 4, 1993.
Passmore, J., “Philosophy” in, Paul Edwards (ed.), The Encyclopedia of Philosophy, New York 1967, vol. 6.
Plato, Collected Dialogues, ed. by E. Hamilton and H. Cairns, Princeton 1963.
Popper, K. R., The Logic of Scientific Discovery, London 1959.
- Conjectures and Refutations, N. York 1962.
-  The Poverty of Historicism, N. York 1997 (1957).
Price, H. H., ‘Clarity is not Enough’ in, H. D. Lewis (ed.), Clarity is not Enough: Essays in Criticism of Linguistic Philosophy, London 1963.
Quine, W. V. O., Word and Object, Cambridge 1960.
- “A Letter to Mr. Ostermann” in, C. J. Bontempo & S. J. Odell (eds.), The Owl of Minerva, New York 1975.
- “Epistemology Naturalized” in, H. Kornblith (ed.), Naturalizing Epistemology, Cambridge (MIT) 1994.
Quinton, A., “Philosophy” in, T. Honderich, The Oxford Companion to Philosophy, Oxford 1995.
Reichembach, H., The Rise of Scientific Philosophy, Berkeley 1968.
Reale, G., A History of Ancient Philosophy (translated by J. R. Catan), New York 1987 (4 vols).
Rorty, R. (ed.), The Linguistic Turn: Recent Essays in Philosophical Method, Chicago 1967.
Russell, B., Our Knowledge of the External World, New York 1929 (1914).
Ryle, G., “Systematic Misleading Expressions” in, Proceedings of the Aristotelian Society, 32, 1931-2.
Searle, J., “Proper Names” in, Mind 67, 1958.
- Mind, Language and Society: Philosophy in the Real World, New York 1998.
Scharf, R. S., Comte After Positivism, Cambridge 1995.
Schiller, F. C. S., “Must Philosophers Disagree?” in his, Must Philosophers Disagree? And other Essays in Popular Philosophy, London 1934.
Schlick, M., “The Future of Philosophy” in, R. Rorthy (ed.), The Linguistic Turn: Recent Essays in Philosophical Method, Chicago 1967.
Schmaus, W., “A Reappraisal of Comte’s Three-State Law” in, History and Theory 21 (2), 1982.
Smart, J. J. C., “The Province of Philosophy” in his, Philosophy and Scientific Realism, London 1963.
- “My Semantic Ascents and Descents” in, C. J. Bontempo & S. J. Odell (eds.), The Owl of Minerva, New York 1975.
Strawson, P. F., “Metaphysics” in, D. F. Pears (ed.), The Nature of Metaphysics, London 1957.
- Individuals, London 1959.
- “Analysis, Science and Metaphysics” in, R. Rorthy (ed.), The Linguistic Turn:Recent Essays in Philosophical Method, Chicago 1967.
- Analysis and Metaphysics: an Introduction to Philosophy, Oxford 1992.
Tugendhat, E., Vorlesungen zur Einführung in die Sprachanalytische Philosophy, Frankfurt 1976.
- “Überlegungen zur Methode der Philosophie aus Analytischer Sicht“, in Philosophische Aufsätze, Frankfurt 1992.
Turgot, A-R-J., Réflexions sur la Formation et la Distribuition des Richesses, Heidelberg 1913 (1750).
Waismann, F., How I See Philosophy? New York 1968.
Wittgenstein, L., Tractatus Logico-Philosophicus, edited and translated to English by D. F. Pears and B. F. McGuinness, London 1961.
- The Blue and Brown Books: Preliminary Studies for the Philosophical Investigations, Oxford 1958.
- Philosophische Untersuchungen, Frankfurt 1976, edited and translated to English by G. E. M. Anscombe as Philosophical Investigations, Oxford 1953.
Wright, G. H. v., “Analytic Philosophy: A Historic-Critical Survey” in his, The Tree of Knowledge and Other Essays, Leiden 1993.
Ziman, J. M., Public Knowledge: An Essay concerning the social Dimension of Science, London 1968.