1. Edmund Husserl's Dream of ‘Philosophy as a Rigorous Science’

It is difficult to imagine that a Royal Institute of Physics would organize an annual lecture series on the theme ‘conceptions of physics’. Similarly, it is quite improbable that a Royal Institute of Astronomy would even contemplate inviting speakers for a lecture series called ‘conceptions of astronomy’. What, then, is so special about philosophy that the theme of this lecture series does not appear to be altogether outlandish? Is it, perhaps, that philosophy is the reflective discipline par excellence, so that the very nature of philosophy is a topic that belongs to its domain of investigations? Or is there something more serious at issue? Let me first tell you how the question concerning the nature of philosophy arose for me as a young student, and how I tried to answer it, quite naively, by endorsing Edmund Husserl's programme of turning philosophy into a rigorous science. In a second section, I shall offer three diagnoses of the fact that all such attempts seem to have failed. Finally, I'd like to say a few words on my own research agenda in philosophy and on a public function of philosophers.

When I studied philosophy on the European Continent during the years 1969–1976, in Leiden, Paris, and Cologne, the dominant school of thought still was the phenomenological movement, inaugurated by Edmund Husserl. Yet there was no lack of rivalry with other traditions. In The Netherlands, we also read Oxford philosophers, such as Austin, Dummett, Ryle, Strawson, or Hare, and Americans such as Quine or Davidson. We engaged in the fierce debates on the ‘open society’ between Karl Popper and the Frankfurter Schule, while the Continental – Analytic split was often discussed. In Paris, two triumvirates were adored by the masses of students of La Sorbonne (Paris I, Paris IV). One was the ‘holy trinity of suspicion’, consisting of Marx, Nietzsche, and Freud. However, many of the more self-consciously serious students preferred the other trinity: ‘les trois H's’: Hegel, Husserl, Heidegger.

You will be able to imagine the abyss of intellectual confusion young students plunged into, when they eagerly read works of all these different philosophers, and attended lectures as diverse as those by Michel Foucault at the Collège de France, Emmanuel Levinas at Paris IV, or Jacques Derrida's interminable ruminations at the École Normale Supérieure. What was it all about? One was supposed to study philosophy, but was there a unified discipline bearing that name? When I moved from Oxford, where I had spent a Hilary Term during the winter of discontent (1973–1974), to Paris, it seemed to me that I had migrated from one intellectual universe to another. Whereas in Oxford one had to write essays and argue with one's tutor about quite precise philosophical problems, what counted in Paris was a vast intellectual erudition. Students were given the impression that one was without intellectual worth as long as one had not read all great philosophers from Plato to the present.

By hindsight I think that this type of methodological pluralism and academic peregrination produced a mental independence that perhaps one could not have acquired otherwise. As was the case in ancient Greece, critical thought often starts with travelling and making the discovery that different ideas about what counts as self-evident dominate the intellectual scene elsewhere. But at the time, the state of intellectual bewilderment and confusion I suffered from merely raised urgent questions to which I did not have an answer. Is philosophy a unified field of investigation? Do philosophers have at their disposal sound methods by means of which they can acquire real knowledge? Should one adopt the descriptive methods of phenomenology and endorse Husserlian Wesensschau, or was the austere analysis of language à la Austin the only – and quite modest – thing philosophers could achieve? Edmund Husserl once formulated the question as follows: can philosophy be a rigorous science (in the broad sense of Wissenschaft)? Or should we rather conclude that what is called ‘philosophy’ is a loosely knotted tapestry of opinions and schools, nothing but a contest of world views (Weltanschauungen), with no consensus about methods allowing us to decide who is right and who is wrong?

In Paris one could not escape the impression that the second conclusion was apt. I once stood up in one of the great Sorbonne amphitheatres during a plenary lecture by Henri Birault, a follower of Martin Heidegger, who had been talking for a while about Nietzsche's doctrine of the Will to Power. Exasperated by his lengthy and rhetorical exposition, I asked him whether he thought that the Nietzschean teaching that everything is will to power was true, or false, or meaningless. He answered me by addressing the audience at large, saying that ‘votre camarade Hollandais’ was asking a naive question, since Nietzsche had showed that the notion of truth is deeply suspect. Although I was impressed by French historians of philosophy such as Martial Gueroult or Ferdinand Alquié, whose seminars on Spinoza's Ethics I attended with enthusiasm, many philosophical celebrities of the day seemed to be more interested in dazzling rhetoric and in the alleged political implications of their views than in the truth of what they were saying.

In this condition of intellectual turmoil, I felt attracted to the humourless philosophical Ernst of Edmund Husserl, the founder of the phenomenological school to which my ‘directeur d’études' Emmanuel Levinas claimed to belong. The French translation of Husserl's Logische Untersuchungen (1900/01) had been published by the Presses Universitaires the France in 1959–1963, and Levinas devoted a seminar to this seminal work in 1974–75.Footnote ¹ The more my teacher criticized Husserl's texts, often taking Derrida's early book on Husserl La voix et le phénomène as his point of departure, the more I came to admire the rigour of Husserl's arguments, which could be defended quite easily against Derrida's and Levinas's criticisms.Footnote ² I decided to write a doctorate on Husserl's early philosophy of logic, a decision which stimulated me to study also the important logicians of the turn of the century, such as Schröder, Frege, Peano, and Russell. Accordingly, I first went to the University of Cologne, which harboured a centre of Husserl studies, and then worked for a year in the Husserl Archives in Leuven, Belgium. The title of today's lecture is derived from that of an essay Husserl published in 1910/11, Philosophie als strenge Wissenschaft, which I have put in the interrogative form.

Initially I tended to believe Husserl's claim that philosophy can be a rigorous science or discipline (strenge Wissenschaft) if only one follows what he called the method of transcendental phenomenology. Husserl kept repeating that it is extremely difficult to practice this method, and he devoted all his later books, such as Ideen zu einer reinen Phänomenologie (Ideas Pertaining to a Pure Phenomenology) of 1913 or Formale und transzendentale Logik (Formal and Transcendental Logic) of 1929, to pointing out different ‘roads’ to transcendental phenomenology, attempting to convince the readers that they had to perform what he called ‘phenomenological’ or ‘transcendental reductions’. Yet the basic idea of Husserl's method is easy to convey, if one takes Kant's transcendental philosophy as a point of departure.

Kant held that mathematics and physics partly consist of synthetic a priori principles, which provide information about reality although they are not based upon experience. In order to explain how this is possible, Kant postulated that such propositions can be known a priori because they are generated by cognitive mechanisms in the knowing subject, and that they can be informative about empirical reality because the empirical world is subjectively constituted out of sense impressions or sensations (Empfindungen), which are processed by the very same subjective mechanisms that yield the synthetic a priori propositions. This celebrated ‘Copernican revolution’ effected by Kant implied that natural science is not about the world as it is in itself (an sich), but about a subjectively constituted phenomenal world. The theory also implied an embarrassing contradiction concerning things in themselves, which allegedly cause the sensations in us. Since Kant held that the principle of causality is a synthetic a priori proposition, its validity should be restricted to applications within the phenomenal world. How, then, could one claim that our sensations are caused by things in themselves?

Like many critics of Kant in a period when neo-Kantianism was fashionable, Husserl argued that the Kantian notion of a thing in itself is incoherent. As a consequence, he ended up with the Fichtean view that the subjectively constituted phenomenal world is the only material world there is. Since this world is constituted by transcendental subjectivity, it is ontologically dependent on it, so that Husserl became a radical transcendental idealist. His philosophical method can now be characterized as follows. Husserl pretended that by performing a number of ‘transcendental reductions’, we can become reflectively aware of the intentional mental acts by which our transcendental ego constitutes different types of entities, such as objects of sense perception, other persons, geometrical configurations, or numbers. By describing these mental acts of ‘constitution’ and their complex interweavings, we can elucidate the ‘sense’ in which entities of different types exist (their Seinssinn). By so doing we allegedly can solve all traditional ontological problems in philosophy.Footnote ³

One cannot accuse Husserl of modesty. In his lectures on First Philosophy of 1923–24 and later in his essay on the Crisis of the European Sciences of 1936, he claimed that his transcendental phenomenology was the telos or goal of the entire tradition of Western philosophy, because it finally realized its dream of becoming a rigorous science (strenge Wissenschaft). Furthermore, transcendental phenomenology allegedly is foundational with respect to mathematics and all positive sciences, since it elucidates the ontological status of their objects. It is ‘first philosophy’ in Aristotle's sense and provides a Letztbegründung (ultimate foundation) to scientific knowledge. Finally, transcendental phenomenology also provides a philosophical justification of our longing for immortality, because, Husserl argued, the transcendental stream of consciousness goes on indefinitely.Footnote ⁴ Hence, transcendental phenomenology was held to deliver European culture from the crisis of relativism, positivism, and scepticism into which it had plunged during the interwar years.

You will wonder why I am telling you so much about a philosopher who published his most influential works between 1900 and 1936 and who is largely unknown in the United Kingdom. The reason is not only that my infatuation with Husserl produced my first great disappointment concerning a dream of a rigorous method in philosophy. The popularity of Husserl in Paris of the 1970s also tells you something about the influence of politics and fashion on even the most abstruse domains of philosophical thought. Although Husserl was a model assimilated Jew in Germany, having adopted the Christian faith in his student days and having been something of a conservative Bismarckian nationalist during the First World War, he suffered deeply from the ascent of Nazism between 1933 and his death in 1938. This latter fact explains why, after the Second World War, Husserl could be staged as a ‘good’ German philosopher. Husserl's library and the bulk of his stenograph manuscripts (some 45,000 pages) and letters had been rescued in 1938 from Germany and taken to Leuven, Belgium, by a courageous Belgian priest, father H. L. Van Breda, who founded the Husserl Archives in 1939. After U.N.E.S.C.O. decided to give financial support for the edition of Husserl's collected works, of which the first volume was published in 1950, a small Husserl industry flowered on the European Continent, and Husserl's ideas were widely discussed.

My infatuation with Husserl's philosophy did not last long, however. While I was still studying the many thousands pages of Husserl's works, I was beset by doubts about his methodological pretensions. Let me mention four points. First, Husserl assumed, like Kant, the logical positivists, and many others, that the constitution of the perceived empirical world starts with sense impressions, sensations, or sense-data (Empfindungen), which we simply find in our stream of consciousness. Husserl argued, contrary to Franz Brentano, that perceptual sensations (colour-sensations, sound-sensations, smells, tactile sensations, etc.) are non-intentional mental occurrences, which we ‘interpret’ by an ‘objectifying’ intentional mental act, from which results a perception of something in space and time. But do we really find such perceptual sensations when we reflect on our stream of consciousness? Or is the very notion of perceptual sensations rather a posit of ‘scientific’ theories of perception first proposed in the seventeenth century?

Inspired by two books, Merleau Ponty's La phénoménologie de la perception (1945) and Austin's Sense and Sensibilia (published posthumously in 1962), I came to doubt the very meaningfulness of the traditional notion of perceptual sensations as the elementary building blocks of the perceived world. Sensations in the ordinary sense of the word are things that we feel, such as itches or pains, and there does not seem to be any such thing in visual perception, for example. But this doubt destroyed both the basis of Husserlian constitution, since Husserl argued that the objectifying intentional interpretation of sensations is the very basis of (nearly) all mental acts of constitution, and my belief in his reflective and purely descriptive method of phenomenology, for what he claimed to ‘find’ in the stream of consciousness was in fact a posit of dubious traditional theory construction.

A second point of doubt concerned Husserl's notion of eidetic intuition or Wesensschau, which he originally introduced in his Logical Investigations with regard to logic and mathematics. Assuming, like so many philosophers of mathematics, that this discipline acquires knowledge about entities of some type, he argued like Frege that these entities could be neither physical nor mental. They had to be essences, grasped by a special kind of mental operation which he called the intuition of essences (Wesensschau). Husserl then generalized these notions, like Plato had done, and argued that his transcendental phenomenology obtained knowledge of non-mathematical essences, which is synthetic a priori. Inspired by both Popper's and Wittgenstein's criticisms of essentialism, I came to reject this central tenet of Husserl's method. Attempts by Kripke, Putnam, and others in the 1970's, to rehabilitate the notion of essences as structures to be discovered by empirical science never convinced me for a number of reasons, which I shall not spell out here. In other words, I do not think that it is ‘anachronistic to dismiss essentialism as anachronistic’.Footnote ⁵

This second point of doubt automatically led to a third. Husserl had assumed that all problems of theoretical philosophy can be resolved by describing what he called the constitution of entities of different kinds, such as other minds or numbers. But this view seems to presuppose a specific interpretation of our use of numerals or of the word ‘mind’, which in the wake of Wittgenstein came to be called the ‘Augustinian picture’ of language or the ‘referential fallacy’. Surely, when we say that there are ten people in this room, we refer to the room and to people. But are we also referring to yet a third type of entity, the number ten? Are words always used to refer? Surely not. I came to share the view of the later Wittgenstein that many philosophical problems are generated by misunderstanding our uses of language, and that we should deconstruct, rather than answer, philosophical questions such as ‘what kind of entities are numbers?’

A fourth reason for rejecting Husserl's approach to philosophy was inspired by post-Kantian philosophy of science. Like Aristotle, Descartes, and Kant, Husserl had assumed that philosophy has the task of providing a foundation for the special sciences. But this view was completely outdated already in the 1920s, a point I come back to below. By the time I had completed my doctorate in 1982, I had rejected most of the claims Husserl made for his transcendental phenomenology as a result of these and other criticisms. The monumental edition of Husserl's collected works called Husserliana now seemed most impressive if interpreted as a memorial to illusions and wishful thinking in philosophy.

2. Philosophy as a Boulevard of Broken Dreams: Three Diagnoses

Edmund Husserl is not the only philosopher who claimed that he was the first to transform philosophy finally into a ‘rigorous science’. As you all know, this pretension of being the first to found philosophy as a rigorous discipline is a recurrent theme or trope in the history of modern philosophy. We find it in Descartes' Meditations, in Hume's Treatise, in Kant's first Critique, and in Hegel's Phenomenology. The nineteenth century is full of attempts, now largely forgotten, to put philosophy on the secure path of a science, for instance by philosophical naturalists or by positivists in the wake of August Comte. And the rhetoric of founding philosophy as a scientific enterprise for the first time in history did not disappear in the Twentieth century. Compare, for example, Hans Reichenbach's popular treatise on The Rise of Scientific Philosophy, published in 1951.

Although on the European Continent the English are still renowned for their understatements, the rhetoric of finally founding philosophy as a rigorous theoretical enterprise has become popular in Oxford quite recently, where it is used by occupants of the Wykeham Chair of Logic. My precursor in this lecture series, Peter Hacker, called it with his usual wit the ‘Wykeham chair gambit’. Michael Dummett, Freddie Ayer's successor in the Wykeham Chair, proclaimed some thirty years ago that ‘philosophy has only just very recently struggled out of its early stage into maturity’. He argued that philosophers have to construct collectively a systematic theory of meaning, which allegedly ‘is the foundation of all the rest of philosophy’. And he claimed that ‘the search for such a theory of meaning can take on a genuinely scientific character’.Footnote ⁶

Dummett's successor but one in the Wykeham Chair, Timothy Williamson, professes in a similar vein that, perhaps, we have just arrived at ‘the end of the beginning’ of philosophy, pleading for rigorous methodological constraints in philosophical theory construction. He characterizes the ‘opponents of systematic philosophical theorizing’, who claim that we should not construct theories for answering questions if these questions are conceptually incoherent, as making ‘a feeble and unnecessary surrender to despair, philistinism, cowardice or indolence’.Footnote ⁷ But he repudiates Dummett's idea that the primary task of philosophy is to construct a systematic theory of language.

Endorsing criticisms of the linguistic turn in philosophy such as those put forward by Quine and Kripke, Professor Williamson also rejects the view that philosophy should be restricted to resolving philosophical problems by analyzing our linguistic forms of representation. To those who wonder by what methods philosophers might be able to acquire knowledge that goes beyond a reflective awareness of the logical grammar of our language, he answers by pointing to the recent revival of essentialism and claims:

Unfortunately, however, none of these attempts to put philosophy on the secure path of a rigorously scientific discipline ever acquired general endorsement in the community of philosophers. Even worse, none of them survived the demise of its author more than a few decades. We may predict the same fate for Professor Williamson's claim that we can acquire ‘new knowledge’ about ‘how things are’ by mere thinking or thought experiments. Indeed, the historical track record of mere thinking as a reliable method in philosophy for acquiring knowledge about how things are is not impressive. Williamson's stress on thought experiments strikingly resembles Husserl's method of Wesensschau by imaginative variation. Is it surprising that one gets a déja-vu experience when one reads such things?

It seems, then, that we may safely perform a pessimistic induction and conclude that philosophy will never be established as a rigorous science. The aspirations of philosophers of the past to transform philosophy into a decent scientific discipline, when collected in a historical survey, now seem to us nothing but a boulevard of broken dreams. Understandably, this pessimistic conclusion has inspired a general scepticism with regard to philosophy as such, even among philosophers. For example, Richard Rorty concluded from his critical analysis of the views of Sellars and Quine that epistemology-centered philosophy should not be conceived of as a distinct discipline modeled upon the paradigm of empirical sciences or mathematics. Although Rorty saw no ‘danger of philosophy's “coming to an end”’, he did not tell us either what philosophy would be like in the future.Footnote ⁹

The German philosopher Odo Marquard drew an even more radical conclusion in a lecture with the telling title ‘Inkompetenzkompensationskompetenz?’, held in Munich in 1973. He imagined a contest between Chinese executioners. Who could perform the most perfect beheading? The last finalist had to surpass an immaculate decapitation performed by his rival. He swept his sword through the neck of the condemned convict faster than anyone could perceive. Surprisingly, however, the head did not fall and the convict looked up with amazement. Then the executioner smiled affably and said ‘Nicken Sie mal’ (just nod). According to Marquard, present-day philosophy is like the thoughts of this convict before he nods.Footnote ¹⁰

How can it be, one might wonder, that in a time of unsurpassed scientific progress and rapid specialization, philosophy is in a permanent crisis about its own identity? How should we explain the fact that the numerous attempts to put philosophy upon the secure path of a scientific discipline have failed? Let me briefly mention and evaluate three different diagnoses of this phenomenon.

A first diagnosis is the mother-of-the-sciences conception. The word ‘philosophy’ originally meant simply each and every search for truth. Whenever this search in a particular domain acquired methodological maturity and results began to accumulate, that domain split off and became an independent scientific discipline. As a consequence, it is not true that philosophy never became a rigorous science. What is true is that whenever a part of philosophy became a scientific discipline, it was not called ‘philosophy’ any more. This happened to physics in the seventeenth century, to chemistry in the eighteenth, to psychology in the nineteenth, and to cosmology at the beginning of the twentieth. So, the term ‘philosophy’ as we use it is just a label for immature proto-scientific speculations. All issues that are ‘philosophical’ today will duly be transformed into scientific questions. Marquard would add, however, that by now the womb of philosophy as the mother-of-the-sciences is quite empty.Footnote ¹¹

There is much truth in this first diagnosis. For example, philosophers before Kant used to speculate about the age and size of the universe. Although Kant argued critically that such issues are beyond the reach of human knowledge, they became scientific questions when Einstein, De Sitter, and Lemaître started to apply general relativity theory to the universe as a whole. Nevertheless, there are three reasons why the mother-of-the-sciences diagnosis cannot be the complete story.

First, not all philosophical questions can be transformed into scientific ones. Take, for example, questions of normative ethics, which is a sub-discipline of philosophy. Science can tell us much about what should be our instrumental norms, that is, what means we should use in order to attain specific morally valuable ends. But can questions about ultimate moral ends or values also be transformed into scientific questions? Most philosophers deny this on conceptual grounds, even though the sociobiologist Edward O. Wilson once wrote that ‘the time has come for ethics to be removed temporarily from the hands of the philosophers and biologized’.Footnote ¹²

Second, the sciences that emancipated themselves from philosophy gave rise to new philosophical questions, which are not and cannot be fully absorbed into these scientific disciplines. Physical theories such as general relativity or quantum mechanics sparked off intriguing questions about the status of postulated entities, such as relativistic space-time. And formal logic, once absorbed into mathematics, generated puzzling questions about the relationship between formal-logical calculi on the one hand and natural languages on the other hand. As a consequence, many special sciences are accompanied by philosophical sub-disciplines such as the philosophy of psychology, the philosophy of logic, or the philosophy of law.

The third reason for thinking that the mother-of-the-sciences diagnosis cannot be the whole story is that one of its essential assumptions is false. This assumption is that the distinction between philosophy proper and scientific disciplines such as physics, astronomy, psychology, or theoretical linguistics was made only quite recently, that is, during and after the scientific revolution in the seventeenth century. In fact, however, this distinction was made already by Aristotle in his Posterior Analytics and elsewhere.

Aristotle distinguished between particular scientific disciplines, such as physics, astronomy, or psychology, which he called ‘second philosophies’, and philosophy proper, which he called ‘first philosophy’, and which his successors baptized ‘metaphysics’. This Aristotelian distinction and his concomitant conception of first philosophy, which allegedly is both a universal discipline and the foundation of the special sciences, deeply influenced the philosophical tradition up till Descartes, the early Hume, Kant, the neo-Kantians, Edmund Husserl, Bertrand Russell, the early Heidegger, and even Michael Dummett, who claimed that a general theory of meaning is ‘the foundation for all the rest of philosophy’. But if the distinction between (first) philosophy and the special sciences (second philosophies) is as old as Aristotle, it cannot be the case that the mother-of-the-sciences diagnosis is correct for philosophy proper, that is, first philosophy or metaphysics.

This is why a second diagnosis is needed of the fact that philosophy has never managed to become a rigorous science: a diagnosis that explains why first philosophy or metaphysics has never managed to do so. To put it briefly, this second diagnosis is that the very idea of first philosophy or metaphysics is mistaken, among other reasons because it is based upon an inadequate conception of the special sciences. According to Aristotle and many later philosophers, theories in the special sciences ideally should have the logical structure of an axiomatic-deductive system, such as that of geometry in Euclid's Elements. It was further assumed that our knowledge of the axioms should be more secure than that of the theorems derived from them, and that apart from domain-specific axioms or postulates, each scientific discipline also presupposes some general axioms about reality as such. First philosophy or metaphysics was conceived of as a theory about reality as such, which would function as a foundation for the special sciences, providing them with their general axioms. Philosophers often thought that they had reliable non-empirical methods at their disposal that would enable them to construe such a general theory of reality without having to engage into empirical research. Methods such as Cartesian clear and distinct intuitions, Kantian transcendental reconstruction, Husserlian Wesensschau or Professor Williamson's thought experiments.

However, this conception of first philosophy is flawed, as was brought to light by developments both in mathematics and in physics. Why exactly it was flawed has been shown extensively by logical positivists such as Carnap and Reichenbach, who argued against the German neo-Kantians. In mathematics, the construction of consistent non-Euclidean geometries in the nineteenth century deeply changed our understanding of the epistemological status of axiomatic systems. We do not think any more that, for example, the axiom of parallels in Euclidean geometry is a necessary truth which holds for space as such, as Newton and Kant had assumed. Rather, together with the other axioms it defines a certain conception of space, or a representational system for spatial relations, which can be applied to reality only if it is provided with physical interpretations of its geometrical terms. And in physics, the theoretical revolutions of general relativity and quantum mechanics showed that metaphysical dogmas of philosophers can be disregarded by physics, dogmas such as the truth of Euclidean geometry or a deterministic principle of causality. In other words, the theories of physics do not stand in need of a philosophical foundation, which provides them with metaphysical first principles. And the very idea that philosophers can acquire substantial knowledge about reality by non-experimental methods turned out to be an illusion.

If we conjoin these two explanations of the fact that philosophy has never become a rigorous science, it may seem at first sight that we get a complete picture. To the extent that philosophy had always been distinguished from the special sciences as first philosophy or metaphysics, its very conception was mistaken. And to the extent that philosophy consisted of proto-scientific speculations, the mother-of-the-sciences diagnosis is correct. Nevertheless, the picture is not complete. For, as we saw, new special sciences often beget new philosophical problems, and old metaphysical questions as to the nature of substance, universals, or numbers pop up again and again. With regard to these philosophical problems, which do not seem to be amenable to empirical methods of investigation, I endorse a third diagnosis. It, and the concomitant positive conception of philosophy, have been defended forcefully in this lecture series by Peter Hacker from St. John's College, Oxford, so that I can be brief.

According to this third diagnosis, traditional philosophers misinterpreted the nature of many paradigmatic philosophical problems:

But if philosophy, as far as many of its core problems are concerned, should not be seen as a cognitive discipline, which aspires to acquiring knowledge about the world, how should it be conceived of?

Inspired by the later Wittgenstein, the logical positivists, and Oxford philosophers such as Ryle or Austin, Hacker defends the view that philosophers should aim at acquiring a specific type of understanding. Their proper activity is to clarify conceptual structures, both of ordinary language and of mathematics and the sciences. Their legitimate aim is to curb the tendency of language users, including scientists, to transgress the bounds of sense. For example, ‘[e]xplaining that alternative geometries are not alternative theories of space but alternative grammars for the description of spatial relationships contributes to a better understanding of the enterprise of geometry’.Footnote ¹⁴ We find many conceptual confusions in the works of cognitive scientists, for instance, where they misinterpret the results of their empirical investigations due to confusions between everyday psychological terminology and the terminology of brain sciences, thereby producing scientific mythologies.Footnote ¹⁵ It is the proper business of philosophy to criticize such mythologies by elucidating the linguistic confusions that produce them.

There are many misunderstandings about this analytical conception of philosophy, according to which philosophy is an a priori conceptual investigation of a specific kind. For example, some critics say that it is an armchair exploration of empirical languages, which should be transformed into a branch of empirical linguistics.Footnote ¹⁶ Others claim that it has been refuted by Quine's criticism of the analytic/synthetic distinction, or that it implies a sterile conceptual conservatism. In his lecture and elsewhere, Dr. Hacker effectively dispels these misconceptions, and I shall not repeat what he said.Footnote ¹⁷ Yet I want to illustrate by one example the inconclusive manner in which a Wittgenstein-style conception of philosophy is criticized by some present-day philosophers.

In a paper called ‘Past the Linguistic Turn?’, Timothy Williamson intended to refute the view that (all) traditional metaphysical questions should be resolved or dissolved by conceptual analysis.

The implied conclusion seems to be that in order to discover whether, in general, there are numbers, we have to investigate some kind of reality instead of analyzing how numerals are used. But Williamson's argument contains a fallacy of ambiguity, since it mixes up two different distinctions. One is the distinction between successful reference and unsuccessful reference. By using the name ‘Timothy’ I may refer successfully to Timothy Williamson, but if Satan does not exist, I cannot refer successfully to Satan by using the name ‘Satan’. The other distinction is the one between referential uses of words and non-referential uses of words. Surely, only if a word is used referentially it makes sense to engage in an investigation as to whether an entity exists that corresponds to the word used in a sentence, with the further aim of discovering whether the assertion made by using the sentence is true or false. Hence, a linguistic analysis concerning the former issue logically precedes an investigation concerning the latter.Footnote ¹⁹ Only if ‘Satan’ is used referentially, it makes sense to engage into an investigation in order to find out whether Satan exists.

Edmund Husserl thought that the words ‘and’ and ‘or’ are used referentially. As a consequence, he looked for elements or aspects of reality corresponding to these words and he tried to trace phenomenologically the ‘constitution’ of these elements. However, the logical constants ‘and’ and ‘or’ are not used referentially. Husserl was mistaken, not primarily about what exists, but about the logical grammar of the words ‘and’ and ‘or’. Quite a few philosophers of mathematics assume that number words such as ‘7’ are used referentially in the sense that they must refer to entities of some kind. Then they go on to wonder what kind of queer entities numbers are, and by what superior intuition mathematicians are able to know these queer entities. But if number words only seem to be used referentially in this sense, whereas conceptual analysis can show that they are not, such questions are misguided products of conceptual confusion.

Many philosophers, such as Karl Popper or John Searle, hold that it is futile to engage in debates about the nature of philosophy. Admittedly, the term ‘philosophy’ does not stand for some essence, so that one may risk becoming embroiled in verbal disputes when one discusses the nature of philosophy.Footnote ²⁰ We should also admit that quite often debates about this theme tend to be fruitless because they merely express preferences for different research agendas.Footnote ²¹ Yet, I would not have accepted the invitation to give this lecture if I had agreed entirely with Popper and Searle. One reason why discussions of the nature of philosophy may be useful is that philosophers often have pursued imaginary research agendas, such as the project of Aristotle, Descartes, and Kant to establish by some non-empirical method a foundational discipline of metaphysics that contains first principles of the empirical sciences. Such imaginary research projects should be criticized. Another reason is that although philosophers may engage in different research projects, they should be clear about the nature of the projects they are engaging in and refrain from pursuing incompatible objectives.

Searle claims, for instance, that what he did in his books was to advance ‘general theories’ about speech acts, intentionality, the mind, rationality, and the nature of society.Footnote ²² But what, exactly, does Searle mean by a ‘general theory’? Is he engaging in proto-empirical theory building, as present-day string theorists are, hoping that his theories will become empirically testable in the future? If so, what are the empirical tests he is dimly envisaging for his theories, and what are the predictions he derives from them? Or is he just proposing a somewhat simplified and systematized picture of how we are using segments of ordinary language, such as words expressing intentions or mental characteristics of human beings? Or, finally, is he engaging in a descriptive study of mental phenomena in the tradition of phenomenologists such as Brentano and Husserl, making some new conceptual distinctions when necessary? But such a descriptive enterprise should not involve any theory-construction, as Husserl used to stress.

Maybe there are yet other interpretations of what Searle is doing. But if, inspired by Kripke or Quine, philosophers claim to advance ‘theories’ about reality without envisaging empirical tests, arguing, for example, that ‘mere thinking’ is a good method for acquiring knowledge, I think that they are under the spell of an illusion. What they are doing in such a case sometimes is advancing conceptual truths in the guise of metaphysical statements. However, when Martin Heidegger said that ‘thinking’ should replace traditional philosophy, he was advancing a religiously inspired world-view, a new myth, with quite disturbing moral implications. Given the deep influence Heidegger had in Germany and on Continental philosophy in general, I devoted some time to developing a properly historical interpretation of his works and to criticizing his views, using the toolbox of analytic philosophy.Footnote ²³

3. Socrates as a Role Model

Let me finish by saying something about my own research agenda in philosophy, however modest it is. Traditionally, philosophy was divided roughly into three broad areas: logic, ethics, and metaphysics or theoretical philosophy. In the philosophical tradition from Aristotle to Kant, metaphysics was conceived as a theory about reality in general, which had a foundational role for the special sciences. I argued above that the idea of such a foundational role was misconceived and based upon an untenable conception of scientific knowledge.

Yet I think that the motivation of traditional metaphysics to acquire an intellectual overview of reality and to determine the ‘place of man in the cosmos’ is of great value. As Wilfrid Sellars once wrote, ‘[t]he aim of philosophy is to understand how things in the broadest possible sense of the term hang together in the broadest possible sense of the term’.Footnote ²⁴ Roughly speaking, there are two different responsible ways in which such a large-scale picture can be obtained. One is in the material mode, to use Carnap's phrase, that is, by constructing scientific theories. For example, if one wants to know what is the place of man in the cosmos, one should take the neo-Darwinist synthesis and modern Big Bang cosmology as a point of departure for one's reflections, because they are the best theories available. In this case, the objective may be described as ‘philosophical’ but the manner of achieving it is by doing empirical science and empirical research in the humanities.

The other way is the formal mode, by engaging in conceptual analysis, focused on how different conceptual systems of representation are logically related. One may wonder, for example, whether a scientific analysis of matter and perception implies that material things do not really have properties such as red and blue or hot and cold, as many physicists and philosophers since the scientific revolution have argued. More generally: how do everyday descriptions of the perceived world on the one hand and the scientific analysis of matter and perception on the other hand hang together conceptually? This was one of the foci of Sellars's own work: the celebrated problem of the alleged incompatibility between what he called the scientific and the manifest image. It is also a topic of my own writings, and I would deny the incompatibility. Sorting out such issues by conceptual and logical analysis is the proper business of the philosopher.

Another question that philosophers might raise is to what extent it is legitimate to use successful scientific theories, such as the theory of evolution, in order to explain phenomena that lie outside their domain of origin. For example, can we use the theory of evolution in order to account for cultural phenomena like religion?Footnote ²⁵ Many conceptual points have to be clarified in order to decide this issue. For example, how is the notion of ‘fitness’ defined in such applications of the theory of evolution? Do these applications generate empirical hypotheses that are testable?

Many philosophers today think that they have to construct ‘theories’ about how things hang together. For example, in his early work Paul Churchland developed a theory of language, inspired by Quine, according to which all language is entirely theoretical. He concluded that our everyday ways of describing reality and ourselves are part of ‘folk theories’ such as ‘folk psychology’, which should be eliminated in the course of scientific progress. I have argued elsewhere that this doctrine of ‘eliminative materialism’ is a dogmatic metaphysical myth, which should be dispelled by careful logical and conceptual analysis.Footnote ²⁶ In general, such a critical analysis of philosophical views and ideologies or world views is more to my taste than constructing ones of my own. Like Hacker I hold that the business of theory construction belongs to science rather than to philosophy proper.

However, I also think that the methodological toolbox of the philosopher should contain more techniques than merely the Wittgenstein-style analysis of conceptual confusions. In order to analyze philosophical theories and ideologies critically, one equally needs the instruments of formal logic, an extensive knowledge of informal fallacies, and one has to be acquainted with many different areas of empirical research. Let me illustrate this thesis of methodological pluralism with regard to critical philosophy by saying something about my present topic of research, the philosophy of religion.

When I was a student, I did not think that I would ever be interested in criticizing religious systems of beliefs. Most of us assumed, naively perhaps, that the world would become ever more secularized. Furthermore, it seemed that the philosophy of religion could not yield any new or interesting insights. But today we live in different times, and a critical analysis of religious systems of beliefs is more urgent than ever. There are three reasons for this urgency. One is purely demographic. Since the alarming population growth on our planet takes place predominantly in religious countries, the secularized societies of Western Europe have a dwindling share of the world population. If we want the tolerant tradition of the Enlightenment to survive, we will have to engage in a battle of ideas and to criticize religious and other ideologies to the extent that they are incompatible with the Enlightenment tradition. For this reason, a philosophical critique of religions has great public utility.

Second, because of globalization, air travel, and internet, ideological conflicts between different cultures tend to become frequent and more passionate. Furthermore, because of migration, these conflicts emerge at home, in our own countries. Let me just remind you of the cartoon crisis in 2006, which was essentially a clash between Islamic religious absolutism and the value of freedom of expression. Or let me remind you of the murder of the Dutch film-maker and journalist Theo van Gogh by the Islamist Mohammed Bouyeri on 2 November 2004, because he had made a film with Mrs Hirsi Ali, criticizing the discrimination against women in Islam.

A third and more philosophical reason for engaging in a critique of religious ideologies is that during the last forty years, there has been a surprising revival of apologetic philosophy of religion in the Anglo-American world. A philosopher such as Alvin Plantinga argues on sophisticated epistemological grounds that a religious believer may be entirely within his epistemic rights and fulfill all his intellectual obligations although he does not put forward any argument for his religious beliefs.Footnote ²⁷ And a philosopher such as Richard Swinburne claims that the thesis that God exists can be shown to be probably true on empirical evidence by using exactly the same logical apparatus which scientists use implicitly when they confirm their hypotheses about nature by evidence, to wit, Bayesian confirmation theory.Footnote ²⁸ These and other apologetic philosophers of religion intend to bestow on religious beliefs an aura of reasonableness which these beliefs lack, at least so I suggest. It is a challenging and urgent philosophical task to point out the flaws in the arguments of these and other apologetic philosophers.

In order to be well equipped for a critical philosophical analysis of religious ideologies, one should have more intellectual baggage than the mere ability to perform Wittgenstein-style linguistic analysis of religious texts, however indispensable that ability is. For example, one should be widely read in the anthropology of religions, in order not to be narrow-minded about the manifold manifestations of this cultural phenomenon. One should also be informed about the history of the great clashes between science and Christianity in the West, which show in detail why scientific methods for acquiring knowledge are vastly superior to alleged religious methods such as receiving revelations. Furthermore, one should be aware of the failed empirical attempts to validate by double-blind tests religious methods of influencing the course of events, such as prayer.Footnote ²⁹ And in order to understand what happened in the early stages of Christianity or Islam, one should be informed about the psychology of collaborative storytelling and the research on cognitive dissonance produced in sects whose messianic expectations were refuted by the facts.

When all these trunks of background knowledge are in place, the real philosophical work can begin. I would claim that this work strikingly resembles what Socrates did in Athens according to Plato's early dialogues. Let us confront the defendant of a religious ideology with two very simple questions. One is: what, exactly, do you mean by what you are saying? The other question is: how, exactly, do you know what you are claiming to know, or why do you believe that what you believe is true? I predict that after hours of discussion, or after hundreds of pages, in which the answers are analyzed meticulously, the outcome will be one of two. Either the defendant of a religious ideology has to admit that what he was saying does not make sense, after all. Or, if it makes sense, he will have to admit that very probably, what he was saying is false. And of course, we should apply this same type of Socratic and analytical critique of ideologies to other world-views than the religious ones, such as reductionist scientistic ideologies.

Let me end, then, by quoting Socrates, who exemplifies the type of critical philosophical analysis that I want to practice. For it is Socrates whom we have to imitate, rather than those philosophers who pretend to turn philosophy into a rigorous science.

I suppose you all know what the consequence was.