I.C. Jarvie

York University, Toronto

Specially written in 1991-1992 for a planned memorial volume for W. W. Bartley, III. It is much to be regretted that the volume has not appeared.

This paper was stimulated by the work of the late W.W. Bartley, III, whose ideas influenced me more than those of any other contemporary.  My way of honoring his memory is to think through some issues he raised, in lieu of the lost opportunity to discuss them with him.

My aim is to derive from Bartley's work a fresh characterization of philosophy as diverse traditions of debate stemming from a small number of linked problems.  Such a characterization would expose the internal logic and the internal relations of some debates we for convenience label ‘philosophy’.  This dialectical model would not presume to be exhaustive or prescriptive, of course, for the selection of problems and of their debating links would be subject to debate.  By contrast, standard introductions to philosophy characterize it either as a series of great authors and their ideas, or as a set of topics, or both.  Insofar as questions of how this set of authors and ideas came to assembled and of why these topics and not others are addressed at all, it is with the intention of making the practice of philosophy as the writer understands it seem unobjectionable.  The introducer seeks to tell a story which leads up to and naturalizes his own practice.  From Bartley we get a way of throwing into question the presuppositions of our practice which allows us to be more deeply critical of our present as well as of our past efforts at philosoph.

Before plunging in to my fresh endeavor to characterize philosophy in a general way, briefly consider the purpose of such exercises and the dangers to be avoided.  The purpose of such general characterizations is, I suppose, to alleviate the bewilderment experienced by those entering the subject and by some within it who find themselves lost.   A general characterization should simplify the overall picture and thereby create a setting in which beginners and others can try to find their way by debating goals and standards; that in turn should help foster creativity and problem-solving.  Although this sounds like an innocuous program, it can mislead.  There can be factual misdescriptions of what philosophy is or what philosophers do.  A more serious pitfall is the normative or stipulative aspect of all such exercises in characterization.  As we characterize something we correct and recommend as well as describe.  Characterization thus easily decays towards 'persuasive definition.'  We do better to try to persuade by arguments, not verbal formulas.

Consider the manner in which characterizations of philosophy so often resort to formulaic definitions which either claim to specify what 'philosophy' really means (or will by the writer be taken to mean), namely (i) the verbal characterization of philosophy, or to capture 'what philosophy really is', namely (ii) the characterization of the essence of philosophy.  The objection to these traditional formulas is that 'philosophy' is a rubric comprehending plural and diverse activities, so specifications of what the word really means (or will be taken to mean), or what philosophy really is, are prescriptive and restrictive.  Not only are many and diverse things done under the name of 'philosophy', but lots of people have a rough and ready idea of what it amounts to.    If we attempt to accommodate these facts we reach (iii) the pluralist characterization of philosophy which lists the actual issues that are discussed.  If we add to it people's notions of what it should be discussing we reach (iv) the intuitive characterization of philosophy.

Further, two quite different approaches to characterization respond to the fact that one institution and more than one tradition claim to know what philosophy is, indeed, to embody it.  The institution claiming philosophy is the academy, which gives us (v) the academic characterization of philosophy; the tradition claiming philosophy is the history of philosophy - with branches East and West - which gives us (vi) the historical characterization of philosophy.  Like diversity and intuition, these claims on behalf of an institution and a tradition should be accommodated so far as possible, limited only by their persuasive attempts to restrict philosophy to their characterization of it.
Having criticised verbal, essential, academic and historical characterizations, my own effort, (vii) the problems characterization of philosophy should, to avoid inconsistency, itself be seen as problematic and hence open to criticism - this perhaps is (viii) the critical or criticist characterization of philosophy.
Consider again the purpose of these characterization exercises.  If we manage to formulate a view that reconciles or at least accounts for so much, then we might hope it will help rather than baffle students, which is the pedagogical aim of the exercise, and also be a stimulus to the more adept as well, which is the fruitfulness aim of creativity and problem-solving. 

In approaching the fresh characterization of philosophy we could ask that it take account of the diverse characterizations, and seeks fulfil the two purposes.  Together these amount to a set of negative and positive desiderata.  To summarize, the fresh characterization should:
(i)   not be merely verbal or definitional;
(ii)  not be essentialist;
(iii) be pluralist, i.e. respectful of the diversity of philosophical endeavour;
(iv)  and of people's diverse intuitions about it;
(v)   make sense of the academic embodiment of philosophy without wholly respecting its restrictive claim to proprietorship of the real thing;
(vi)  make sense of and respect the history of philosophy without wholly acquiescing in the academic version of that history;
(vii) be of pedagogical value;
(viii)be fruitful in some recognized sense;
(ix)  be open to critical discussion.

Accordingly, the characterization offered centres on the thesis that philosophy is critical debate about a small number of problems (to be specified later) and proposed solutions to them.  Once such debate is initiated parties or schools of thought cluster around the solutions put forward and engage in dispute.  There may be more than two parties, but debate regularly gets bilaterally polarized between a solution and its contrary, or a solution and its contradictory.   Decisive results or the weight of opinion or other factors can render one line of thought dominant, but as such hegemony rarely lasts very long, it should never be treated as a fixture.  The achievement of decisive results with a problem has repeatedly led to the creation of specialized studies, the sciences and the humanities, the existence of which invites specialists to pursue those results and the new problems that are seemingly autonomous or at least free from philosophical origins.   That this autonomy is illusory becomes apparent when those philosophical origins are adverted to in introductions to the specialty, in textbooks, in histories of it, and in the invariable manner philosophical issues come to the fore when an intellectual crisis is in the making within the specialty in question (such as a scientific revolution or a 'crisis' in the humanities).

Just what are the problems around which philosophy clusters?  The view that takes philosophy to be universal and comprehensive would say that all problems are originally philosophical.  This view, advocated by Russell, may be true, but the effort of tracing the origins within philosophy of such problems as, 'what is the best crop rotation?', or, 'why is English not inflected?', is not especially enticing.  Is the only alternative to such a universal claim the rather untidy conclusion that philosophy is the debate about certain problems but by no means all?  The degree of untidiness will depend upon whether the set of problems specified has internal affinities.  Formalists and other extreme rationalists would expect a fully spelled out statement of the affinity.  Historians and others may be inclined to favour such permissive formulas as Whitehead's, “The safest general characterization of the European philosophical tradition is that it consists of a series of footnotes to Plato”.
A formally neat and tidy picture was not among the desiderata listed above, and for a very obvious reason.  As Einstein said,  formal presentation is always very much post hoc: it cannot respect history and so is incompatible with desideratum (vi) except in those rare cases where the formalism is itself the outcome of a prolonged historic process.  History is usually both untidy and dependent on our creativity, including the creativity involved in simplifying and rendering ideas formalizable.  There remain limits to what can be formally put.  In particular, the very elevation of a problem to the centre of attention is a creative act that cannot be formalized.  Much of the history of discussion of philosophical problems is dependent on such creative acts, as well as the more obvious creative acts involved in solving and debating the highlighted problem.

So a polarization between desirable formality and disorderly happenstance is to be avoided.  Yet, characterizing philosophy in a way that displays some order is, I take it, preferable to characterizing it as devoid of order, at least from a pedagogical point of view.  Gradually rendering a sketch more messy and confused is easier than trying to bring order into confusion.  Thus, within the constraints of an historist rejection of formalism, some degree of tidiness and order is to be sought.  In seeking such a general characterization which yields an order, even if a not wholly tidy one, I follow through on some suggestions of Bartley.

The thesis being maintained is that philosophy clusters around a very few, central problems.  Many of the troubles philosophers get themselves into stem from failing to attend carefully to the formulation of those problems.  Such failure makes it difficult to keep track of the logic of the debate and makes us prey to the tendency to fuse criticism with justification, as Bartley put it.  What he meant was this.  To rebut criticism of one solution is not to justify another, unless they are contradictories, and to criticise a rival view is not to strengthen one's own - except in the rare cases where the list of alternative views is complete: it includes all the logical alternatives.  Criticism is creative and can take the negative form of showing an idea does not solve the problem to which it is addressed.  Such criticism offers no justification of anything.  Bartley said the check of the problem was the last test to apply to ideas under debate.   I sometimes think it should be the first.

Popperians and Problems

Problemstellung is a central though as yet underdeveloped topic within the Popperian approach to philosophy, and one about which Bartley and I wholly agreed, an agreement shared with a small and select group, most of them Popperians.  What we shared was an outlook that those of us close to Popper learned from the horse's mouth when the horse had not written very much about it.  The outlook centred on the idea that rational inquiry began from, and ended with, one's being possessed by interesting problems.  Popper trained us by example to develop a nose for problems, one able to distinguish between genuine ones and pseudo-ones, between interesting ones and trivial ones, discriminations which turned on a sensitivity to history, to the logic of debate, to the implications of ideas, in sum, to what we called the problem-situation.

Talk of and concern with 'problems' was central to the Popperian school and distinguished its members sharply from other philosophers.  Of course, the notion of 'problems' is widely bandied about by other philosophers in a manner that fails to distinguish problems from questions and both from contradictions and almost never refers to any rational criteria for importance or ranking.  Russell's admirable The Problems of Philosophy, for example, addresses a question (viz. “Is there any knowledge in the world which is so certain that no reasonable man could doubt it?”) and does not map this to a problem, and ranks questions as worthy of discussion solely by whether he “thought it possible to say something positive and constructive”.  Not only the role but also the nature of problems were conceived of differently by Popperians.

In his classic paper, “The Nature of Philosophical Problems and their Roots in Science”  Popper had maintained first that there were philosophical problems, this in the face of the disclaimer fashionable at the time (1952) that philosophy was largely therapeutic and had no intellectual content.  More interesting in the longer term was the thesis, “Genuine philosophical problems are always rooted in urgent problems outside philosophy, and they die if these roots decay”.   Popper gave examples of intellectual crises in mathematics and science such as the Greek discovery of irrational numbers and the resultant crisis in geometrised atomism, and the clash between Newton's theory and epistemological ideas that made his theory seem an impossible achievement namely, synthetic a priori knowledge.

Two of Popper's most gifted disciples were Joseph Agassi and W.W. Bartley, III.  By his own account Agassi was attracted to study with Popper by the very paper just mentioned.  However, he did not always agree with his teacher.  Agassi made the criticism that Popper's major work on the philosophy of science, The Logic of Scientific Discovery,  took too little account of the true history of science and, in particular, of the unruly role of metaphysics in the choice and treatment of problems.  Agassi saw the influence of metaphysics as pervasive on both sides of the science/non-science demarcation line.  Science and philosophy, in history and in fact, were for Agassi inextricably intertwined.  The problems of metaphysics were a stimulus to science and often responsible for setting its priorities.  To study the problems and debates of the one invariably took one over into the problems and debates of the other.  Science grew out of metaphysics, but scientific ideas and crises then stimulated changes in metaphysics and vice versa.

Bartley made different critical claims and, regardless of whether one accepts them or not, succeeded in reformulating some of Popper's central ideas in a way that revealed how Popper's work transformed the problem-situation in philosophy, just as it enabled Bartley to formulate what he considered a logically impeccable position, first christened “comprehensively critical rationalism”.   The central point Bartley made was that Popper had developed the very first non-authoritarian epistemology, one which recast the notion of debate over problems from seeking justification to seeking criticism.  Thus debate about problems could be construed as a process of genuine discovery about the structure of the problems and about the implications and hence further problems generated by each solution.

My argument in the present paper was stimulated not by Bartley's superb The Retreat to Commitment, where the views just outlined are argued, but by a subsequent paper entitled “Rationality versus the Theory of Rationality” which he published two years later in Mario Bunge's volume honoring Popper, The Critical Approach to Science and Philosophy.   In this contribution to the first volume of essays ever dedicated to Popper, Bartley opened in magisterial vein:

The three principal problems of philosophy are the problem of knowledge, the problem of rationality, and the problem of reconciling knowledge and rationality

In a footnote he added:

Almost all other philosophical problems are directly related yet subordinate to the three fundamental ones

As far as I know these provocative claims, which were not developed in the paper, were not elsewhere followed up.  None of Bartley's later writings attempted to carry through this bold unification of philosophy.  My present aim is to sketch the program of this powerful insight and to test it.

Has Philosophy a Subject-Matter?

A striking idea Popper put forward in his lectures and publications was a critical attitude to what is generally called 'topics' or, synonymously, 'subject-matters', namely the naturalistic idea that underlying any academic topic or subject there was a pre-existing domain which it was 'about'.  The metaphysical objection was that this assumed the world was prepartitioned into natural major categories of natural kinds to the study of which academic divisions should correspond.  Such an assumption had infected science, such as the idea that celestial, terrestrial, subterranean, and so on were inherently different realms where different principles of order prevailed.  Refutation of metaphysics was impossible, but the demise of scientific theories inspired by particular metaphysics suggested that the metaphysics was, if not mistaken, unfruitful.  Popper held that such metaphysical ideas should be brought out as part of the overall problem-situation and its formulation to the extent possible, and that the overthrow of scientific theories also had the effect of driving us back to uncover the underlying metaphysics that had been invisible to us, after which we could assess it and revise it.

He did not think there were given domains of objects (celestial, terrestrial, subterranean or philosophical) to which corresponded the fields of study astronomy, geography, geology or philosophy.  He thought, rather, that astronomy, geography, geology and philosophy were classifications of the phenomena of experience buried in theoretical constructions which the bureaucratic language of academe tended to reify.  The theoretical constructions were devised as solutions to problems, these solutions then took on a problematic fascination of their own and, insofar as they envisaged the entities of this world in a particular way, say as molecules, or as living systems, or as money transactions, they were declared to be 'fields' that 'dealt' with a certain partition of the world (viz. chemistry, biology, economics).  Philosophy was a good case in point.  To all attempts to say what philosophy was about that specified some subject matter - ultimate principles, say, or deep thought - Popper would produce counterexamples in an effort to show that subject-matters were conventional not natural.

Bartley himself drew attention to a passage in the 1959 preface to the translation of Logik der Forschung where Popper took on his fellow London Professor A.J. Ayer on the subject. In the preface to his The Problem of Knowledge (1956) Ayer had written that what distinguished philosophy was not its subject matter - here the logical positivist agreed with Popper, but its method.

'It is by its methods rather than its subject-matter that philosophy is to be distinguished from the other arts and sciences.'

Less than a mile south of University College where Ayer was Professor was the London School of Economics, where Popper operated and wrote in flat contradiction in 1959:

"Philosophers are as free as others to use any method in searching for truth.  There is no method peculiar to philosophy."

This thesis had been propounded after a paragraph attacking Ayer's then-cohorts the language philosophers:

"Language analysts believe that there are no genuine philosophical problems, or that the problems of philosophy, if any, are problems of linguistic usage, or of the meaning of words.  I, however, believe that there is at least one philosophical problem in which all thinking men are interested.  It is the problem of cosmology: the problem of understanding the world - including ourselves, and our knowledge, as part of the world."

Interestingly enough, in his attempt to unify the problems of philosophy quoted above, Bartley had exempted “certain cosmological problems”.   (Let me demur on whether this qualification was in order.)

Popper's problem-centred vision of philosophy was shared by all of his close students.  It meant a head-on clash with the then-influential language philosophers and their preoccupation with philosophy as a treatment for those bewitched by language.  (Connoisseurs of folly can look up their 'dissolution' of the free will problem with the so-called paradigm case argument. )  It also meant dissenting from the positivist view that the only serious problems were scientific and that most philosophy was simply empty talk.  But it also left us in a more serious jam.  For it followed, as Popper himself repeatedly pointed out in his lectures, that the history of philosophy would best be reconstructed as a series of addresses to the problems which have been identified as philosophical (for whatever reason), the proposed solutions to these problems, and the dialogue between all concerned (overt or veiled as the case may be).  But such a history of philosophy did not exist at the time Bartley and I began our studies with Popper (1958): those histories that were not plain one-damn-thing-after-another textbooks, were whiggish offerings; that is, they defined philosophy according to current views on its 'subject matter', and then used that as a criterion for selectively rummaging the past and connecting the result up chronologically.

When I read Bartley's pronouncement in the Bunge book quoted above I hoped that Bartley might see his task as sketching at least some of the outlines of a Popperian, problem-oriented history of philosophy.  History of philosophy was not his specialty, but he knew some areas of it pretty well.  The prospect was exciting because the abolition of subject matters, and the waywardness of problems, ensured that a genuinely responsive account would follow the problem wherever the debate went, and that might entail pursuit into science, through metaphysics, into fiction and back to philosophy again.  To this day no such Popperian history has been ventured, even in J.N. Hattiangadi's exciting “The Structure of Intellectual Problems” (soon to be discussed) which throws out many hints and certainly ventures across putative demarcation lines between philosophy and science in its brief case studies.

The view of philosophy as a complex set of interlocking and overlapping debates involving participants who paid no regard to whether they had proper credentials (indeed, before this century, without a 'credentials committee' to face), was a very exciting one for another reason: debates themselves can be exciting, provocative and emotional experiences, hence the history of philosophy could be expected to become many overlapping narratives in which great events of science and politics and personal life became part and parcel of debate.  Philosophy would be anything but a dessicated scholastic subject under such a construction, but rather part of the grand universal endeavour to understand the human condition and our predicament.

Popper himself had sketched the history of pre-Socratic philosophy in this light, and received little sympathy from the  specialists in the pre-Socratics.   He had also tried to treat Plato and Marx partly in this way, but not in a manner which satisfied the historians and the specialists.   One reason for his not pursuing it, of course, was that a corollary of his view of philosophy was that philosophy as an academic subject was unimportant; what was important were problems.  Hence Popper spent his life working on first order problems, not working on second-order questions such as how to present philosophy as a set of debates about first order problems.

The present paper no doubt fails to meet that exacting standard of confining oneself to first order problems.  Bartley hinted at his view of a hierarchy of problems in The Retreat to Commitment where he maintained that the problem of “showing it is possible to choose in a nonarbitrary way among competing, mutually exclusive theories” or even 'ways of life' was more fundamental than that of the demarcation between and non-science because it pre-dated and was broader than the latter, which was a special case of it.   He connected the importance of the problem of competing and exclusive ways of life to the post-Renaissance situation of competing candidates for political, religious and intellectual authority.  I would add to what Bartley said that another contribution to the Renaissance problem-situation he described was the reception given at that time to proto-anthropological accounts of alternative ways of life (not buried in the dust of classical literature but) discovered live and well on the fringes of the known world, especially among the 'Indians' of the new world.  The competing political, intellectual and religious authorities made their own adjudications of these newly discovered 'ways of life', but they in turn were being scrutinized as competing authorities.

Soon after his The History of Scepticism from Erasmus to Descartes was published, in 1960, Richard Popkin became the favorite historian of philosophy of the Popper circle.  Not noticeably influenced by Popper, Popkin's project was not problem-oriented either: it was originally an investigation into the influence of ancient scepticism on seventeenth and eighteenth century philosophy which, as he pursued the great ideas of the past and the problems they gave rise to, took him into the history of religious disputes and their political and social correlates.  Thus, though not himself problem-oriented, Popkin presented the history of philosophy as such.  His immensely illuminating work, he reports, is sometimes denigrated by philosophers:

"For decades, brash, crass teachers in philosophy departments have asked me if what I am doing is philosophy.  I do not know, and frankly I do not care.  Obviously the answer depends on how one defines “philosophy”.  I firmly believe that what I have been doing for over forty years is, or should be, of importance to anyone concerned with understanding what Jack Randall called “the making of the modern mind”, with understanding the unfinished product, the intellectual world that we are involved in.  Whether this understanding properly belongs in a philosophy department or some other department I do not know.  Years ago...Kristeller...told me that we historians of philosophy should not give up our posts in philosophy departments until we are thrown out by the so-called “real” philosophers.  I do not know whether he was right or wrong."

Problem-oriented history and history of problem-orientation whether by Popper, Agassi, Bartley or Popkin can be exciting and fruitful (viii).  We can postpone no longer saying something to clarify just what problems are.

Problems, Questions, Contradictions

'Problem' is almost as promiscuously used as 'theory'.  Since for me it is a term of art I want to say something about it.  A problem is not to be equated with a worry.  Some problems are worrying, but mostly they are intriguing and challenging.  A problem is also not to be equated with a question.  Questions are a very large category of which problems are only a small part.  There are many questions asked in philosophy which are not serious problems, including: what is philosophy, what is justice, what is freedom - indeed virtually any 'what is ...' question.  These questions are either verbal, and merely seek elucidation of the word and its various senses - something which, inexplicably, some philosophers and much of the public take as illuminating; or they presuppose that there is a specifiable something that constitutes the 'what is' of whatever is completing the question.

If the question asked is 'what is freedom?' a verbal answer might be, 'the absence of constraints on action, as when we say of a person completing a gaol sentence, “they got their freedom today”'.  The essentialist answer might be, 'the essence of freedom is the absence of coercion'; or, 'the essence of freedom is when the soul can realize itself to the greatest extent'.  Neither of these approaches to 'what is freedom' treats it as a serious problem.  For example, Russian Communists and their fellow-travelling allies argued for many years that the capitalist west did not consist of free societies but of societies in which capitalism exploited and oppressed the workers with the help of much empty talk of a 'freedom' that was purely formal.  Life under the Soviet constitution, by contrast, which guaranteed social and economic freedoms, was truly free.

Treating these assertions at face value is to raise a serious issue: when we demand political freedom what are the sorts of things we demand and did the Soviet Union provide them while western nations did not in, say, the nineteen-fifties?  Verbalism, essentialism and conceptualism evaded and trivialized the issue; formulated in this way it is a serious problem.

Most philosophers who think of philosophy as centreing round certain problems fail to distinguish problems from questions.  This a Popperian is forced to do because essentialist, 'what is' questions, verbal, 'what is the definition of' questions, not to mention conceptual, 'what kind of a concept is' questions do not evoke problems and deserve to be set aside as uninteresting.  The thesis, after all, is that philosophy centres on interesting problems, not just on problems.  Problems are certainly usually posed as questions, and interesting problems are a sub-class of that sub-class of questions.
What makes a problem interesting?  It arouses our curiosity.  Something has been said or seen or reported that challenges, i.e. contradicts, some expectation we have, whether straightforwardly factual, whether scientific, or whether metaphysical.  Thus there is a problem of choice.  Both sides of a contradiction cannot be true, so, which is false?   Even the possibility of the contradiction arouses this curiosity and intellectual excitement at the prospect of such a choice.

We reach at this point the appropriate place to bring in the work of J.N. Hattiangadi, whose ideas on problems show the deep influence of Popper.   Hattiangadi noted that etymologically the word 'problem' has more connection to obstacle or difficulty than to question.  His argument was that problems are difficulties that can ultimately be traced back to contradictions in the set of statements that we entertain.   From a contradiction any statement follows, including the paralyzing injunction to follow and not to follow a certain course of action or of thought.  Hattiangadi gives an elaborate logical structure to problems only to conclude that the deepest and most interesting problems (matter and motion, light, and the issues surrounding Newton's synthesis are cases he considers) have idiosyncratic structural features that derive from the history of debate surrounding them.  (Problems that we treat while ignoring their historical structure if any, or which we even treat ad hoc, are practical and shallow.)  His ideas make very good sense of Popkin's practice.

Roughly speaking, Hattiangadi's schema for identifying the historical structure of deep or intellectually interesting problems is this.  To a problem one or more solutions may be proposed.  Each solution will be the object of debate as to its viability and its success in solving the problem.  Partisan allegiance may develop around a solution, leading in effect to a research program to develop and foster that solution and to explore whether it in its turn contains problems.  Thus there will be debate around a solution in relation to the problem, and there will be jousting debate between rival solutions, especially seeking further consequential problems the rival cannot solve.  A line of thought leading through a solution and beyond should dominate a rival line to the degree it encounters no new problems of its own, and to which it solves problems found common to both lines.

Hattiangadi elegantly tries to show that explanation is not an aim but a by-product of this problem-oriented view of science, and that science itself has no aim, but is an unintended consequence of the clustering of our problem-solving activities and our development of social institutions to foster them.  (Hattiangadi is studiedly indifferent to the labels 'science' and 'philosophy' and the demarcation between them.)

Once problems are known, once solutions to them have been proposed, once a library of arguments has been accumulated, there  exists a highly structured problem-situation that both assists and constrains our efforts to cope with standing problems and newly-discovered ones.  Yet it is hard to articulate problem-situations and it takes time.  Like comedy, it only looks easy.

Philosophy Coheres Around Problems

In a bold unifying move, akin to Bartley's, Hattiangadi suggested that the fundamental problem of epistemology was formulated in Plato's Meno as a paralyzing intellectual conundrum: if we know something we will not seek it, if we do not know it, we will not know what we seek.  It thus seems that there is no way to construe our intellectual quest for more knowledge.  Two basic lines of thought flow out of this conundrum: one is that we find some way to build out or extrapolate from what knowledge we now have; the other is to hold that the world somehow causally imputes knowledge into us.
Without going further into it, let us take it that this is one candidate for Bartley's 'the problem of knowledge'.  The problem of rationality then would be, faced with the problem of knowledge, does either alternative show that the very hope for rational, non-arbitrary knowledge is forlorn?  Is the problem of knowledge a special case of a contradiction that pervades the effort to solve the problem, 'how can we be rational?'  The difficulty is that a rationalist tries to hold all ideas open to critical debate, but can he consistently hold that rationalist idea itself open to critical debate?  If not, then does not his rationalism rest on a contradiction and hence fail to be rational by its own standard?
The third problem Bartley ventured as fundamental was 'the problem of reconciling knowledge and rationality'.  One way he put this was, does our solution to the problem of knowledge make rationality impossible; does our solution to the problem of rationality make knowledge impossible?  In other words he saw a way of connecting these two problems together at a fundamental level, since solutions to each could easily contradict a chosen solution to the other.

Bartley's effort was strikingly anticipated by an attempt in Kant's lecture notes on logic to unify philosophy through four questions:

For the science of the relation of all knowledge and every use of reason to the ultimate end of human reason, to which, as supreme, all other ends are subordinated, and must be combined into unity in it.

The field of philosophy, in this sense, may be reduced to the following questions: -
1.  What can I know?
2.  What ought I to do?
3.  What may I hope?
4.  What is man?

The first question is answered by Metaphysics, the second by Morals, the third by Religion, and the fourth by Anthropology.  In reality, however, all these might be reckoned under anthropology, since these first three questions refer to the last.

The philosopher, therefore, must be able to determine -
1.  The sources of human knowledge.
2.  The extent of the possible and useful employment of knowledge; and lastly -
3.  The limits of reason.

The last is the most useful, but also the most difficult, and is one about which the philodoxus  does not concern himself.

We see at once that while the first four questions are simply questions that could be fobbed off with brusque answers (viz. 1. very little; 2. whatever you want; 3. nothing; 4. a means for propagating DNA strings), the second three stipulations tie directly back to Bartley's effort.  1. points to Meno's problem, 3. to the problem of rationality, 2. to the problem of the relation between rationality and knowledge.

Philosophy today consists of a lot more than the problems of knowledge and rationality.  We have logic and metaphysics, philosophy of science, philosophy of art, of morals, of religion, of politics, of law, of technology, and so on.  Yet in so far as any of these ask questions of knowledge (say about god, or morals, or fundamental ontology) they are special applications of debating lines stemming from Meno's problem which differentiate so as to see if the preferred answer works when the general question, how can we learn?, is rendered specific, becoming, how can we learn about god/morality/substance specifically?  Either we know it already so the inquiry is idle, or we do not know and so do not know what we seek.  Similarly, the problem of how rationally to justify rationality in art, morals, politics, law, etc., amounts to testing the preferred solution to the general or basic problem against what are seen as the special demands of reasoning by art critics, moralizers, political theory and legal reasoning.  In effect each field differentiates not a new set of debates, but particular problems re-written as variants of it in varied problem-situations, where each of them develops historically a structure of its own, but in a line of debate that is logically subordinate to the fundamental line of debate in the sense that we can trace a line of descent one way (deductively) but not the other (inductively).  (No one has ever presented a theory of induction where lines of debate are given general importance.)

Bartley's problem of the reconciliation of knowledge and rationality can readily be applied to all of the instances mentioned, since it is the problem of the reconciliation of the theories of knowledge and of rationality as these are applied to art, science, law, etc.  Inquiry at this depth is very seldom carried out.  Russell, the supreme rationalist, tried and failed.  He was unable to give an account of knowledge and rationality in morals resting on anything other than feelings of repugnance.  Near the end of his life, he wrote bitterly of his inability to reconcile himself to such a milquetoast position:

"I am not, myself, satisfied with what I have read or said on the philosophical basis of ethics.  I cannot see how to refute the arguments for the subjectivity of ethical values, but I find myself incapable of believing that all that is wrong with wanton cruelty is that I don't like it."

The fruitfulness of the problem-centred characterization of philosophy can be further illustrated by looking at a reverse case: not philosophical specialties and their unifying roots but at supposedly distinct non-philosophical subjects and their roots in philosophy.  My own special study has been the roots of anthropology in the problems of knowledge, rationality and the reconciliation of knowledge and rationality.  It is not hard to show that information about the diversity of mankind (whether accepted or denied) had an impact on debates about these problems. For one thing it prompts a line of criticism of the very formulation of the problems.  Take knowledge: 'if we know something we will not seek it, if we do not know it we will not know what we week'.  This relies heavily on the undiscussed 'we'.  Does this problem change if we substitute 'they' for 'we'?  One solution and its line of debate explores the idea that 'we' and 'they' are more or less the same and hence the problem of knowledge is universal; another solution and line of debate asserts that 'they' are different, either in their ways of knowing (relativism) or in not knowing at all (racialism and other doctrines about the demarcation of cognitive mankind).  Material for this discussion long antedates the emergence of academic anthropology in the nineteenth century, even long antedates the seventeenth and eighteenth century speculations on the origins of man and society.  It is my contention that none of the divisions in anthropology between various theories, theoretical orientations or schools of thought can be historically or logically understood without some sense of this background.  Good examples are the technical debate about whether the Trobrianders studied by Malinowski did or did not 'know' the connection between copulation and pregnancy, and the meta-debate about whether or not Margaret Mead's first field trip to Samoa was undertaken to buttress the nurturist views of her teacher, Franz Boas, in the nature/nurture controversy.

To come back to philosophy proper.  We have looked briefly at some of its subdisciplines, but we have not looked at one which seems to be technical and not problem-driven, logic.  Logic is an intriguing special case, because, in the very process of isolating and identifying problems as intellectual difficulties stemming from a contradiction in statements entertained, logic already is presupposed.  Logic is all-too-often presented as an attempt to 'analyze' arguments and devise rules to capture the valid among them.  This superficial text-book approach conceals that logic is the sediment from some very bold and false hypotheses.  These include the hypotheses that logic is the key to the rationality of the mind (laws of thought), that arithmetic and the rest of mathematics are built on the foundation of logic.

However construed, is logic the heart of philosophy?  Certainly it is treated that way in many academic environments, where it is a required course.  The greatest logician of our times changed his view on this.  Popper was fond of reminding us that at one point Russell thought logic was the essence of philosophy (Our Knowledge of the External World, 1914) while at another he held that logic was by no means the essence of philosophy.   If the problems of philosophy are general, underlying not only all the subfields of philosophy but also those of the other subjects taught in academe, then logic has an even greater claim to generality since from it we learn the consequences of inconsistency and hence the force of criticism.

Reflexive Self-Questioning

The discussion above has advocated that view of philosophy which characterizes it with the notions of problems, problem-situations, proposed solutions and critical debate about them.  While showing that this view satisfies the criteria I stated at the beginning, I want now to stress the way it satisfies the last of them, namely (ix): it is a characterization of philosophy which describes and stipulates and can, on both counts, be critically discussed.

Does the present paper in effect claim that the problem-characterization constitutes the essence of philosophy (ii); or, is it an arbitrary stipulative definition that says this is what I am going to mean by 'philosophy' (i)?

The answer to both accusations is 'no'.  In the list of desiderata there were embodied some reasons for raising the question of characterizing philosophy at all, without aspiring to its essence (ii) or a convenient definition (i).  The idea rather was to attempt to capture history (vi), practice (v), intuition (iv) and even popular understanding (v) and (vi) of this activity in a way that would make sense of them and subject them to critical discussion (ix).  The Philosophy of Drinking and Smoking  and The Philosophy of the Turf  conceivably can be connected to more academic conceptions of philosophy.

The question is how we do this.  In the notion of a hierarchy of problems, some related to but subordinate to others (and that hierarchy itself being subject to criticism and revision), we have our connection.  Just as Bartley challenged Popper's claim that “the central problem of the theory of knowledge” was the problem of the demarcation between science and metaphysics with the argument that there were prior and broader problems which involved adjudication between competing ways of life, thus suggesting a broadening and reordering of the hierarchy of problems, so he was able to replace Popper's claim with one that is a self-exemplifying case of the process said to be at work.  His claim that the fundamental problem of modern philosophy was that of “defeating the tu quoque by showing that it is possible to choose in a nonarbitrary way among many competing, mutually exclusive theories, and - more broadly speaking - among competing 'ways of life'“ purports both to re-order and to solve.

A related objection might claim that the focus on a small number of central problems is another way of claiming that these perennials are philosophical and that, despite the separation of specialized sciences, they remain the heart of philosophy and they connect it to everything it has spawned.  In reply I would say that historically these are the central problems, and that logically they do underlie most of the activity they have spawned, whether philosophical or not.  But it is conceivable that further research will show they are not all historically central, or that they are not the logical underpinnings, or both.  In other words, the answer to the essentialists is not to deny that one is making a definite claim for the centrality of these problems, only to deny that this claim is the last word, the definitive pronouncement.

An example of where the problem-centred approach would have made in inquiry deeper and more fruitful is Hacking's book Why Does Language Matter to Philosophy?.  It is an attempt to show why the problems of language have come to the top of the philosophical agenda, but also to show how many traditional problems are related to and subordinate to central problems of the philosophy of language, such as reference, meaning, rules, etc.  The book fails because it is whiggish: it uses history to legitimate the present.  Even if there is a language philosophy solution to Meno's problem it must be considered to be only one solution among many, and its adherents one party among others.  At best language philosophy should either show a problem to be merely verbal and so not genuine, or else show that translating a philosophical problem from object- to meta-language is generally advantageous.  The first option is traditional Wittgenstein and rejected by Hacking.  The second admits that language as such is the vehicle of philosophical debating lines and not the solution to the problems debated.  Hence it is not that language matters 'to philosophy', it is: what is the appeal of solving problems verbally in sophisticated ways?  This may be thought of as an extra-philosophical question It is notable that in his book on language philosophy, Words and Things, Gellner found that a sociological account of the training of philosophy students and the milieux in which they moved was an essential explanatory component.   Unlike Hacking's book, Gellner's does not whiggishly see history as progressing towards the present, and does not respect boundaries (v) and (vi).  His attempt to show that the problem-situation cannot be adequately described without extra-philosophical factors is the single most controversial aspect of the book.  Of course, in a problem-centred and criticist view of philosophy, controversy is always a hopeful sign.

It is important to note that the view outlined here means there can be no definitive history of philosophy.   The historian is reconstructing when he focuses his historical accounts round a problem and its lines of solution and debate.  His reconstruction can be criticised on various grounds.  It can be found simply implausible.  It can be found to lack explanatory power, that is, to not adequately explain the undisputed historical facts.  It can be challenged by facts that do not easily 'fit' the reconstruction.  Above all it can be challenged by rival reconstructions that discern other problems or other debating lines in the evidence.  One of the thrilling experiences of reading Popkin is to see his reconstructions sending him in search of evidence in unlikely places.  Time and again he came not only upon evidence that served to solve mysteries, but upon new evidence that is recognizable and valuable to those with a narrower construction than he of what characterizes 'philosophy'.
Philosophy and its problems are historically structured, but what its problems are and what the structure is are matters open to free debate and discussion.  This ensures that desideratum (iii) is met.  The diversity of the endeavour is met by insisting that any putative order attributed to philosophy is tentative and subject to the precise criticism that it fails to account for this or that endeavour which plausibly counts as philosophy (iii) and (iv).

The present characterization of philosophy is criticizable in general and in detail, since it makes many claims that are far from trivial.  This criticizability of everything, although not all at the same time, is a central tenet of the Popperian outlook, and one of great power.  Popper wrote in The Open Society that our attempt to be critical rationalists is limited in the sense that we ultimately are making a rationally unjustifiable commitment to reason.  Bartley's point was that Popper's concession to the fideists was an unnecessary relic of the justificationist viewpoint.  HIs immense achievement the argument that the choice of the criticist viewpoint need not be irrational, and hence is not, crowns Popper's system of ideas:  A system that sees problems as the starting point, and problems as the point of return.