PHENOMENOLOGICAL INQUIRY AND PHILOSOPHICAL SELF·REFLECTION · enterprise, the most philosophically sophis-ticated development of 'conceptual analysis', might beseen tobe included.

Journal of tile British Society for Phenomenology, Vol. 10, No.3, October 1979.

PHENOMENOLOGICAL INQUIRY ANDPHILOSOPHICAL SELF·REFLECTION

K. DAVIES

This essay is intended as a step towards aphilosopher's Holy Grail. towards that ultimatestarting-point, presupposed by all other beginningsin philosophy, which has been the goal of somuch past searching inquiry. It takes the formof an examination of the necessary conditions ofthe meaningfulness of phenomenological discourse,and a meta-theoretical discussion of the natureof philosophical self-reflection, i.e. of philo-sophical reflection on philosophy

Merleau-Ponty remarks, concerning the Cogito,that "Descartes, and a fortiori his reader, begintheir meditation in what is already a universe ofdiscourse."? This indicates a direction for philo-sophical activity, in reflection on the conditionsof the existence of philosophy itself. which isdeveloped below; but first, to pre-empt someinitial misconceptions of the status there ascribedto questions of meaning and language, I shallbriefly distinguish that direction from otherswhich may come to mind.

I. It is not a direction which can be followedwithin the horizon of contemporary semantictheory which, as manifest in the work ofDa vidson, Grice and ·their many associates,tacitly bases itself on presuppositions whichlimit, in advance, its possible relevance tofundamental philosophy.

II. It is not a quasi-phenomenological descrip-tion, or 'showing', of language - gamessituated in 'given' forms of life so as toelicit what can and what cannot meaning-fully be said.

III. It is not, at least primarily, a theory of thatstructure of the conceptualisation of experi-ence necessary to the existence of anythingwe would call 'experience'; though this

enterprise, the most philosophically sophis-ticated development of 'conceptual analysis',might be seen to be included.

IV. It is not, though it is a prologue to, a dis-cussion of the essential relations betweenlanguage, man, and world. It pushes thetraditional demand for definition of termsto its limit, asking .for ·the necessary condi-tions of any meaningful discourse whatever.Only by first pursuing this demand can anysuch 'deeper' discussion acquire philo-sophical validity.

While none of these differential claims can beexplicitly made out here, they may gain in sub-stance from what follows. Two more preliminaryremarks. First: I shall be concerned with philo-sophical discourse in a way which might seemto beg the question of the possibility of non-discursive philosophical thought, This is for thesake of simplicity, since even if a case can beentertained for the possibility of private 'marksof concepts', as many past great philosophers(including Kant) seem to have done, the dualrequirements, explained below, of structure andcontent still hold: without them there is nothought at all. Hence, when speaking of 'philoso-phical thought' I have it in mind as embodied inlinguistic terms. and shall speak of it, as well asits linguistic expression. as meaningful. meaning-less, and so on. What is not presupposed here isthe existence of a full-blown linguistic communi-cation context, an embodied speaking subjectmouthing to an embodied hearing subject in theordinary world: for the time being Husserl'sassertion that the real physical existence ofan expression is not essential to the expressionas such, can be accommodated. Second: Philoso-phers of language have, since Frege, increasinglyfocussed on the sentence, rather than the word, as

1. M. Merleau-Ponty, The Phenomenology of Perception, Trans. Colin Smith. London: Routledge & KeganPaul, 1962. p. 401.

172

the basic unit of utterance, and, in a sense, ofmeaning. I shall speak directly of the meaningsof words, rather than of sentences containingthose words, but this is not important in anyway for my discussion.

What, then, are the necessary conditions of theexistence of philosophical thought or discourse?This might not seem to be a very promisingbeginning. for the very nature of philosophy hasbeen at issue in much of 'the history of philosophy',and any arrogation on my part of the right topronounce on its essential character seems un-seemly. Hence we begin with philosophicalthought as thought rather than as philosophical,and ask after the necessary conditions of itsexistence. An obvious start might appear to bethe inference to a thinker of the thought How-ever, in deference to what I take as justifiableHumean bewilderment, and with regard to theimmense complexity of the problem of the natureof such a 'thinker' (one of the central questionsin modern Western philosophy), I hold back fromthis move. There remains one simple essentialfeature of thought as such which signposts theway forward: thought must have meaning. Bythis I mean simply that a thought must thinksomething in order to be a thought, and that thisis only possible if the words which articulate ithave meaning. If they do not, nothing is said,or thought (or whistled) - there is no thought atall, nothing of cognitive value has occurred.s Thecharacter of this situation can be more strikinglyelicited by considering its explicit denial. "Thisthought is meaningless" can be ascribed the truth-value 'true' only on pain of accepting it as a pieceof gibberish. meriting no truth-value whatever.and little enough of our attention. It can becoherently evaluated only as false-

What, ,then, is involved in the meaningfulnessof philosophical thought? Firstly, I should saythat the confusion between semantic and prag-matic meaningfulness. which occurs for example

at key points in the later philosophy of Wittgen-stein, is to be avoided. Only the strict semanticmeaning of the linguistic articulation of thethought is to be considered. Secondly, to assertthat thought, to be thought, must be conceptuallyarticulated is just to reject, with philosophersfrom Socrates to Hegel and after, the notion ofpure knowledge as simple, unmediated, un-structured 'awareness' of the infinitely rich mani-fold of experience,' an 'awareness' which precludesany determinate consciousness, the thinking or say-ing of anything specific, i.e. of anything at all; andwhich leaves only the possibility of some mysticalinexpressible union with the world. Even philo-sophical knowledge, as Hegelian 'reason' whichgoes beyond the limits of the mere understanding,is determinate and conceptual in nature.s Thirdly:the linguistic expressions of these concepts, to bemeaningful and hence capable of articulatingmeaningful thought, must have both structureand content. Two aspects of meaning are impliedhere:

1. A word has meaning only in differentialrelation to other words in a system. It wasthe recognition of this diacritical structureof language (at various levels) that providedmuch of the impulse behind the "Struc-turalist Movement". Here, however. I wantto concentrate on a different aspect of mean-ing. For while the above assertion may statea necessary condition for the meaningfulnessof linguistic expressions. it is not a sufficientone. It is insufficient because words couldnot have meaning only by virtue of theirrelations with other words=-there would bea bare formal structure enabling nothing tobe said. thoughts embodied in suoh barren'expressions' would be about nothing. Hencethe second point:

n. A word has meaning only in relation tosome extra-linguistic realm. While some ex-pressions can be given their meaning wholly

2. Much has been made of certain paradoxes arising from attempts at philosophical self-reflection; that onehas to say what cannot be said, with Wittgenstein, think what cannot be thought with Kant. deny synthetica priori knowledge in a synthetic a priori proposition with the logical positivists, and so on. Rather thanmake a cult of such paradoxes, which are in each case founded on a prior metaphysical stance. I wouldstudy the necessary conditions of the philosophical discourse within which that metaphysics is originallyexpressed.

3. Reference might also be made to strains of Zen Buddhism. where the ultimate comprehension is attainedonly by first pushing the understanding to its limits.

173

4. E. Husserl, preface to the English edition of Ideas. Trans. W. R. Boyce Gibson. London: Allen & Unwin,1931. p. 27.

5. Op, cit. p. 196. Shorter Oxford Dictionary, Third edition. Oxford, 1959. p. 2239.

in terms of connections with other expres-sions-c-Bachelor' can be explained in termsof 'unmarried man'-it is notoriously thecase that not all words can be so fortunate.At least at the Quinean 'periphery' of thelinguistic framework there must be somedirect connection with a realm which con-fers content on basic expressions. and hence,via the interconnections, with higher-orderexpressions. Without such content we are leftwith an abstract sterile framework throughwhich, since it is connected with nothing,nothing can be said or thought.

So muoh, I hope, is ordinary and controversial.Where the difficulties emerge is with the questionsof the nature of this content-conferring realm,and 0'f the relation that holds between it and theexpressions upon which content is confer:red-the semantic relation between language and theworld. It is the first question which needs investi-gation here, leaving indeterminate. for the time,the precise nature of the semantic celation.

At this stage we might be wondering aboutthe possible relevance of this question for pheno-menology. Fortunately, the case is clear. For anabiding aim of Husserlian phenomenology, as acritical philosophy, was at least the examinationand clarification of all presuppositions whichcould not actually be dispensed with. Schuppehad already interpreted "freedom from pre-supposition" as acceptance of only those pre-suppositions which were necessary conditions ofthe possession of meaning and content by thephilosophical enterprise itself, and Husserl echoesthis as late as the preface to the English editionof Ideas, asserting that: "Philosophy can take rootonly in radical reflexion upon the meaning andpossibility of i,t1sown scheme";.l it reverts to "thatwhich is already presupposed implicit in all pre-supposing and in all questioning and answering,and herewith of necessity exists already, imme-diate and persistent. This is the first to be freelyand expressly posited ... "5 The necessary pre-supposition of a semantic relation between the

174

terms of phenomenological enquiry and some, asyet unspecified, extra-linguistic realm, is anobvious candidate for such reflexion. This isimportant for phenomenology because whateveris necessarily involved as a condition of themeaningfulness of the phenomenological enquiry,might subvert the results and the very method ofthat enquiry: a crucial possibility is that the sus-pension of the "natural attitude", of the naive"belief-in-the-world", could be undermined if theexistence of that world was necessarily pre-supposed as a condition of the meaningfulness ofthe enquiry itself.

Even as it begins the work of reduction, which isintended t0' prepare the way for the clarificationof all assumptions, phenomenology needs clari-fications of its assumptions: this vaguely para-doxical remark is to indicate that, prior to theinvestigation of transcendental subjectivity, inwhich terms alone true phenomenological explica-tion is to be sought, there is need of a form oftranscendental argument which traces back thechain of necessary conditions of the meaningful-ness of any phenomenological statement orthought. The first link in this transcendentalchain has been presented above, and it is theproblem of accommodating some of its conse-quences within the Husserlian framework that Ishall now examine.

Let me begin with some banal remarks concern-ing the explanation of ordinary words, To explainthe meaning of "tree" one could mention itsmeaning-connections with other expressions - atree is a specific sort of plant, a perennial planthaving a self-supporting woody main stem, andgrowing to a considerable height and size6 - but,as mentioned above, this form of explanationcannot be general, somewhere a direct connec-tion between the linguistic expression of a conceptand its instances must be established. With "tree"this would perhaps be done by pointing at varioustrees while uttering "That is a tree", and at other,vaguely tree-like, objects (telegraph-poles, etc.)while uttering "But that is not a tree". Whether

an adequate model of language -Iearning can beelaborated along such lines is not in questionhere, the point is only that, to explain the mean-ing of a word, only the indirect or, ultimately,the direct approach can be taken. Now, in theordinary everyday practice of meaning - explana-tion, a tacit assumption is that of the existenceof a transcendent, independent. objective world,related to our words, such that, in using them, wecan think and speak about that world and thethings within it. These include trees, lightningflashes rock music, people, colours, emotions,magnetic fields, after images and much more. Itis here that a difficulty is presented to pheno-menology. For if acceptance of the existenceof such a world were an essential presupposi-tion of phenomenological investigation itself,the transcendental-phenomenological reductionwould have none of that metaphysical powercustomarily ascribed to it Phenomenology wouldbecome the study of the essences within a broadrealist framework, the experiential version oflinguistic analysis. While such an apparentemasculation of the skeptical forces in pheno-menology might not unduly depress the manycritics of the later developments of Husserl'sthought, it does involve me in my concern toassess the difficulties facing a self-reflexive criticalphilosophy, faa: a phenomenology so modifiedwould hardly merit the epithet 'critical' in itscapitulation to a naive realist metaphysic. Thealternative then, the only means of critical sur-vival, is to arrive at a rival account of the natureof that content-conferring realm, semantic rela-tion with which is a necessary condition of themeaningfulness of phenomenological statement;an account which does not immediately pre-suppose the existence of the everyday trans-cendent world and its entities

How can phenomenology arrange this? Thesolution requires recourse to the phenomenologicalreduotion. Various interpretations of the reduc-tion exist, and it is not my aim here to fullytheorize its nature, but the following points seemcentrally involved. In the need to use only

apodictically given evidences, to preserve theabsolute purity of the phenomenological project,all presuppositions of experience and judgementmust be elicited, explicated and evaluated, nothingbeing taken for granted, and this involves thesuspension of the immanent claims of experienceto be experience of actually existing objectivitieswhich transcend that experience. This, in turn,involves the suspension of that general, implicit'belief-in-the-world' which is man's fundamentalcognitive orientation. This 'belief' is deactivatedby the reduction, enabling the phenomenologistto discover the transcendental ego as the truesubject of the stream of pure experience, and tofind the world and things within the world asphenomena. Objects are considered, not asactually existing independent realities, but onlyas intended in conscious experience. The correla-tion between every conscious experience of some-thing, and that of which there is, intentionally,experience - the noetic-noematic structure ofexperience - is brought to our attention, andthis enables eidetic intuition of the essences ofobjects, and the analysis of their intentional hori-zons, within the phenomenological understandingof objectivities as transcendentally constituted inand through the pure stream of experience. Nowthe 'objective' aspect of an experience, the in-tended as such, the noema, has a crucial role inHusserl's theory of meaning. Much has beenmade of the relation between Husserl's conceptof 'noema' and Frege's concept of 'Sinn? and asalient point emerging from 'Such discussion isthat the meaning of a linguistic utterance, forHusser], is the 'noematic Sinn' of an underlyingact of consciousness which is indicated by theutterance, and from which it gains its meaning.sLinguistic meaning, then, is the direct expressionof the 'noematic Sinn' of an underlying meaning-giving act, and this seems to provide a solutionfor the initial problem. Phenomenology escapesthe necessarry presupposition of the actual exist-ence of the world of the natural attitude as thatwhich gives content to its concepts, and hencemeaning to its discourse, by recourse to the idealrealm of noematic Sinne, the core of the 'object'

7. For example, of their 'ideality', being reducible neither to 'real' elements of subjective consciousness, nor toreal elements of the physical realm; and of their function as a form of mediation of the object to theknowing/speaking subject.

8. See, for example, Fellesdal, "Husserl's notion of noerna", in the Journal of Philosophy. 1969; and McIntyre& Smith. "Husserl's Identification of Meaning and Noema" in the Monist 1975.

175

----~-------------=~~--'------~-- -~--

end of pure conscious acts, of transcendentallypurified experience. The meaning of the expres-sion "tree", if it is to be given directly rather thanvia its relations to other expressions, is given interms of the noematic poles of various experi-ences which are experiences, intentionally, oftrees: phenomenology avoids the threat of ground-floor incoherence.

Unfortunately, we cannot pack up and go homejust yet. For what at first looks like a solutionproves. on further investigation. to be merely adeepening of the problem. If we take the call forthe examination of presuppositions seriously, wemust ask: of phenomenology an explanation ofthe meaning of "noema" or of "experience", orany of the other theoretical terms used in theinitial 'solution'. so as to clarify whatever pre-suppositions might be involved as necessary con-ditions of the meaningfulness of these terms. Thisposes. as I hope to show, an intraotable difficulty.

Let this difficulty be first manifest in connec-tion with "experience" (standing in also for "pureconsciousness", "conscious act" and related for-mulations). How can the phenomenologist explainthe meaning of this term? Not this way: "experi-ence is what happens when human (and. perhaps,other animal) subjects encounter the world; seeingis a kind of experience; hearing, feeling and soforth"; for unless the terms used here. "human","seeing". etc.. presuppose the existence of theordinary world of the natural attitude as theirultimate. meaning-giving realm - which is whatthe phenomenologist wants to avoid - they canonly be explained. as above, in terms of "experi-ence", or its "noematic" aspect. itself - the veryterm which is to be explained. Such an indirectexplanation of "experience" in terms of its rela-tions with these other terms results either in thefundamental self-defeat of the phenomenologicalprogramme. or in a vicious circle - "experience"being explained, ultimately, only in terms of"experience". The same goes even fOT such atheoretical indirect explanation of "experience"as "the object of phenomenological reflection":for it the latter is to be understood as other than"reflection on pure experience", which gives usback the problem, it can be understood only as"reflection on the residue of the phenomenological

176

reduction". This last can be explained only interms of the suspension of objectivity-claimsimmanent to experience and judgement, and theseterms can be explained only by presupposing theexistence of the ordinary world of the naturalattitude. or by referring to the noematic Sinne ofexperience again. Either way there is no escape.

The only alternative to this dilemma of self-defeat or vicious circularity is to try to explainthe meaning of "experience" directly; the pheno-menologist will have to point to instances of"experience" such that explication of the natureof these cases does not reveal them as essentiallymundane, but as independent of the presupposi-tion of the actual existence of the ordinary,mundane world. And this alternative seems.initially. fairly plausible. Having performed thereduction, after all, the pure stream of experiencein its living presence is presumed to be accessibleto the reflective gaze. What could be simpler thanthe mental ostension of one or more of theseliving moments of pure consciousness coupledwith the exclamation: "That is an experience't-e-thus establishing, by direct relation betweenword and 'world', the meaning of "experience".To focus on the real problems here, let us firstwaive any objections that might arise concerningthe presuppositions possibly involved in the per-formance of the reduction itself. and othersconcerning the linguistic abilities already pre-supposed in any 'ostensive definition' as such; for,without arguing the point here. I feel that mostof these objections could be met fairly comfort-ably within the problematic of transcendentalphenomenology. Let us also waive objectionsconcerning the immediate accessibility of pure'reduced' experience itself - which arose as earlyas Bishop Huet's criticisms of Descartes doctrineson this score - and concerning the possiblevalidation of the claim to apodictic, certainevidence on the part of reflection on such experi-ence; for these objections, though far moreserious, are not as fruitful, nor as demonstrablyoritical. as that with which I am here concerned.

For even apart from such problems, thereferent of the demonstrative "that" in the would-be ostension "That is an experience", or of anyother indexical element which is to play the same

9. Compare the case of "That is kind", said of an action, to explain the word "kind". For such explanationto succeed a prior understanding of the referent of "that" as an action, rather than, say, the colour of thegift, is necessary.

role, is not yet determinate, and cannot be madedeterminate without a characterization of itwhich will beg the important question. To eluci-date: in absence of additional elaboration, towhat can "that" refer in this context? What amI pointing out, ostending? Not a specific content,for experience as such is no 'thing', it is experi-ence of 'things' and not, for Husser! or any otherrelatively aware philosopher, another 'thing'alongside them. Nor is it some kind of semi-transparent container of contents, such as couldbe perceived independently of the contents. Whenthe phenomenologist turns 'from ordinary worldlythings to those things as phenomena nothingchanges: only transcendently existential judge-ments are suspended. There is no 'more' toexperience than its contents and structure, noother 'thing' to serve as the victim of an ostension."That" can succeed in its explanatory role onlyif the nature of its referent is already implicitlyunderstood. This pre-understanding, as a neces-sary presupposition of the success of the ostensiveexplanation of "experience", must be explicatedby phenomenology: but, to enable the appropriatedetermination of the referent of the demonstra-tive, this understanding must already be of thething pointed to, as not itself a content of therealm with which the ostension is apparentlyconcerned, but as a unifying awareness of suchcontents. That is, the ostension requires, as anecessary condition of its success, a pre-understanding of the ostended 'object' whiohalready makes the distinction between conscious-ness and that of which consciousness is conscious-ness, between experience qua experience, and itscontents. Nothing less will suffice to determinethe referent of "that" as an experience as such9(rather than, say, a specific content of an experi-ence); and this gives us right back the problemof explaining what is meant by "experience",since in explicating this presupposition of thesuccess of this attempt to explain "experience",the phenomenologist must already make use ofthe term "experience" or its phenomenologicalcognates ("consciousness", "awareness", etc), andhence. remains in the dilemma previously out-lined - either he presupposes the actual existence

of the mundane. or his circular "explanations"leave his statements without meaning, strictlynonsense.

Perhaps, though, there is a clue to a possiblesolution to be found in the rejection of this lastline of argument, For if ostension most naturallyseems to point out the content of an experience,and can be understood to refer to an experienceitself as such only if a prior understanding of thenature of experience exists, and if that content,after the reduction, is thereby the noematic aspectof the experience, then could not the meaning of"noema" thus be directly explained, and themeaning of "experience" also, as the correlationalunity of noema and 'noesis? Thus, the pheno-menologist says, or thinks, "That is a noema, andthat; that too" and so on. The problem here isnot, obviously at least, that the pre-understandingrequired of the ostended object already involvesan understanding of the nature of a noema, forthe ostension can be understood as ostending thatwhich is present before the (reflective) gaze andthe explication of this understanding does notseem to immediately entail a break with the basicphenomenological problematic. Other difficultiesemerge very quickly, however, and destroy ourhopes of success with this gambit: I shall focuson two.

To begin with, the applicability of "noema" isuniversal in the sense that every time one ostendsthe (intentional) object of an experience, "noema"will successfully apply. Such success, however,brings with it problems. For if "noema" is to beunderstood as applying to any (intentional) object,to anything experienced (and so to anything thatcan be ostended), then this must involve thedifferential understanding of "noema" as notapplying just to some specific kind of object:generality can only be understood in contra-distinction to specificity, and this requires the pre-understanding of objects as, say, trees, rocks,beetles, etc, whose elucidation and explication bythe phenomenologist will lead back to the sameold problem; "noema" is understood in (negative)relation to specific kinds of objects, trees, etc., and

177

r

the word "tree" is understood either by pre-supposing the existence of the mundane realm ofthe natural attitude, or circularly 'understood'via the concept "noema".

Secondly, "noema" has to be understood in away which enables it to playa role in the explica-tion of "experience", and this can happen only ifit is distinguished from other terms of similarlygeneral application, like "something". In short,"noema" must be understood as applying toevery intentional object rather than to everyobject, tout court since otherwise we can notsecure the desired freedom from mundane. non-transcendental presuppositions. But such under-standing could be gained only through apre-understanding of the reduction itself (whichhas so far been assumed to have been. de facto.accomplished) - elucidation of which wouldagain return to us the initial problems - orthrough some other understanding of the distinc-tion between consciousness and being whosepossibility, in its transcendental purity. is now inquestion. Even after all this remains the questionof explaining "experience" as the correlationalunity of noesis and noema. How could we explaint·he character of this unity; how to' explain themeaning of "noesis" - it seems clear that "ex-perience" cannot be understood in terms of theconcepts of the constituents of experience, butvice versa; and still no satisfactory way of explain-ing this term has been found.

The upshot of this wrangling is that the neces-sary condi tions of the meaningfulness of pheno-menological discourse cannot be accommodatedwithin the confines of the critical phenomeno-logical step of the transcendental reduction:ordinary terms are meaningful ultimately in rela-tion to the pure realm of transcendental subjec-tivity, but the understanding of the vocabulary ofthe <transcendental in which such an account isformulated is either circular - hence no under-standing at all- or involves reference to thedecidedly non-transcendental mundane world ofordinary experience.

The reduction now seems a peculiarly futile.

self-defeating gesture, since the meaningfulness otits formulation. and of the discourse which isenabled by it, requires the existence of thatordinary world of the natural attitude, belief inwhich is to be disconnected in the reduction itself.The meaningfulness of the claim that with thereduction to pure consciousness and the trans-cendental ego the phenomenologist gains accessto' absolute existence, presupposes the existenceof that ordinary world Such, at first sight, arethe fruits of philosophy; but before undertaking adeeper examination of the conclusions at whichwe seem to have arrived, it is necessary to con-sider what HusserI and his more faithful followerswould see as a crucial objection to the argumentsabove.

A vigorous formulation of this objection,underwritten by Husserl himself, appears in theKant-Studien article "Husserl's Philosophy andContemporary Criticism", by his assisant EugenFink. For Fink. and so for HusserI, pheno-menology has from the beginning "a certainunintelligibility precisely because it cannot, inprinciple, be grasped with reference to' mundaneproblems; with reference, that is, to questionswhich stand in the horizon of the 'natural atti-tude'. Its basic problem is concealed in this way:it is at first not an unsettling problem which issomehow present before phenomenological theory,so that by virtue of its threatening character itcan serve to' provoke philosophical reflection. Itfirst originates as OJ problem in and through thephenomenological reduction itself, which isalready the first step to be taken in masteringit." 10

Mundane forms of the question of the origin ofthe world can be at most "symbolic anticipations"of the transcendental problems, of the world. whichitself can be posed only by transcending theworld through the phenomenological reduction.This performance of the reduction, then, cannotbe adequately comprehended as a possibility ofhuman existence, nor as motivated by any prob-lem given within the horizon of the naturalattitude. This rather excessively esoteric situa-tion involves the phenomenologist in a certain

10. E. Fink. "The Phenomenological Philosophy of Edmund Husser! and Contemporary Criticism". Translatedin R. O. Elveton. The Phenomenology of Husserl. Chicago: Quadrangle Books, 1970, p. 101.

178

c

pedagogic paradox. Since, as Fink continues, "thereduction i~ its own presupposition insofar as italone opens up that dimension of problems withreference to which it establishes the possibilityof theoretical knowledge," any attempt to expli-cate the phenomenological reduction to someonewho has not himself performed it, as an attemptto lead out of and from the natural attitude to' thetranscendental attitude, is in a unique way false:"no exposition which takes the natural attitude asits starting-point can satisfactorily explicate thatreduction comprehension of which can emergeonly after its performance."ll Here is the pheno-menological equivalent of Hegel's Owl of Minerva.The most a phenomenologist can do for the un-enlightened is to point, vaguely, to symbolicallyanticipate, in an attempt to effect some sort ofleap from the natural to the transcendental atti-tude.

If, as successful noviciates, we manage theserites of passage, we are then in a position tounderstand this didactic difficulty as a special caseof what Fink leaves to us' as "the paradox of thephenomenological statement." For the pheno-menologist, there is both a mundane and a trans-cendental meaning to his words. The crucialterms "existence", "experience", "ego", "inten-tional", etc., have a meaning acceptable to' thosewithin the natural attitude, and another accessibleonly to those who have achieved, via the reduc-tion, the transcendental viewpoint. His problem,then, is to express the results of his phenomeno-logical analyses when the only concepts at hisdisposal are worldly, mundane concepts, andFink concludes that "For this reason no pheno-menological analysis . . . is capable of being pre-sented adequately.v-s This seems to remain, forFink, essentially only a pedagogic problem. "Thisinadequacy Qf aLI phenomenological reports,caused by the use of a mundane expression fora non-worldly meaning, also cannot be eliminatedby the invention of a technical language. Sincephenomenological communication is chiefly acommunication to the dogmatist, such a languagewould be devoid of meaning. Phenomenologicalstatements necessarily contain an internal conflict

1 J. Op, cit. p. 105.12. Op, cit. p. 143-413. Op, cit. p. 144.

between a word's mundane meaning and thetranscendental meaning which it serves to indi-cate. There is always the danger that the dog-matist will grasp only the mundane meaning ofwords and overlook their transcendental signifi-cance to such an extent that he will imagine hismistaken explication of phenomenology to becorrect and capable of calling upon the text forits justification."19 This passage is revealing in anumber of ways. Firstly, the difficulty is greaterthan Fink suggests: there is not merely a dangerthat the dogmatist will overlook the transcendentalfor the mundane significance of the phenomeno-logical word, but a necessity; for the dogmatist.as such, has nO' access to' the transcendental realmof meaning - the only communication worth theeffort seems to be that inductive of the perform-ance of the reduction, only then can the audiencegrasp any of the real meaning of phenomeno-logical utterances. More important, here though,is the question of the possibility of a purelytechnical language, adequate to the expression oftranscendental meaning, available for the self-expression of, or communication between, pheno-rnenologists themselves; and we already havereason for suspicion on this point. FQr even if weallow that an understanding of the transcendental,and of the mundanity of the natural attitude assuch, can come only from a standpoint exteriorto that attitude; and even if we accept thatsuch a move from the interior to the exterior isin some sense a possibility - though comprehen-sible as such only from the position of thetranscendental attitude - how could even thethe strictest phenomenologist explain his trans-cendental vocabulary? A phenomenologicalthought must be specifiable in terms of its differen-tial relations with other possible thoughts, and beindirectly or directly given some content. With-out these aspects of structure and content it is nota thought, however transcendental its import. Butwe have already seen some of the snags inattempts to' explain the meaning of the terms inwhich the phenomenologist thinks and expresseshis thoughts. Without that reliance on the mun-dane world of the natural attitude which thephenomenologist seeks to eschew, the only way

179

of giving content to any of the expressions of thetranscendental already presupposes a comprehen-sion of the transcendental. Nothing but a purelycircular conceptual explanation is available; onewhich makes no contact with anything beyond thatcircle, with any extra-linguistic realm - not eventhat of transcendental subjectivity itself, since tounderstand any 'ostensive definition' of a trans-cendental term again already presupposes anunderstanding, if only implicit, of the transcen-dental as such. The only escape route now seemsto be a recourse to an almost Tractarian mysti-cism; expressions of mundane concepts throughwhose semantic connection with the mundaneworld the transcendental terms gain their content,are used, like Wtittgens.tein's ladder. to reach astage of ineffable and strictly unthinkable. un-sayable, 'insight' which is then rendered indepen-dent of the means through which it was achieved-the ladder can be thrown away.

If this state of inarticulable meaningless 'aware-ness' is the end result of the Fink-Husserlcharacterization of the nature of the reductionand of the transcendental attitude - despite theirdisavowals of mysticism - it might be more fruit-ful to follow other paths. But where might thesebegin, and where could they lead? Are we justled to a grateful recognition of the status ofa realism to which we return, humbled andchastened, no longer naively unaware of criticalphilosophy, but with a mature understanding ofthe value of the natural attitude from which werebelliously attempted to free ourselves? For-tunately not. Or do we depart from Husser! withHeidegger, who, with his intuition of the depend-ence of the theory of transcendental subjectivityon the ordinary world of everyday experience,sought a hermeneutic explication of the essentialprimordial structures of that Being-in-the-worldfrom which the pure transcendental ego was aderivative abstraction? Again. the answer is no.and for essentially the same reason. For despitethe height at which Heidegger, in philosophicaland methodological rigour and insight, stands inrelation to his Anglo-American contemporaries.he treats the transcendentalism of oritical philo-sophy as essentially only a philosophical digres-sion. heuristically but not dialectically relevantto philosophical progress. But such a summary

180

dismissal leaves us with a broad realism which.if not, as Husserl commented cn Heidegger'swork, merely naive. still leaves many importantquestions unanswered.

Even if the transcendental positivism of Husser-Han phenomenology must be rejected, and with itthe possibility of general scepticism which attendsthe universal epoche, the negative critical tendencyof that philosophy remains almost entirely un-affected. For if general scepticism concerning theexistence of the 'external world' is no longertenable, since strictly either- meaningless or false.there never-theless survives an infinity of possibleparticular scepticisms. possible limited epoches.Every aspect of belief and knowledge can besubject to the critical-sceptical response. withinthe bread horizon of the acceptance of the exist-ence of that 'external world' of the naturalattitude. Some of our concepts, it seems, musthave genuine application, to provide for the verypossibility of meaningful statement, but we haveno principle for deciding which are the privilegedones: certainly no general principle of arguingfrom paradigm cases is acceptable. and any speci-fic claim to knowledge is left under the threat ofsceptical attack. Some of the force of the argu-ments against the phenomenological projectseems thus to have dissipated. Its methodologicalequipment is intact: the phenomenologist nowworks within the bounds of specific epoches,using his eidetic intuition, undertaking constitu-tional analysis and so on. producing all thesubstantive results of phenomenology whileabandoning the embarrassing transcendentalmetaphysics. Or. in a different philosophicalclimate but within basically the same theoreticalframework, the 'linguistic phenomenologist' pur-sues the description of language-games and itsimplications about what can or cannot sensiblybe said when playing them. These reactions tothe defeat of critical phenomenology are certainlyseductive of those who could not accept Hnsserl'stranscendental interpretation of his earlier investi-gations, and of those who altogether dislike thevery possibility of metaphysics; but such conquestis gained only at the price of having little ornothing to say about the essential relations be-tween existence and essence, between languageand the world between the knowing subject and

the known object, between theory and practice,and so on - questions to which transcendentalphenomenology at least addresses itself and offerssome important answers, but which the gratefulacceptance of the broad unselective realism I havedescribed permits only those formulations and'solutions' visible from its own inadequate andpre-given metaphysical stance.

We seem to have lodged in a dilemma: eithera philosophically impoverished realism, or arigorous but ultimately incoherent transcen-dentalism. In the second part of this paper I shallconsider the way out, and examine a fundamentalpossibility of contemporary philosophy as itappears in this context.

II

Here I shall return to the question of philoso-phical self-reflection posed at the beginning ofthis paper. First, the history of modem philosophy.Since Descartes philosophy has mostly begunwith reflection on something other than philosophyitself, namely experience. By such means anapodictically certain starting-point is to be esta-blished, from which much else is to be derived:the existence of God and the external world inDescartes, the necessary forms of objects ofthought ill Kant, the location of all modality inthe subject again in Hume, resulting in a catholicscepticism; the internal contradictions leading in-evitably to the self-knowledge of the Absolute inHegel; the transcendental phenomenology ofHusserl, the neutral monism of James, Mach andRussell, the positivist empiricism of Ayer; pheno-menalism, even the hermeneutic phenomenology ofHeidegger - all these have a positive beginning inreflection on experience.r- Naturally an adequateevaluation of these frameworks and their failuresis somewhat beyond the scope of this essay, but aproblem that accompanies any such positive start-ing-point is that of the theorization of the starting-point itself, Where is the justification for theclaim to the apodicticity of reflection and reflec-tion and experience? How can we be certain thatwe can be certain here? Answers cannot be foundafter the beginning. since they will presuppose

that beginning in its indeterminate epistemicmodality. But how could they be given priorto the beginning? What alternative could therebe to critical reflection on experience as thestarting-point of philosophical enquiry, if thatundiscriminating thick-skinned realism referredto above is to be avoided? It is here that theproject of philosophical self-reflection recommendsitself. but what must be meant by this?

Not, surely. just looking at other philosophers.at the 'content' of their work. but rather elicitingthe essential features of philosophical thought.and thence the necessary conditions of its exist-ence. But surely philosophy has to exist. beforewe reflect on it; don't we need something therebefore us to work on? A purely self-reflexivethought. for example, "This thought is beingthought" has no essential determination, in termsof what the thought is about. or thinks and canonly enter a infinite process of self-reference. Anydetermination of the thought must be external,for example, "The thought being had at 1.30 p.m.on Tuesday", but here the identity of the thoughtis fixed at the cost of adulterating its purity as aphilosophical beginning. The thought itself hasno specific content, it forever refers to itself with-out any concrete determination of what 'itself' is;and in being thus indeterminate, contentless, it isnot a thought. as such at all. The point hereis that the self-reflection of philosophy cannotbe pure, immediate. it must be mediated.But in what way? And if reflection onphilosophy itself is to be mediated by reflectionon something other than that reflection, are wenot back with the problem of a purely positivebeginning for philosophy? In which case is notphilosophy condemned to reflection on that posi-tive moment, and reflection on that reflection, andso on, caught in an endless orgy of eternal return.unable to achieve that absolute yet mediated self-reflection which ties up aU the ends and establishes.philosophy for once and for all? It is to thesevery abstract questions that I hope to sketch apossible approach. The heuristic guideline is thenotion of a form of inquiry whose findings willapply to that inquiry itself: the essential featureof any inquiry is that it be meaningful, and this

14. I realise that these brief characterizations, especially those of Hegel and Heidegger, beg certain centralquestions; for now,at least, they are left begging.

181

is all that can be specified in advance concerningthe philosophical inquiry.

Let us begin, then, with the question of thenecessary conditions of meaningful discourse, andsee how this question fulfils the requirements forthe beginning of philosophy. How does it avoidthe need to establish the existence of that of whichit seeks the necessary conditions? Talk of thenecessary conditions of experience can only pro-duce interesting conclusions if we have alreadyassured ourself of the existence of experience,which involves the difficulties already referred toconcerning the positivity of such a beginning. Thequestion of the necessary conditions of meaning-ful discourse is in the same logical situation, buthere the existence of meaningful discourse is pro-vided for by the question itself, and secured bythe fact that there can be no coherent denial ofthe existence of meaningful discourse, since anysuch denial must, as such, presuppose its ownmeaningfulness. The initial attractiveness ofthis question has of course been recognized byphilosophers, mainly in the present century, butwithout being afforded the rigorous examinationit requires. IA second essential characteristic ofthis question is its self-closure; that is, any dis-cussion of the meaning of the first question itself,or of its constituent terms, will follow the samecourse as that undertaken and followed by theinitial question. Any other questions, includingother philosophical questions, the relation betweenmind and body, between knowledge and action,even between language and the world are broughtback to the beginning by strict consideration oftheir meaning and the necessary conditions oftheir meaningfulness.15

At least two queries might be raised here.First, is this investigation of the meaning ofphilosophical questions not merely a preliminarymethodological step, establishing the meaning ofyour terms as a precondition of fruitful argument?Second, how is the sceptico-critical attitude inphilosophy to be accommodated by the projectso far sketched? The replies are interrelated. The

sceptical attitude in philosophy is rooted in therecognition of the possible disparity between howthings are experienced and how they are: so thatwhat we may be utterly convinced of might notreally be the case. This possibility is first broughtto bear, in the context of the search for theconditions of meaningfulness, on the nature ofthat extra-linguistic realm which provides contentfor linguistic expressions; and the first part ofthis paper was concerned to' show the ultimateincoherence or unintelligibility of the positivemetaphysic, as well as the sustained force of thenegative critical challenge. once limited to ageneral acceptance of the existence of a non-immanent 'content-conferring realm'. No area ofphilosophy can be adequately grounded withouttaking up this challenge from the beginning, sinceotherwise it runs the risk either of underminingits own results through the exigencies of justify-ing its progress against scepticism, or of restingon a naive or dogmatic foundation which vitiatesany claim to philosophical truth. The only wayto avoid both this, and the Ieap into transcendentalaphasia associated with the positive aspect ofcritical philosophy, is to elicit the necessary con-ditions of the possibility of critical enquiry itself.This will involve, essentially, discovering thenecessary conditions of the meaningfulness of theterm "experience", or "consciousness" or any ofthe other terms (e.g. "cogitatio", "impression","idea". "sense-datum") which have played thesame role in philosophy.

H this can be achieved philosophy will havebeen inoculated against. scepticism, since any suchattack would be an attempt to' question the neces-sary conditions of its OWiD meaningfulness, whichis not a coherent strategy. Given the earlierarguments concerning Husserlian transcendentalphenomenology it can be seen that such assur-ance can be bought only at the price of breakingout of the critical philosophical problematic: butthe fact that the critical philosophy will havebeen not merely dismissed, but dialectically trans-cended, means that instead of the hapless realismdiscussed earlier, a determinate denial will be

15. These two characteristics of the initial question are closely related to that 'self-referentiality' which R.Bubner sees as essential to transcendental philosophy. See R. Bubner, "Is transcendental hermeneuticspossible?" in Manninen & Tuomela (Eds.), Essays on explanation and understanding. Dordrecht, Holland:D. Reidel, 1976, pp. 59-77.

182

I

effected. That is, rather than a pure, starkrejection, .a dialectical overcoming of the criticalapproach will advance to something important inits own right: the negative moment in Westernphilosophy will have been taken up and trans-

formed into something positrve: namely, thatstructure of the world necessary to the meaning-fulness of all discourse, and to the fullness of allthought.

University of Durham

183