Last time I attempted to sketch the general pattern of the human ‘plight’ or ‘predicament’, which we connected, largely in order to draw profit from traditional associations, and with an eye particularly to the later, harder part of our task, with the Platonic myth or metaphor of the cave. Odd as it may seem, we cannot help seeing our whole experience, with its complex net of real, personal and intermediate fixtures, as being but one alternative among others, an alternative that we can describe with varying types and degrees of dismayed or merely neutral suprise, much as we can record the odd qualities and dimensions of an empirical object. Only whereas, in the latter case, it is always possible to say with just what alternatives we are contrasting it, and in what ways it is remarkable, it is not easy to say why we feel this in the case of some basic feature of the human predicament. Plainly we ought to feel ourselves so much part and parcel of ordinary this-world situations that we should barely be bothered to speak about their fundamental features: to speak of them as a ‘predicament’ has every mark of being a confusion or an affectation.

However this may be, we seem always to have before us factors in experience which, though intimately engaged and combined, none the less have much that makes them seem mutually resistant and disparate, and which therefore often make us feel that we should like one or other of them to be wholly excised from the picture, or to be shown to be a mere offshoot or disguise of the others. ‘A is only a trivially different case of B’: that seems the solution for which we are always sighing, and for which we sigh in virtue of its real logical value. We live in a world where the smooth face of public reality suffers from the strange mirrorings and duplications of contrasted privacies, which force us to conceive the same things a million times over, where the smooth life of the schizoid ego suffers irruptions that it can never completely enamel over with its rational nacre, where abstracta and verbal formulae, the most derivative of things, none the less hold the high stars in fee, everything being geared to preserve the most abstruse scientific constancies, and where things are alike surprising in their fulfilment and frustration of our most deeply cherished presumptions and demands. All these things, it would seem, could be wholly otherwise, and, if they were otherwise, they would also be less perplexing and less oppressive. And because, it seems, they could be otherwise, we can also by contrast describe them as they are, and it is just this sort of ‘description’ of the various fixtures of the human predicament, and of the way in which they fit together, or blatantly fail to fit together, that we shall now attempt to carry, out in more detail, as a possible preparation for a final solution which may well carry us beyond all such fixtures.

But before we embark on such detailed ‘description’, we must to some extent explain and justify the methods that we propose to follow. This is needful, because they differ considerably from the argumentative, conscientiously commonplace methods at present much in vogue in British philosophy, and because what we have to say might otherwise seem unduly poetic, dogmatic or oracular. Philosophy, on our view, cannot be conducted without a creative use of words, the joint issue of the phenomena and ourselves, and it also requires certain initiatives in the inferential transformation of our assertions. That such moves and procedures can be deeply responsible, and not merely extravagant, is what it is important to establish.

We said in our first lecture that we proposed to investigate the human cave in a phenomenological manner, that its very character as a cave meant that it must be investigated phenomenologically. This means that we must say what factors and objects and principles really count in human experience, and precisely as what each of them is experienced, before we offer any analysis, let alone any explanation, reductive or otherwise, of them. A phenomenological treatment may therefore be equated, in what, by persuasive definition, we may call the ‘true’ sense of ‘empirical’, with an empirical treatment, and this we may equally oppose to a phenomenalist treatment, on the one hand, or to a realist treatment, on the other. It is empirical, because it studies how things actually do come before us in experience, and not how, if we accepted certain analyses or followed certain difficult procedures, they could or should come before us, and because, also, it studies things as they do come before us, and not as, in their intrinsic being, we think they really must be.

A phenomenological treatment is opposed to a phenomenalist treatment in that the latter has a sophisticated theory of ‘immediate data’, of ‘what is immediately given’, of ‘direct experience’, which is open to much objection, and which is arguably based on an illicit hypostatization of a more or less obscure abstraction. For the phenomenalist the data, the things given, are not the matters we see before us, the matters we imagine, or the matters we conceive of or believe in: they are a somewhat bare residuum, attained by peeling off from what we have before us whatever is not grossly palpable, all that does not impinge on us in a peculiarly pictorial, illustrative, concrete manner. For the sofa actually in front of us they substitute an immediate datum which covers only the side of the sofa turned towards us, together with some index of the angle and distance at which we see it, for the recalled mountains, ships and lakes of our recent holiday they substitute a set of strengthless, depthless ‘images’, for the situations we read of in some work of fiction or history, and which are perhaps more vivid to us than the objects around us, they drag out, perhaps, a quite unnoticed array of half-formed, pictured words, for the fully-fledged attitude we note as taking place within us, they perhaps substitute a series of throbs, aches and thrills on which a certain diagnosis has been put, and so on. Philosophers who shun this sort of reductionism in the case of sensibly perceived bodies, are quite ready to adopt it in the case of our dealings with remote and abstract objects, or in our awareness of our own inner attitudes and emotions. It is not our aim here to undermine the distinction between what is grossly, illustratively present before us, and what is before us in what we can only characterize negatively as a non-intuitive, impalpable manner. Difficult as it may sometimes be to apply such a distinction with precision, and hesitant as may be our relegation of given features of objects to one or other side of it, it is none the less a distinction quite firmly drawn in certain paradigm cases, and then, on reflection, and with varied unsureness, extended to others. No one should doubt that there is a big and important difference between the fully painted, seeing manner in which the front side of a sofa comes before us, and what we may by contrast call the ‘skeletal’ manner—since we say we feel it ‘with our bones’—that we ‘know’ it has an inside and a back. But a sofa partly painted over in rich detail and partly felt in or with our bones, is not two sorts of sofa more or less loosely associated, not an interpretation added to a datum, not a concept ‘applied’ to a fully-fledged ‘intuition’. It is like the superimposed contributions which make up a Japanese colour-print, the one yielding an outline which the others fill in with various colours, and the first being intrinsically apt to receive the others and the others intrinsically fitting into the first. What one has before one (except in a case of gross mistake), is always a sofa or some genus of which a sofa is a species, and what differs in successive situations is the degree or precise character of the illustration, the concrete ‘carrying out’, which this sofa receives. The phenomenological genius of Husserl has here coined the fine word ‘fulfilment’ (Erfüllung) to which the term ‘emptiness’ yields a suitable antithesis. Something may come before us in empty or fulfilled fashion, and almost all things in fact come before us in a combination of the two manners. A totally unillustrated presentation is probably an Unding, as a fully illustrated, comprehensively seen object probably also is. The mistake, however, is to try to tear illustrative material from the thing or situation it imperfectly illustrates, to make it a thing or situation in its own right: this is more absurd than trying to tear words from the paper on which they are written and whose blanks they fill in. What one can significantly abstract from the sofa that stands before one is the sofa qua illustrated, but this is plainly a mere unsaturated (to use Frege's word) ‘side’ of the phenomenon before us, and not anything independently given. If such abstracta are not even genuine data, they are even less to be regarded as independent existents out of which common-or-garden realities can be ‘constructed’. This is not, however, at present the issue, but that the genuine phenomena are things as they come before us, and not merely artificially abstracted aspects of things. Thus the sofa comes before as having a rear and an interior, which may, according to our knowledge, be highly definite or largely indefinite: in either case it leaves room for subsequent fulfilment and illustration. What we have said are the merest commonplaces in the writings of Husserl, though not yet fully understood and adopted in our British thought.

If the phenomenological method is not phenomalist, it is even less realist or reist. It is correct to hold that no real object, no res, can be perfectly determinate in certain respects while only determinable in others: the very idea of reality, or at least one of the many ideas of reality, is such as to exclude such a possibility. But the things in the cave, qua things in the cave, qua phenomena before us, are all thus merely determinable: they are given as being definitely this or that, but for the rest as having a mere ‘horizon’ of determinable properties which subsequent discovery or specification will render more definite. There are certain philosophical conundrums here which it is not opportune to probe at the moment. I am not, however, committed to saying that the object that comes before us, that is experienced, is necessarily other than some real, fully determinate object: what is only determinate in one context may be fully determinate in another, and determinableness in the one context in fact includes as part of its content that it will or could be determinate in the other. Common sense realizes that an object ‘as we know it’ may be less than an object ‘as it really is’, without therefore amounting to another quite different object. We shall return to the difficult logic of this situation at a later point in our treatment: here we are only concerned with our right to place among the indefeasible phenomena, things which, to the extent that they do come before us, cannot be specified as full realities, capable of independent being. This applies not merely to the determinable objects we have been mentioning but to many objects to which it is doubtful whether an independent status should be conceded. There can, it may be argued, be no detached universals: this does not prove that being red as such is not something that can be independently dwelt on, that the Cheshire cat's grin is not, qua phenomenon, detachable from the cat. We do not assent to Hume's principle that what is impossible in reality must also be impossible in idea. There can, it may be argued, be no negative facts: this does not prove that Mother Hubbard and her dog did not encounter a most distressing negative phenomenon when they opened the cupboard, which admits neither of metaphysical nor psychological reduction. Heidegger's withers are unwrung by all those sunny analyses which prove that nothing, or the total absence of anything, is not a genuine object, and that it is not therefore possible to feel dread in the face of it. Nothing or the total absence of anything is a genuine object of contemplation and of varied emotional attitudes—much of the exquisite culture of Japan, for instance, seems to be built around it—and it is much more certain that this is so than that some piece of analysis is a correct one. I am not here maintaining that statements as to what appears or is given to us are incorrigible: I think it very likely, from my own experience, that there are appearances of appearances to quite a number of removes. Sometimes I am so sceptical about something, that I not only want to say that there only seem to be cases of it, but even that it has only seemed that there seemed to be such cases. What I am here maintaining, however, is only that there are some appearances so indubitable that they cast doubt on any theory of consciousness or of language which proves that there cannot be such appearances. (Thus the facts of three-dimensional vision disprove Berkeley's non-phenomenological arguments that it is impossible to see a direction at right-angles to the fund of the eye, and not vice versa.)

The phenomenological method therefore unveils much that is not, in the narrowest acceptation, present to our senses, and it may unveil much that, on reflection, is held not to have being or to be capable of having being. That all this involves deep puzzles is undoubted, and these are among the cave-problems that we shall have, at a later stage, to unravel. We turn, however, to another feature of the method of Husserl which is firmly part of our method, and that is its universal or eidetic character. Phenomenology is of types or ειδη, and never of individual specimens. Whether the experienced features we consider are as generic as those usually considered by Husserl or as specific as those considered by later phenomenologists such as Heidegger or Sartre, it is never the individual instance which counts, but only its ειδος or principle. This is of course true of the older phenomenology of Hegel who, when discussing Stoicism, e.g., as a mental phenomenon, is discussing it as a typical attitude of mind, and not merely as a particular historical attitude which was widespread in the early years of the Roman Empire. This means that our ‘descriptions’ of phenomena should be of typical or essential features, that they should not be, in the ordinary sense, descriptive at all. In delineating the human cave we shall want to sketch Body as such and the role played by Body: we shall not get down, except by way of illustration, to anything so specific as felspar, much less to an individual conglomerate such as the Giants’ Causeway. We shall likewise be interested in the genus Natural Genus as such, rather than in any specific natural genus such as Hydrogen, we shall be interested in such things as Vital Spontaneity or Temporal Continuity or Significant Reference or Arbitrary Choice or the Worshipful as Such. We shall be interested in all these as Ideas, rather than in any particular example of such types. Even if like Sartre we carry our phenomenology so far as to deal with a fictional situation in which one eye spying through a keyhole meets another eye spying back upon it, we shall still be studying the species rather than the individual. It will be the horrible general possibility of persons thus mutually intruding on each other's privacy that will concern us, and that as a peculiar limiting case of our life with others, rather than the real danger of particular incidents of the kind in question. The notion of an eidetic description, of a pure envisagement of essence, is, however, fraught with profound difficulties: it seems, on the one hand, to involve an arbitrariness which is quite alien to the nature of description, and, on the other hand, a necessity which seems equally alien. How can we be said to describe, to characterize what we largely make up, and how can we be said to describe or characterize what cannot be otherwise, what is a presupposition of meaningful discourse in a certain field?

The basic problem of ειδη, of generic and specific essences, is one which the ancients never properly tackled, and which even Platonizing thinkers like Husserl have left practically in their virgin state. This is the problem as to how we distinguish a genuine ειδος, a unique, indispensable, fundamental type from thought-contents which are merely factitious or complex or hybrid or derivative, or which represent a merely deviant or imperfect ‘case’. In dealing with the structure of our experience we shall again and again have to say that this or that is not part of the ‘idea’ of something, or that the ‘idea’ of something requires or excludes or favours or is unfavourable to the instantiation of some other ‘idea’. How shall we distinguish authentic ideas, which ought to concern us, from counterfeit ones which ought not, if in all cases we can shape our thought-contents and general meanings quite arbitrarily?

This is of course the problem that the Platonic Academy investigated under the title ‘Of what things are there no forms?’ We are told fairly reliably that they concluded by holding that there were no forms of composites, hybrids, artefacts, accidents, inessential relations, negations, things mutable and things due to choice or chance, and that they restricted the eidetic status to such substances as occur in nature—presumably the geometric forms were included among these—and to their various natural excellences. Of these doctrines there is no explicit statement in the Platonic dialogues, though the practice of some dialogues, e.g. the Timaeus, clearly implies them.

Aristotle likewise built upon the notion of an ειδος, and his whole physics and natural history involves that things have ‘forms’ or ‘essences’ in terms of which their main behaviour can be understood. Little time is, however, devoted to the epistemology of such forms or essences, to telling us how they come to be picked out, and how they come to be distinguished from the odd assemblages of accidentally connected meanings, e.g. ‘white man’ or ‘goat-stag’, which we can always, if we like, make to be the senses of specially chosen words. It is only by implication that we gather that νουσ, Aristotle's supreme cognitive faculty, perhaps aided by preliminary επαγωγη and assisted by αγχινοια can discover what forms there really are and how they are to be defined: possibly the υποθεσεις of which Aristotle speaks perform the former function, by which there also comes to be a distinction between real and merely nominal definitions. At any rate the whole matter is wrapped in such obscurity that some commentators have seen no special connection between definitions and the operations of νους. Yet if forms occupy the central position in Aristotle's thought, then the hypotheses that there are such and such forms, and the definitions which state their make-up, are the central operations in thought and knowledge, beside which all postulation of axioms and all deductive demonstration pales into insignificance. And while Aristotle certainly implies that this is the central noetic performance, he says extremely little about it, particularly in its negative aspect. He does not instruct us how to distinguish a mere collocation of accidents, or a form plus an accident, from a true form.

Descartes in his Rules for the Direction of the Mind makes a certain sort of deep integrity and simplicity the hall-mark of those ideas which we do not ourselves arbitrarily frame, as opposed to those which mere experience and the course of nature have contingently assembled. It is this deep integrity which distinguishes the idea of God from that of a centaur, the idea of a circle from that of some disorderly squiggle, and the idea of a thinking, or the idea of an extended being from the idea of something which combines them both. Descartes did not, however, make his distinction really clear, and the notion of ‘simplicity’ can only provide leaking coverage for a would-be privileged set of instances.

Later empiricist thought gradually blurred the distinction between ‘real’ essences and arbitrary or ‘nominal’ ones, though it never failed to appear in their practice, which always went for fundamental notions and presumed they had a stateable structure. Only in recent times has so fundamental a denial of essence been possible as is represented by Wittgenstein's doctrine of ‘family relations’, it being held that all the important words in our language cover a ‘family’ of cases having no genuinely ‘common’ element which runs through them all, but only a series of overlapping resemblances and analogies, so that A and B have certain resemblances, B and C only some of these resemblances and again others, C and D again only some of the resemblances of B and C and again others, until in the end we come to terms so separated that they have nothing at all in common. It is far from clear what this celebrated doctrine really means. Does it mean to say that all the general terms in our language merely cover ‘family resemblances’, and that they cannot possibly do otherwise? In this case it is an incoherent doctrine. For in addition to denying the possibility of a genuine community of meaning, which now appears to be a meaningless restriction, community of meaning being either nothing at all or identical with what are called ‘family relations’, it also makes use of the notion of genuine communities of meaning in talking of the small overlapping strands which make up a series of ‘family relations’. If we are not to be involved in an infinite regress, the small strands of meaning cannot themselves be made up of family relations without end, and if there are definite communities of meaning in the small strands than there is no reason why they should not also be found over wider ranges. Possibly, however, the doctrine merely states a fact about our actual use of words, in which case the philosopher is quite free to say that he does not care for ‘family relations’ in philosophy and prefers to proceed by way of genuine communities of meaning.

The modern rejection of ‘essentialism’ has, further, made a practical appearance in such notions as Nelson Goodman's ‘bleen’ and ‘grue’, ‘bleen’ being a word applied to anything that is blue up to a certain arbitrary date and green thereafter (or so at least we should describe its meaning in our language, which has no prerogative for the pure semanticist), while ‘grue’ is used in a precisely contrary manner. ‘Bleen’ and ‘grue’ are plainly cases of arbitrary notions of which in Platonic terms ‘there are no forms’, but for Goodman they are not at all different from ‘blue’ and ‘green’. It is all for him a question of the language one chooses to speak. Unfortunately, the new notion raises inductive problems, since no one would argue that because all observed bleen things have hitherto been sweet, therefore it is reasonable to suppose that all bleen things without exception will henceforth be sweet. These curious inventions of Goodman plainly have no correspondence with any genuine essences, and have no function in science or philosophy, except perhaps, by their sheer absurdity and uselessness, to make us ask what a genuine form or essence might be. (See N. Goodman: Fact, Fiction and Forecast.)

We may here regret, as remarked before, that Husserl has had absolutely nothing to say on the matter. He simply devotes no space to considering what marks off a genuine ειδος or fundamental meaning from an arbitrary conceptual combination, though he always in practice shows a fine flair for the fundamental as opposed to the arbitrary. Husserl has made use of Platonic language with even less explanation and justification than Plato, who at least shows, in the Phaedrus and elsewhere, that he does know that there is such a thing as ‘division at the joints’.

Our own position will here be that for philosophy genuine essences, genuine pervasive communities, are an object of serious and necessary, if not of exclusive pursuit. There may, in many contexts, be a place for fairly loose concepts such as the ‘terms’ and ‘propositions’ of many logicians, which stray happily from semantic to linguistic contexts and vice versa—but in general it is the task of the philosopher to try to pick out simple, salient, independent peculiarities, that run, or could run, through a whole range of cases, possibly present without change in all of them, possibly assuming various specific or systematically modified forms, and possibly also being ‘reduced’ or ‘distorted’ in various ways. This last means that we do not in fact have a pure or full case of the peculiarity in question, but only one which in some degree approximates to being such. The really interesting Platonic forms of course belong to the third type: they are patterns that really have no instances, only approximations to such instances. It is legitimate, e.g., for the phenomenological philosopher to form the idea of a purely arbitrary choice, a choice between alternatives in which we have absolutely no reason to prefer one alternative to another: whether or not any case of choice perfectly illustrates this idea, all choices, in virtue of their genuine hesitation among more or less balanced alternatives, come more or less close to it. The case where we pick something quite at random, if we ever do so, is the perfect case of such arbitrariness, and it is valuable to group a whole set of possible practical responses in relation to such an ideal limit. On the other hand this is not the only sort of ειδος that it is valuable to pick out and to work with. If the idea we have been considering is so finely abstract that it is doubtful whether any cases exemplify it, there are other ideas so rich and full, that such true exemplification is likewise in doubt. We need not go far afield from our previous example: the notion of fully considered, rational choice affords an excellent example, involving a catalogue of conditions so numerous and so readily inviting addition, that we seldom or never have a really adequate illustration of the idea in question. These two types of idea, the excessively spare and thin, and the extremely rich and full, are the sorts of ideas that it is most profitable to have in philosophy, and of the latter ideas those are best which gather the largest number of spare, thin ideas into a more or less close unity and which are not merely a loose conglomerate.

The ideas that are important in philosophy admit of yet a further antithesis: they may, on the one hand, be extremely clear-cut at the edges and rigidly maintained throughout a piece of philosophical discourse; they may, on the other hand, be maintained in a certain state of determinability, so that more and more accretes to them as the discussion proceeds, and they may also be allowed a certain amount of shifting or free play which is not thought to be contrary to their identity. These last-named features of ideas will concern us in the next lecture, where dialectical development is our main topic.

What are really opposed to the genius of philosophy are, however, ideas of the family-relations type described by Wittgenstein, in which the shift of sense is confined by no limits and obeys no principle. That such ideas are opposed to the genius of philosophy is shown, if one needs to show it, by the fact that the philosophers who recognize such ideas never use them in their philosophizing. Thus, the idea of family resemblances is itself a higher-order idea of a common-or-garden philosophical type, and so is that of the passion for generality which is held to afflict most philosophers: neither idea covers a loosely overlapping series of cases. Wittgenstein even shows his approval of this passion for generality by the fact that he regards his family-relations discovery as more or less an expose of our ordinary notions, and he does not in his own philosophizing use words in a merely loosely shifting way. He talks about looseness, but not loosely, or not deliberately so.

Much the same may be said of Austin's wonderful treatment of words in their actual use: he may be said to have shown that something like Wittgenstein's doctrine really holds of the working of our ordinary expressions, of such expressions, in particular, as fall under the vague rubrics of ‘reality and appearance’, ‘the voluntary and the involuntary’, ‘the determined and the free’, etc. The expressions falling under these headings do not conform to the tidy paradigms philosophers set for them, they work very differently in different contexts, so differently in fact that it would be vain to try to comprehend their Protean modifications in a limited number of clearly stated ‘senses’. Austin has further shown that the conceptions philosophers attribute to the ‘ordinary man’ are stated in words so uncommon, or words so uncommonly used, that they do not accurately represent those conceptions at all. Ordinary men, for example, do not operate with the unqualified, blanket notion of ‘reality’ which philosophers so readily make use of, nor do they conceive of certain ‘directly perceived’ things called ‘appearances’ which may or may not be ‘veridical’. ‘When the plain man sees on the stage the Headless Woman’, Austin tells us, ‘what he sees (and this is what he sees, whether he knows it or not) is not something “unreal” or “immaterial”, but a woman against a dark background with her head in a black bag.’1 Probably Austin is right as to what the plain man would probably say he saw, and how he presumably conceives the whole matter, and the philosopher is plainly wrong in suggesting that the plain man would at all willingly say that what he saw was something ‘unreal’ or ‘immaterial’. But all this does not affect the fact that the philospher may still be right in finding the philosophically noteworthy element in the whole situation in the ‘presence’, in some sense of ‘presence’, to the plain man of at least one feature of the material woman before him which is not part of the physical situation, namely this woman's headlessness, and in the assimilation of this case to other cases where what is ‘seen’ is really ‘conjured up’ by the brain, and has, in its totality, no place in the natural order at all. The philosopher may be right, for his purposes, in setting up as a type, the illusive or delusive situation in which something, whether a whole object or situation or a feature of it, is in some sense ‘before us’ while in another sense it is not ‘there’. This may be the odd, really problematic thing that he feels to be lurking behind mirages, mirror-images, hallucinations, misperceptions, perspectival variations, and to keep it muffled up in the swathings of actual usage may simply be to refuse to do philosophy.

Philosophy, in short, may simply be the sort of activity in which we replace ordinary notions and usages by various clear types which bring out interesting issues of principle that the instances covered by these notions and usages can be seen to exhibit in varying degree. Philosophy may in fact be essentially revisionary and creative, though it builds, and must build, on the notions and usages embedded in ordinary thought and speech. And that it is right to regard it as revisionary and creative may be shown by the various fruit of Austin's own researches, valuable notions like that of the performatory, the illocutionary and the perlucutionary, etc., notions which certainly never occurred to any ordinary user of language, and which are as shiningly clear-cut in their content as the activities they cover are shuffling and obscure. The marvellous merit of Austin, we may say, is that he has brought unphilosophical discourse under the yoke of philosophy, that he has framed concepts to cover the ways the ordinary speaker really adapts his expressions to fit changing situations, rather than distorted these ways to fit various stratified, predetermined concepts. But in this subjugation of ‘unphilosophy’ to philosophy, it is philosophy that has won the day, not unphilosophy. For Austin's practice has given us not a shadow of a reason for holding that the philosopher ought to imitate the ordinary man in the style of his notions. Rather it has suggested that the philosopher, while paying the closest attention to actual usage, should be as daringly unordinary and clear-cut in his philosophical notions as Austin has been in his.

If philosophers then are right in looking for ‘ideas’ and ‘essences’ that cover wide ranges of individual and specific cases, and that cover them in a different manner from ordinary notions, what is it shall we say that makes ideas and essences ‘genuine’, which distinguishes them from idle constructions and confabulations which reflect nothing but caprice or random experience? It seems that we shall here have to recur to another notion suggested by Plato: that the true ειδος is one that has a connection with the Good, that embodies certain values that the false, fake ειδος simply lacks. It is not simply something in their sheer make-up as thinkable contents which makes ειδη genuine or fakes: it is the relevance of certain normative standards, of certain high-level prescriptions, which they fulfil or flout. These standards or prescriptions are not connected with our practical conduct in so far as this goes beyond thought and utterance, nor are they connected with what is good as a goal that our practical effort might seek to compass. We are not here taking a leaf from a superb part of the Phaedo, where it would appear that Plato is connecting an ειδος with some manner in which it would be good for things to be. The excellences that are here in question are merely logical excellences, excellences of conceptual thinking, and that ειδος will be a true one that exemplifies them most fully. It will not therefore be by some mere process of passive looking around that the true essences of phenomena will dawn upon us, and Husserl's term reine Wesensschau has the demerit of suggesting that this will be the case. We shall have to search actively for the notions that satisfy our logical requirements and which will thus embody corresponding logical excellences. Only certain ways of looking at the phenomena of the human cave will be philosophically valuable ways, and these will in general require some revision of our ordinary notions and some coining of new and perhaps mystifying expressions.

We must here content ourselves with using ordinary, sufficiently understood expressions, and with saying such things as that true ειδη must gather very diverse cases together in an illuminating manner, that they must provide us with standards in relation to which very varying materials can be assessed, that they must be free from internal obscurity, discrepancy or patched-up compromise, that they must have the clearness and distinctness so much praised by Descartes and so on. Obviously logical values lie in a variety of directions, and are not necessarily always compatible: the spareness and the fullness that we noted above, as also the sharp closure and the openness of content, are obviously excellences that compete with one another, which require sacrifice of one to another. It is not our task, at this point, to go into the general problems raised by values and by normative prescriptions, nor in particular with the problems raised by their connection with particular traits such as those we have just listed. One thing only requires to be said: that it is absurd to identify logical values merely with what we find convenient or useful in thinking and speaking, or with what we simply happen to prefer. Here as elsewhere we are in a region where agreement is in general as impressive and cogent as it is unspecific in detail: we all attach importance to clarity, simplicity, explanatory unity, etc., though we may not all agree where it is most strikingly present, nor in what order such points of value are to be preferred. We may here comment generally on the fact that any philosophical investigation, even an investigation of logic and concepts, soon lands us in an issue of values and that if there is no clarity to be had in regard to values there is no clarity to be had on any philosophical topic.

Another modern obstruction, however, comes up seriously at this stage. We shall have to make clear, as Husserl never troubled to make clear, how we establish understandings in regard to the various ideas used in cave-description, how we make sure that others have the same notions as ourselves. We do not, alas, live in some monumental avenue of types which we can point out to one another as if on some social promenade. The mere use of words is no sufficient criterion of communication, for obviously no clear sense may attach to such use, or if a clear sense, then no shared, intersubjective sense. Here, of course, our first obligation is to follow the practice of Moore, and to introduce ideas, in the first instance, by connecting them closely with ordinary expressions, used scrupulously as they are in fact ordinarily used. Our established linguistic usages have at least this merit: that they have connections with definite features in our experience, and with features in regard to which we already have understandings with one another, or that, if they do not have such connections and do not embody such understandings, it is unlikely that any devised by philosophers will be more successful. Once we have successfully ‘pinned down’ the sort of idea we are trying to put forward by assembling the right set of ordinary expressions and usages, it is possible for us to go on to carve out an idea which is not the same as any present in ordinary usage. It may have a more striking outline, it may be smoother, it may be richer in differences and so on; it may even represent an analogical leap beyond the confines of primary discourse, which brings into our ken matters that before lay entirely outside of it, and towards which others find themselves able to make similar flights. It is obviously not necessary that we should here follow Moore's view and regard the new philosophical idea which emerges as a mere ‘analysis’ of the ideas previously present, which would give the philosopher no more than a rearranging, displaying function. We may very well credit the philosopher with what may be called a ‘taking off’ function, a rising out of the plane of ordinary usage on to a new level of sense.

Certain philosophers are not, however, satisfied to start with ordinary notions and usages, and to take off from them in the revisionary flights of philosophy. They refuse to fly till guarantees of air-worthiness have been given. As such guarantees must for them rest on performances done on the ground, it is not remarkable that they never receive the right guarantees. They have to be quite sure in advance that expressions really have been given a sense before they will allow philosophers to use them, and the mere fact that they are employed by a set of people who feel they understand one another is for them no proof of their right to employ them. The first of these critical scrutineers of ideas were the traditional empiricists who worked on the assumption that for every legitimate idea there must be a genealogy leading each of its elements back to some source in sensation or reflection. If no such genealogy was forthcoming, as seemed to be the case in regard to our ideas of substance, cause, the conscious self, etc., then the idea in question was not genuine but spurious, a mere shadow cast by our use of language. In more recent times those inclined to a similar empiricism have disavowed laws of this Nuremberg strictness: they have declared, not quite truly, that they do not care where our ideas originate. The new empiricism, however, shows its derivation from the old in that a large part of its researches and its arguments consists in enquiring into the way in which certain meanings were taught, the situations in which they were learnt, the situations, e.g., in which we learnt to use numerical expressions, in which we learnt to obey and utter commands, in which we learnt to talk about our dreams, our inner feelings, our memories, etc. The ground covered in Locke's Essay is traversed anew, only from a linguistic rather than an ideal angle, and with a sketchiness and an inventiveness quite foreign to the plodding care of Locke. What is, however, implied by all such genealogical investigations is that if, perchance, it is not possible to show just how we were taught to use a certain expression significantly, or if the circumstances are not patently such as fully to explain how we came to give it a certain sort of sense, then the expression cannot have the sense with which we credit it: either it will have no sense at all or a sense different from what we should ordinarily say it had. Thus Malcolm, accurately and conscientiously developing the teaching of Wittgenstein, argues that, since we can never have been taught to use the expression ‘I am dreaming now’—dreamers, being sleepers, are automatically neither speakers nor auditors, so that no one can be taught to say, nor to understand anyone saying, ‘I am dreaming now’—we can not now understandingly use the expression whether in ordinary talk or philosophy. The wide use of the notion of a dream by sceptical or idealist philosophers is therefore invalidated: Descartes cannot suppose that he is dreaming now, since he can never have been taught to use the corresponding utterance.

It is not possible, in a course of lectures with an aim as wide as the present, to give all the reasons why we should not load ourselves with the difficulties and restrictions of either the old or the new empiricism. Obviously there are many soi-disant notions in regard to which the question is very important as to how we came to form them, or to give meaning to the words which express them. In the case of a vast number of empirical or quasi-empirical concepts, concepts which cover the contingent stuff of the phenomena before us, it is plain that a proof that we did not acquire such concepts in a regular manner constitutes a proof that they are not really concepts at all, only the putative meanings of nonsignificant words. But that some soi-disant notions can be unmasked in this manner does not prove that all may be so, and while the question how we acquired a notion or gave meaning to a set of words is always important, the answer may sometimes be that nothing in the manner or circumstances in which a meaning has been taught or communicated will fully explain how it has been so. When people are ready to understand certain things—a fact well known to educators—almost anything will suffice to make them understand it: a nod, a gesture or a metaphorical expression may make the all-important communication.

The whole stress on processes of teaching the use of expressions is thoroughly misguided. That a man knows French is shown by the fact that he uses it understandingly and that we who know French understand his use of it: it is not important to ask whether he learnt it in the right school or by an approved method. If he learnt it by an injection into the frontal lobes of his brain, this would be a perfectly satisfactory method of learning French. In much the same way the fact that we have certain ideas, and understand the expressions used to cover them, may come out sufficiently in the fact that we engage in intelligent dialogue involving such expressions and that others successfully join in this with us. It cannot be shown that we lack certain ideas, or that we cannot share them, because the circumstances in which we were taught the words that express them did not involve any sensuously observable public object that corresponded to them. Nor can it be argued that because certain sensuously observable public matters did enter into the situation in which we were taught the use of certain words, and perhaps were instrumental in the teaching, such observable matters also entered into the contents communicated. The circumstances which trigger off understanding or communication are not necessarily part of the content grasped or imparted, and may in fact pass wholly unnoticed: it is the content, the meaningful idea, that one man introduces to another, just as if it were a publicly observable object, while the features that trigger off the understanding of this content wait modestly in the wings. And even when such triggering features do not pass unnoticed, but are seen as signs of the content communicated, and perhaps rightly count as ‘criteria’ of it, they need still not seem part of this content. All we have said may be applied to our understanding of understanding itself. The only way to understand anything is to understand it, in such triggering circumstances, or by the use of such criteria (if any), as are necessary or relevant. In the same way the only way to understand understanding and to recognize its presence in ourselves and others, is to understand it, again in such triggering circumstances, or by the use of such criteria (if any), as are necessary or relevant. In the understanding of the understanding of others, triggering circumstances or criteria are admittedly necessary, but in the case of the understanding of our own understanding, there are at least some cases where we may say, following Spinoza, that our understanding of something, our own good idea of it, is its own criterion. Or, if we do not like such language, we may say that in these cases no criterion of the presence of understanding is necessary, though superfluous signs of its presence may no doubt be multiplied. There are, of course, paradoxical cases where seeming understanding, internally vouched for, breaks upon the test of later understanding or of performance judged inadequate. Paradoxical cases are, however, only possible provided all cases are not paradoxical, and understanding may be said to authenticate itself in the person who has it, just as doing and saying the right thing often enough sufficiently attests understanding in other people.

We may now go on to make a general pronouncement in which we shall part company with many philosophers, and regrettably with Edmund Husserl. A careful examination of what we ourselves know and understand, and what our consideration of the rational utterances of others leads us to be rationally sure that they too know and understand, makes us sure that in every field there are ideas or sides of ideas that we perfectly understand, and can successfully communicate to each other, which are nevertheless not such that we were ever shown instances of them by anyone, or could ever show instances of them to anyone, and which we apply with confidence, or use in assessment, though they have never been concretely illustrated in what came before us, and can in some cases never be so illustrated. We are rejecting the old empiricism with its hindward glance towards sensation, and the new empiricism with its hindward glance towards ostension. We are taking our stand on the ground which led Plato to his theory of reminiscence. There are profound difficulties in this situation, perhaps such as to justify something like Plato's heroic inferences; we shall consider these at a later stage, but they do not affect the truth of what we say. Not only are there such familiar Platonic instances as those of the straight, the circular, the equal, etc., which presented instances do not in fact perfectly illustrate, and do not usually even seem to illustrate perfectly, but we have many other much more fundamental ideas whose very content precludes the possibility that they should ever be adequately illustrated, though they can, no doubt, be led up to by many helpful devices. Here we have such ideas as that of open universality, of being true in absolutely all cases of a certain sort, which is involved in all general judgements, as also that rudimentary form of the idea of infinity which we express by the phrases ‘and so on’ or ‘and so on for ever’. Even children readily grasp the momentous meaning of this latter phrase when told one of those teasing, reentrant stories which are readily seen to be in principle interminable—they find no difficulty in what perplexes mathematical philosophers—yet nothing is of course clearer than that such interminability is unillustrable. The notions likewise of the independent reality of bodies, of their being there even when no one observes them, is not rendered inadmissible in virtue of its necessary unillustrability, and the same is the case with the system of contrasting personal privacies, whose unillustrability is part of its very notion. It is not the case that we cannot make sense of such ideas because they are unillustrable: it is because we can make sense of them, and in fact perfectly understand them, that we see them to be unillustrable. It would seem that, if our unillustrable ideas are not vastly numerous, they are none the less important and fundamental, and condition our view of the world at all sorts of crucial points. It is arguable that they condition the general form of the world-picture within which experience, in the sense of individual encounter, deposits its findings, and that without them no experience of anything would have illumimation for us at all.

All this will, however, be best argued as we develop our detailed view of the human cave. It is possible, further, that the unillustrability or imperfect illustrability of some of our fundamental notions is one of the things that make the human cave a cave, and will perhaps finally force us beyond it towards something more intelligible. The policy we shall adopt for the present is not to require ideas to present passports, whether British or Viennese, not to question their existence or their significant content because it is doubtful where, if anywhere, we acquired them, or because we cannot point with our fingers to what justifies their application. If there are unillustrable notions, ideas whose true application lies beyond observable public situations, then we are committing a vast begging of the question or a circle in proof if we reject them because they are not thus ostensively applicable, or if we substitute other ideas for them. Moore used the fact that certain epistemological theories cast doubt on our certain knowledge of certain truths, to discredit those epistemological theories themselves: in much the same way we may use the fact that we certainly do understand certain possibilities and can talk about them to others, to refute all such theories of meaning as render it doubtful whether we can understand or communicate the possibilities in question. We are not, in all this, at war with what we may call a genuine empiricism, one that takes the world as we find it, with all that we feel in our bones about it, as well as all that we grossly observe. We are only at war with forms of empiricism that commit the old error of thinking that superior clarity and certainty attaches to the so-called data that some special theory distinguishes.

It is time to bring this widely wandering lecture to a close. We have explained what we mean by a phenomenological method of investigating the human cave, that it is not a phenomenalist method which deals in data more certain than the persons, things, etc., of ordinary experience, and that it is not an explanatory method which goes behind phenomena to what really is. We have explained that it is a general or eidetic method that investigates the general essences or species of things rather than the individuals that fall under them. We have dwelt on the difficulty of distinguishing a genuine from a factitious essence, and have considered Greek, Cartesian and modern views on the point. We ourselves have placed the marks of a true ειδος in the direction of values. We have lastly considered the problem of communication, how we bring philosophically significant ideas to the attention of others, and pin them down for them as for us. Modern difficulties and restrictions have been considered, but we have considered them mainly question-begging and muddled. In our next lecture we shall continue to investigate the methods of exploring the human cave.

• 1.
  Sense and Sensibilia, p. 12.