Metaphysics as a Guide to Morals

by Iris Murdoch

  Wittgenstein, Heidegger and Derrida are all anti-Cartesian (Wittgenstein said his work brought the Cartesian era to an end) but with different methods and conclusions. They agree in rejecting the philosophical concept of the autonomous 'I’ which discovers one primal immediate truth in its momentary present experience. (‘This at least is indubitably true.’) Argument against this position will go along lines of saying that there can be no such thing as momentary knowledge, knowledge requires concepts which develop in an extended 'world’ and cannot be the property of an instant flash of awareness. How do I reach the world? You are in the world, it is your world. (For this relief much thanks.) But what does this mean and what exactly follows? Derrida, in placing Heidegger as ‘last metaphysician’, attacks particularly what he calls the ‘metaphysics of presence’, the use of concepts of present being, consciousness, experience. He, Derrida, following 'logical’ implications of the rejection of cogito, concludes that if there can be no solitary knower (since knowledge and meaning depend on conceptual networks, that is language) then there can, really, be no knower, only a network of meanings (the infinitely great net of language itself) under which there is nothing. (See a philosophical discussion of these matters in my first novel, Under the Net.) Because of the vast extent of language and the way in which meanings of words and concepts are determined by innumerable relationships with other words and concepts, no individual speaker can really 'know’ what he means, we are unconscious of the immense linguistic beyond which we think that we ‘use’ when really it is using us. That is, in rejecting Descartes’ argument, Derrida rejects also the concept of the autonomous individual; and with it the (ordinary) concept of truth. Derrida shares with Wittgenstein the ‘discovery’ (made earlier by the latter) that words are not names of things, predeterminedly affixed to particular objects (the fallacy which Derrida calls ‘logocentrism’), but have meaning through relations with other words. But Wittgenstein would add that meaning comes through use and forms of life, presumed to be, in a common-sense sense, the local property of individuals. Wittgenstein offers no further theory here. The individual is ‘saved’ by (early) Heidegger in the concept of Dasein taken as primary, human being, or human reality, thrown into a world and existing immediately in it as its relations to it. Dasein (being-there) is In-der-Welt-sein (being-in-the-world). Thus the old Cartesian problem is thrown away. The Heidegger of Sein und Zeit has been called an existentialist, and is certainly a defender of (takes as his ‘main character’) the existing individual. He calls this individual the Shepherd of Being, the creator of spaces and clearings where Being (as in some sense transcendent truth and reality) can manifest itself. (This imagery can be readily understood in religious or aesthetic terms.) Later Heidegger seems to reverse the roles, picturing man more as the victim of Being than its shepherd. (Being plays its own secret game.) In the Tractatus Wittgenstein ‘saves’ the individual by metaphysical decree: ‘I am my world.’ I discuss later what happens in the Investigations.

  Saussure separated language conceived of as a general system from its particular local use by individuals. He retained however the idea (which belongs with ‘presence’, consciousness, experience) that speech, not writing, was the basis of language and meaning. Plato expresses this intuition in the Phaedrus and the Seventh Letter. Spoken words are understood in their proper meaning in the presence of the recipient. In the Seventh Letter Plato seems to speak particularly of philosophical conversation, but the parable in the Phaedrus (274ff.) can apply to all speech; as opposed to speech, writing cannot answer back, may be misunderstood, is at a remove from original meaning. It is a secondary technique. Saussure agreed; real speech and clear meaning belong to speaking persons. Derrida reverses this. What is primary is writing, thought of in a (metaphysical) sense as a vast system or sign structure whereby meaning is determined by a mutual relationship of signs which transcends the localised talk of individual speakers. This postulated transcendent, or basic, language Derrida calls archi-écriture or ‘primal writing’. We make use of local codes which derive from an area of similarities and differences which transcends our awareness. In this sense language speaks us. Derrida has fortified his theory by inventing the word différance, a variant of ordinary French différence which means 'difference’. The verb différer means to differ, and also to defer, and Derrida’s coinage includes both meanings. That which determines meaning, since it is a function of the vast linguistic region of archi-écriture, is something infinitely deferred; any meaning is marked by ‘traces’ left by other, perhaps extremely remote, similarities and differences. So in an important sense we are not in control of our meaning, we are unconscious of how language itself, at a deeper level, makes meaning. We do not know what we are uttering. This image of the whole of language determining every part of it is awesome and impressive. It may undermine our confidence in what we took to be our 'mastery of language’ and our ability to say anything clearly, or to say what is true. Here truism, half-truth, and shameless metaphysics join to deceive us. Yes, of course language is a huge transcendent structure, stretching infinitely far away out of our sight, and yes, when we reflect, we realise that often we cannot say quite what we mean or do not quite know what we mean. Common-sense does not usually take the trouble to reflect as far as this, or if it has done so realises that nothing is really being changed and meaning and truth are what they have always seemed. (A study of philosophy may be likened to a catharsis, like that of the Zen Buddhist who begins with rivers and mountains, doubts rivers and mountains, then returns to rivers and mountains.) Suppose we worry about the discovery that our tables and chairs are made of atoms. We are being frightened by a picture, in the case of tables and chairs a scientific picture, in the case of archi-écriture a metaphysical picture. Anyway (if we want to proceed) we may ask, if all that is so, how does Derrida know what he means? The answer rests in the technique of ‘deconstruction’, which it must be assumed the theorist can apply not only to others but to himself. This is a technique, most generally on view in literary criticism, whereby texts are minutely studied in a manner which reveals deep meanings of which the writer is unaware. This may seem fair enough. Literary critics, especially since Freud, may do this as part of their general task, and writers, as ordinary human beings, might agree that their ‘text’ can be shown to indicate (for instance) prejudices of which they were less than fully aware. This would be compatible with what is expected of a literary critic who would, we hope, be a well-read scholarly person, with wide tastes and a knowledge of languages and history, who appreciates and enjoys all kinds of literature and is able to write about it with insight and balanced judgment. But ‘deconstruction’ is something more radical and no friend of the old-fashioned or traditional critic just described. Our whole concept of what literature is is here in question. This quest for the hidden deep (primal-language) meaning of the text (to use the jargon) is now said to be the main and essential part of the critic’s task. The ‘old’, and in my view good, proper, literary critic, approaches a literary work in an open-minded manner and is interested in it in all sorts of ways: which certainly does not exclude treating a tale as a ‘window into another world’, reacting to characters as if they were real people, making value judgments about them, about how their creator treats them, and so on. Here the enjoyment, or otherwise, of the critic is like that of the layman, only generally (one hopes) well informed and guided by a respect and love for literature and a liberal-minded sense of justice. He will beware suitably of his own prejudices, but will not be chary of speaking his mind. Literary critics are speaking as individuals and not as scientists speak. Literature is a vast scene of confusion, that is of freedom. Of course, critics who say that they are 'deconstructionists’ often felicitously lapse into the ‘old habits’. But the ideal deconstructionist is more like a scientist who shows that things are absolutely not what they seem (they really are made of atoms). He will tell us that the literary object, as we have hitherto understood it, is pure phenomenon, below which
lies something quite different from what its naive creator believed and intended, or what the naive reader imagines he perceives. In fact the real work of literature is what the critic produces. The deconstructed work is the real work.

  The notion that every sentence we use bears, indeed consists of, the invisible traces of other meanings created by a vast non-human system, carries serious implications. We seem to be losing our concept of the individual. Literature, as we ordinarily see it, is full of values, and we the clients receive it as such and consider it in the light of our values. The consumption of literature involves continual (usually instinctive) evaluation, of characters in stories, content and quality of poems, skill and intentions of authors, etc. etc. etc. Value, morality, is removed by the structuralist picture if taken seriously. This removal of value is, in a quiet way, characteristic of this age. Heidegger’s book What is Metaphysics? is partly concerned with showing how the general idea of value (morals) is a superficial phenomenon. Behind this new 'revaluation of all values’ by Heidegger and by Derrida lies the (metaphysical) concept of a vast superhuman area of control: Heidegger’s later concept of Being, and Derrida’s theory of Language. These systems represent new forms of determinism. Determinism is always reappearing in new forms since it satisfies a deep human wish: to give up, to get rid of freedom, responsibility, remorse, all sorts of personal individual unease, and surrender to fate and the relief of 'it could not be otherwise’.

  Derrida, trying to explain his idea of archi-écriture, admits that:

  'Quant au concept d‘expérience, il est ici fort embarrassant. Comme toutes les notions dont nous nous servons ici, il appartient à l’histoire de la metaphysique et nous ne pouvons l‘utiliser que sous rature. "Experience” a toujours désigné le rapport à une présence, que ce rapport ait ou non la forme de la conscience. Nous devons toutefois, selon cette sort de contorsion et de contention a laquelle le discours est ici obligé, épuiser les ressources du concept d’expérience avant et afin de l’atteindre, par déconstruction, en son dernier fond. C’est la seule condition pour echapper à la fois à “l’empirisme” et aux critiques “naïves” de l’experience. Ainsi, par exemple, l’expérience dont “la théorie”, dit Hjelmslev, “doit être indépendente” n’est pas tout de l‘expérience. Elle correspond toujours à un certain type d’expérience factuelle ou régionale (historigue, psychologique, physiologique, sociologigue etc.), donnant lieu à une science elle-même régionale, et, en tant que telle, rigoureusement extérieure à la linguistigue. Il n’en est rien dans le cas de l’expérience comme archi-écriture. La mise entre parenthèses des régions de l’expérience ou de la totalité de l’expérience naturelle doit découvrir un champ d’expérience transcendantale.’

  (De la Grammatologie, published 1967, p. 89.)

  ‘As for the concept of experience, it is certainly an embarrassment here. Like all the ideas we are dealing with, it belongs to the history of metaphysics and we can only make use of it under erasure. [Sous rature.] “Experience” has always indicated a relationship to something present, whether or not this relation takes the form of consciousness. We must always, in accordance with the contortions and contentions which our discourse is here obliged to adopt, exhaust the resources of the concept of experience, before, and with the intention of, reaching, by deconstruction, its deep foundation. This is the only way in which we can keep clear both of “empiricism” and of “naive” accounts of experience. For example, when Hjelmslev speaks of an experience “the theory of which must be independent”, the experience in question is not the whole of experience. It always corresponds to some particular type of factual or regional experience (historical, psychological, physiological, sociological, etc.), occasioning a science which is itself regional, and as such strictly outside the domain of linguistics. This is not the case with experience conceived as archi-écriture. The putting into brackets of regions of experience, or of the totality of natural experience, must uncover a transcendental field of experience.’

  (Hjelmslev, a follower and critic of Saussure,

  Prolegomena to a Theory of Language, 1943, trans. 1953.)

  This passage (taken out of context but I think fundamentally informative) refers to Derrida’s programme for a transcendental deduction. His theory follows Kant and Husserl in seeking to uncover an all-inclusive ‘transcendental field’, actualising Husserl’s dream of a philosophical discovery more radical than any science. This field or network is sought in language, not in local individual uses, but in the postulated primal language system. This system is not in any ordinary sense experienced, nor can it be indicated by traditional metaphysical concepts of experience; its discovery involves a deconstruction of ‘experience’ so radical that the concept can only be used sous rature, under erasure. This ‘embarrassing’ device, of using a crossed-out word, is intended to alert us to an unusual, stripped, deconstructed sense of a concept which no better words can at present be found to exhibit. It is a term of art designed to be stronger and more specialised than what is often achieved by putting a word in inverted commas. This ‘contortion’ should enable us, alerted, to avoid thinking of archi-écriture either naively, as something we can be at some moment ‘aware of’, or in terms of traditional empiricism, say of the Kantian or Humian or ‘ordinary language’ variety. To understand what archi-écriture means we are not to think about ordinary, personal and local, everyday experience, nor about any historical, psychological, sociological or (previous) philosophical explanations or analyses of such experience. The totality of these regional and natural experiences, whether naive or worked by science, must be set aside or put in ‘parenthesis’ (a concept borrowed from Husserl) if the deep basic transcendental field is to be found. Well, philosophy is to some extent a foreign tongue. The deconstructed field of archi-écriture is uncovered by the post Saussurian analysis of ‘language’ as basic (non-individual, non-experienced) human activity. Saussure, as an aid to thinking about language, made a distinction between parole, actual historical and various speech uses, and langue, the abstract notion of language as, at any given moment, a single system. Language as system, the subject matter of a possible semiology or science of signs, was said by Saussure to rest upon two intimately related ideas or principles. Linguistic signs are arbitrary, words are not names vertically posed upon pre-existing, already fixed (as if by God), things or qualities, this is not how language has meaning. Language has meaning through lateral horizontal networks or groupings of signs related by systematic differences. Both these ideas, that words are not names and that concepts gain sense through use in relational groups, have been expressed by Wittgenstein and belong to a rejection of Cartesian or Humian starting points which is common to modern philosophies. Saussure, who still thought that a sign or signifier should signify or point to something signified, detached verbal signs from a one-one naming connection with entities in the world, and attached them instead to mental concepts in the mind. Wittgenstein (I discuss his view later) attacked ideas of meaning as mental process, or use of private ‘inner contents’, suggesting that there were many ways in which language related to the world, that concepts (meaning) depended on public rules and forms of life. He pointed out mistakes but did not create a general counter-theory. At a certain point he left things alone. Saussure’s structuralist successors also pointed out that the same error (that of assuming entities of definite predetermined quality) was committed by postulating inner things instead of outer things. They then made a further step, not made by Wittgenstein, in deciding that therefore the signified, whether thought of as external entity or mental datum, was otiose, and that meaning was entirely enclosed in the self-referential system of language. This is the crucial move which, in the structuralist theory, separates meaning from truth, outlaws the idea that truth rests on some kind of relation with a non-linguistic reality, and in effect removes the concept of truth altogether. The removal from language of any reference except to other parts of language sweeps away not only the correspondence theory,
but any theory, of truth. Meaning, then, is an internally self-related movement or play of language. Le jeu des signifiants. The word ‘play’ (or ‘game’), often used by structuralists (and by late Heidegger and by the presocratic philosophers), is here seen in a fundamental context. Here ‘language’ must of course not be thought of in terms of effective individual usages, but as a postulated totality of system which transcends these. What is ‘transcendent’ is not the world, but the great sea of language itself which cannot be dominated by the individuals who move or play in it, and who do not speak or use language, but are spoken or used by it. This is the ‘transcendental field’ which is revealed when ‘experience’, as something regional or natural, is put in brackets. On this view, almost all language-use is an unconscious subjection to system. Only at some points (in the activity of some minds) can language be seen to emerge as conscious play. Such playing of self-aware system could be either science or art. The central structuralist ideas, closely integrated with each other, could be said to be aspects of the same idea. It is interesting to see how a number of half-truths, which might be expressed as intelligent suggestive hermeneutic observations about language, are made into what appears to be a closely knit metaphysical argument. Of course, in an obvious sense, language transcends its user, meanings are ambiguous, words are clarified through discrimination, and so on. But the amassing of such general considerations is remote from a conclusion to the effect that ‘really’ no final statement which can be said to be true or false is ever made. Here, as in other metaphysical ‘totalities’, system obliterates a necessary recognition of the contingent. What is left out of the picture, magically blotted out by a persuasive knitting-together of ideas and terminology, is that statements are made, propositions are uttered, by individual incarnate persons in particular extra-linguistic situations, and it is in the whole of this larger context that our familiar and essential concepts of truth and truthfulness live and work. ‘Truth’ is inseparable from individual contextual human responsibilities. The ingenious continuous weave of structuralist generalisation checks reflection at the point where it should be most industrious, and makes important problems invisible and so undiscussable by removing familiar distinctions and landmarks. Structuralism can be attractive because it seems to exhibit or point out much that is new in our now so fast-moving world; and so it can look like the perennial philosophical study of conceptual change. But structuralist theory disables this study (and in this respect indeed can claim to have parted from traditional philosophy) by its removal of the ‘old’ idea of truth and truth-seeking as moral value. If all meaning is deferred our ordinary distinctions, for instance between what is clearly true and what is dubious and what is false, are removed and we begin to lose confidence (as structuralists urge us to do) in what is made to seem the simple, old-fashioned, ordinary concept of truth and its related morality. If, in some ‘deep sense’, it cannot be finally established whether or not the cat is on the mat, then how can we have the energy to trouble ourselves about the truth or falsity of more obscure and difficult matters ? In this frame of mind we may feel that we can understand and welcome this new doctrine as ‘the philosophy’ of our age, and embrace the idea of language as stirring ‘play’ whereby we are incited to discover ‘meanings’ which are independent of truth. We can see at this point how naturally structuralism preys upon literature and indeed inspires a new kind of literature which seems to prove its point. Literary forms, fictions, texts, recognised as such, exhibit the ways in which language, no longer thought of as a ‘picture of reality’, can liberate the mind into a creative enjoyment of unfettered meanings. (Of course this game may produce good writing, but need not obliterate all other literary forms!) Not the least damage done by the doctrine is that students spend their time studying the very obscure books which propound it, rather than reading the great works of literature.