On Using Metaphors in Philosophy
This paper deals with the question, of whether the cognitive content of metaphors can be put to use in philosophy, and, if so, what cognitive or methodological place metaphors have within philosophical discourse. Three philosophical attitudes toward metaphors can be distinguished: First, the various arguments for rejection of metaphors in philosophy. Second, the unrestricted affirmation of metaphors, taking "absolute metaphor" as the replacement of metaphysics. The third position can be described as the restricted affirmation of metaphors.
1. The rejection of metaphors in philosophy
The rejection of metaphorical language in philosophy can take any one of five forms: first of all against confounding metaphors and concepts or arguments, secondly, against a purposeful blurring of metaphors and concepts, thirdly, against metaphors in general, fourthly, against using metaphors too often, and fifthly against using metaphors in special functions.
In my opinion, the first reproach, the reproach of exchanging or confounding metaphorical and conceptual discourse, is the most common. But I am also of the opinion, that in a great number of cases the interpreter can be blamed: it can be the readers fault, if he is not able to identify a metaphor as metaphor, for example, out of its context. The author on his side can signal the right interpretation. The decision, if an expression is to act as metaphor or as concept depends on the mode of usage. If the philosopher, for example, emphasizes a metaphor by situating it in an explicit theoretical context or characterizes it with relevant epithets or quotation marks, the danger of hidden metaphors is small.
The second case, the willful blurring of concepts and metaphors, in order to persuade, for example, is less a philosophical than a rhetorical and stylistical problem. The fourth type of criticism, the injunction, not to use metaphors too often, or the fifth type, when making use of metaphors, to do so in special functions and to avoid others, are restricted affirmations of using metaphors. And they may be classified as weaker forms of the third type of critique, the rejection of the philosophical use of metaphors on principle: The main argument against metaphors as adequate articulation is based in an inaccuracy and ambiguity in regard to that, which substitutes for a metaphor. Briefly summarized the reproach is: Instead of explicating something, metaphor only indicates it and this in an equivocal way. Instead of differentiating, metaphor aims at global coherence. Instead of grasping intrinsic structures, metaphor stays extrinsic or even veils such structure. According to this conception - beginning with Aristotle (1) - metaphors are understood first as a rhetoric figure or as a poetical instrument of expression, but not as a philosophical kind of articulation. Compared with concept, sentence, argument or syllogism as the basic philosophical operations, metaphor is thus marked by a lack of strength and conclusiveness. (2)
2. The Value of "ABSOLUTE" metaphors in philosophy
In contrast to this conception and to the attempt to maximize the explicit character of theoretical discourse, advocates of metaphors underline their ability to open and to enlighten precisely because of their ambiguity. According to this point of view metaphors are able to indicate contexts, which are so global and deep, that they defy theoretical comprehension. This conception has a long philosophical tradition as well, particularly since the eighteenth century and Giambattista Vico. (3) In addition to some postmodern trends (and much ealier), Hans Blumenberg can be named as one of the main representatives of this approach with his outline of a so called "metaphorology". I will thus take his approach exemplary.
Blumenberg holds an extreme position in his advocacy of metaphors. In his opinion, metaphors are fundamental elements of philosophy. It is not possible to go behind them, to translate them into a conceptual language or to analyze them. He calls these irreducible philosophical basics "absolute metaphors". (4) In Blumenbergs conception they take the place of the "theoretically non-fulfillable". That philosophical discipline, in which this deficiency is most pronounced, is the realm of metaphysics. We must thus construct another philosophical discipline to complement metaphysics: the realm of "metaphorics". Its function is even compensatory. Blumenberg writes: "The loss of metaphysics calls for metaphorics again on its own place". (5) Here is not the place to criticize Blumenbergs judgement about contemporary metaphysics. I mentioned it merely in order to show the central status of metaphors in his conception: for him a metaphor is a constitutive medium to articulate a cognitive attitude, an attitude, which has not only the same status as metaphysics - as the traditional center of philosophy - but is even superior to it.
The conceptual inaccessibility of metaphors and their central position in philosophy must be studied and differentiated. This study leads to a less affirmative evaluation of metaphors in philosophy. According to Blumenberg metaphors play a compensatory role in philosophy, when its special modes of articulation fail, that is, when concepts, sentences, arguments, and explanations are not able to fulfill their theoretical tasks. This raises the question, in respect to what they cannot fulfill their theoretical tasks and what one demands of them. The idea, that theoretical cognition cannot fulfill its tasks implies in principal an intended object of cognition, which is outside the theoretical realm, outside the realm, of what can theoretically be comprehended. Behind such a conception lies the basic conviction that there exist - perhaps independent - objects of knowledge or a dimension of objects of knowledge thoroughly independent of theoretical knowledge, which are accessible to one kind of knowledge and articulation but shut off from the other. But then it would follow that we are committed to assume something ontologically inaccessible on the opposite side of the epistemologically unachievable, something accessible via metaphor. This however means the abrupt end of any further discourse about this theoretically inaccessible, since it eo ipso can not be determined by theory, or. with Nicholas Rescher : "Inexpressible questions, however, cannot only be not answered, they rather cannot be asked at all [ ]". (6)
It is therefore consistent when Blumenberg doesnt tell us anything about what it is that - compared to theoretical knowledge - absolute metaphors are able to grasp. Instead he gives a description that is purely extrinsic: "Absolute metaphors answer those supposedly naive and unanswerable questions which are only relevant because they cannot be dismissed, because we do not ask them, but find them already asked in the ground of being." This circumscription of absolute metaphors as answers remains vague; it leaves open, what absolute metaphors are answers to, that is (1) which questions we "find [ ] already asked", i. e. what is asked, (2) who has asked them, (3) how they are asked, that is, whether and how these fundamental questions are being formulated, and, as a consequence, (4) how they can be answered at all or by means of metaphors. In other words: neither the syntactic relationship between question and answer by metaphor nor the semantic bearing on something being asked or something that has to be answered metaphorically are further explained. (7)
But, as I said, the intrinsic indeterminacy of the semantic scope of reference of the concept of the theoretical inaccessible is implied in that very concept. Hence, thus Blumenberg: "[In the process of] doing metaphorology we have already deprived ourselves of the possibility of finding answers to those unanswerable questions in metaphors". (8) But how is it then possible to state that metaphors replace metaphysics? And furthermore, how is this to take place: Is metaphorics (as, for instance, art in Schellings Philosophy) developing into an organ of philosophy? And finally: Who claims and is able to prove that there really is something theoretically absolute inaccessible? (9) The implications of the concept of absolute metaphor make clear the contradictions a theoretical text becomes entangled in, when it grants non-theoretical means of expression, a compensatory cognitive function of knowledge and, as a consequence, grants them access to an absolute truth that surpasses theoretical dimensions on principle.
It is up to further research to demonstrate how the absolute metaphor works as an answer. This research must adress the syntactic elucidation of its answering function as well as the pragmatic question, how absolute metaphors are used since a description of one sort of use may lead to a more exact definition of its compensatory character, such as: always then, when the linguistic or cognitive situation x obtains, the absolute metaphor must (or can) be employed. However, to such an inquiry it might be replied with Blumenberg "that a metaphorology can of course not lead to a method for using metaphors or for handling questions being expressed in them". (10) If, however, metaphorology is not able to set norms on the use of absolute metaphors, than all that remains is the particular use of ungeneralizable metaphors. In which sense, however, is it then possible to speak of a metaphorology? What is achieved by such a discipline, and to what purpose, if it neither lays down and gives reasons for philosophical rules of using metaphors nor determines the dimension of their objects of reference? (11) How is for instance the recipient of metaphors able to know, whether he touches the "inaccessible", or whether he understands (if only approximately) what is meant by them? (12)
Yet, in spite of those difficulties, we can claim: What is meant by "absolute metaphors" in Blumenberg could be something like an indirect grammatical echo of relations, which are too comprehensive and/or too deep to be made explicit directly by means of literal language. We find these relations, namely as questions, asked in the ground of being, and in a way obviously existential. This expression points to the fact, that the conceptually incomprehensible relations (or just the theoretically inaccessible) are nonetheless intelligible intentional objects of man in a broader sense and even such as to carry in them the continuous demand to be known. But how they are to put forward this demand or how we find them as demanding at all is uncertain.
One final explanatory approach is the analysis of the kind of metaphorical reference to theoretically inaccessible contexts. That is, the substitution-relation could promise an explanation of absolute metaphor. Characteristically, the relevant thoughts in this connection can be found mainly in a context that deals with the "truth of the metaphors themselves". (13) According to Blumenberg the concept of truth or true - "as result of a methodically established procedure of making something true", in the sense of assigning truth values for instance - can, of course, not be applied to metaphors. Nevertheless, absolute metaphors, or what is expressed by them, correspond (in a way) to the theoretical inaccessible, meet it, grasp it adequately. Therefore Blumenbergs considerations are based allbeit, implicitly and vaguely, on a relation of adequation or correspondence theory of truth. When he speaks of the "value of evidence" of metaphors and then later speaks of "their historical truth" it becomes even explicit that the absolute metaphor is to work as a historically relative bearer of truth. And this is even necessary with regard to the status of a metaphoric: As a compensation of metaphysics within philosophy the metaphoric in Blumenbergs theory must be susceptible to claims of truth. But how can the truth of metaphors, the correspondence to the theoretically inaccessible in the case of the absolute metaphor, be understood more accurately and, at the same time, as different from theoretically comprehensible constellations of truth? Obviously as a historic truth of correspondence, insofar as metaphors aim at the fundamental questions of their respective period and mediate them for the "seeker of a historic understanding"; and furthermore as verité à faire or pragmatic truth in a broad sense as well, insofar as metaphors or their content" give an orientation of behaviour in their historic period. "Metaphor as topic of metaphorology [ ] is mainly a historical object, so that their value as evidence presupposes that the person who makes an assertion didnt herself have a metaphorology". (14) If the correspondence of an absolute metaphor expresses itself only in action and behaviour of the respective user of metaphors of a certain historical period, and if the correspondence, as a matter of principle, approaches its singular object in a historical review, then it is deprived of the ability to show a truth thus understood. The assignment of a claim of truth to an absolute metaphor remains a hypothesis. (15) Failing to give an exact elucidation, Blumenbergs approach is in one sense consistent with his metaphysic-and-theory-compensatory reading of metaphors. But at the same time it is inconsistent, insofar as it tries to explain this by means of discourse.
In spite of the critique, the strong version of affirmation of metaphors brings to light, as Blumenbergs example demonstrates, some characteristics of metaphors, which suggest an restricted affirmation of metaphors in philosophy.
3. Functions of metaphor in philosophical contexts
First of all we have to revise the absolute knowledge claim of metaphors in philosophy, which has more or less as its basis the hypothesis of something theoretically absolutely unfulfillable and unobtainable. One cannot exclude the possibility that something can be made theoretically accessible in principle or in some far future.
But aside from such basic considerations, assuming such a theoretically unfulfillable and correspondingly metaphorically fulfillable can lead to an attitude of theoretical resignation, which finally leads to self-contradiction. If certain aims of knowledge are not accessible in any case, then a premature retreat from a theoretical and philosophical disputation can appear warranted. Theoretical striving and the Hegelian "effort of the concept" is replaced by appeal to metaphors, which do not give or allow an analysis of intrinsic structures, and even veil them, but are supposed to open up holistic contexts. If these were never and in principle are not accessible, if the substitutes of metaphors can therefore never be translated into concepts, then they too can be explained only by metaphors, and these again can be explained only by metaphors, and these again can be explained only by metaphors, and so on.
Then, however, they could not be recognized as metaphors, but only as figures of speech referring to figures of speech, referring to figures of speech, and so forth. Consistently philosophizing in metaphors would always remain on the same level of talking in figures or pictures of speech. Hence an accurate substitutional theory contains the correct description of the structure of metaphors and is therefore, contrary to other assumptions, indispensable - but only on condition that it contains complex connections of the substituent with neighbouring picture or association fields and - corresponding - a conceptually not strictly determinated field of the substitute. Poetry, therefore, has the most complete understanding of metaphors in its use them as independent and constitutive means of expression. (16) Thus I affirm a form of the so called structure of substitution of metaphor, which is quite compatible with Max Blacks "theory of interaction" of metaphor. Black contrasts the interaction-theory with the theory of substitution and its special case, the "comparison view" of metaphor. I think, Blacks rejection is based in the false presupposition, that "substitution" implies total replacement. Black too has namely to assume indirect elements of comparison - between the so called "implication complexes". He must assume a kind of isomorphy as the relation of similarity between the metaphorical expression and the implication complex of another realm, so that the "interaction" comes about, that is, so that the metaphor actually functions as metaphor. (17)
Thus even in philosophical contexts we have to agree to the affirmation of metaphors in two points: In the first place, metaphors, in their historical and individual usage, cannot be completely explained theoretically. Because they do not literally refer, but also have, per definitionem, the characteristic property of referring to something else, in such a way that they can refer to many things, metaphors are not only in need for interpretation, but are also ambiguous. This implies semantical plenitude as well as semantic imponderability, as opposed to strictly defined concepts, exact rules of predicates, etc. in fact a particular metaphor always evades in part. Secondly, metaphors can only have an independent function in philosophy, when they cannot be grasped intrinsically in their individual use, that is, when it is not possible to translate them completely. Otherwise either they would be obsolete or dispensable, because they are conceptual at least in principle, that is to say, they can be replaced by the common and most precise medium of philosophy. Or, from a semantical point of view, they would function as mere redundant decoration: In this function they would, taken strictly, not be allowed to occur in non-aesthetic texts. (18) However, decorative metaphors that are relatively clear and consequently to a great extend explicable, as far as their communality and their familiarity is concerned, can have a high value of mediation in a pedagogic-didactic respect.
The vagueness of metaphors is also a reason for their extrinsic character. Because they represent, so to speak, comprehensive but inexplicit connections from outside, they are well suited for a regulative reference to an orienting frame of special theoretical analyses. Metaphor then might be taken as legitimizing proof of reference to something that is not yet explicit. For this reason metaphors are at the same time precluded from articulating the intrinsic and precise structure of what they ambiguously substitute. They have, therefore, no value for explanation and reasoning. But because of the plenitude also implicit in the ambiguity they have a high value for implication and innovation: Metaphors can therefore be used as means of anticipatory indication of something not yet accomplished by means of theory, and this in different respects. It is thus possible and sensible to take metaphorical formulations as scientific beginning and as intuition for constituting a hypothetical framework for a theoretical process. In another case, complex perspectives resulting from theoretical investigations can be shown by metaphors to be not yet fulfilled. They serve, therefore, not only as justifications of an author, who has not been able to comprehend the aforementioned perspective, but are at the same time a sign of the desideratum of research which has already been recognized. The thus implied impulse of the author to attain the not yet fulfilled or the not yet fulfillable has as its counterpart a need for interpretati on, which is entailed in the structure of metaphors.In addition to their function of semantical indication metaphors indicate within philosophical texts a demand for conceptual explication and theoretical study.
Hence, in their philosophical use metaphors can be characterized in a more narrow sense as non-philosophical, but nevertheless theory-constitutive forms of articulation, and in three respects: they are paraphilosophical, in so far as they express what they aim at in the shape of figures of speech or even pictures. They are periphilosophical in their extrinsic circumscription of what they aim at philosophically. They are, and this is their crucial aspect, proto- or prephilosophical, in so far as they anticipate in a characteristic articulation what is philosophically not yet fulfilled or fulfillable. (19)
(1) Cf. Aristoteles, Topik (Organon V), Übers. u. mit Anm. hg. v. E. Rolfes, Hamburg ²1922 , ND 1968, 158b, 139b, - 140 a, 133b: "Jede Metapher ist undeutlich"/ "Every metaphor is indistinct".
(2) Cf. for instance Nieraads description and critique of this position; B. J. Nieraad, Bildgesegnet und Bildverflucht. Forschungen zur sprachlichen Metaphorik, Darmstadt 1977, bes. 90 f.
(3) See. Giambattista Vico, Die neue Wissenschaft über die gemeinschaftliche Natur der Völker (1744), dt. v. E. Auerbach, mit e. Essay v. E. Hora, hg. v. E. Grassi, Hamburg 1966 bes. Kap. II, Abschn. 2, 2 u. 4.
(4) Paradigmen zu einer Metaphorologie, in: Archiv für Begriffsgeschichte 6, 1960, 7-142, 9; cf. ibid., Ausblick auf eine Theorie der Unbegrifflichkeit (1979), in: A. Haverkamp 1996, a.a.O., 438-454. The expression "absolute Metapher" was already used by Hugo Friedrich, Die Struktur der modernen Lyrik (1956), Hamburg, 8. erw. Aufl. 1977. He understands the "absolute" literally. "Absolute metaphors" are, how one could say, language-pictures, constituting their own "world", totally resolved of the realm, they (may) refer to. I think, the problem is the same as in the conception of Blumenberg: in which way are metaphors then still metaphors ?
(5) Blumenberg 1960, 142
(6) Nicholas Rescher, Die Grenzen der Wissenschaft (1984), trans. by K. Puntel, with an introduction by L. B. Puntel, Stuttgart 1985, 81; for the entire subject of a theoretically inaccessible see especially chap. VIII and IX.
(7) Blumenberg, 1960, 19; the claim of metaphors to be a compensate for metaphysics as well as the reference to a negative theology give the hint, that whats at issue are for instance problems of the philosophical doctrine of god. Cf. Blumenberg, 1979, 445.
(8) Blumenberg, 1960, 19
(9) See Rescher, 1985, especially chap. IV, VII and VIII.
(10) Blumenberg, 1960, 19.
(11) In what Blumenberg says about the "paradigms of a metaphorology" he leaves open the exact meaning of "metaphorology", as for example the question of its status (as a science?) and whether its place is within or outside philosophy.
(12) The strategy of a closer determination of the absolute metaphor by way of a syntactic elucidation of metaphors as answers is of little help if one takes into account, that in one and the same context Blumenberg holds that absolute metaphors "are answers to unanswerable questions, that someone who has knowledge by metaphors, however, cannot find these answers to unanswerable questions in them, and, finally, that metaphors express questions. See Blumenberg, 1960, 19.
(13) ibid. 19/20 ff. One problem of Blumenbergs writings, that cannot be dealt with in this argumentation, is that he confounds two levels: Chapter II, the origin of the above passages, has as its heading "metaphoric of truth and pragmatics of knowledge", that means it will talk about metaphors or their truth, i.e. about "metaphors of power or powerlessness of truth" (p. 19). In what follows he then says that those kind of metaphors cannot be verified theoretically, and he asks therefore "for the truth of the metaphor itself". (My italics, C. P.)
(14) ibid., (my italics, C. P.)
(15) It becomes obvious, that it is in principle questionable to claim the correspondence relation of truth or truth values at all for metaphors, for how can a truth value be found out? For that reason Nelson Goodman suggests making their "correctness" dependent on the fact how "useful, instructive and informative" they are; see Nelson Goodman/ Catherine Z. Elgin, Revisionen. Philosophie und andere Künste und Wissenschaften, dt. Frankfurt a.M. 1993, 32; see Weisen der Welterzeugung, dt. Frankfurt a.M. 1990, 32 128 ff.
(16) Cf. the capable substitution-theory of metaphor of Karl-Heinz Stierle, which is based on the reciprocal influence of metaphor and its context; Karl-Heinz Stierle, Aspekte der Metapher, in: Text als Handlung. Perspektiven einer systematischen Literaturwissenschaft, München 1975, 181 f., 183 f., 185 f.; see e.g. Nieraad, 1977, 13 f. and Max Black, Die Metapher (1954), in: Theorie der Metapher, hg. v. Anselm Haverkamp, Darmstadt, 2. um ein Nachw. u. einen bibliograph. Nachtrag erg. Ausg. ergänzte Neuaufl. 1996, 55-79 [Black 1996 a]; ibid., Mehr über die Metapher (1977), ibid., 379-413 [Black 1996 b], bes. 392/393.
(17) Cf. to the "theory of interaction" Black, 1996 a und b; against Blacks interaction theory, Robert Fogelin endorses the comparative approach, when he defines metaphors as "figurative comparisons". But he do so in the context of supporting an approach, which, like Davidsons theory, takes metaphors literally (which I do not agree with); c.f. Fogelin, Figuratively speaking, New Haven 1988, 28, and c.f. Fogelin, Metaphors, Similes and Similarity, in: J. Hintikka (ed.), Aspects of Metaphor, Dordrecht 1994, 23-39; c.f. to the literal view Donald Davidson, What metaphors mean, in: Inquiries into truth and interpretation, Oxford 1984, 245-264; see also Stierle, 1975, 152; to avoid the obviously misleading terminus "substitution" I therefore propose a substitution-theory, which might be called "Theory of interactive indication" of metaphor.
(18) "aesthetic" in the sense of "sensuous-(artistic)"; when we mean the adjective to aesthetics-as-theory, we use the construction "aesthetic-theoretical".
(19) See Constanze Peres, Antizipation. Spektrum und Struktur, in: F.Gaede/ C. Peres (Ed.), Antizipation in Kunst und Wissenschaft. Ein interdisziplinäres Erkenntnisproblem und seine Begründung bei Leibniz, Tübingen 1997, 19-33; Pegasus und Einhorn. Antizipation in Kunst und Wissenschaft und ihre Begründung bei Leibniz und Goodman, ibid. 47-72.