Metaphor has long been under suspicion as a mere embellishment, a rhetorical flourish, a linguistic device, a way of pulling the wool over the eyes of the reader or listener. The dismissive phrase 'a mere metaphor' is indicative. Argument has raged since Aristotle as to whether metaphor is merely ornamental to already existing thoughts or words, or actually creative and constitutive of thoughts and words. Writers in the classical tradition, together with analytic linguistic philosophers, have inclined to the 'ornamental' or 'substitutionary' view of metaphor. The Romantics, followed by many literary critics and some philosophers of science, have held to a 'constitutive' or 'incremental' view. These two schools correspond broadly to what we earlier identified as the 'analytic or naturalistic' and 'synthetic or fiduciary' traditions. Let us examine more closely the argument between these two assessments of metaphor, beginning with some definitions.
First we note that metaphor itself has become the battleground for these two fundamental views of language. Although there are numerous figures of speech in the grammatical repertoire—hyperbole, metonymy, catachresis, oxymoron, to name but a few—metaphor has tended to absorb the others. Metaphor is no longer one figure among others, but 'the figure of figures' (Culler, 1981, p.189). Metaphor is widely recognised as the operative factor in language. Ricoeur quotes Shelley that language is 'vitally metaphorical' and himself claims that metaphor is the constitutive form of language (Ricoeur, 1978, p.80). The classical scholar Bedell Stanford writes:
Metaphor is the vital principle in all living languages. It is the verbal expression of the process and products of the imagination with its powers of creative synthesis. Metaphor is thus the dynamic, synthetic and creative force in language.
Second, metaphor is notoriously difficult to interpret. We are hard put to say how exactly it works. As Bedell Stanford remarks in his study of Greek metaphor, 'a fine metaphor is one of the hardest things in the world to rationalize'—and the problem becomes particularly acute when translating a metaphor from the original language into another. Reflecting on the problems of translation, George Steiner has observed that we cannot fully imagine what it must have been like to be the first to compare the colour of the sea with the dark of wine or to see autumn in a man's face. Such figures are new mappings of the world, they reorganise our habitation in reality.
Several writers have emphasised that the meaning of a metaphor depends almost entirely on the immediate context: a metaphor is justified solely by its use (see Searle, 1979; Davidson in Sacks, 1979, p.41; Swinburne, 1992, p.48).
Third, we need to try to clarify the two components that make up a metaphor. Aristotle said that metaphor consists in giving a thing a name that belongs to something else. A basic definition of metaphor, along these lines, is that metaphor describes one thing in terms of another, it joins together two perceptions, an immediate or primary perception and a borrowed or secondary perception. Max Black suggests that metaphor 'selects, emphasizes, suppresses and organizes features of the principal subject by implying statements about it that normally apply to the subsidiary subject' (Black, 1962, p.44). I.A.Richards distinguished between the 'tenor' and 'vehicle' of metaphor, the tenor being the conceptual meaning and the vehicle being the concrete comparison (Richards, 1965, pp.96f.). However, it does seem to me that these terms could be reversed, for the image is often the tenor or substance of the thought and the original occasion for the metaphor simply the vehicle for delivering the insight. These ambiguities suggest to me that it is unhelpful to attempt to decide which of the two components of metaphor is primary and which is secondary. It may suffice to speak of the 'event' of metaphor in which two aspects are fused together instantaneously. If we need to distinguish them conceptually, for the purpose of analysis, we can call them the 'occasion' of the metaphor and the 'image' through which we view the occasion. Together they comprise the total event of metaphor. Let us now try to elucidate these aspects by means of examples.
Shakespeare's paradoxical metaphor 'light thickens' in
Light thickens, and the crow
Makes wing to the rooky wood;
Good things of day begin to droop and drowse,
While's night's black agents to their preys do rouse.
comprises the 'occasion' (the perception of the gathering dusk) and the 'image' (the perception of a substance something like porridge that progressively thickens as it settles). There are also sinister overtones: as 'light thickens', so 'the plot thickens' and so also spilt blood coagulates.
To stay with nocturnal metaphors for a moment: Dylan Thomas' night 'starless and Bible-black' in Under Milk Wood (Thomas, 1985, p.1) combines the occasion, the physical perception of the night, with the image of the Bible, carrying resonances derived from the child's associations of the family Bible—not only black, but great, heavy, oppressive and awesome. 'Black' is probably the most frequently used adjective in the play and in combination produces a dazzling range of images of night: 'the sloeblack, slow black/crowblack, fishingboat-bobbing sea' (op. cit.); 'the black, dabfilled sea' (ibid., p.2); 'the slow deep salt and silent black, bandaged night' (ibid., p.3).
To move from night to day and from darkness to light: when Louis XIV was called 'le Roi Soleil', the occasion was the perception of the monarch's glory, the image that of the sun in its splendour.
While we are considering metaphors of darkness and light—one of the earliest and most basic insights of human experience—let us look at Blake's tiger, which has already been mentioned as an exceptional example of the pre-Romantic genre of the awe-inspiring 'sublime'.
Tyger, Tyger, burning bright In the forests of the night.
The occasion is not just seeing, but facing or confronting the tiger in the imagination (Blake had seen caged tigers in London, but when he tried to draw a tiger it looked like a rather foolish, friendly pet dog). The beast suddenly manifests itself through the thickets of the jungle. The image is of fire, or rather 'burning': is it two eyes glowing menacingly like red-hot coals? Or are we seeing the animal sideways, with its stripes like flames flickering as its supple muscles ripple along its coat? The tiger burns the more brightly, of course, for being set against the dark backcloth 'the forests of the night'—not just physically dark, but sinister with the impartial cruelty of amoral nature. It is because we already know the second verse that we picture the glowing eyes:
In what distant deeps or skies Burnt the fire of thine eyes!
In the fourth verse the primary metaphor of this poem ('burning bright') is extended, but without losing—it seems to me—its character as metaphor: the 'furnace' is the unfathomable, all-creating, all-consuming mind of its Maker:
What the hammer? What the chain, In what furnace was thy brain? What the anvil? What dread grasp, Dare its deadly terrors clasp?
Blake is here the begetter of one of the most powerfully sustained extended metaphors in literature.
For the Romantics, as we have already noted, the Bible and the natural world were excelled as sources of the sublime by the human mind. The visions and prophecies of scripture became merely the types of the visions that appeared to the Romantic imagination (here Blake surely cannot be denied the name of Romantic). The wonders of natural scenery were simply an analogue of the landscapes of the mind and provided a hermeneutical key to mapping them. Wordsworth and Coleridge conducted a rigorous exploration of the psyche in its heights and depths, not just as they walked the fells but precisely by walking the fells. The Romantics' love-fear relationship with the sea supports this interpretation. Coleridge in 'The Ancient Mariner', Shelley in dicing with death on the fickle waves of the Gulf of Spezia and Byron in swimming the Hellespont, respectively knew by instinct what later depth psychology has abundantly shown: the sea is one of the primary symbols of the unconscious. Coleridge united all three sources of the sublime—the imagery of the Bible, of precipitous nature so enthralling to the Romantics and of the turbulent yet creative psyche—when he said that the poetic mind of Wordsworth was 'a rock with torrents roaring'. Gerard Manley Hopkins receives and is almost overwhelmed by the dark side of Romanticism in his 'Terrible Sonnets', but he needs the imagery of landscape to name the experience of desolation that engulfs him:
O the mind, mind has mountains; cliffs of fall
Frightful, sheer, no-man-fathomed.
Hold them cheap
May who ne'er hung there.
The occasion, if we must be pedantic, is Hopkins' depression or 'dark night of the soul'; the image that we see is physical, momentous, vertiginous. They are united in the single event that is gifted or inspired metaphor.
Fourth, as will by now hardly need stating, metaphor is a matter not just of words but of thoughts. Metaphor is generated in the drive to understand experience. 'When new, unexploited possibilities of thought crowd in upon the human mind, the poverty of everyday language becomes acute. Apprehension outruns comprehension so far that every phrase...has a vague aura of further significance' (Langer, 1956, p.149). In fusing together two perceptions, metaphor has a stereoscopic function. Each image carries with it a host of associations which interact and fertilise each other. Metaphor is not just naming one thing in terms of another, but seeing, experiencing and intellectualising one thing in the light of another. To evoke M.H.Abrams' work on the Romantic imagination, metaphor is not just a mirror but a lamp (Abrams, 1953). I.A.Richards suggests that a metaphor is not so much an interaction between words as 'a borrowing between, and intercourse of thoughts', for it is primarily thought that is metaphoric and individual metaphors of language derive from it (Richards, 1965, p.94). Middleton Murray, in Countries of the Mind (1937)—the title itself a legacy of the Romantics—supports this interpretation. Metaphors, he says, belong to 'the primary data of consciousness'; they are 'as ultimate as speech itself, and speech as ultimate as thought'. Metaphor is the 'instinctive and necessary' way in which the mind explores reality and orders experience, using intuitive comparisons whereby the less familiar is assimilated to the more familiar and the unknown to the known. Metaphor belongs to 'the genesis of thought' and opens up reality in ways that are 'hazardous, incomplete and thrilling' (Murray, 1937, pp.1f.). It is important that we locate metaphor in this context of knowing. A metaphor is not just a word, not even an image, but a conjunction of discourses, a joining of worlds.
Fifth, though we sometimes loosely say 'image' when we mean 'metaphor', a metaphor is much more than an image. As P.N.Furbank puts it, a picture is not at all the same as a comparison. In metaphor we envisage one thing in terms of another, bringing two images into juxtaposition. An image is somehow static, while a metaphor is 'an invitation to an activity, ending in an impossibility' because the two pictures or images that comprise a metaphor do not mesh: 'Macbeth has murdered Sleep.' Both painting and sculpture deal in images, but in poetic metaphor it is the words, the syntax and the rhythm that do the work (Furbank, 1970, pp.1-12, 23).
Finally, we must clarify the difference between metaphor and simile. Fogelin has suggested that 'similes wear their comparative form on their grammatical sleeves, and metaphors...differ from similes only in a trivial grammatical way: metaphors are similes with the term of comparison suppressed, they are elliptical similes' (Fogelin, 1988, p.25). But if the grammatical difference between metaphor and simile is 'trivial', the difference of impact or import is great. Metaphor and simile belong to two different psychological worlds. They arise out-of different experiences or perceptions; they generate different experiences and perceptions. Fogelin takes up Aristotle's example from The Iliad—'the lion leapt' (metaphor); 'Achilles leapt like a lion' (simile)—and claims that the difference is slight. Analytically it may be slight, but in terms of the perception, the experience and the effect it is considerable. Homer's metaphor presents, virtually simultaneously, two images—Achilles and the lion—and it is hard to say which one is dominant. Translated into simile, however, the dominant image is definitely Achilles, with the lion comparison following rather doubtfully behind. If we return to some of our earlier illustrations of metaphor, centred on the basic images of darkness and light: the differences between these metaphors and the same comparisons in the form of similes is not only the formal elaboration of the thought—the night is getting darker; the night is as black as the Bible; the King of France is resplendent like the sun—but also the loss of immediacy and spontaneity. Wilson Knight, the Shakespearean critic, calls metaphor 'a concrete image. fused with a spiritual vitality' (Knight, 1933, p.35).
The debate between comparisonists and non-comparisonists is beside the point. There is a comparison involved in metaphor but it has a spontaneity, immediacy and vividness that is absent in simile. Metaphor has been called 'a condensed comparison by which we assert an intuitive and concrete identity' (Ullmann, 1964, p. 180). Identity is surely the key to the nature of metaphor. Mallarmé spoke as a true poet when he said: 'I have struck out the word like from the dictionary' (ibid., p.181). Ricoeur suggests that metaphor is located not in the word, the sentence or even the discourse, but in what is elided—'the copula of the verb to be. The metaphorical "is" at once signifies both "is not" and "is like".' Hence, for Ricoeur, metaphors are (equally with symbols) vehicles of 'tensive' truth (Ricoeur, 1978, p.7).
Ornamental or incremental?
I use the terms ornamental and incremental to designate the two main rival traditions in the assessment of metaphor. The first regards it as a rhetorical flourish to increase effect, the second as a source of insight; the former naturally tends to regard it as dispensable, the latter as unsubstitutable. I.A. Richards observes that 'throughout the history of Rhetoric, metaphor has been treated as a sort of happy extra trick with words .a grace or ornament, or added power of language, not its constitutive form' (Richards, 1965, p.90). It is thanks to the former tradition that metaphor is generally suspect in argument and carries a 'logical taint', as Middleton Murray puts it (Murray, 1937, p.4). We may take Aristotle and Samuel Johnson as representative of this reductionist school of metaphor.
Aristotle is, in effect, the source of the ornamental view of metaphor. While his observations are occasional and therefore fragmentary, they have exerted a profound influence on the assessment of metaphor. In the Poetics, he states that metaphor cannot be borrowed from someone else and is an indication of genius, for 'to observe metaphors well is to observe what is like' (Aristotle, 1987, p.28:59a5). The Rhetoric adds little: metaphor is 'in the highest degree instructive' and the most successful simile is one that comes closest to metaphor (Aristotle, 1886, 3.10.11). Bedell Stanford judges Aristotle's statements on the nature of metaphor to be confused and unhelpful (Stanford, 1936, pp.512), but the philosopher set the trend for the 'classical', ornamental view of metaphor.
Boswell records Dr Johnson pronouncing that metaphor is 'a great excellence in style, when it is used with propriety, for it gives you two ideas for one;—conveys the meaning more luminously, and generally with a perception of delight' (Boswell, 1953, p.855). In his celebrated Dictionary, Johnson defined a metaphor as 'the application of a word to an use to which, in its original import, it cannot be put' and summed it up as 'a simile comprized [sic] in a word' (Johnson, 1755, 'Metaphor'). Johnson seems to regard metaphor as a matter of artifice, to be produced when required in order to achieve an intended effect. As we have seen in an earlier chapter, the rationalistic strand of the Enlightenment, influenced by Descartes' insistence of clear and distinct ideas, was suspicious of metaphor. Hobbes and Locke, with their 'counter' theory of language, regarded metaphor as a form of deception.
Romanticism reinstated metaphor as a vehicle of insight; reflection on Romantic literature has given impetus to the constitutive or incremental view. Coleridge's distinction between the contrivances of Fancy and the creative power of Imagination seems to correspond to the difference between simile and metaphor (though Coleridge preferred to discuss the more inclusive concept of symbol).
In his early work The Birth of Tragedy and before he had adopted his later jaundiced view of language and its images, Nietzsche recognised the importance of metaphor: 'For the true poet,' he wrote, 'the metaphor is not a rhetorical figure but a representative image that really hovers before him in place of a concept' (Nietzsche, 1993, p.42). The depth psychology of both Freud and Jung revealed the symbolic nature of unconscious thought processes and the determinative role of mental images in shaping our perception of reality.
Philosophers of science, including practitioners such as Einstein and Michael Polanyi, have asserted the heuristic power of metaphor to explore reality and order experience. Leatherdale (1974) provides an overview of the role of metaphor, analogy and models in science, based on a cognitive view of metaphor. Mary Hesse in particular insists that poetic metaphors are cognitive, not just expressive, and intend a 'redescription of reality' (Hesse, 1966, p. 176). She affirms a continuity between poetic and scientific metaphors, arguing that 'rationality consists just in the continuous adaptation of our language to our continually expanding world, and metaphor is one of the chief means by which this is accomplished' (ibid., p.164). She points out that acceptance of the view that metaphors are meant to be intelligible implies rejection of all views that make metaphor a wholly non-cognitive, subjective, emotive or stylistic use of language (cf. Hesse, 1988). Hesse is indebted here to Max Black's 'incremental' theory of metaphor.
Black made a notable contribution to the reclaiming of metaphor by developing I.A.Richards' 'interaction' theory into an incremental theory of metaphor which stresses that metaphor is neither decorative of thoughts already chosen nor a substitute for them, but creatively constitutive of new ways of seeing the world. Living metaphors give emphasis and so awaken insight; they give resonance and so invite implicative elaboration (Black, 1962, pp.44f.; Black in Ortony, 1979, pp.19-43).
A similar approach has recently been developed by Kittay whose 'perspectival' theory is broadly equivalent to Black's 'incremental' concept. Kittay insists that metaphor has cognitive value and that this stems not from providing new facts about the world but from a reconceptualisation of the information that is already available to us. In this sense, metaphor actually gives us 'epistemic access' to fresh experience and, to the extent that we have no other linguistic resources to achieve this, metaphor is 'cognitively irreplaceable' (Kittay, 1987, pp.39 [cf. 303], 301). This incremental view of metaphor spans literature, science and theology.
Among theologians, Jungel is notable for his positive view of metaphor. Jungel's presupposition is one that could not be held by a theologian in the analytical tradition, that God becomes a reality in the world through language. He shares the incremental view of metaphor that has been promoted among theologians by Ricoeur. Jungel grounds human transcendence in the capacity for making symbols, culminating in language: the creative freedom evinced in language is the guarantee of ultimate human freedom.
The metaphorical mode of language has ontological relevance insofar as through it a new context of being is disclosed, grounded in a gain to language. The new (metaphorical) use of a word gives this word a new meaning and with this new meaning new being is brought to speech.
Soskice advocates a form of critical realism in which metaphorical description, in science or theology, has cognitive value but is not (of course) claimed to be veridical; as she puts it, it refers and depicts, but does not claim to define. This implies for theology that talk of God, which cannot be abstracted from the world of religious images, can depict reality without claiming to describe it definitively (Soskice, 1985, pp.140f.). Davis has seconded this position, claiming that 'irreducibly metaphorical utterances can themselves state truths about the world'. Inadequate and inaccurate though they may be, they can still successfully refer to reality (Davis, 1989, pp.10, 13; see also Pickstock, 1998, pp.169ff.).
When we want to say, with emphasis, that something is 'really' true, we sometimes claim that it is 'literally' true. This is perhaps pardonable in common speech, though it is unhelpful and actually meaningless, but in theology, where it is not unknown, it is a classic faux pas. Where the metaphors of religious confession are not recognised as such, the qualifier 'literally' produces some interesting biological and cosmological situations with which, I guess, most of us have at some time been regaled by fervent fundamentalists: 'Jesus was literally God's Son. He literally came back to life on the third day. He literally sits at God's right hand. He will literally come again on the clouds.' We will be returning to this problem of misplaced literalism in our final chapters when we consider the implications of metaphor, symbol and myth for the truth of Christian beliefs, but I mention it now to suggest that our approach to metaphor so far leads to a critical reevaluation of the received distinction between metaphorical and literal language.
There is a danger in exalting metaphor that we forget that not all metaphors are full of spontaneous creativity and open new horizons for human knowing. The language is full of metaphors that are routine, hackneyed or inept. Fogelin makes the point that 'many metaphors are lame, misleading, overblown, inaccurate' and advises: 'It is important to calm down about metaphors. Some are good; some are bad. Some are illuminating; some are obfuscating' (Fogelin, 1988, pp.98f.). Most of our language is composed of 'dead' metaphors (which is a good metaphor in itself) but metaphors they remain. The notion that it is possible to 'translate' a metaphor into 'literal' terms and that this is an advance in understanding is the literalistic fallacy.
The literalistic fallacy is found in some of the most respectable places. The philosopher W.M.Urban, with whom we have frequently made common cause, claims: 'Just as to call all knowledge symbolic without qualification, is to make the notion of symbolism meaningless, so to call all language symbolic is to make unintelligible the distinctively symbolic uses of language.' But Urban concedes that it is impossible to find any use of language that is wholly non-symbolic (Urban, 1939, pp.412, 435). Our likeliest candidate for non-symbolic language is perhaps 'scientific' language of (supposedly) pure description of bare sense-data. But we need to recognise that mere sense-data cannot become the subject of linguistic expression until they have been processed by conceptual thought—a powerfully symbolising event. So Urban admits that 'the only literal or non-symbolic element in scientific knowledge is. precisely that which is not knowledge at all, namely the mere sense-datum' (ibid., p.544). The best solution, he concludes, is to regard the designations 'literal' and 'symbolic' as relative, as limiting notions, as abstractions. Statements which relate more closely to 'sensuously observable entities' are appropriately labelled 'literal', while statements which relate more closely to interpretative constructions, which have to be translated back into so-called literal statements in order to be verified, are more appropriately called symbolic (ibid., pp.423f., 544).
However, I would go further than this and ask whether the distinction between literal and metaphorical (or symbolic) usage can be rigorously maintained at all. Barfield suggests that literal and metaphorical are relative terms existing in tension and lending meaning to each other. So-called literal language is comprised of metaphors that have lost their living force—dead or extinct metaphors. 'Literal' and 'metaphorical' are merely limit concepts on a sliding scale of imaginative investment—the imaginative investment in so-called literal speech being minimal. It is therefore a fallacy to assume that the original state of language was unremittingly literal and that metaphor represents a deviation from this. A hard and fast distinction 'is based on the premise that literalness of meaning is some kind of unclouded correspondence with a mindless external reality, which was given from the start'. On the contrary, Barfield argues, what we call literalness is a late stage in historical linguistic development. It is impossible to imagine the birth of the first metaphor in a wholly literal world. Metaphor is the primary and inescapable constituent of language as it is of thought. Consciousness and symbolisation (including metaphor) are actually simultaneous and correlative. Literalness, Barfield concludes, is 'a quality which some words have achieved in the course of their history; it is not a quality with which the first words were born' (Barfield in Knights and Cottle, 1960).
Whereas the naturalistic or analytical view of language takes literal truth as its ideal 'speech situation', regarding metaphor and symbol as concealing or distorting the truth, the fiduciary view of language holds that symbolic modes of speech, particularly metaphor, are not a mere adornment to be stripped away in order to reveal the reality underneath, but themselves truly participate in the reality that they seek to convey and induct us into it. The first approach entails the fallacy that reality can be known independently of language, that intuition is separate from expression. The second approach involves a personal commitment to understanding what linguistic symbols are capable of telling us: it is the approach of hermeneutics. Theologians above all need to recognise that any quest for greater and greater degrees of literalness is a wild-goose chase. All the significant assertions of theology are expressed in language that is irreducibly metaphorical.
Was this article helpful?