Startside Op Born Busch Dines Grodal Grund Halskov Hansen Jorgensen Klitgaard Lauge Lundquist Nielsen Siggaard Sondergaard Svendsen Forord Tekster Deltagere


Intentionality, Food and Music -
A Fictionalist Approach

Cynthia M. Grund(1)

In this article, it is suggested that theories of metaphor be augmented with the notion of fiction so as to provide the sort of premetaphorical counterfactuality which is required for the direct understanding of certain concepts. Vaihinger's treatment of fictions is presented in detail and the relevance of fictions to the metaphor theories of Lakoff, Turner, Johnson, Levin and Hausman is discussed. The article concludes with an examination of the suggestion that the "sound-food" fiction be entertained as constitutive of our concept of music.

Introductory remarks

The role played by metaphor in the elaboration and exploration of concepts has been widely discussed, as is indicated throughout the course of this essay. If one chooses to express a given metaphor in the canonical form "A is B," the usual tack is to say that "A" is either directly understood, by which is meant "on its own terms," free of metaphorical mediation, or that the understanding of "A" which has relied on metaphorical mediation can be scaled away by a deconstructive recursion which identifies these anterior metaphors. If we envision this as an n-step process, it will end with some first member. Call it "An is Bn" in canonical form. "An" is thus the component of "A" which is directly understood, viz. is understood "on its own terms" and is free of metaphorical mediation. This can of course, be achieved with many concepts. (Later on in this essay we will see how Lakoff and Johnson regard the understanding of "light" and "death" in this fashion.) There is thus either direct understanding, or metaphorical mediation.

Are there not, however, concepts which require for even the most direct, nuance-free kind of comprehension a component which involves the understanding of one domain in terms of another? With regard to such concepts, it seems paradoxical to maintain that they are inherently metaphorical, or something along those lines, because the concept itself is not available for metaphorical reconfiguration until the interdomainal understanding indicated in the above has been achieved. To capture such a concept at all, to grasp it at its most basic level, requires the sort of counterfactual transfer from one domain to the other which is characteristic of metaphor, but for which it is simply too early to introduce metaphor onto the scene.

At least one such group of concepts can be strongly suspected to be of this sort: those for which even direct, nuance-free understanding requires a relationship between an agent and some constitutive part of the concept, i.e., where the concept itself is not determined until an interpretive act upon its constitutive part(s) has been carried out by an agent and where this interpretation involves some form of counterfactual transfer from one domain to another. These postulated concepts are thus highly intentional in nature; that is to say, any attempts to delineate them without attention to features of the mental states of an agent who is somehow placed in an interpretative relationship with some of their constitutive parts will result in an underdetermination of the concepts themselves which is of such a magnitude as to render them unintelligible.

Many examples from the area of aesthetics are of this sort. It is patently clear that to understand what dance, painting, or music are requires some transformational, intentional component at the most basic, direct level of understanding. There is no level of understanding of these terms which is truly a level of understanding, of even knowing what they are, without this component. It must be present at the level of basic conceptual comprehension before any metaphorical attribution can take place. One could put it this way, that the level of intentionality required for comprehension-as (e.g., seeing-as or hearing-as) which underpins virtually all concept formation and understanding, requires a supplementary level of intentionality in which comprehension-as-if (e.g., seeing-as-if and hearing-as-if) are central components.

How should we then characterize this transformational, intentional component of direct understanding? The suggestion which will be examined is that the supply of tools employed in the analysis of concepts within the context of metaphor theory needs to be expanded to include fictions. By fiction here is meant the technical philosophical device expounded by Vaihinger (and, in an earlier form, by Bentham). By using music as an example of the sort of intentional concept indicated in the preceding, we will examine the role of fictions as opposed to the role of metaphors in the constitution of at least some intentional concepts. The temptation to conflate fictions with the results of metaphorical activity will be explained by examining the role which counterfactual conditionals play in the analysis of both fictions and metaphors. We also will see how the importation of some of the analytical tools employed in recasting metaphors as counterfactuals into the understanding of fictions in terms of counterfactual conditionals can help in the framing of the sorts of fictions which can be useful for capturing the intentionality inherent in some concepts.

In this paper, I put forth and elaborate upon the suggestion that interesting perspectives may be provided upon music and music aesthetics by means of seriously entertaining the highly implausible counterfactual statement (C) about a given sonic sequence u:

(C) If foodstuffs were sonic sequences, then u would be food.

The roles which (C) might play within theorizing about music and the aesthetics of music are discussed. One such - admittedly controversial - role is that the set U consisting of all such sequences u would be - music! (C) would thus be the main constituent in a proposed definition of what music is, where suitable relativizations regarding hearers and their background assumptions are taken into account by the underlying semantics for statements like (C).


(C) represents a concrete suggestion for the characterization of music as sound heard as if it were something else, namely, foodstuffs. Readers who have embraced Lakoff and Johnson's well known ideas about cognitive metaphor in Metaphors We Live By (1980) and Women, Fire and Dangerous Things (1987) and Lakoff and Turner's treatment of it in More Than Cool Reason: A Field Guide to Poetic Metaphor (1989) might already at this point feel compelled to interject: Let's look at the way music is talked about in our culture and others, and see if it is characterized in food-terms. If we find enough such examples, then we may conclude that the conceptual - and if we're lucky, even the functional - embodiment of the concept 'music' is structured by the 'Music is food' metaphor (2).

An extremely important, but subtle move is almost ignored in all of this. As seductive as all of this is, in the hypothesized basic conceptual metaphor "Music is food," it is already implicit that there is a sonic category - music - which is already a conceptually staked-out domain previous to its employment in the metaphor. Lakoff and Turner acknowledge this with regard to basic conceptual metaphors when they remark:

In the case of profoundly conventionalized conceptual metaphors, such as the basic metaphors we discuss in this book, aspects of one concept, the target, are understood in terms of nonmetaphoric aspects of another concept, the source. A metaphor with the name A is B is a mapping of part of the structure of our knowledge of source domain B onto target domain A (Lakoff and Turner 1989:58).

For the purposes of much of my investigation, which is simply to get a grip on how that delineation occurs, the level at which "Music is food" is located is simply too high, having overshot the whole question of discriminating music from sound by taking for granted that we already have a working concept of "music" at hand. Of course, we will be challenged to think more about what this concept is like when we are confronted with "music is food," and, if we agree that there is something "true" about this metaphor, then we will have augmented our insight into the nature of music as a concept which we already have available, it can eventually steer the criteria according to which we judge and enjoy music, accept new pieces into the canon, etc., etc. The upshot of all this is "Music is food" relies on some more basic characterization of what music is in order for "Music is food" to make any sense as a metaphor (3).

Let's look once again at

(C) If foodstuffs were sonic sequences, then u would be food.

Recall that u denotes some sonic sequence or other. Therefore, "u is food" is arguably at the level of poetic metaphor for Lakoff and Turner, self-conscious and non-automatic as it is: ". . . poets lead us beyond the bounds of ordinary modes of thought and guide us beyond the automatic and unconscious everyday use of metaphor. What makes poetic metaphor noticeable and memorable is thus the special, nonautomatic use to which ordinary, automatic modes of thought are put" (Lakoff and Turner 1989:72). It is Lakoff and Turner's view that the reason we are able to understand poetic metaphor is that it builds, by means of various devices, upon a tacit understanding of a myriad of basic conceptual metaphors, the presence of which may be unearthed in a culture by empirical study of the way concepts are talked about.

Now, to get back to the "u is food" in (C), its assertion is conditional upon the antecedent "Foodstuffs are sonic sequences." I think that it is far-fetched, to say the least, that "Foodstuffs are sonic sequences" is a basic conceptual metaphor, at least within any cultural contexts with which this writer is familiar. Let's say it's a poetic metaphor. On what basic conceptual metaphor is it based? Isn't it simply an invitation to enter into a highly deviant world, where food is sound, and then to use whatever intuitions we can muster from our attempts to wrap our minds around "sound-food" to decide whether or not the sound we now are confronted with is food on this implausible hypothesis? It is clear that we must use imagination to transfer the grid we usually have laid down within the food realm upon the sonic one.

Counterfactuals, metaphors and Tormey

We do, indeed, seem to be in real metaphor territory here. It does not, however, seem to be territory for which Lakoff-Johnson-Turner (hereafter abbreviated as LJT) provide us with any readily useful map. We are, however, within a realm in which analyses of counterfactual thinking do. Indeed, as I have indicated elsewhere (4) and will reiterate within the course of this essay, thinking about counterfactual thinking can, in fact, be extremely illuminating with regard to a host of issues regarding metaphor, and as has been adumbrated, matters of musical ontology and interpretation as well. By examining the way in which counterfactual thinking expressed in terms of counterfactual conditionals also provides a context for understanding the creation of fictions and their use in the constitution of concepts, we will see more clearly how fictions play a role which is misleadingly like that of metaphor on the LJT program, but which is actually on the level of non-metaphorical concept constitution, on the level of understanding a concept on its own terms. The fact that implausible counterfactuals also underlie metaphors on our analysis explains how and why this confusion may occur: Metaphors are analyzed as implausible counterfactual conditionals. Alan Tormey has introduced this approach into the literature dealing with the philosophical analysis of metaphor (see Tormey 1983). He analyzes "Juliet is the sun" as

If Juliet were a celestial object, she would be the sun.

Tormey pointedly refers to this as the "counterfactual parsing" (Tormey 1983:240) which "recasts" (Tormey 1983:237) the original metaphor. He does this in order to highlight the fact that he is not paraphrasing the metaphor when he recasts it as an implausible counterfactual. What is commonly understood as the attempt to paraphrase a metaphor is, for Tormey, better reconstrued as an attempt to provide supportive grounding for it (Tormey 1983:242). The relocation of attempted paraphrases as attempts at providing supportive grounding is an important and helpful aspect of Tormey's approach, and we will look at its bearing upon the themes of the present essay a bit later on (5).

Let's now look more carefully at the suggestion which I made earlier that fictionalism supplies a missing link in the LTJ theory of metaphorical v. non-metaphorical understanding of concepts. In short, the creation of fictions, we will see, can be analyzed as a form of counterfactual thinking which is expressed by counterfactual conditionals highly remininiscent of the ones which underlie metaphors. We will examine this in terms of the example of the "sound-food" fiction which was suggested earlier as a means of infusing the concept of music with intentionality at the level at which, as Lakoff and Turner often say in More Than Cool Reason, the concept is understood "on its own terms."

Fictionalism and Vaihinger

We begin with a discussion of Vaihinger's brand of fictionalism (6).

There are five cardinal points:

  1. The practical value of thought ranks above knowledge.
  2. Artifices of thought are distinguished from rules of thought and the former are allocated a very important role.
  3. The artifice which Vaihinger wishes to examine further is that of fiction.
  4. Fictions are divided into two kinds: real fictions and semi-fictions.
  5. Fictions must be distinguished from hypotheses.

(1) underscores some of the similarities which fictionalism has with pragmatism (7). Vaihinger's position regarding "objective reality" is interesting, in that it underscores (1) and discloses some attitudes which are shared by LJT when they criticize "the objectivist paradigm" (8)


Since, however, we do not know objective reality absolutely but only infer it (and this is also an ordinary scientific view) we must revise our statement and say that thought has fulfilled its purpose when it has elaborated the given sensation-complexes into valid concepts, general judgments, and cogent conclusions, and has produced such a world that objective happenings can be calculated and our behaviour successfully carried out in relation to phenomena. We lay most stress on the practical corroboration, on the experimental test of the utility of the logical structures that are the product of the organic function of thought. It is not the correspondence with an assumed "objective reality" that can never be directly accessible to us, it is not the theoretical representation of an outer world in the mirror of consciousness nor the theoretical comparison of logical products with objective things which, in our view, guarantees that thought has filled its purpose; it is rather the practical test as to whether it is possible with the help of those logical products to calculate events that occur without our intervention and to realize our impulses appropriately in accordance with the direction of the logical structures (Vaihinger 1935:3).

The mention of "the organic function of thought" in the above is interesting, in light of LJT as well, since it bears striking resemblance to LJT's ideas about the "automatic and unconscious everyday use of poetic metaphor" (Lakoff and Turner 1989:72). Compare with the following remark by Lakoff and Johnson

Basic conceptual metaphors are part of the common conceptual apparatus shared by members of a culture. They are systematic in that there is a fixed correspondence between the structure of the domain to be understood (e.g. death) and the structure of the domain in terms of which we are understanding it (e.g. departure). We usually understand them in terms of common experiences. They are largely unconscious, though attention may be drawn to them. Their operation in cognition is mostly automatic. And they are widely conventionalized in a language, that is, there are a great number of words and idiomatic expressions in our language whose interpretations depend upon those conceptual metaphors. But there are no words or idiomatic expressions in our language whose meanings depend upon a conceptual connection between death and a banana (Lakoff and Turner 1989:51).

The issue at hand here is the automatic use of LJT's basic conceptual metaphors and the conceptual embodiment which they structure, and to compare and contrast this with Vaihinger's ideas about the organic function of thought. The following passage by Vaihinger is indicative of a point which will reoccur in our discussion, namely, the relatively free nature of the use of fictions in conceptual formation, since the only constraint which is to be observed is that of "useful organic elaboration of the material of sensation." If death-as-a-banana were to serve this end, that would be all well and good. Degree of linguistic conventionalization, which is extremely important for LJT's theory of basic conceptual metaphors, is largely irrelevant to theoretical (and practical!) employment of fictions. Vaihinger writes:

The psyche then is an organic formative force, which independently changes what has been appropriated, and can adapt foreign elements to its own requirements as easily as it adapts itself to what is new. The mind is not merely appropriative, it is also assimilative and constructive. In the course of its growth, it creates its organs of its own accord in virtue of its adaptable constitution, but only when stimulated from without, and adapts them to external circumstances. Such organs, created by the psyche for itself in response to external stimuli, are, for example, forms of perception and thought, and certain concepts and other logical constructs. Logical thought, with which we are especially concerned here, is an active appropriation of the outer world, a useful organic elaboration of the material of sensation. Logical thought is therefore an organic function of the psyche (Vaihinger 1935:2).

The references to logic in the above may seem to be leading us astray from the metaphor theme of LJT, but shortly, Vaihinger comes to the distinction which was labelled (2) a few pages back; namely, the distinction between rules and artifices of thought:

We make a distinction between rules and artifices of thought. In other functions also this distinction is of value; the rules are the totality of all those technical operations in virtue of which an acitvity is able to attain its object directly, even when more or less complicated. In logic too we call such operations, and in particular those of induction, "rules of thinking". The artifices, on the other hand, are those operations, of an almost mysterious character, which run counter to ordinary procedure in a more or less paradoxical way. They are methods which give an onlooker the impression of magic if he be not himself initiated or equally skilled in the mechanism, and are able indirectly to overcome the difficulties which the material in question opposes to the activity. Thought also has such artifices; they are strikingly purposive expressions of the organic function of thought (Vaihinger 1935:11) (9).

It might seem tempting to draw a direct comparison between the rules to which Vaihinger refers and to whatever cognitive operations we employ in understanding a concept on its own terms and to continue by comparing artifices with the use of metaphor in understanding a concept more fully. This is, however, misleading, since the case which I am making here is that both rules and artifices - for us and for Vaihinger, the latter are most interestingly represented by fictions - are on the level at which a concept is understood on its own terms. The fact that fictions might have the aura of metaphoricity about them, once again, is due to the contrafacticity inherent in them. We will look more closely at this with regard to our discussion of (4) below.

As stated in (3) above, the artifice which Vaihinger wishes to examine further is that of fiction:

We are therefore dealing with a peculiar kind of logical product, a special manifestation of the logical function. We have already seen that this peculiar activity is expressed in what we call artifices, that its products are artificial concepts. We would here, anticipating the outcome, substitute other terms for these expressions: our subject is the fictive activity of the logical function; the products of this activity - fictions.

The fictive activity of the mind is an expression of the fundamental psychical forces; fictions are mental structures. The psyche weaves this aid to thought out of itself; for the mind is inventive; under the compulsion of necessity, stimulated by the outer world, it discovers the store of contrivances that lie hidden within itself. . . .

Meanwhile, in the interests of greater clearness an intelligibility we may premise the following remark:

By fictive activity in logical thought is to be understood the production and use of logical methods, which, with the help of accessory concepts - where the improbability of any corresponding objective is fairly obvious - seek to attain the objects of thought. Instead of remaining content with the material given, the logical function introduces these hybrid and ambiguous thought-structures, in order with their help to attain its purpose indirectly, if the material which it encounters resists a direct procedure. With an instinctive, almost cunning ingenuity, the logical function succeeds in overcoming these difficulties with the aid of its accessory structures. The special methods, the by-paths, of which thought makes use when it can no longer advance directly along the main road, are of many different kinds, and their explanation is our problem. They often lead through thorny undergrowth, but logical thought is not deterred thereby, even though it may lose something of its clearness and purity. It is relevant also to remark here that the logical function, in its purposeful instinctive ingenuity, can carry this fictive activity from the most innocent and unpretentious beginnings on through ever finer and subtler developments right up to the most difficult and complicated methods (Vaihinger 1935:13).

This idea that the fictive activity can be carried from "the most innocent and unpretentious beginnings on through ever finer and subtler developments right up to the most difficult and complicated methods" (10)

echoes some of the considerations to which Lakoff and Turner give voice regarding the extensions enabled by "the mode of metaphorical thought that poets use and invoke in their readers" and which carries us "beyond ordinary metaphoric thought" (Lakoff and Turner 1989:71). Note, however, that in Lakoff and Johnson, these extensions are, once again, on the level of metaphoric rather than non-metaphoric understanding of a concept, whereas my point is that fictions are on the level of our nonmetaphorical understanding of at least some concepts, among which music is one.

With regard to (4) in the above, Vaihinger's comments are important in recognizing the sort of contradictory nature which he means that fictions should have. Needless to say, this is useful from the point of view of our analysis, since, as the following passage suggests, fictions are tailormade objects for analysis with the help of implausible counterfactual conditionals:

Before entering on our task it is necessary to make a distinction that will subsequently assume considerable importance. Ideational constructs are in the strict sense of the term real fictions when they are not only in contradiction with reality but self-contradictory in themselves; the concept of the atom, for example, or the "Ding an sich." To be distinguished from these are constructs which only contradict reality as given, or deviate from it, but are not in themselves self-contradictory (e.g. artificial classes). The latter might be called half-fictions or semi-fictions. These types are not sharply divided from one another but are connected by transitions. Thought begins with slight initial deviations from reality (half-fictions), and, becoming bolder and bolder, ends by operation with constructs that are not only opposed to the facts but are self-contradictory (Vaihinger 1935:16).

It is here we can begin to see the sort of role which fictions can play above and beyond the sort of role which metaphor plays in LJT, whether it be of the basic conceptual or the poetic kind: they create in a more radical fashion.

This brings us back to the whole discussion which was started with regard to (C) above: in LJT, as in many other theories of metaphor, the "creating" done by the metaphor is of the nature of the structuring or restructuring of one conceptual realm in terms of the structuring found in another. In LJT, the objects of thought, to use Vaihinger's term, are at hand. The metaphors can help to flesh out our intuitions about them, make them more three-dimensional, and, in general, serve to bring them into focus. This is where fictions go one step beyond - they serve to create objects of thought, which then serve as linchpins in the creation of our world view:

Logical processes are a part of the cosmic process and have as their more immediate object the preservation and enrichment of the life of organisms: they should serve as instruments for enabling them to attain a more complete life; they serve as intermediaries between living beings. The world of ideas is an edifice well calculated to fulfil this purpose; but to regard it for this reason as a copy is to indulge in a hasty and unjustifiable comparison. Not even elementary sensations are copies of reality; they are rather mere guages for measuring the changes in reality (Vaihinger 1935:16).

Lakoff and Turner do, indeed, point out that: "To understand what is metaphorical, we must begin with what is not metaphorical. In brief, to the extent that a concept is understood and structured on its own terms - without making use of structure imported from a completely different conceptual domain - we will say that it is not metaphorical" (Lakoff and Turner 1989:57). They present both death and light as examples of concepts that are, at least in part, understood directly. Regarding death, they remark: ". . . death is in part understood directly as well: when one is alive, one is functioning; when one is dead, one is not functioning. This is a nonmetaphorical aspect of our understanding of death. As such. it can be used as the source domain for other metaphors" (Lakoff and Turner 1989:58). In the case of light: "We perceive light, react to it emotionally, and know that it allows us to see things. . . . The life as light metaphor depends . . . on certain nonmetaphorical knowledge about light: that it promotes growth, that it makes us happy for the most part, that it allows us to see and gain the knowledge necessary for our survival, and so on" (Lakoff and Turner 1989:58). With respect to light, they do also comment that our scientific understanding of light employs "two common scientific metaphors for light: as waves and as particles that move faster than anything else in the universe." Thus, the common sense understanding of light is direct, whereas the scientific one employs metaphor.

This last point is a crucial one. Note that the scientific understanding of light is an understanding of something which was understood and structured on its own terms in a direct, non-metaphorical way. The cardinal point which I am making with respect to our concept of music is that, on the most direct understanding of what it is, an intentional component is required, a component which builds in, at the most basic level, a relationship between hearer and heard. My suggestion for this relationship is counterfactual hearing. On this proposal, counterfactual hearing becomes part and parcel of the understanding and structuring of music on its own terms, of our non-metaphorical understanding of it. Of course there are acoustic, rhythmic and dynamic properties which are essential to music, but they are also essential to a host of other sonic phenomena as well, and I have argued in detail elsewhere that they are not enough to capture the direct, non-metaphoric understanding of music. So, how to articulate this intentional component? The explanation which we will advocate is that something related to metaphor is being done in (C) in that (C) is an implausible counterfactual which looks very much like the implausible counterfactuals which underlie metaphors. Note, however, that what (C) is doing is to create a fiction, that of sound-food. It cannot be proven that music is sound-food; recall (5) above - fictions are not hypotheses. One can, however, provide reasons for the employment of one fiction rather than another, to explain why one fiction is truer as a criterion for music than another. Here we may borrow something related to the notion of grounding which Tormey discusses as part of his proposal for the recasting of metaphors in terms of implausible counterfactuals. The notion of the grounding for such a counterfactual is a presentation of what it is that sustains the counterfactual. What we need in the case of the creation of a fiction, such as sound-food in (C) we shall call fictional grounding, because what we are interested in is what justifies the entertainment of (C). Can such a justification, for example, sustain the definition-like statement

(M)u is music iff (C).

This essay will conclude with a suggestion for a fictional grounding for (C), and an associated discussion of (M)'s suitability, which, given that we are using it to define when a sound sequence is or is not music for some given group of agents along with the background assumptions which they share, is tantamount to a definition of music relativized to that group and those assumptions. (11)

We have noted that, since fictions are not hypotheses, they cannot be subject to proof. Another interesting feature of fictions which Vaihinger points out is the type of provisional nature which they have and the manner in which they may be discarded:

The function of an hypothesis is, of course, only provisional - but the goal which it has ultimately in view is to be theoretically tested and established by the facts of experience. The hypothesis has also to be discarded, but this is because the hypothetical idea has become fully qualified for admission into the circle of what is accepted as real. The provisional object of the fiction is quite different; for the fiction, in so far as we have termed it a provisional auxilliary construct, ought to drop out in the course of time and make way for its real function; but in so far as it is a pure fiction, it ought, at any rate logically, to disappear as soon as it has done its duty (Vaihinger 1935:87).

I think there are two ways in which this comment applies to the suggestion that counterfactual hearing and its attendent use of fictions, such as sound-food, be incorporated as an inextricable, intentional component in our characterization of music at the non-metaphorical level. One is one which does not have anything much to do with its nature as a fiction nor with the deeper meaning of the remark, the other one does. The first one is that the 'sound-food' need not be present as a conscious criterion for the presence of music everytime on listens to a piece of music. The concept of music is one that we will borrow and learn how to use from the linguistic group of which we are a member, without subjecting ourselves to the sound-food test every time. This is no more or less remarkable than the fact that, although "being able to support some agent's weight" is more or less uncontroversially a basic feature of chairs, we do not present ourselves with this everytime we identify something as a chair.

The second point is more interesting. In the sense that 'sound-food' is being proposed in order to capture some inextricably intentional feature of music, its presence at the level of non-metaphorical understanding of the concept of music can provide a location in our theory for addressing the perception of a paradoxical sort of permanent novelty in works of music, at least in good ones. If, indeed, a case can be made for the pretense of a constituting component in our concept of music which is of a nature such as the "unreachable" concept of sound-food - please see the discussion of Levin and Hausman which immediately follows - there will be a sort of built-in openendedness in music.

We have taken pains to distinguish between the role played by fictions in the process of concept construction and the role played by metaphors in concept enrichment or exploration. Two metaphor theorists who contribute to the conflation of these two endeavors are Samuel Levin and Carl Hausman. It is of interest to examine their work in this context, since I believe that the drawing of the fiction-metaphor distinction will prove to be useful in reconfiguring some of their theories, which upon reconfiguration, have some things to say which are directly relevant to the themes of this essay.


Samuel Levin has the following to say about Lakoff and Johnson:

Among the examples that they adduce are "wasting time," "attacking positions," "going our separate ways." These expressions, they say are 'reflections of systematic metaphorical concepts that structure our actions and thoughts. They are 'alive' in the most fundamental sense: they are metaphors we live by. The fact that they are conventionally fixed within the lexicon of English makes them no less alive'([Lakoff and Johnson] 1980:55).

This last statement has about it an air of paradox, if not indeed inconsistency. One normally assumes that to the extent items are conventionally fixed within the lexicon their meanings are normalized, and thus rendered stable. . .

These conclusions highlight the strongly conceptual nature of Lakoff and Johnson's theory of metaphor. For them the "vitality" of a linguistic expression is not determined by the status of its elements in the lexicon and the role played by those elements in grammatical arrangements; it is determined, rather, by the role those elements play in our conceptual system and by the significance of their function in the conduct of and talk about our daily lives (Levin 1993:120-121).

Levin contrasts this approach with his own view of literary metaphor. Literary metaphor is, of course, explicitly expressed, and not, like the above metaphors which we presumably are living by, buried among the conceptual foundations of our world view. As such, literary metaphors are invitations to contemplate states of affairs at deviant worlds by means of the production of a "progressively lapsing impression" which he terms a conception. In contrasting acts of imagination with acts of conception, Levin remarks: ". . . even for 'farfetched' imaginative acts the presentation can take the form of a clear image, whereas in acts of conception (in my terms) this does not occur. Entertained instead is a progressively lapsing impression which the conceiver, bent on its realization in consciousness, impels and urges to attainment. Although attainment is not reached, that is, no definite presentation is achieved, there stands behind and activates the process what may be regarded as a schema of the representation" (Levin 1988:53). Levin succinctly contrasts his view with that of Lakoff and Johnson by remarking: "A series of fundamental differences between our respective theories derives from the fact that whereas the conceptual metaphors of Lakoff and Johnson express concepts, mine express conceptions" (Levin 1988:6).

It might, at first glance, seem as if Lakoff and Johnson have already anticipated Levin's comment in the following passage from More Than Cool Reason: A Field Guide to Poetic Metaphor: "Poets can appeal to the ordinary metaphors we live by in order to take us beyond them, in order to make us more insightful than we would be if we thought only in standard ways. Because they lead us to new ways of conceiving of [my italics, CMG] our world, poets are artists of the mind" (Lakoff and Turner 1989:215). Conceiving of is precisely the phrase which Levin reserves for the cognitive stance one takes toward conceptions, which are of such a character that the mental space which is prepared by them is of such a nature that it cannot be filled, but the best one can hope for is a "progressively lapsing impression" of what would be the case at the deviant world which must be construed in order to accomodate the state of affairs indicated by metaphors which express conceptions. For Levin, it would be overkill to conceive of anything which properly belongs to our world, rather than to a world which is so deviant as to only allow of cognitive approximation. In this context, Levin employs ". . . 'concept' to designate the mental image produced by a linguistic expression and 'conception' to designate a mental function performed preparatory to the formation of such an image" (Levin 1988:70). Since Levin allows that "The terms concept and conception (like the terms meaning, sense, reality, and so forth) figure at several theoretical intersections . . . " (Levin 1988:70-71), he also accepts that concepts may be spoken of "as the constituent senses of a dictionary entry (in this concurring with Katz) and conceptions as the set of general beliefs held by individuals in respect to an object." He describes the latter characterizations as being "drawn against a background that was primarily linguistic" (Levin 1988:72), whereas he sees the former characterization as a phenomenological approach "where concepts and conceptions are considered not as static counters in some theoretical framework but aspects of dynamic and individuated mental functions" (Levin 1988:72). The following passage is helpful here:

Thus, to continue with our own line of investigation and try to draw more sharply the phenomenological difference between concepts and conceptions, we shall say that to conceive something, say x, is to have a clear and distinct image or idea of that x, whereas to conceive of x is as it were to prepare a mental space into which that x might have been placed. If x is a flying horse then to conceive it means to have before one a clear and distinct image of a flying horse; to conceive of it means to allow for the possibility that such an image might be produced. In the same way, to conceive a golden mountain means to have before one a clear and distinct image of a golden mountain; to conceive of a golden mountain means to allow for the possibility that such an image might be produced. In both our examples, where x is either a flying horse or a golden mountain, the mental space prepared by the concept can in fact be filled - that is, both objects can be conceived. A flying horse and a golden mountain can be conceived because, even though neither exists, the elements out of which they are composed are physical characteristics, and those elements can be combined, both in a conception and as the components of a concept.

In the two examples thus far considered, the distinction between conceiving and conceiving of (and the correlative one between a concept and a conception) has had no operative role to play; the objects in both of our examples (and the types that they represent) can both be conceived of and conceived. The utility of the distinction emerges, however, when we consider examples in whose composition there figures an attribute of affection or emotion. Earlier we asked whether we can conceive a tree as sad or the sea as laughing. I would now say that we can conceive of such 'objects' but we cannot conceive them. In terms of our distinction we can focus on an area in our minds such that it delimits the space into which the concept of a sad tree would fit, but we are unable to fill the space with a concept. In the process of focusing on that area, however, we project a schema, an abstract model or framework which, given the purpose of our exercise, we take to be an implicit or potential representation of that 'object'. This schema or model conveys the sense in which I am using the notion of conception (Levin 1988:69-70).

It should be clear by now that the use of conceived of in the quote from More Than Cool Reason is not carried out in the same spirit as that preferred by Levin, since poets are credited with leading us to "new ways of conceiving of our world," and that which is properly conceived of, according to Levin, belongs to a deviant world. There is no proper distinction between conceiving and conceiving of in the Lakoff and Johnson program, at least as it is laid out in Metaphors We Live By; Women, Fire and Dangerous Things; and More Than Cool Reason. In More Than Cool Reason, Lakoff and Johnson do submit that:

[Poetic metaphors] allow the use of ordinary conceptual resources in extraordinary ways. It is by these means that poets lead us beyond the bounds of ordinary modes of thought and guide us beyond the automatic and unconscious everyday use of metaphor. What makes poetic metaphor noticeable and memorable is thus the special, nonautomatic use to which ordinary, automatic modes of thought are put (Lakoff and Johnson 1989:72).

It might seem as if we should somehow relate Levin's concept/conception distinction to Lakoff and Johnson's automatic-mode-of-thought/nonautomatic-use-of-automatic-mode-of- thought distinction. I think this is questionable for two reasons: First, it is part and parcel of Lakoff and Johnson's program to make poetic metaphor a sort of self-conscious variant of their metaphors we live by, which, it is fair to say, do, indeed express concepts, as concepts are delineated by Levin. Secondly, and very importantly, another reason one feels reluctant to draw the concept-conception distinction in the Lakoff-Johnson context, is that, for all their merits, in none of the three books referred to here - Metaphors We Live By; Women, Fire and Dangerous Things; and More Than Cool Reason - do Lakoff and Johnson define what a concept is! Some important points are, to be sure, made: We are told, for example, in More Than Cool Reason that "Concepts are cognitive in nature; that is, they are part of human cognition. On this view, it is contents, not words or phrases, that have meaning" (Lakoff and Turner 1989:111). Also, a myriad of combination terms in which the word "conceptual" appears are defined, and the word is subject to constant use. Neither "concept" nor "conceptual" however, are ever singled out and defined. The attitude seems to be that there is enough intuitive understanding of what a concept, so that corrective comments, such as the previous one about being part of human cognition, are sufficient for circumscribing what is meant.

All of this leads me to suspect that, even after More Than Cool Reason, the aims of the Lakoff-Johnson and Levin programs are still sufficiently different as to matter. Perhaps one could choose to regard "Music is food" as a poetic metaphor á la Lakoff and Johnson and then proceed to see which basis conceptual metaphors (those ones we live by) serve to support it. Returning also to my remarks at the beginning of the article, interesting things would certainly turn up on a study of metaphors used to talk about music, or even as the result of an empirical search for the "Music is food" metaphor in our linguistic habits. I would, however, like to place myself between the Lakoff-Johnson and Levin chairs by advancing the view that Levin's outlook on metaphor can be combined with what he calls Lakoff and Johnson's metaphysical argument, so that, indeed, metaphors which express conceptions in his sense can also accorded a role in which they too "define an outlook on 'truth' and reality" (Levin 1979:120). The means of doing this, I submit, is by shifting the discussion so as to deal with fictions and not metaphors: I believe that Levin's conceptions - "progressively lapsing impressions" - are often best understood as fictions. We will take a closer look at this after examining Hausman's position on creative metaphor.


Carl Hausman is a philosopher whose approach to metaphor is largely directed by concerns connected with the philosophy of creativity. Metaphors are often viewed as suffusing our discourse with novelty and as being creative in some sense. Hausman observes ". . . metaphors are sometimes creative and at the same time appropriate or adequate to something in 'the world' " (Hausman 1983:181). He views creative metaphors ". . . as naming or reference-fixing expressions that give birth to the referents they fix" (Hausman 1983:181). In honing in on what he means by the "referent" of a creative metaphor, Hausman applies a narrow use of "reference" to creative metaphors as follows: "A metaphorical expression functions so as that it is creative of its significance, thus providing new insight, through designating a unique, extra-linguistic and extra-conceptual referent that had no place in the intelligible world before the metaphor was articulated. There are three key terms in the proposal, 'creative,' 'unique,' and 'extra-linguistic' or 'extra-conceptual' (I use the last two expressions interchangeably here). These key terms indicate that in being creative, a metaphor must meet two considerations: uniqueness and extra-linguisticality" (Hausman 1983:186). The criterion of uniqueness is mandatory in order to insure that the referent of a creative metaphor is new. Hausman feels that extralinguisticality is necessary ". . . to justify saying that a creative metaphor is 'appropriate,' 'faithful,' or 'fits the world.' " (Hausman 1983:186). Hausman observes that in this sense, such metaphors are "true" (Hausman puts the word in quotation marks) in a manner similar to the manner in which literal expressions are true, but that the combination of the uniqueness and extra-linguisticality conditions are what is special in the metaphor case: "There is something to which the expression is appropriate, some resistant or restraining condition; yet this condition is new" (Hausman 1983:186). Using our old friend, "Juliet is the sun" as an example, Hausman illustrates his view as follows:

What then is the referent or object that uniquely satisfies the sentence as it is understood within its poetic context? It should be clear that this question cannot be answered by offering a description of the referent, unless the description is the metaphor itself. For if the referent is new, a description would have only previously established linquistic means, that is, antecedent senses, to rely on. At best, then we might say that the new referent is Juliet-the-sun. As with metaphors that establish themselves in language or theory, insofar as the metaphor makes a lasting impact on our understanding of human beings and nature, the names 'Juliet' or 'the sun' or both might come to be used for this new referent without recurrent recognition that it first occurred as an expression with a metaphorical function (Hausman 1983:189-190).

Hausman understands Juliet-the-sun as an individual. In discussing the ontology which results from this, Hausman, while recognizing the problematic nature of some of Peirce's view of individuality, and "although proposing something Peirce might not have intended" (Hausman 1983:192) chooses to

. . . pursue his suggestion that individuals are real in a way other than as spatially determinate entities. An individual is something singular in experience that reacts - that acts to some extent against our will, that acts so that just any interpretation of it is not necessarily faithful or fitting. Thus, as said earlier, an individual manifests its reality as a confluence of insistent contraints, as a unique focal point of resistance in experience. And as new, referents of creative metaphors are centers or cores of possibilities. They are focal points of relevance for qualities (e.g., Juliet-the-sun), and in turn newly created senses that would give absolute determinateness to the focal center if they were fully actualized (Hausman 1983:192).

Hausman sums up by saying:

The implied ontology is a realism, patterned in part on suggestions by C.S. Peirce, that leaves room for unique individuals that are centers or foci of relevance. These resist and constrain the appropriateness of qualities and properties to one another, as these are apprehended in terms of senses. Individuals sustain clusters of properties and, in turn, of senses. Thus, the significance of creative metaphors is not reducible to one or more systems of senses. And their significance is not limited to linguistic or conceptual conventions. They can be creative in interacting with an independent reality which they create - with individuals that constitute a dynamic evolving world (Hausman 1983:194) (12).

Levin, Hausman and fictionalism

Staying with "Juliet is the sun" as our example, I think it is now possible to say the following with respect to Levin's and Hausman's approaches: For Levin Juliet-the-sun represents a conception, a schema which initiates a cognitive process which might be characterized as striving for convergence at a sort of asymptote - Juliet-the-sun - which it never can reach, since Juliet-the-sun resides at a world which is very deviant with respect to our own. The ontological status of Juliet-the-sun is not at issue for Levin, other than as a schema. Hausman, on the other hand, posits the existence of this Juliet-the-sun asymptote, but as a Peirce-inspired dynamic individual. I think that Juliet-the-sun is most interestingly understood as a sort of Vaihingerian fiction, which, in the manner discussed in the above presentation of Vaihinger's view, helps us to indirectly reach an object of thought by means of the artifice of fiction. Even though Vaihinger is concerned with scientific theorizing in the following passage, I think that the definition of the purpose of thought supports this interpretation of Levin and Hausman and nicely captures some of their salient intentions with their respective theories:

If sensations are the starting-point of all logical activity and at the same time the terminus to which they must run, if only to render control possible (and as we remarked above, it must remain undecided whether we must regard the logical functions between these two points as having some inherent purpose), then the purpose of thought may be defined as the elaboration and adjustment of the material of sensation for the attainment of a richer and fuller sensational life of experience (Vaihinger 1935:6).

Vaihinger sees fictions as important artifices for the "elaboration and adjustment of the material of sensation for the attainment of a richer and fuller sensational life of experience." As was pointed out in the presentation of Vaihinger's thought earlier on, Vaihinger sees fictions (1) as mental structures, a view which connects with Levin's schema notions; (2) as organic creations of an organic psyche, a view which coalesces with the LJT view of meaning as embodied; and (3) as tools for indirectly obtaining objects of thought, a view which takes into account much of the matter which concerns Hausman, but with, it seems to me, fewer and less problematic ontological commitments.

Note that the preceding passage from Vaihinger provides a sort of criterion for deciding whether or not (C) is worth adopting: the extent to which it allows us to elaborate and adjust "the material of sensation for the attainment of a richer and fuller sensational life of experience." Relativized to the discussion at hand, three questions are relevant: (1) Does (C), indeed, provide us with an object of thought which aids us in the elaboration and adjustment of the material of audio sensation, so as to attain a richer and fuller sensational life of experience?; and (2) Does the sonic material so chosen coincide intuitively with what we usually regard as music? (This is essentially the question of whether or not (M) is a good definition of music.); and (3) Does (C) help us to reorient and direct our thinking about music in such a way as to understand this concept better?

Notice that each of the above questions can be addressed independently of the other two. It is possible that (1) receives a positive answer, while (2) and (3) receive negative ones. It may be so that the counterfactual hearing suggested by (1) strikes at least some readers as an interesting way to try to stucture audio experiences "for the heck of it," but that no connection to music is perceived. It may also be the case that it is acknowledged that (C) does, indeed, pick out music in a manner which coincides with our intuitions about what music is, but does it in such an ad hoc, irrelevant way that it does not teach us anything about what music really is (13) and neither does it boost our sensational experience. Lastly, it may do something to direct our thinking about what music is so as to give us some insights into the concept which are neither sufficient for defining it, nor of such a nature as to boost the quality of our audio experience to yield richer or fuller experiences of this kind. (I think that it is more difficult to maintain the independence of (3) from both (2) and (1) than the independence of (1) from (2) and (3), or the independence of (2) from (1) and (3), but it is, nevertheless, plausible.)

Grounding , fictions and sound-food

Although I think it is arguable that a positive answer to any one of (1), (2) or (3) would justify entertainment of (C), it is my belief that (1), (2) and (3) all are answered in the positive. We shall now embark upon the enterprise of looking for reasons for entertaining (C). Note that this is a project quite different from that of going on a hunt for basic conceptual metaphors which support (C). This is rather an enterprise which is closely related to what Alan Tormey calls the supportive grounding for a metaphor, when the metaphor has been recast as an implausible counterfactual. Tormey remarks that "what distinguishes metaphors from other subspecies of counterfactual will be found largely in the difference in their grounding" (Tormey 1983:242). He points out that

Thoroughly plausible counterfactuals - for example, "If this piece of platinum had been heated, it would have expanded" - are reasonably supposed to be sustained or supported by relevant causal laws or nomological relations. Metaphors, as implausible counterfactuals, can expect no support from this quarter. . . . Lawlike relations are usually irrelevant to the probity or aptness of metaphor - namely, providing a "paraphrase" - is actually a procedure for grounding or sustaining the metaphor (Tormey 1983:242).

Tormey submits that paraphrases do not translate metaphors any more than the enumeration of inductive evidence translates the causal laws which are sustained by it:

We can understand a metaphor without knowing what sustains it, just as we can grasp the meaning of the conditional statement 'If you drink the water in Naples, you'll become ill' without knowing its grounds (namely, that the water contains, among other things, high levels of sodium trioxide).

What have been commonly mislabeled as paraphrases of metaphor are not crude and misguided efforts at translation, but serious attempts to support or sustain the metaphor by citing, inter alia, shared properties and analogical resemblances. We have been victimized by a sort of logical dislocation in our readines to promote the evidence or grounds for the aptness of metaphors to the position of a semantic surrogate for the metaphor. But metaphors do not mean whatever it is that sustains them. (My bolding, CMG.) 'Man is a wolf' does not translate: 'Man is voracious, cunning, crafty, or cruel' (and so on), though these supposedly shared properties may be good reason for advancing the metaphor in the first place. Thus, on the view that metaphors are elliptical counterfactuals, there is no danger of commiting the dread 'heresy of paraphrase,' for there is no paraphrase. To the extent that it is amenable to analysis, the meaning of a metaphor may be brought out by recasting it in counterfactual form, and purported paraphrases may be relocated where they belong, providing grounds for the metaphor and supplying arguments for its appropriateness (Tormey 1983:242-243) (14).

Earlier we pointed out that in order to distinguish the sort of grounding given for metaphors which have been recast as counterfactuals from the grounding provided for fictions understood as counterfactual constructs, we would refer to the latter procedure as fictional grounding. We return to (C) and the matter of whether or not (1) - (3) receive positive answers. Recall that (1) - (3) were as follows:

(1)Does (C), indeed, provide us with an object of thought which aids us in the elaboration and adjustment of the material of audio sensation, so as to attain a richer and fuller sensational life of experience?

(2)Does the sonic material so chosen coincide intuitively with what we usually regard as music? (This is essentially the question of whether or not (M) is a good definition of music.)

(3)Does (C) help us to reorient and direct our thinking about music in such a way as to understand this concept better?

The fictional grounding for (C)

Let us now look at at least some of the points which should be included in a fictional grounding for (C):

  • The suggestion that the sound-food fiction yields a characterization of music accounts for universality which is always rearing its head among our intuitions about music, while allowing for enormous variation among practitioners with respect to both taste and practice.
  • It allows for the "inheritance" by music of a great many qualities which we take for granted in the food model, but which seem to be highly problematic on other theories of music and its ontology:
    1. its naturalness;
    2. the nature of our need for it, which often seems to be appropriately conceptualized in terms of hunger and satiation;
    3. the feeling that some is more natural than others within and across style types;
    4. its importance and immediacy;
    5. freedom to describe it in terms which we see fit. The conflicts about what is the proper way to talk about music assume an appropriately silly character when we reflect on the enormous freedom allowed in the verbal expression of gustatory experience. To be sure, both areas are possessed of a professional, technical vocabulary, but no one would dream of discrediting a person's ability to appreciate a gustatory experience, simply because he or she could not describe it in technical terms or rewrite the recipe.
  • We are given an intuitive model for distinguishing music from "sonic art," which can be theoretically and practical useful. Traditionally, we have had recourse to the distinctions broadly provided by speech, music and noise, where music has been a sort of catch-all for sonic phenomena which were somehow created and somehow structured, but were neither speech nor noise. On the suggestion presented in this paper, interesting formal and creative things may be done with sounds, so that they may comprise a form of sonic art, but if the proper sort of counterfactual hearing is not involved, what is at issue is not music, simply by the default of being neither speech nor noise.
  • "Sound-feeling" would be the fiction of choice for many, based on the work, for example, of Susanne Langer, and the nature of Western musical aesthetics ever since the Baroque period's fascination with music and rhetoric. Some afterthought, however, reveals that the sound-food fiction provides a more direct model of an experience of something external which is internalized. "Direct" is a key word here: one of the problematic qualities of feelings is that they themselves are most often about or caused by something else. There is not space to pursue this here, but it could be an interesting avenue of inquiry to examine to what extent the reformulation of our fixation with sound and feeling in terms of counterfactually hearing music as "sound-feeling" could account for the persistent notion that music is about something, rather than tending to see this as a sign of music's inherently linguistic nature.
  • The old and tired debates about scores get put into new perspective. With some luck, the notion of score-as-recipe could have a healthy effect on the preoccupation with scores as identifiers of musical works or as their essential component.
  • The serious entertainment of the sound-food fiction highlights the intentionality of both food and music and can reorient our attitudes toward the intentions of composers and artists in a more fruitful direction. Not all stuctured non-speech and non-noise is music in a culture, and neither is all prepared nourishment food there.
  • We are provided with a possible candidate for a generalized theory of music for sentient, intelligent beings. The discussion of what music might be for aliens and androids takes on an added dimension, as long as they have to engage in activities which we can understand as akin to eating and as long as they have organs which can process soundwaves (15). Note that since the creation of a fiction is involved, this approach posits language as prior to music.
  • Serious eating (tasting) provides a useful model for serious listening.
  • The aesthetic of disinterestedness, which has come down to us from the eighteenth century, is currently being questioned. For example, Arnold Berleant writes: "Disinterestedness is historically important for having helped us recognize the distinctiveness of aesthetic experience, but it is misleading in claiming its separateness, not just from the other areas of experience, but from the very person of the observer" (Berleant 1992:158). When the sound-food fiction is taken seriously, two important things occur: distinctiveness is assured, due to the inherent counterfactuality of the enterprise, yet disinterestedness is out of place, except as a pathological condition, since disinterest as an attitude toward food is usually a sign of ill health.


In a recent article, Peter Kivy laments the "crypto-linguistic" model of musical understanding as it has been applied ad nauseum to absolute music:

But the question we should be asking is not: "How does absolute music, appearances to the contrary, manage to represent?" That is the evasion. Rather, we should be asking: "What is it about us, and about our world, that has made the pure contentless art of musical design so important for our lives in the past one-hundred-and-fifty years, but not before? What needs of ours does it serve? And why have other peoples not felt these needs, or we until so recently in our history? What distinguishes sonic from visual design, that seems to make the former play so different a role in our lives from the latter? What does it all mean - but not in the semantic sense?..."

I think this is a great challenge, and will spare us another dreary round of philosophical books and papers trying to understand absolute music as a representational or linguistic art. The books and papers get more and more clever, more and more philosophically sophisticated, and, for me at least, less and less plausible. The lady doth protest too much, methinks. I believe that, until we meet this challenge head on, and try philosophically to understand pure music on its own terms - which is to say as a pure decorative art - we will remain with an important and, to date, mysterious aspect of our artistic natures unexplained and, worse still, unexamined (Kivy 1991:553-554).

The sound-food fiction can help to solve many of the problems posed in this passage. The way in which this fiction puts distance between itself and any form of crypto-linguistic model for musical understanding is beautifully underscored in the following remarks by Susanne Langer:

Another recommendation for words is that they have no value except as symbols (or signs); in themselves they are completely trivial. This is a greater advantage than philosophers of language generally realize. A symbol which interests us also as an object is distracting. It does not convey its meaning without obstruction. For instance, if the word "plenty" were replaced by a succulent, ripe, real peach, few people could attend entirely to the mere concept of quite enough when confronted with such a symbol. The more barren and indifferent the symbol, the greater is its semantic power. Peaches are too good to act as words; we are too much interested in peaches themselves. But little noises are ideal conveyors of concepts, for they give us nothing but their meaning. That is the source of the "transparency" of language, on which several scholars have remarked. Vocables in themselves are so worthless that we cease to be aware of their physical presence at all, and become conscious only of their connotations, denotations, or other meanings. Our conceptual activity seems to flow through them, rather than merely accompany them, as it accompanies other experiences that we endow with significance. They fail to impress us as "experiences" in their own right, unless we have difficulty in using them as words, as we do with a foreign language or a technical jargon until we have mastered it.

But the greatest virtue of verbal symbols is, probably, their tremendous readiness to enter into combinations. There is practically no limit to the selections and arrangements we can make of them. This is largely due to the economy Lord Russell remarked, the speed with which each word is produced and presented and finished, making way for another word. This makes it possible for us to grasp whole groups of meanings at a time, and make a new, total, complex concept out of the separate combinations of rapidly passing words (Langer 1957:75-76).

It is interesting note that Suzanne Langer, a well-known proponent of the music and emotions linkup - she regarded music as a "graph of the emotions" - unwittingly supports the notion of food as an ontogenetic fiction for music which will free music from being misunderstood as a literary art. Note how well what she says in the above serves to underscore the kind of attention that would need to be paid to sound in order to hear it as if it, were, say, a ripe, succulent peach, and how different this attention is to that which is paid to an essentially linguistic construct. One point on which she is mistaken, however, is her low estimation of the way in which musical sounds could combine in order to function as language. I have given an example, in the guise of a thought experiment, of a culture which uses what we call music exclusively for the encoding of and communication of propositional thought, i.e. as language. That this example may seem strange to people is, I suggest, due to considerations along Langerian lines, not to any a priori impossibility or even implausibility. (16)

* * * * *

The suggestion that sound-food should act as the fiction which appropriately intentionalizes our concept of music is, to be sure, a controversial one, and it has been the modest aim of this article to try to point out that there are enough reasons for integrating this approach to intentional concepts within the philosophical discussion of metaphor and of aesthetics in general. In this case, the proof of the pudding is in the hearing. As with any meal, this discussion will now come to an end, since there is no point in serving more once the level of satiation has been reached, regardless of the number of dishes remaining, so further discussion will be saved for future occasions.


I wish to express my gratitude to Veikko Rantala and Finn Collin for valuable comments and criticism with regard to the present paper and to Anders Engstrøm for the same with regard to an earlier version.


Berleant, Arnold (1992) The Aesthetics of Environment. Philadelphia: Temple University Press.

Currie, Gregory (1993) 'Aliens Too'. In: Analysis, Vol. 53 (1993):116-118.

Grund, Cynthia M. (forthcoming) 'From Speech Act to Music Act: Some Thoughts on Intentionality and Music.' Contribution to the proceedings of the Fourth International Congress on Musical Signification, Université de Paris I - Pantheon-Sorbonne, Institut d'Esthetique et de Sciences de l'Art & Institut Finlandais en France, 9-13 October 1994.

Grund, Cynthia M. (1996c) 'Kierkegaard: Metaphor and the Musical Erotic'. In: Danish Yearbook of Philosophy, Vol. 31:65-88.

Grund, Cynthia M. (1996b) 'Fictionalism: A Neglected Context for Studies in Musical Signification'. In: Contemporary Music Review 1996, Vol 16, Part 7:119-128.

Grund, Cynthia M. (1996a) 'Jeremy Bentham's Theory of Fictions: Some Reflections on Its Implications for Musical Semiosis and Ontology'. In: Musical Semiotics in Growth. Edited by Eero Tarasti. Bloomington and Indianapolis: Indiana University Press: 55-71.

Grund, Cynthia M. (1995) 'How Philosophical Characterizations of a Musical Work Lose Sight of the Music and How It Might Be Put Back'. In: Musical Signification: Essays in the Semiotic Theory and Analysis of Music. Edited by Eero Tarasti. Berlin/New York: Mouton de Gruyter: 63-79.

Grund, Cynthia M. (1988) 'Metaphors, Counterfactuals and Music'. In: Essays on the Philosophy of Music. Acta Philosophica Fennica, Vol. 43. Editors: Veikko Rantala, Lewis Rowell and Eero Tarasti. Helsinki: The Philosophical Society of Finland: 28-53.

Hausman, Carl (1983) 'Metaphors, Referents and Individuality'. In: Journal of Aesthetics and Art Criticism, Vol. XLII, No. 2, Winter 1983:181-195.

Hausman, Carl (1989) Metaphor and Art: Interactionism and Reference in the Verbal and Nonverbal Arts. Cambridge: Cambridge University Press.

Kivy, Peter (1991) 'Is Music an Art'. In: The Journal of Philosophy, Volume LXXXVIII, Number 10, October 1991:544-554.

Lakoff, George (1987) Women, Fire, and Dangerous Things. Chicago: The University of Chicago Press.

Lakoff, George & Mark Johnson (1980) Metaphors We Live By. Chicago: The University of Chicago Press.

Lakoff, George & Mark Turner (1989) More Than Cool Reason. Chicago: The University of Chicago Press.

Langer, Susanne (1957) Philosophy in a New Key: A Study in the Symbolism of Reason, Rite and Art. 3rd ed. Cambridge, Massachusetts: Harvard University Press.

Levin, Samuel (1979) 'Standard Approaches of Metaphor and a Proposal for Literary Metaphor' in Metaphor and Thought, edited by Andrew Ortony. Cambridge, UK: Cambridge University Press: 124-135.

Levin, Samuel (1988) Metaphoric Worlds: Conceptions of a Romantic Nature. New Haven: Yale University Press.

Levin, Samuel (1993) 'Language, concepts, and worlds: Three domains of metaphor' In: Metaphor and Thought, 2nd ed. Edited by Andrew Ortony. Cambridge, UK: Cambridge University Press: 112-123.

Richards, I.A. (1936) The Philosophy of Rhetoric. London: Oxford University Press.

Stecker, Robert (1996) 'Alien Objections to Historical Definitions of Art'. In: British Journal of Aesthetics, Vol. 36, No. 3, July 1966:305-308.

Tormey, Alan (1983) 'Metaphors and Counterfactuals'. In: Essays on Aesthetics: Perspectives on the Work of Monroe C. Beardsley. Edited by John Fisher. Philadelphia: Temple University Press: 235-246.

Vaihinger, Hans (1935) The Philosophy of 'As if .' Trans. by C.K. Ogden. 2nd ed. London: Routledge and Kegan Paul Ltd.

1. Cynthia M. Grund, FD, (b. 1956) is an ekstern lektor (adjunct associate professor) in philosophy in the Department of Education, Philosophy and Rhetoric at the University of Copenhagen and a forskningsadjunkt (researcher) in the same department on a three-year NOS-H (Joint Committee of the Nordic Research Councils for the Humanities) grant. This article is the last chapter in her doctoral dissertation in philosophy for Tampere University, Tampere, Finland, entitled Constitutive Counterfactuality: The Logic of Interpretation in Metaphor and Music. Tilbage

2. Of late, much stress has been put upon the extent to which linguistic meaning is embodied. By this is meant that the ways in which we form concepts and then interpret the world conceptually is a direct function of the fact that the "I" which is involved is an "I" which is in a human body, and which is obliged to interact, in all its characteristic physicality, with the world. Lakoff provides the following definitions:

Conceptual embodiment
The idea that the properties of certain categories are a consequence of the nature of human biological capacities and of the experience of functioning in a physical and social environment. It is contrasted with the idea that concepts exist independently of the bodily nature of any thinking beings and independent of their experience.

Functional embodiment
The idea that certain concepts are not merely understood intellectually; rather, they are used automatically, unconciously, and without noticeable effort as part of normal functioning. Concepts used in this way have a different, and more important, psychological status than those that are only thought about consciously (Lakoff 1987:12-13).

In addition, the connection between counterfactual conditionals like (C) and metaphors is something about which I have a good deal to say elsewhere - see Grund 1988 and Grund 1995. Tilbage

3. Within musical semiotics, related problems arise regarding the nature of the musical sign. See Grund 1996b. Tilbage

4. See Grund 1988 and Grund 1995. Tilbage

5. Actually, the counterfactual analysis addresses many of the issues raised by Lakoff and Johnson in More than Cool Reason. See, for example, p. 5 ff. in More than Cool Reason: "If a lifetime is a day, then the setting sun is old age." Tilbage

6. For more background on both Vaihinger's and Bentham's treatments of fictionalism, see Grund 1996b and Grund 1996a. Tilbage

7. See Grund 1996b:7-8. Tilbage

8. See, for example, Lakoff 1987:269 ff. Tilbage

9. Of course, the fact that this discussion is couched in terms of logic is only gratifying for this writer, since the treatments which I have suggested for analyzing metaphors do, indeed, rely on a form of intensional, possible-worlds semantics. Tilbage

10. This is also foreshadowed in Bentham's writings. See Grund 1996a.

Given that the present paper is concerned with both comparing and contrasting the ways in which metaphors and fictions function, the following remarks by I.A. Richards are of interest:

Throughout the history of Rhetoric, metaphor has been treated as a sort of happy extra trick with words, an opportunity to exploit the accidents of their versatility, something in place occasionally but requiring unusual skill and caution. In brief, a grace or ornament or added power of language, not its constitutive form. Sometimes, it is true, a writer will venture on speculations that go deeper. I have just been echoing Shelley's observation that "Language is vitally metaphorical; that it, it marks the before unapprehended relations of things and perpetuates their apprehension, until words, which represent them, become, through time, signs for portions or classes of thought instead of pictures of integral thoughts: and then, if no new poets should arise to create afresh the associations which have thus been disorganised, language will be dead to all the nobler purposes of human intercourse." But that is an exceptional utterance and its implications have not been taken account of by rhetoricians. Nor have philosophers, as a body, done much better, though historians of language have long taught that we can find no word or description for any of the intellectual operations which, if its history is known, is not seen to have been taken, by metaphor, from a description of some physical happening. Only Jeremy Bentham, as successor to Bacon and Hobbes, insisted - with his technique of archetypation and phraseoplerosis - upon one inference that might be drawn; namely, that the mind and all its doings are fictions. He left it to Coleridge, F.H. Bradley and Vaihinger to point to the further inference; namely, that matter and its adventures, and all the derivative objects of contemplation, are fictions too, of varied rank because of varied service (Richards 1936:90-91).

Although I sense that this quote displays the sort of tendency towards the conflation of fictions and metaphors against which I warn in the present paper, it reveals interesting insights into the importance of paying attention to the work of Bentham and Vaihinger when dealing with the constitutive aspects of language. Tilbage

11. An important aspect of Tormey's treatment is that truth and grounding are in many ways independent of one another. I think that this is a good thing. Whereas Tormey is very pessimistic about providing truth conditions for the sorts of implausible counterfactuals which underlie metaphors on his view, however, I propose truth conditions in terms of an intensional, possible worlds semantics in Grund 1988. A provocative upshot of all of this that if we regard the acceptance on the part of an agent of the counterfactual underlying a given metaphor as a belief, allow that that counterfactual may be true or false, and moreover regard its grounding as justification for the belief, we have that a metaphor can, indeed, represent justified, true belief, and thus, at least in the Socratic tradition, can be said to represent knowledge which the agent possesses. Tilbage

12. Hausman (1983) has been the source for quotations in the present essay, since the presentation there is succinctly relevant to the matters being dealt with here. Readers who would like more insight into the broader context of Hausman's program are referred to Hausman (1989). Tilbage

13. Sometimes such procedures are, in fact, valuable. During the seminar Metaforer i kultur og samfund, held October 24-25 in Copenhagen and sponsored by the Danish Network for Metaphor, Culture and Cognition, a presentation was given by Carsten Hansen and Peter Widell in which they presented a fascinating ad hoc way of identifying metaphors within open, machine-readable text corpora, which in essence exploited only syntactic and frequential properties of words in the texts. The frequency of correct "hits" when identifying metaphors was astoundingly high, even though it was clear nothing about what metaphors really are was involved. Tilbage

14. The Grounding Hypothesis in More Than Cool Reason must from the outset be distinguished from what Alan Tormey means when he writes of the grounding for a metaphor, which has been recast as a counterfactual conditional. Lakoff and Turner advocate the Grounding Hypothesis as a correct alternative to what they dub the Literal Meaning Theory, which they believe "not only is incorrect but also leads to many other fallacies" (Lakoff and Turner 1989:114). They sum up the Literal Meaning Theory as follows:

    The Literal Meaning Theory

  • If an expression of a language is (1) conventional and ordinary, then it is also (2) semantically autonomous and (3) capable of making reference to objective reality.
  • Such a linguistic expression is called "literal."
  • No metaphors are literal (Lakoff and Johnson 1989:114-115).

Now, for Lakoff and Johnson, the Grounding Hypothesis

addresses the question of how metaphorical understanding is possible at all. Generally, it states that metaphorical understanding is grounded in nonmetaphorical understanding. But, because of the complexity of metaphorical understanding, it must be stated more precisely than that.

    The Grounding Hypothesis

  • Many conventional concepts are semantically autonomous or have aspects that are semantically autonomous.
  • Semantically autonomous concepts (or aspects of concepts) are  grounded in the habitual or routine bodily and social patterns we  experience, and in what we learn of the experience of others.
  • Semantically autonomous concepts (or aspects of concepts) are not  mind-free. They are not somehow given to us directly by the objective world. They are instead grounded in the patterns of  experience that we routinely live.
  • The source domain of the metaphor is characterized in terms of  concepts (or aspects of concepts) that are semantically autonomous.
  • In this sense, metaphorical understanding is grounded in semantically autonomous conceptual structure (Lakoff and Johnson 1989:113).

Lakoff and Johnson define semantic autonomy as follows: "An expression in a language is semantically autonomous if it is meaningful completely on its own terms. It follows that any expression that is semantically autonomous does not derive any of its meaning from metaphor. Nor does it derive its meaning through other conceptual relationships that stand outside of classical logic, such as metonymy, irony, conversational principles, and so on" (Lakoff and Johnson 1982:111). Tilbage

15. Some thoughts about aliens and music appear in Grund 1995, but it is interesting to note that they have also made their entrance into the forum of debate over the definition of works of art in Currie 1993 and Stecker 1996. Tilbage

16. See Grund 1995. Tilbage