27
How Metaphor Reveals the Limitations of the Myth of Objectivism
The heart of the objectivist tradition in philosophy comes directly out of the myth of objectivism: the world is made up of distinct objects, with inherent properties and fixed relations among them at any instant. We argue, on the basis of linguistic evidence (especially metaphor), that the objectivist philosophy fails to account for the way we understand our experience, our thoughts, and our language. An adequate account, we argue, requires
---viewing objects only as entities relative to our interactions with the world and our projections on it
---viewing properties as interactional rather than inherent
---viewing categories as experiential gestalts defined via prototype instead of viewing them as rigidly fixed and defined via set theory
We view issues having to do with meaning in natural language and with the way people understand both their language and their experiences as empirical issues rather than matters of a priori philosophical assumptions and argumentation. We have selected metaphor and the way we understand it from among the possible domains of evidence that could bear on these issues. We have focused on metaphor for the following four reasons:
In the objectivist tradition, metaphor is of marginal interest at best, and it is excluded altogether from the study of semantics (objective meaning). It is seen as only marginally relevant to an account of truth.
Yet we have found that metaphor is pervasive, not
p.211
merely in our language but in our conceptual system. It seems inconceivable to us that any phenomenon so fundamental to our conceptual system could not be central to an account of truth and meaning.
We observed that metaphor is one of the most basic mechanisms we have for understanding our experience. This did not jibe with the objectivist view that metaphor is of only peripheral interest in an account of meaning and truth and that it plays at best a marginal role in understanding.
We found that metaphor could create new meaning, create similarities, and thereby define a new reality. Such a view has no place in the standard objectivist picture of the world.
Many of the facts that we have discussed have long been known in the objectivist tradition, but they have been given an entirely different interpretation from ours.
The conventional metaphorical concepts we take as structuring our everyday conceptual system are taken by the objectivists to be nonexistent. Metaphors, for them, are matters of mere language; there are no such things as metaphorical concepts.
Words and expressions that we have taken as instances of metaphorical concepts (e.g., digest in "I can't digest all those facts") would be taken by objectivists as not being instances of live metaphor at all. For them the word digest would have two different and distinct literal (objective) meaningsdigest i for food and digest! for ideas. On this account, there would be two words digest which are homonyms, like the two words bank (bank of a river and bank where you put your money).
An objectivist might grant that digest an idea was once a metaphor, but he would claim that it is no longer metaphorical. For him it is a "dead metaphor," one that
p.212
has become conventionalized and has its own literal meaning. This is to say that there are two homonymous words digest.
The objectivist would probably grant that digest i and digest 2 have similar meanings and that the similarity is the basis for the original metaphor. This, he would say, explains why the same word is used to express two different meanings; it was once a metaphor, it became a conventionalized part of the language; it died and became frozen, taking its old metaphorical meaning as a new literal meaning.
The objectivist would observe that the similarities upon which the dead metaphor was based can in many cases still be perceived today.
According to the objectivist account of metaphor, the original metaphor was a matter of use and speaker's meaning, not literal objective meaning. It would have to have arisen by the general speaker's meaning formula applied to this case (where digest referred only to food):
In uttering a sentence S (S = "I couldn't digest his ideas") with literal objective meaning M (M = I couldn't transform his ideas, by chemical and muscular action in the alimentary canal, into a form my body could absorb), the speaker intends to convey to the hearer the speaker's meaning M' (M' = I couldn't transform his ideas, by mental action, into a form my mind could absorb).
Two things have to be true in order for this objectivist account to hold. First, the intended speaker's meaning M', referring to ideas, must be an objectively given meaning, having objective truth conditions. In other words, the following must be objectively true of the mind and ideas by virtue of their inherent properties:
Ideas must, by virtue of their inherent properties, be the kind of thing that can have a form, be transformed, and be absorbed into the mind.
The mind must, by virtue of its inherent properties, be the kind
p.213
of thing that can perform mental actions, transform ideas, and absorb them into itself.
Second, the metaphor must have been originally based on preexisting similarities between M and M'. That is, the mind and the alimentary canal must have inherent properties in common, just as ideas and food must have inherent properties in common.
To summarize: the dead-metaphor account of digest would claim the following:
The word digest originally referred to a food concept.
By a "live" metaphor, the word digest was transferred to a preexisting objective meaning in the realm of ideas, on the basis of preexisting objective similarities between food and ideas.
Eventually the metaphor "died," and the metaphorical use of digest an idea became conventional. Digest thus obtained a second literal objective meaning, the one occurring in M'. This is seen, on the objectivist account, as a typical way of providing words for preexisting meanings that lack words to express them. All such cases would be considered homonyms.
In general, an objectivist would have to treat all of our conventional-metaphor data according to either the homonymy position (typically the weak version) or the abstraction position. Both of these positions depend on the existence of preexisting similarities based on inherent properties.
As we have just seen, the objectivist account of conventional metaphor requires either an abstraction view or a homonymy view. Moreover, the objectivist account of both conventional and nonconventional metaphor is based on preexisting inherent similarities. We have already presented detailed arguments against all of these positions. These arguments take on a special importance now. They show not only that the objectivist view of metaphor is in- p.214
adequate hut that the entire ohjectivist program is based on erroneous assumptions. To see just where the objectivist account of metaphor is inadequate, let us recall the relevant parts of our arguments against the abstraction, homonymy, and similarity views as they pertain to the objectivist account of conventional metaphor.
We saw in our discussion of the ideas are food metaphor that, although the metaphor was based on similarities, the similarities themselves were not inherent but were based on other metaphorsin particular, the mind is a container, ideas are objects, and the conduit metaphors. The view that ideas are objects is a projection of entity status upon mental phenomena via an ontological metaphor. The view that the mind is a container is a projection of entity status with in-out orientation onto our cognitive faculty. These are not inherent objective properties of ideas and of the mind. They are interactional properties, and they reflect the way in which we conceive of mental phenomena by virtue of metaphor.
The same holds in the case of our concepts time and love. We understand sentences like "The time for action has arrived" and "We need to budget our time" in terms of the TIME IS A MOVING OBJECT and TIME IS MONEY metaphors, respectively. But on the objectivist account there would be no such metaphors. Arrive and budget in these sentences would be dead metaphors, that is, homonyms, deriving historically from once-live metaphors. These once-live metaphors would have to have been based on inherent similarities between time and moving objects, on the one hand, and time and money, on the other. But, as we have seen, such similarities are not inherent; they are themselves created via ontological metaphors.
It is even more difficult to make a case for an inherent-similarity analysis for expressions involving the concept love, such as "This relationship isn't going anywhere,"
p.215
"There was a magnetism between us," and "This relationship is dying." The concept love is not inherently well defined. Our culture gives us conventional ways of viewing love experiences via conventional metaphors, such as LOVE IS A JOURNEY, LOVE IS A PHYSICAL FORCE, etc., and our language reflects these. But according to the objectivist account (based either on dead metaphor, weak homonymy, or abstraction), the concept love must be sufficiently well defined in terms of inherent properties to bear inherent similarities to journeys, electromagnetic and gravitational phenomena, sick people, etc. Here the objectivist must not only bear the burden of claiming that love has inherent properties similar to the inherent properties of journeys, electromagnetic phenomena, and sick people; he must also claim that love is sufficiently clearly defined in terms of these inherent properties so that those similarities will exist.
In summary, the usual objectivist accounts of these phenomena (dead metaphor, homonymy with similarities, or abstraction) all depend on preexisting similarities based on inherent properties. In general, similarities do exist, but they cannot be based on inherent properties. The similarities arise as a result of conceptual metaphors and thus must be considered similarities of interactional, rather than inherent, properties. But the admission of interactional properties is inconsistent with the basic premise of objectivist philosophy. It amounts to giving up the myth of objectivism.
The Objectivist Default: "It's Not Our Job"
The only remaining alternative for the objectivist is to give up any account of any relationship between the food and idea senses ofdigest in terms of similarity (including denial that there was ever a metaphor at all) and to turn to the strong homonymy position. According to this view there is one word digest with two entirely different and unrelated meanings---as different as the two meanings of punt (a kick
p.216
in football and an open, flat-bottomed boat with square ends). As we have seen (in chapter 18), the strong homonymy position cannot account for:
Internal systematicity
External systematicity
Extensions of the used portion of the metaphor
The use of concrete experience to structure abstract experience
The similarities that we, in fact, see between the two senses of digest, based on metaphorically conceptualizing ideas in terms of food.
Of course, an objectivist philosopher or linguist could grant that he cannot adequately account for such systematicities, similarities, and ways of understanding the less concrete in terms of the more concrete. This might not disturb him in the slightest. After all, he could claim, accounting for such things is not his job. Such things are matters for the psychologist, the neurophysiologist, the philologist, or someone else. This would be in the tradition of Frege's separating off "sense" from "ideas" and Lewis's separating off "abstract semantic systems" from "psychological and sociological facts." The homonymy view, they could claim, is adequate for their proper purposes as objectivists, namely, to provide objective truth conditions for linguistic expressions and to give an account of literal objective meaning in terms of them. This, they assume, could be done independently for the two senses of digest without having to account for systematicity, similarity, understanding, etc. Relative to this conception of their job, conventional metaphorical uses of digest involve merely homonyms and not metaphors at all, dead or alive. The only metaphors they recognize are nonconventional metaphors (e.g., "Your ideas are made of cheap stucco" or "Love is a collaborative work of art"). Since these, they would claim, are matters of speaker's meaning, not the literal objective meaning of a sentence, issues of truth and meaning arising from them are to be handled by the account of speaker's meaning given above.
p.217
In summary, the only internally consistent objectivist view of conventional metaphor would be that the issues we have been primarily concerned with---the properties of conventional metaphors and the way we use them in understanding---are simply outside their purview. They would insist that they are not responsible for such matters and that no facts of this sort concerning conventional metaphor could possibly have any bearing on the objectivist program or on what they, as objectivists, believe.
Such objectivists might even grant that our investigations of metaphor correctly show that interactional properties and experiential gestalts are, in fact, necessary to account for how human beings understand their experience via metaphor. But even granting this, they could still continue to ignore everything we have done on the following grounds: they could say simply that experientialists are merely concerned with how human beings happen to understand reality, given all of their limitations, whereas the objectivist is concerned not with how people understand something as being true but rather with what it means for something to actually be true.
This objectivist response perfectly highlights the fundamental difference between objectivism and experientialism. Such an objectivist reply boils down to a reaffirmation of their fundamental concern with "absolute truth" and "objective meaning," entirely independent of anything having to do with human functioning or understanding. Against this, we have been maintaining that there is no reason to believe that there is any absolute truth or objective meaning. Instead, we maintain that it is possible to give an account of truth and meaning only relative to the way people function in the world and understand it. We are simply in a different philosophical universe from such objectivists.
The Irrelevance of Objectivist Philosophy to Human Concerns
We are in the same philosophical universe as, and have real disagreements with, those objectivists who think that there
p.218
can be an adequate objectivist account of human understanding, of our conceptual system and our natural language. We have argued in detail that conventional metaphor is pervasive in human language and the human conceptual system and that it is a primary vehicle for understanding. We have argued that an adequate account of understanding requires interactional properties and experiential gestalts. Since all objectivist accounts require inherent properties and most of them require a set-theoretical account of categorization, they fail to give an adequate account of how human beings conceptualize the world.
Objectivist Models Outside of Objectivist Philosophy
Classical mathematics comprises an objectivist universe. It has entities that are clearly distinguished from one another, e.g., numbers. Mathematical entities have inherent properties, e.g., three is odd. And there are fixed relationships among those entities, e.g., nine is the square of three. Mathematical logic was developed as part of the enterprise of providing foundations for classical mathematics. Formal semantics also developed out of that enterprise. The models used in formal semantics are examples of what we will call "objectivist models"models appropriate to universes of discourse where there are distinct entities which have inherent properties and where there are fixed relationships among the entities.
But the real world is not an objectivist universe, especially those aspects of the real world having to do with human beings: human experience, human institutions, human language, the human conceptual system. What it means to be a hard-core objectivist is to claim that there is an objectivist model that fits the world as it really is. We have just argued that objectivist philosophy is empirically incorrect in that it makes false predictions about language, truth, understanding, and the human conceptual system. On the basis of this we have claimed that objectivist philosophy provides an inadequate basis for the human sciences.
p.219
Nonetheless, a lot of remarkably insightful mathematicians, logicians, linguists, psychologists, and computer scientists have designed objectivist models for use in the human sciences. Are we claiming that all of this work is worthless and that objectivist models have no place at all in the human sciences?
We are claiming no such thing. We believe that objectivist models as mathematical entities do not necessarily have to be tied to objectivist philosophy. One can believe that objectivist models can have a functioneven an important function---in the human sciences without adopting the objectivist premise that there is an objectivist model that completely and accurately fits the world as it really is. But if we reject this premise, what role is left for objectivist models?
Before we can answer this question, we need to look at some of the properties of ontological and structural metaphors:
Ontological metaphors are among the most basic devices we have for comprehending our experience. Each structural metaphor has a consistent set of ontological metaphors as sub-parts. To use a set of ontological metaphors to comprehend a given situation is to impose an entity structure upon that situation. For example, love is a journey imposes on love an entity structure including a beginning, a destination, a path, the distance you are along the path, and so on.
Each individual structural metaphor is internally consistent and imposes a consistent structure on the concept it structures. For example, the argument is war metaphor imposes an internally consistent war structure on the concept argument. When we understand love only in terms of the love is a journey metaphor, we are imposing an internally consistent journey structure on the concept love.
Although different metaphors for the same concept are not in general consistent with each other, it is possible to find sets of metaphors that are consistent with each other. Let us call these consistent sets of metaphors.
Because each individual metaphor is internally consistent, each
p.220
consistent set of metaphors allows us to comprehend a situation in terms of a well-defined entity structure with consistent relations between the entities.
The way that a consistent set of metaphors imposes an entity structure with a set of relations between the entities can be represented by an objectivist model. In the model, the entities are those imposed by the ontological metaphors, and the relations between the entities are those given by the internal structures of the structural metaphors.
To summarize: Trying to structure a situation in terms of such a consistent set of metaphors is in part like trying to structure that situation in terms of an objectivist model. What is left out are the experiential bases of the metaphors and what the metaphors hide.
The natural question to ask, then, is whether people actually think and act in terms of consistent sets of metaphors. A special case where they do is in the formulation of scientific theories, say, in biology, psychology, or linguistics. Formal scientific theories are attempts to consistently extend a set of ontological and structural metaphors. But in addition to scientific theorizing, we feel that people do try to think and act in terms of consistent sets of metaphors in a wide variety of situations. These are cases where people might be viewed as trying to apply objectivist models to their experience.
There is an excellent reason for people to try to view a life situation in terms of an objectivist model, that is, in terms of a consistent set of metaphors. The reason is, simply, that if we can do this, we can draw inferences about the situation that will not conflict with one another. That is, we will be able to infer nonconflicting expectations and suggestions for behavior. And it is comfortingextremely comfortingto have a consistent view of the world, a clear set of expectations and no conflicts about what you should do. Objectivist models have a real appealand for the most human of reasons.
We do not wish to belittle this appeal. It is the same as the
p.221
appeal of finding coherence in your life or in some range of life experiences. Having a basis for expectation and action is important for survival. But it is one thing to impose a single objectivist model in some restricted situations and to function in terms of that model---perhaps successfully; it is another to conclude that the model is an accurate reflection of reality. There is a good reason why our conceptual systems have inconsistent metaphors for a single concept. The reason is that there is no one metaphor that will do. Each one gives a certain comprehension of one aspect of the concept and hides others. To operate only in terms of a consistent set of metaphors is to hide many aspects of reality. Successful functioning in our daily lives seems to require a constant shifting of metaphors. The use of many metaphors that are inconsistent with one another seems necessary for us if we are to comprehend the details of our daily existence.
One obvious utility for the study of formal objectivist models in the human sciences is that they can allow us to understand, in part, the ability to reason and function in terms of a consistent set of metaphors. This is a common activity and an important one to understand. It can also allow us to see what can be wrong with imposing a requirement of consistency---to see that any consistent set of metaphors will most likely hide indefinitely many aspects of reality---aspects that can be highlighted only by other metaphors that are inconsistent with it.
One obvious limitation of formal models is that, so far as we can imagine, they provide no means for including the experiential basis for a metaphor and therefore provide no way of accounting for the way in which metaphorical concepts permit us to comprehend our experience. There is a corollary of this that has to do with the issue of whether a computer could ever understand things the way people do. The answer we give is nosimply because understanding requires experience, and computers don't have bodies and don't have human experiences.
P.222
However, the study of computational models might nevertheless tell us a great deal about human intellectual capacities, especially in the areas where people reason and function partly in terms of objectivist models. Moreover, current formal techniques in computer science show promise of providing representations of inconsistent sets of metaphors. This could conceivably lead to insights about the way that people reason and function in terms of coherent, but inconsistent, metaphorical concepts. The limits of formal study seem to be in the area of the experiential bases of our conceptual system.
Summary
Our general conclusion is that the objectivist program is unable to give a satisfactory account of human understanding and of any issues requiring such an account. Among these issues are:
---the human conceptual system and the nature of human rationality
---human language and communication
---the human sciences, especially psychology, anthropology, sociology, and linguistics
---moral and aesthetic value
---scientific understanding, via the human conceptual system
---any way in which the foundations of mathematics have a basis in human understanding
The basic elements of an experientialist account of understanding---interactional properties, experiential gestalts, and metaphorical concepts---seem to be necessary for any adequate treatment of these human issues.