The Metaphor as a Mediator Between Semantic and Analogic Modes of Thought [and

Comments and Reply]

Author(s): Brenda E. F. Beck, Mary Douglas, M. S. Edmonson, James W. Fernandez, István
Fodor, Ivan Fónagy, Yalçin İzbul, Stephen C. Levinson, Franklin Loveland, William C.
McCormack, Robert A. Randall, J. David Sapir and J. Christopher Crocker
Source: Current Anthropology, Vol. 19, No. 1 (Mar., 1978), pp. 83-97
Published by: The University of Chicago Press on behalf of Wenner-Gren Foundation for
Anthropological Research
Stable URL: http://www.jstor.org/stable/2741153
Accessed: 03-04-2016 14:25 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Wenner-Gren Foundation for Anthropological Research, The University of Chicago

Press are collaborating with JSTOR to digitize, preserve and extend access to Current Anthropology

This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
 CURRENT ANTHROPOLOGY Vol. 19, No. 1, March 1978
? 1978 by The Wenner-Gren Foundation for Anthropological Research 0011-3204/78/1901-0004$01.75
The Metaphor as a Mediator Between
Semantic and Analogic Modes of Thought
by Brenda E. F. Beck
THE USE OF METAPHOR is a fundamental aspect of the human
communication process. More importantly, metaphors play a
vital role in the fine tooling of the concepts that govern our
communal life. Metaphors can be used both to shore up weak-
ening cultural constructs and to change them. As a result, these
seeming asides or adjuncts to communication actually provide
a fundamental data set for anthropological analysis. I shall
attempt to show how one can use these colourful additions to a
given communication code as keys to help unlock the secrets of
human culture-building processes.
Having claimed so much for the metaphor, I must now at-
tempt to define it and to describe the processes involved in its
formation. According to Ogden and Richards (1960), a meta-
phor is a primitive abstraction. It involves referring to a set of
concrete relationships in one situation for the purpose of
facilitating the recognition of an analogous set of relations in
another situation. Metaphorical language thus borrows a
part of one context, by identifying its abstract form, and
applies it to another group of things that is less easily under-
stood (p. 213). Metaphors can thus "provide . . . new sudden
and striking collocations of references for the sake of the
compound effects of contrast, conflict, harmony . . . or [be]
used more simply to modify and adjust emotional tone . ..
through association" (p. 240). The essential nature of the
BRENDA E. F. BECK is Associate Professor of Anthropology at the
University of British Columbia (Vancouver, B.C., Canada V6T
1W5). Born in 1940, she was educated at the University of
Chicago (B.A., 1962) and Oxford University (B. Litt., 1964;
D. Phil., 1968). She was Assistant Professor at the University of
Chicago in 1968-69. Her research interests are social organization,
regional cultures in relation to national traditions, folklore,
ceremonial, and the use of metaphor in building a social identity.
Among her publications are Peasant Society in Kohku: A Study of
Right and Left Subcastes in South India (Vancouver: University
of British Columbia Press, 1972); "Centers and Boundaries of
Regional Caste Systems: Towards a General Model," in Regional
Analysis, edited by Carol Smith, pp. 359-99 (New York: Aca-
demic Press, 1976); "The Symbolic Merger of Body, Space and
Cosmos in Hindu Tamilnad" (Contributions to Indian Sociology
10:213-43); and "The Concept of Kingship in a Contemporary
South Indian Epic," in Kingship and Authority in South Asia,
edited by John Richards (Madison: University of Wisconsin
South Asia Center, in press). She is at work on a detailed analysis
of an oral epic of contemporary South India with special attention
to the use of pan-Indian imagery and metaphoric formulae to
bind together local history and national traditions.
The present paper, submitted in final form 10 VI 77, was sent
for comment to 50 scholars. The responses are printed below and
are followed by a reply by the author.
Vol. 19 * No.
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
metaphor, then, is that it juxtaposes elements of a concrete
image in order to formulate some set of more abstract relation-
ships. A metaphorical comparison is always limited to certain
relevant parallels, the irrelevant features of the two sets being
understood to cancel one another out (p. 214).
Metaphors need not always have a verbal form; in ritual, for
example, they may be expressed in actions alone. Indeed,
anything built or designed by man can have metaphoric as well
as literal qualities. Thus a fence, for example, can serve the
direct function of bounding something or of keeping the un-
wanted out. At the same time, a fence can have a metaphoric
function, such as providing a concrete statement about the
owner's general economic status and/or social attitudes.
While it is easy to identify some metaphors, others lie hidden
in the communication process. Furthermore, the boundary
between literal and metaphorical usage is often difficult, even
impossible, to draw. We often do not know the extent to which
a statement is consciously intended to be an alternative or
indirect expression (as opposed to a direct one). These problems
should not deter us, however. As in so many issues of definition
and identification, boundary discriminations (drawing the line
between what is and what is not a metaphor) are less important
than the need to identify typical or clear examples of the
phenomena under study. Many processes and products studied
by "scientists" cannot be strictly delimited, and need not be so
to be understood.
One final point concerning definitions is worthy of discussion
before we proceed. Metaphors are often used to describe
ultimate truths as well as mundane ones. Indeed, a metaphor
in a religious ritual can be used to describe the nature of a
cosmic force. Such metaphors lead us towards an understanding
of a particular world view. In ancient Greece, for example,
philosophers described reason as "the king of heaven." Earlier
in the same cultural tradition, we find repeated references to
an oath or contract thought to govern strife between opposing
anthropomorphized forces (Lloyd 1966:219-20).
Plato was the first to contrast the use of images or myths with
nonfigurative reasoning. Though exhibiting a high regard for
the latter technique of discourse, he clearly thought the former
essential to the description of the very highest truths (p. 300).
The situation is similar today. Our best physicists and astron-
omers continue to give metaphorical descriptions of puzzling
phenomena. "Black holes" in the outer universe and the
"charmed" behaviour of quarks are two well-known examples.
Such terms may come to have very technical, literal meanings
with the accumulation of scientific experience, but they still
show that modern researchers, like ancient ones, have difficulty
1 * March 1978 83
in conceiving of basic principles without recourse to figurative
aids.
Metaphors are thus a part of our scientific discourse, just as
they are a part of religious belief. Furthermore, the metaphor
is not a random analogy. Its imagery builds on clusters or sets
of underlying associations which anthropologists can and should
closely examine. Nonetheless, we must not give too much
importance to any single metaphor in our delineation of a
cultural world view. Such sets of images can be cumulative,
but, at the same time, strings of metaphors are often comple-
mentary in their qualities. The use of a range of metaphors,
then, should be examined, not just for common elements, but
also for genuine diversity. Anthropologists are gradually learn-
ing that a world view is seldom a unified conceptual order
(Barth 1975). The very diversity of metaphors in use in a given
culture points to precisely such in-built flexibility. The study of
metaphor, then, can help us to define the subtle quality of a
culture's penultimate concepts more directly and concretely
than has hitherto been possible.
The central reason that the study of metaphor promises to
provide new insights is the fact that metaphors mediate between
our abstract and our more concrete thoughts. A metaphor
points to the existence of a given set of abstract relationships
hidden within some immediately graspable image. By doing
this, it helps to ground our conceptual structures in the reality
of concrete experience. Anthropologists, like philosophers, have
long debated the status of the "world out there." Does it
ultimately determine our mental state, or do preexistent
structures in the mind determine the nature of what we come
to "know" through our senses? I need not detail that debate
here. It will be well known to readers already. My point is that
focusing on the metaphor can help us to see our way around (or
perhaps through) this false dilemma.
We know that every social group gradually develops a
"culture" or a shared set of interrelated categories and con-
cepts. This culture clearly affects both the group's common
experiences and the actions of individuals within it. At the
same time, however, there is always a reality "out there" to
cope with. The mental set shared by group members must
continually adapt to changing ecological, economic, and social
conditions. Gradually the common code is forced to adjust to a
new environment. The metaphor is one of the simplest and
most important mechanisms by which such a shared mental
framework can be kept in touch with what lies "out there."
By studying the use of metaphors and their shifting content,
we can thus monitor that very central process of adjustment
which any cultural code must undergo when faced with a
changing environment. Through recording metaphors in the
context of their use, we can literally come to "see" that process
at the very moment of its occurrence.
Fernandez (1974) has pointed to some basic processes that
characterize metaphoric thought:' (1) a movement from an
abstract concept to a concrete image, (2) a movement that
entails a reference to affect and/or to perceptual experience,
(3) a movement that bridges logical gaps, (4) a movement that
relates a part to some larger whole, (5) a movement that helps
to map out a nonverbal phenomenon or behaviour. Fernandez
(1975) further describes the metaphor as a process or turning
"from one domain to another" in order to accomplish "a
creative transcendence." He suggests that such "shiftiness" is
elemental to human thought processes and may be bound up
with the very nature of human language. I shall now pursue
this idea about metaphoric "movement" in an attempt to
clarify why it is that a metaphor accomplishes a shift or a
transcendence.
Considerable progress has recently been made in our efforts
to understand what metaphors do and what kinds of logical
1 Fernandez speaks of "seven missions of metaphor." I am respon-
sible for this reduction (by recombination) to five.
84 CURRENT ANTHROPOLOGY
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
processes they depend upon. Many writers, for example, have
discussed the operation of metaphor in the construction of
riddles and (to a lesser extent) proverbs (Arewa and Dundes
1964; Kongas Maranda 1971a, b; Milner 1969). It has also been
noticed that metaphor is central to some aspects of drama and
to ritual activity (Barth 1975; Blacking 1961; Peacock 1968;
Turner 1974). The fact upon which many of these workers
stumble, however, is that the verbal metaphor draws on a
universe of semantic structures. Metaphors, when they occur
in verbal form, must be word-based images. Does this mean that
they are governed by our everyday system of semantic cate-
gories? I think not.
Part of the basis for my belief is found in the study of syn-
aesthesia, perception by a sense other than the one directly
stimulated. In linguistic terms, this means the description of
something experienced by one sense organ by adjectives whose
primary referent is another, as in saying that a color is "warm"
or a noise is "sharp." Ullmann (1957) finds a hierarchy in such
"transfers," remarking that, of the six senses, sight and touch
are most frequently used as metaphors for the other sense
domains. Pesot (1975) builds on Ullmann's observations but
objects, with Kronasser (1952), to the idea that any system of
truly uni-sensoral experience exists. Both argue that the
phenomenon of synaesthesia rests, not on "transfers" (and
hence on the logic of metaphor), but rather on a protosynthesis
of sensations at a more primary level of motor and emotional
consciousness. Thus adjectives like "sharp" may display a set
of polysemantic referents at the verbal level, but this seeming
multidimensionality rests on a more profound unisensate base.
Work with language learning in young children (Piaget and
Inhelder 1967) and in chimpanzees (Davenport 1977, Savage
and Rumbaugh 1977) seems to confirm the existence of a
relatively undifferentiated prelinguistic pool of motor and
emotive experience on which semantic codes are later built.
Such research leads us to think of conscious sense perceptions
as resting on a mesh of associations at the motor and emotional
level that builds up with experience over time. This system of
sensory logic remains highly individualized throughout life in
comparison with semantic codes. Nonetheless, each culture
seems to direct its participants' focus towards certain specific
highly valued, sense-based configurations. A given individual's
sensory network, then, will have certain specific, culturally
induced "nodes" at which given sets of associated qualities
will be especially powerful or salient. Such nodes have a certain
universal form in that they seem to cluster around body-based
imagery. Indeed, the cross-cultural regularity with which body
metaphors are found central (Douglas 1973, Needham 1973,
Onians 1954) points to the existence of certain basic or "mac-
ro" metaphors underpinning semantic elements in our thought
processes generally (Guiraud 1967:195).
Sensory experience is difficult to categorize verbally. How
can one separate the experience of something smelled (say)
from the experience of something tasted, touched, seen, or
heard? As we have seen, input from several sensory systems is
normally merged before it ever reaches the level of conscious
recognition. We can, of course, experience something with only
one sense, as, for example, by smelling something at a distance,
in the dark, or by seeing something that makes no sound and
has no smell. When the input is so limited, however, our other
senses rush in (with "memories" of past associations) to fill the
gap. Thus one tends to picture mentally what one is smelling
and to remember a taste or a texture not immediately given.
This happens in the very process of perceptually identifying
what it is that is being smelled. Similarly, an item seen can be
touched in one's imagination, and so forth.
Levi-Strauss has argued that, because our sensory codes are
interchangeable in this way, they must all be structured
according to a uniform and simple logic. Indeed, he has based
much of his mythological research on this assumption, arguing
that our sensory logic is basically binary in form (1970:153-54;
also Leach 1976:11, 41). Thus Levi-Strauss suggests that all
sensory codes are isomorphic and that they can be reduced (at
least in a given culture) to one general table of equivalences.
From personal experience, however, I find that sensory thoughts
lead in a hundred directions. Images that follow upon the
stimulus of a given prior image need not describe a "logical"
sequence of categories and relationships. Sensory thought is
"shifty" or fluid. We can invent a binary logic for it, but by
so doing we transform it into a "language" governed by verbal
rules. In my view, the structure of sensory logic is governed by
nodes at which multidimensional spatial and emotive images
dominate complex network movements. Such is not the logic
of a detached, categorizing mind. Sensory thought is motor-
based, egocentric, and affective in character.
What I hope to have established is that a preverbal level of
human thought processes exists. On this level, we are not
dealing with clear categories, but are rather led into a world
where reasoning is by analogy, where thoughts seem to move
along paths in a dense network of prior associations. Such
associations occur in a multidimensional space. They may be
associations of colour, shape, texture, smell, type of movement,
and so forth. Many associations are possible, and they may
lead outward in many directions-or in clusters-to new,
stable foci. Since our perceptions of objects normally involve
many dimensions at once, no clear boundaries can be estab-
lished in this structure. One does not cross from category to
categc'ry but moves, at best, along clines or gradients. And not
only are there no clear-cut distinctions of the (a) or (-a) type,
but there are so many dimensions to a set of sensory associations
that even the most sophisticated poet can provide only an
approximate verbal description of a given experience.
A verbal metaphor can now be understood as a device whose
function is to inject the results of analogic reasoning processes
into the semantic domain. As Fernandez has said, a metaphor
bridges gaps. We can now understand this as a process whereby
images and experiential associations that develop at a level
where a network of sensory associations prevails are transferred
to a level where thoughts are ordered according to a logic of
verbal categories. Metaphors cross over such categorical
divides as animate/inanimate, cosmic/biological, human/animal
by recourse to associative and sensory logic.
Returning to Fernandez's ideas about metaphoric processes,
we can see how all the kinds of movement he mentions can in
fact be understood in terms of this one description. The intro-
duction of sensory logic at the semantic level entails a movement
from abstract to concrete. It also involves the introduction of
affect and the notion of perceptual qualities, thus bridging the
logical gaps that separate object categories at the linguistic
level. The same sensory logic also entails a shift from a focus
on a single aspect or quality of an object to a more generalized
sense of the phenomenon as a multifaceted whole. This whole-
ness is a result of the fact that at the sensory level an experience
is normally recorded in several different ways at once.
In sum, we must come to recognize the domain of metaphor
as a domain in which a special kind of preverbal logic of ex-
perience substitutes for the logic of set theory which orders our
verbal codes. Metaphor is an exciting domain to study precisely
because it is here that we can observe our own procedures for
shifting between these two distinct modes of reasoning. Prov-
erbs, riddles, and ritual metaphors thus serve as important
"mediators" between our analogic and our semantic thought
frameworks.
This insistence on there being two levels or types of mental
processes is not, of course, an original suggestion. As Munz
(1976) points out, there have been numerous attempts to
describe human thought in binary terms. Frobenius spoke of
passion versus reason, Worringer of gothic and geometric, and
Northrop of organic versus linear thought; Marcuse, Eliot, and
others, furthermore, have dealt with the general issue of
opposing thought types. There has also been considerable
Vol. 19 * No.
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
Beck: METAPHOR AS MEDIATOR BETWEEN THOUGHT MODES
research into the question of whether these two distinct types
of logical processes may be laterally distinguished by having
their primary locales on separate sides of the brain. The degree
of association that actually exists between these polar types of
thought and the "split" physiological form that sustains them
is still to be firmly established (Martindale 1976).
I will now consider the work of two well-known anthropolo-
gists, Levi-Strauss and Leach, insofar as they have discussed
the logic of sensory reasoning at a more general level.
One or the main themes that runs through Levi-Strauss's
The Savage Mind (1966) is the distinction between metaphor
and metonym. A metaphor, he suggests, exists in a public arena
of discourse and is embedded in the very language of social
relations. It results from holistic thinking and involves a logic
that relies on abstract, formal relationships. A metonym, on
the other hand, involves reasoning which is concrete and
progresses by means of contiguity and substantive analogy.
Levi-Strauss has argued that the first constitutes a process of
abstraction requiring sensory impoverishment, the second a
process which is full of sensory imagery but which (by implica-
tion) entails a logical impoverishment (p. 229). Metonyms,
therefore, entail a largely "private" reasoning process and are
lacking in "good sense" (p. 229). In six of the seven places in
which metaphors are discussed in The Savage Mind, they are
paired with metonyms. Levi-Strauss clearly sees these two
processes as operating together. They are almost necessary
complements to one another. While he generally supports the
view that "savage" minds, like more "civilized" ones, are
governed by logical (binary) reasoning processes, in the final
paragraphs of his work he also tentatively suggests that the
proportion of abstract and concrete material varies in these two
contexts. Primitive thought, he argues, takes the physical as its
starting point and then works towards abstract relations.
Modern scientific thought moves in the opposite direction, from
abstract to concrete. Hence his view appears to be that all
human minds use concrete imagery to a degree, but that those
without scientific training use such images with greater fre-
quency, and as their starting point.2
Whereas Levi-Strauss equates metaphor with a "higher"
(more abstract or scientific) type of thought and metonym with
a "lower" one, Leach (1976) inverts this (implicit) rank order.
He suggests that the most elementary stage of human under-
standing involves the coding of sensory input into a set of
binary categories (p. 19), linked to a sense image. Leach calls
this sense object "arbitrary, at least to some extent," and
suggests that it bears a metaphoric (an arbitrary or "non-
sense") relation to the external world (p. 22). The second (and
by implication the "higher") stage in the reasoning process,
according to Leach, is to make metonymic sense out of these
object metaphors by abstracting the sets of intrinsic relation-
ships that exist between them. It is through this latter process
that he sees us as arriving at concepts and at a more complex
and abstract level of mental functioning (p. 19). Levi-Strauss
and Leach do at least agree, however, that metaphoric and
metonymic logic are interdependent.
It seems to me that, in seeking to distinguish metaphor from
metonym, both writers have incorrectly formulated the prob-
lem. Their concern for this distinction seems to arise from their
devotion to the principle that elements in a cognitive system
can never have meaning in themselves (see, e.g., Leach 1976:13,
49, 95). In such a view, a metaphor can be defined only by
contrasting it with something that it is not-e.g., metonym-
2 A preferable position, in my view, is that taken by Paredes and
Hepburn (1976:740), who suggest that the "primitive and civilized
both think in a number of equally complex-'secondary'-ways, but
the conditions which evoke the several different kinds of thinking of
which both are capable vary with the particular characteristics of the
cultural environment."
1 March 1978 85
because meaning exists only in relationships. An element
cannot exist outside a system, and a system always contains a
set of logical contrasts between its elements that allows no
element an ontological status by itself. A further limitation
these writers place on their analysis is that any given set of
mythic or ritual materials must contain one single, central
message (Leach 1976:41; Levi-Strauss 1970:164). This uniform
message, they argue, can be unambiguously deciphered, but
only if one uses the correct decoding procedures.
Fernandez, in contrast, stands closer to the modern theologi-
cal tradition. Heidegger (1957), Ricoeur (1974, 1975), Crossan
(1973), and Perrin (1976), for example, have suggested that
human symbolic systems normally contain paradoxical ele-
ments. If this be true, then central mythic or ritual messages
cannot be reduced to a single, simple idea, and we can expect
to find not only various levels to cognitive structure, but also
basic contradictions. In such a view, anthropological structures
behave like the outline of a cube, in which what is seen as a
foreground plane one minute is seen as a background plane the
next, or like light, which scientists say behaves like a particle
in one experiment and like a wave in the next.
Why should symbolic structures be so "shifty"? Perhaps it is
because of the very nature of symbolic thought. According to
Leach (1976:49, 95), a symbol is defined by its relationship to
other symbols at an equivalent level in a given system. In my
view, a symbol can evoke associations in its own right; it does
not depend on a system of contrasts for its meaning. To return
to the idea of two levels of thought, I prefer to speak of a
contrast between a sense-dominated level and a second which
is verbal and logical in the common sense of the term. At the
first of these levels, objects and the sensations they engender
have a reality in themselves. At the second level, however,
thought treats classes of objects and the contrasts which
separate and oppose them in various ways. At the first level we
have a proclivity for symbolic thinking and at the second for
structural thinking. The first deals with central foci, nodes, and
symbolic forms, the second with boundaries and the logic of
separation and contrast.
In the sensory medium, our thoughts appear to be analogic,
energized and fluid. Images at this level do not form neat
categories; instead, they tend to transcend verbal norms. Such
"transformations"-the joining of what should not be joined
(according to a verbal code of logic)-can be profoundly
disturbing. Such thoughts are also open-ended; they seem to
extend reality, as does a dream. The second level of thought,
by contrast, is relatively restful and orderly. Our verbal habits
tend to circumscribe images and ideas semantically and keep
them in their proper place. The reason that cognitive structures
contain contradictions, then, is that they must take account of
this primordial base in which experiences evade verbal circum-
scription. At the level of semantic codes, the concept of paradox
is the only way to represent the quality of this other type of
thought experience.
What is needed is some kind of mechanism for communication
between these two levels of cognitive activity. This is where
the metaphor comes into play. Metaphors effect the needed
transfer in a relatively orderly way. They are not "of" one kind
of thought or the other; they are go-betweens. Both holistic
logic and partial associations are needed to effect this transfer,
since what is taken to be a whole at the first level will be taken
as a part of a system at the second. There can be no profit in
rigorously separating the metaphor, therefore, from the
metonym.
An example may help to clarify what I have in mind: At the
first level of thought, an apple is an immediate object. It can
be held, tasted, smelled, etc. We need no name for it to appre-
ciate it, and some idea of its qualities can be formed without
contrasting it with, say, an orange. For an adult of our own
culture, an apple will (can) evoke memory images of taste,
color, and shape that are part of encounters with previous
86 CURRENT ANTHROPOLOGY
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
apples. More importantly, it can evoke associations with
apple-bobbing at Hallowe'en, or with the brown-bag lunch, or
with the gift slipped into the Christmas stocking. By its round-
ness, its color, its weight, etc., it can also evoke associations of
entirely different kinds, becoming, for example, a profoundly
disturbing symbol of forbidden sexuality (by association with
the Garden of Eden story) or of the ripening of some mysterious
vegetal force (by association with the "fruiting" processes of
plants).
At the second level of thought, the apple is contrasted with
other "things." Apples are fruits, not leaves. Apples belong to
the genus Malus, not Citrus. Apples can be distinguished from
oranges by their colour, by the climate in which they grow best,
by their surface texture, and so forth. We can thus sort apples
from other things in many ways. We can talk about the larger
patterns of elements in which they find a niche.
Now, if we want to "stir up" our categorical thinking about
apples a little and produce a surprise, we can introduce a
metaphor, for example, "to be the apple of one's eye." We here
introduce a new category for the apple, one unlike any of the
others just mentioned. This is a surprising departure because
it refers to a "feeling" about a fine, ripe apple as an object-in-
itself (Level 1) and introduces it as a descriptor of a category of
feelings, desiring (Level 2). The situation is similar with the
metaphor "to be a rotten apple." On the one hand, this con-
jures up an image of a rotting apple-in-itself (Level 1); on the
other, it describes a category of undesirable objects, usually
people (Level 2). A third metaphor, "to polish the apple," is
different only in that it refers to a specific action-in-relation-to
a specific fruit at Level 1 and to a category of actions-in-
relation-to persons at Level 2. The presence of a metonymic
logic of parts can also be clarified when we realize that one may
also "eat" an apple, "cut up" an apple, etc. The apple as a
whole, on the other hand, stands as a metaphor for a person.
But the polishing metaphor is successful, as a statement,
because of the unexpected juxtaposition of these several ele-
ments, not because of the details of its (part/whole) logical form.
Lotman (1972), a Russian theorist who has written about the
structure of literary texts, has stressed that a good writer
creates a lively work by juxtaposing dissimilar elements. He
suggests that the contrasts that ensue energize the whole. He
sees these creative linkages between planes or levels as the
quality that activates an artistic work (pp. 277-78, 393-95).
Ricoeur (1974) is also concerned with this juxtaposition, par-
ticularly as it concerns the metaphor. He speaks of metaphors
as "alive" because of their expressive or emotive force. He also
stresses the irnportance of a tension between the literal meaning
of a term or phrase, as defined by the normal context of use,
and the "more fundamental" modes of reference that are
unlocked by the suppression of this literal element in meta-
phoric discourse (p. 79). This is why, he suggests, "documents
of faith [are primarily composed] . .. in such modes of dis-
course as narratives, prophecies, legislative texts, proverbs and
wisdom sayings" (p. 73).
Crossan (1973:345) concurs in this view that metaphors have
the power to create participation. Metaphors are to be ex-
perienced, the experience being to open up thoughts and allow
movement between normally closed domains via the use of a
powerful symbolic form (p. 350). Pesot (1975:5-6) refers to the
same principle: "traditionally the metaphor is defined as being
a process of the transfer of meaning by the substitution of an
analogy.... There is always an act of comparison having as its
goal the enhancement of expressiveness or a livening-up of the
style or a satisfaction of a deeper need for primitive animism."
We may now return to our point of beginning, Fernandez's
description of the basic processes in metaphoric thought.
Fernandez stresses the idea of movement. Metaphors provide
for movement between partial and abstract principles employed
on a verbal plane and concrete, sensual, holistic images that
thrive on a nonverbal one. Metaphors are thus mediators
between these two modes of our thought. They seem "alive"
because they animate the mind, connecting its two normally
separate domains. Metaphors serve as a bridge between the
"structures" of thought that interest the anthropologist and
the "symbols" of thought that have preoccupied psychologists
and theologians. Each discipline needs to broaden its view.
Perhaps through the study of metaphor a new interface between
these separate perspectives will develop.
I propose, then, that metaphors allow us to introduce
nonverbal material into logically structured, semantic contexts.
Furthermore, the metaphor and the metonym are two sides of
a single device used to bridge the gap between rational and
sensory thought processes. Since this device is most commonly
used at points where our verbal paradigms are in some way
inadequate or weak, its use helps to mark the limits of rational
understanding. In proverbs, in riddles, and in ritual, we are in
a good position to see metaphors in action where they are
needed.
Proverbs are like riddles in the sense that one element of
several, in a larger pattern of relationships, is intentionally
kept hidden. While in the riddle the point is to guess the
hidden term, in the proverb the problem is to guess how several
named elements relate to a hidden or unspecified additional
context. Thus a proverb is a comment on a situation. A proverb
teller hopes that his comment will help to classify that situation,
which (by implication) is one that is not otherwise easily
defined. The proverb places an already given situation in a
new light by referring the observer to a fresh and concrete
sensory image. By use of this technique, the teller points up a
lack in the more abstract moral or rational code previously
used to define the problem. For example, the English expression
"to kill two birds with one stone" and the Tamil expression
"to lick the back of the hand when there is food in the palm"
both employ a physical image of doing something. Each image
is used to explicate an event not itself mentioned. In each case
the expression may well be applied to a situation where the
skillful use of limited resources or energy is at issue, but it will
certainly not be a setting where people are actually killing
birds with stones or actually licking their hands. The concrete
reference to bird killing will be used as an analogy for another
problem, from which the commentator has abstracted a prin-
ciple (the conservation of given resources or energy). It will be
applied in a situation in which someone is seen to be attacking
a problem rationally. The expression about licking one's hand,
by contrast, will be applied where the opposite abstract prin-
ciple (lack of economy or lack of rationality) is seen to operate.
In these examples, sensory reasoning has been used as a
support for some rational (abstract) principle. In each case the
listener is asked to transfer an initial sensory image to another
level of thought, a level where binary distinctions are in force.
At this second level, human actions are defined as either
conserving or not conserving energy, as either rational or
irrational means to an end, etc. These contrasts cannot be
made at the level of sensory input. One can throw stones with
more or less accuracy, lick and obtain more or less nutritive
results, see more or less light, etc. But judgements, at the
sensory level, produce only sliding scales. It is the proverb user
who gives the impression of making a categorical judgement.
His logic is: energy is not conserved in Context 1 (the concrete
situation described by the proverb), Context 2 is like Context 1
(the situation the proverb is applied to), therefore energy is not
conserved in Context 2. In terms of set theory this would be:
when a, then c: a = b: therefore when b, then c. But the actual
logic used is not this rigorous. The proverb image does not
mention the abstract principle (b), conservation, at all. It also
makes no reference to the new context (c) that the proverb
user has in mind. Yet the sensory impact of the proverb's
concrete description carries over. If the context of application
is chosen judiciously by the proverb user, then the categorical
judgement the user makes (at a verbal level) receives added
Vol. 19 * No. 1 *
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
Beck: METAPHOR AS MEDIATOR BETWEEN THOUGHT MODES
force because of the sensory impact of the nonverbal image.
Such images thus lend feeling and an aura of truth to our
attempts to define acceptable behavior at the rational, cate-
gorical level, but they do not hold up to the rigours of logical
analysis.
The situation is similar with riddles. Riddles can be viewed
as a kind of joking. Instead of trying to reinforce and shore up
the boundaries of rational thought, riddles release tensions
built up by the presence of semantic boundaries. Riddles always
depend on deception, in the sense that the way a riddling ques-
tion is phrased makes the listener think that the answer lies in
one category while, in fact, it lies in another. For example, when
the riddle "What is black and white and (red) all over?" is told
orally, the listener hears red because the first two terms, black
and white, condition him to think in terms of a set of colour
terms. The trick, in answering correctly, is to hear another word,
namely, read, which leads one to think of the set of things
humans read. Now finding the answer, "a newspaper" (which
is black and white), is easy.
Not all riddles depend on tricks of this kind, however. Many
more depend on a description that sounds human (the use of a
pronoun, a reference to a common human body part or activity)
and require the perception of another set (animals, inanimate
objects, or whatever) that contains some element which (in all
other respects) fits the description given. For example, in the
Indian riddle "A child with one eye," the answer, "A needle,"
depends on a switch from the set "all humans" (only humans
have children) to the set "all inanimate objects." It also
depends on broadening the concept of "eye" so that it refers
not only to a particular body organ but to the idea of a small
opening in an inanimate object. By contrast, the U.S. riddle
"What has eyes but cannot see?" (answer: "A potato") depends
on a transfer from the category "organic, seeing eyes" to the
category "spots or cavities on an object that can be thought
of as eyes."
The riddle is thus a metaphorical "joke" about the conven-
tional categories of language. "Eyes," at the sensory level,
not only see but are deep cavities or spots on the face. "Chil-
dren," at the sensory level, are not only young humans, but
also little things. Riddling is fun because it makes us see the
absurd rigidity of our verbal categories. It shows that we can
also classify phenomena in other ways, using links between
things that exist only at the level of sensory logic.
Finally, we must consider metaphors in ritual. In what way
can we see metaphors of this kind as transgressing the boundaries
of rational thought? As a start, we must restate the principle
that ritual metaphors are concerned with getting beyond the
boundaries of man's rational ability to grasp the complexities
of the human condition. Ritual activity attempts to add to our
understanding of how things happen and why they happen by
recourse to basic sensory experience. Just as in the area of
judgements about behavior (where proverbs are invoked), so,
too, in the area of our understanding of cosmic and biological
processes, we often wish that we saw more clearly or knew more.
The sensory imagery contained in proverbs helps us to reinforce
questionable judgements in the first area. Ritual helps to
reinforce shaky understandings in the second.
As an example here, I offer the common observation that an
egg which hatches (after being covered or concealed in the
ground) produces a bird that flies. Many cultures have picked
up this simple, everyday event and transferred it, via a meta-
phor, to a ritual context. Often this metaphorical transforma-
tion, worked on a basic and incompletely understood process,
remains at a nonverbal level. Even if it does not receive a
verbal "translation," however, the attempt to aid the under-
standing by the use of such ritual metaphors is similar to what
I have already shown to occur in the verbal metaphors em-
ployed by proverbs and riddles.
March 1978 87
 For example, in New Guinea, the egg-to-bird progression is
linked by analogy to ancestral human grave/shrines. The skull
of a deceased kinsman is seen to resemble an egg buried in a
dark place. The spirit of the dead person is then seen to be
released from this "shell" to "fly" upward as a hatched bird
flies after leaving its shell. In South India, the egg-to-bird
process is linked by analogy to a myth about the birth of a
primordial being from an undifferentiated, egglike universe.
In some rituals a round pot is used which, when filled with
water, bears, by analogy, a relation to an egg. When certain
magical verses are recited and special ritual gestures made, it is
said that this pot is transformed into a cosmic body that is
understood to be one with the spirit forces of the world above,
similar to the hatched bird that unites itself physically with the
forces of an upper world.
In these ritual contexts, the participants do something (by
the act of attending the proceedings) that opens them up to a
diffuse set of sensory inputs produced by the ritual activities.
These experiences are of processes that are defined as creative
and energizing and viewed as communicating with a little-
understood upper world of superhuman forces. The rituals do
not explain anything directly. Nonetheless, the metaphors
imbedded in them give the participants a certain renewed feel
for creative and communicative processes at a sensory level.
In sum, metaphors carry our understanding across the
boundaries of verbal thought. Sensory analogy alone is inade-
quate as a conceptual tool for dealing with the many complexi-
ties of human life. The structured paradigms provided by
logically opposed categories of verbal thought are also needed
to order our world. Yet these two modes of reasoning are
most fruitful when they remain interdependent. The use of the
metaphor/metonym device is one important way in which we
maintain the vital link between these two levels of our thought.
Most importantly, the study of specific metaphors, in use, can
provide fresh insight into how cultures adapt to changing
realities. They serve as a primary means by which we may
attempt to overcome the confrontation of first principles which
a radical social or environmental change can produce.
Comments
by MARY DOUGLAS
Department of Anthropology, University College, Gower St.,
London WClE 6BT, England. 18 viii 77
Beck has done a good service in criticising and resynthesising
current work on metaphor, but she pulls her punches. The two
main points made in this article are about the phenomenon of
synaesthesia and the deeply intertwined concepts of metonym
and metaphor.
The discussion of a system of sensory logic or macro-meta-
phors based on associations at the motor and emotional level
is fascinating but too short.
In the citation made of my own work there is a misunder-
standing, since in Natural Symbols I was concerned to point
out that, though the body is a source of possible metaphors,
they are not used in the same way in different societies. Need-
ham, also cited, generally makes the same point. But, of course,
the underlying similarity in a pool of potential symbolic systems
is not disputed. One reason the mapping from one sensory ex-
perience to another is possible is that all except the olfactory
senses are experienced in spatial and/or temporal sequences.
If an event can be seen as a sequence in time, it can be laid out
as a sequence in space, as Wittgenstein illustrated with the case
of the musical score. Smelling and tasting have neither a spatial
nor a temporal structure, and so they deserve the epithets of
fluid, energising, etc., which the author applies indiscriminately
to all sensory exFperience. I believe her to be unduly ethno-
88 CURRENT ANTHROPOLOGY
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
centric in treating the latter ("concrete, sensual, holistic") as
different from abstract principles. It is a pity to try to draw
a sharp line between these two modes of thought. With small
differences of emphasis, the attempt lies straight in the tradi-
tion on which William James built and which he himself traces
with admirable clarity back through Hume to Hobbes. Taking
over this tradition, James was able to use a two-part model of
human thinking, one spontaneous and uncontrolled, the asso-
ciation of ideas (which he saw as rooted in habit and the physi-
ology of the brain), and the other logical and analytical. Fer-
nandez is taking over this same tradition, and so are the struc-
tural analysts this author criticises so aptly, who limit their
tool-kit to two terms only, metonym and metaphor, and in-
extricably mix them up. All those who have tried to separate
contiguity from resemblance (including Frazer) have found
them impossible to distinguish, and James's original criticism
still stands. Metonym and metaphor in the hands of Levi-
Strauss and Leach remain confused and inadequate analytic
concepts. If only the author had gone farther back than James
to Aristotle's Poetics, where the terms were handed out to us
for bricolage, she would have been able to do more analytic work
on recognition of whole-part and resemblance relationships.
The idea that recognizing a resemblance as a basis for a meta-
phor is an immediate experience, correctly to be entitled con-
crete or sensory, seems to me to be very dubious. Recognition
of resemblance requires comparison, assembling a structure of
trans-mappable features. How this faculty deserves to be sepa-
rated from other faculties which are "logical" and "analytical"
beats me. In making this contrast so lightly and ignoring the
puzzles it raises, we are "bricoleurs" of the first rank. James was
after all a 19th-century figure, but Empson's (1967) The Struc-
ture of Complex Words is a contemporary book and very much
to the point. It would have supported the author's implied
view that the usual treatment of metaphors by anthropologists
is extremely heavy-handed and clumsy. Beck is too polite to
carry through a radical criticism of anthropologists' discussion
of metaphor.
by M. S. EDMONSON
Department of Anthropology, Tulane University, New Orleans,
La. 70118, U.S.A. 26 ix 77
The concept of metaphor in one or another acceptation has
attracted increasing attention in recent anthropology without
generating much consensus. Rather generally, however, the
idea seems to be accepted that the concept invokes some order
of distinction between logic and analogy and between sensory
experience and at least some aspects of language. Beyond these
simple assumptions we enter no-man's-land.
Beck identifies logic with cognitive semantics and analogy
with sensory experience, and proposes that metaphor mediates
between them. I disagree. In fact, I think she has it almost
exactly backwards. Binary categorization and the law of the
excluded middle appear to me to rest ultimately on the evi-
dence of our senses-a (logically) primitive perception that
two objects cannot occupy the same space at the same time.
Analogy produces assertions of similarity and dissimilarity in
defiance of both sensory evidence and the law of the excluded
middle. It too is semantic, however, and I do not believe it in
any sense to be "preverbal." On the contrary, I believe that
without human language metaphor would be impossible, and
probably vice versa.
Evidence on this last point is far from clear-cut, but such
data as we have on the ontogenesis and phylogenesis of meta-
phor do not appear to me to encourage the speculation that
it is antecedent to language. Gestural metaphor does of course
occur, but so do cognitively significant gestures, and neither
reaches the fully human level of structural complexity in apes
or infants. The implication that metaphoric relationships are
unclear (because multivalent) is unwarranted. The restriction
of such relationships to sensory linkages is falsified by the
demonstration (by Levi-Strauss and others) that the linkages
themselves may be metaphoric. It is not at all what is experien-
tially raw or cooked that enables the Amazonian Indians to
use those categories (inter alia) as symbols for nature and
culture. In place of Beck's restrictive formula "anything: any-
thing (sensorily mediated)," I would insist that anything may
be likened to anything. Where is the sensory mediation in the
assertion "Politics is a transitive verb"? Is this not metaphor?
I am particularly dubious about the utility of restricting the
domain of semantics to binary categorization. Indeed, I regard
metaphor not only as semantic but as the very essence of hu-
man (as opposed to computer) language. To eliminate metaphor
from semantics implies that metaphor is not part of meaning.
Is this seriously intended?
A great deal has been made of symbols (including metaphors)
as "mediators," and it seems clear that even simple metaphors
take on a mediative structure-a: (is like) b. But the limits of
metaphor are not dictated by the senses (though metaphor
may certainly include sensory imagery). The final limit is what
will communicate, often in highly specific context. In short, the
genesis of metaphor does not lie in some preverbal and non-
semantic stratum: it is born of the highly verbal and supremely
semantic experience of culture itself.
by JAMES W. FERNANDEZ
Department of Anthropology, Princeton University, Princeton,
N.J. 08540, U.S.A. 15 ix 77
It is excellent service to anthropology, whose object is the study
of behavior and thought in situ, for the author to have clarified
the essential "shiftiness" of that behavior and thought-the
decided commitment to categories and discriminations of the
parts of experience, on the one hand, and the feeling of the
essential artificiality of these categories, their betrayal of the
wholeness, multiplexity, motility of experience, on the other.
While we are discomfited by things out of place, as Douglas
(1966) has shown, we are also attracted by their polyvalent
character. We may abuse interstitial categories or use them
for abuse (Leach 1964), but they attract us just the same, for
they suggest, like the fiddle-footed gypsies, a wholer existence
than the one in which our intellectual civility obliges us to live.
It was also Leach's point, and it is recurrent in Levi-Strauss,
that though we must break the continua of experience into
parts, we keep being drawn back to the whole. Beck advances
our understanding in identifying that "whole" with the "uni-
sensate," "undifferentiated," "synaesthetic," "prelinguistic"
pool of motor and enactive experience. Semantic codes are later
built up from this pool, but it constantly reasserts itself, well-
ing up as "intimations" and "recollections," if not of "im-
mortality," at least of transcendent unities. The fact is that
any truly enduring subject of human interest seems to be in-
choate in this way-caught between categories, on the one
hand, and synaesthesia, on the other. That may sound obscu-
rantist, but it is what the anthropology of experience has to
deal with, and Beck gives a good matter-of-fact treatment of it.
Beck is right. We have tended to treat metaphors too much
in verbal terms-as word-based images, and hence too much
governed by the logic of language and set theory: conjunction,
disjunction, implication, contradiction. There has been a sur-
feit of schoolmasterish analyses-rule-bound, rule-seeking,
rule-applying. We have, perhaps, been too careful to avoid con-
tradiction to rephrase Levy-Bruhl's critique of prelogical
thought: that it was "careless about contradictions."
How best can we get at the paradox of metaphor?-for
paradox is what these "primitive abstractions" present. I won-
der if dispelling the metonym-metaphor distinction is helpful.
True, Beck is mainly concerned to collapse it, not reject it. It
is also true that metaphor so easily transforms into metonym
that it is hard to keep these two concepts separate. Still, I find
Vol. 19 * No. 1 *
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
Beck: METAPHOR AS MEDIATOR BETWEEN THOUGHT MODES
Jakobson's (1956) argument-that there must be contiguity be-
fore there can be a relation-helpful to our understanding. It
has been clarified by K6ngas Maranda (1971c) and most re-
cently by Basso (1976). It is pointed out, first, that the asser-
tion of a metaphor rests upon an assessment of the metonymic
relationship of parts within a domain, that is, a sense of the
relevant structure of the domain. Second, the consequence of
a metaphoric assertion is the emergent sense of a transcendent
domain, a "superset," in Kongas's words, with new and dis-
tinctive attributes derived from but encompassing the two
domains brought into relation by that assertion. Let us take a
riddle (a truncated metaphor, as Beck points out): "long legs,
sharp thighs, no neck, big eyes." The assertion of a metaphoric
similarity between scissors and the human body is based on
the accurate assessment of the relationship of parts within the
two domains-upon metonymic awareness. The metaphoric
assertion results in the emergence of a transcendent domain-
that of "articulated entities," including both jointed scissors
and jointed people. The defining attributes of this new domain
derive from but are different from the attributes of the source
domain. In this way, metaphor creatively fills lexical gaps, as
Basso argues.
Of course, Beck's important point is that the assertion of
metaphor takes us back to those preverbal, unisensate modes
of assessment and to possibilities of association that purely
linguistically oriented analysis is likely to overlook. Those
"sharp thighs," for example, evoke all kinds of psychoanalytic
preoccupations to which the structural analysis of metonym-
metaphor relations is oblivious. Perhaps it would be helpful
to distinguish between structural and textual metaphors. The
former rest on similarities of structure and the latter on similari-
ties of feeling tone-on "certain sentimental attitudes," as
Durkheim and Mauss (1963) called them in their search for
the principles by which collective representations were classi-
fied. Incidentally, would not "trope" be the better key term
of reference in this discussion than metaphor-which, with
metonymy, synecdoche, and irony, is a variety of trope?
Further, would it not be preferable to speak of mediation be-
tween analogue and digital, rather than analogic and semantic,
modes of thought? These terms better suggest the direct
"gradient-sensing" and indirect category-assigning modes of
assessing experience. Also, they are body-based, as practically
all enduring metaphors of mind turn out to be. What, for ex-
ample, but our bilateral symmetry keeps bringing us back to
the binary terms that anchor these kinds of discussions?
Beck speaks of "preverbal logic" and "sensory logic," but
doesn't this reintroduce a word-based perspective? I am not
sure what kind of logic this would be unless it were simply the
"logic of vectors," of movement in sensed qualities upon various
gradients or continua. I have been arguing this elsewhere (1977).
In respect to the sense-to-reason ratio, isn't it better to accept
Levi-Strauss's point that the use of sensory imagery requires
logical impoverishment while abstractions require sensory
impoverishment? The challenge to anthropological description
and analysis is to avoid either impoverishment. Our riches lie
in mediation, which may be, indeed, the very metaphor of
anthropology itself.
One final word in favor of methods. In recent years we have
bemused ourselves with the thought that we can sit back in the
armchair and, operating logically on texts, produce all we need
to know about culture. Beck rightly suggests that no amount of
logical analysis will gain for us the plenitude of associations
lodged in those prelinguistic nodes, the central foci of experi-
ence. Only participant observation will lead to the sense of how
the parts of experience participate with each other in a par-
ticular culture. At the same time, there is a challenge to cross-
cultural inquiry-to an effort, let us say, to catalogue the
macrometaphors of cultures in the way that Onians has cata-
March 1978 89
logued them for the European tradition. The impulse to that
kind of inquiry is another consequence of reading the Beck
paper.
by ISTVAN FODOR
Institut fur Afrikanistik der Universitlt zu Koln, Meister-
Ekkehart-Strasse 7, 5 Koln 41, Federal Republic of Germany.
19 viii 7 7
An old topic, discussed since the epoch of classical philosophy
and stylistics, is here raised once again. Although this paper is
too brief to add essentially new elements to the discussion, it
does touch upon some problems worthy of investigation from an
anthropological point of view. I wish to add one remark and to
make a suggestion.
1. The physical and psychobiological bases of metaphors
were discovered by Fonagy (1963) in the course of an explora-
tion of the origin of the technical terms of phonetics, from the
ancient (Sanskrit, Chinese, Greek, Latin) grammatical views
through the national phonetics of the Middle Ages to the
international scientific terminology. This is a fascinating work
that shows the origin and function of metaphors from a general
point of view as well.
2. Beyond doubt, the interpretation of metaphor is a topic
of thought psychology and anthropology that lacks an eth-
nological survey. In my opinion, a cross-cultural study ought
to be made along the lines of Murdock's World Ethnographic
Sample (1957). I am no opponent of quantitative studies, but I
think that, instead of an all-embracing, wearisome, expensive
collection of facts analyzable only by computer, a qualitative
analysis would be better suited to our purposes. We should
concentrate research on a few theoretically selected factors.
First of all, 20-25 societies should be chosen according to
geographical and social characteristics-e.g., in Africa, hunting
(Dorobo), pastoral nomadic (Masai), agricultural (Bemba,
Ewe), and manufacturing (Hausa) peoples, etc. Data from
some peasant communities of the industrial world are also
necessary for this survey, but no fieldwork concerning them is
needed, since a large amount of facts is available in the ethno-
graphic literature. Linguistic differences are also to be con-
sidered in the selection of the peoples (Indo-European, Finno-
Ugric, Semito-Hamitic, Bantu, Polynesian, etc.).
As regards the questionnaires, no more than 25-30 subject
matters should be included, e.g., colors, geometric forms, some
animals and fruits. The questions should be designed to reveal
the character, function, and possibly origin of the metaphors.
For example, with colors, it would be useful to recognize what
kinds of metaphors emerge (e.g., green is the color of hope
among many peoples in Europe), what colors dominate in
popular art, what the proportions of the colors are to each
other. The spectra of color names should be sketched as well
(cf. Hjelmslev 1953).
This research would require international teamwork in-
cluding many institutions, but probably the results would be
better able to elucidate the nature of metaphor than a verbal
discussion based merely on the known facts.
by IVAN FONAGY
1, squ. Claude Debussy, 92160 Antony, France. 11 ix 77
Je me trouve en plein accord avec I'auteur quant au fond des
idees exposees avec clarte et elegance. La meta-phore, comme
le terme indique, est un signe dynamique par excellence (cf.
Quintilian Institutio oratoria 2.13.8 et ss.) qui s'oppose aux signes
stables et qui resume par son mouvement virtuel l'ontogenese et
la philogenese de la pensee conceptuelle a partir de la pensee
preverbale. J'ai tache d'esquisser cette evolution dans des
publications anterieures (ainsi en 1961, 1963, 1965, 1971, 1972,
1973).
J'aimerais, toutefois, mettre en evidence certaines divergences
90 CURRENT ANTHROPOLOGY
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
entre nos conceptions. Je crains que les beaux termes meta-
phoriques, ((sensory logic)), ((sensory thought)), qui designent
et qualifient dans l'article le processus mental qui soutend la
m'taphore, pourraient preter a erreur. Ces termes semblent
suggerer que la pensee metaphorique est etroitement liee au
niveau sensuel et que ses particularites, les ((erreurs)) inherentes
a l'expression metaphorique, s'expliquent par cette limitation.
A mon sens, ce qui distingue l'ideation metaphorique de l'idea-
tion conceptuelle c'est une differente elaboration des memes
donnees sensorielles. La metaphore, en tant qu'erreur volon-
taire, releve d'une analyse plus primitive de ces donnees.
Freud designe par le terme de ((processus primaire)) l'ensemble
des procedes mentaux qui vehiculent le reve, les mythes et
les tropes; il les oppose aux mecanismes qui constituent la
pensee rationnelle, le ((processus secondaire)), susceptible de
donner une image plus objective de la realite (Freud 1940-46
[1900]: 593-614). Le processus primaire constitue le noyau
de la structure mentale et correspond a une vision enfantine,
archaYque du monde, dominee par le principe du plaisir.
L'auteur se laisse parfois entrainer par ses propres meta-
phores (telles que ((sensory logic))). Ainsi, elle conclut par rap-
port aux oppositions conceptuelles binaires que ((these contrasts
cannot be made at the level of sensory input)). II est evident
qu'aucune ideation-d'ordre conceptuel ou metaphorique-ne
pourrait se faire au niveau sensoriel. Je ne crois non plus que
la synesthesie pourrait s'expliquer au niveau physiologique,
sensoriel. L'auteur, en reprenant une these de Herder (Abhand-
lung iuber den Ursprung der Sprache 1770) qui suppose l'existence
d'un sens universel (sensum commune), admet une ((profound
unisensate base)) englobant tous les sens. En realite, la synes-
thesie est un phenomene psychologique complexe: un evenement
concret, preconscient ou inconscient. auquel s'associent differ-
entes sensations permet le passage d'un domaine sensoriel a
l'autre. La metaphore synesthesique est collective si elle releve
d'un acte universel comme dans le cas des voyelles ((claires)) et
((sombres)), prononcees avec un mouvement de la langue vers
avant (vers la lumiere) ou vers l'arriere (vers le noir); elle reste
particuliere, si elle se rattache a un evenement particulier
survenu dans les premieres annees de la vie d'un individu,
comme dans le cas de l'attribution d'une couleur determinee a
une voyelle, une lettre ou a un jour de la semaine (cf. Pfister
1912, Hug-Hellmuth 1912). Dans les deux cas, on doit chercher
la solution au niveau social et non pas physiologique (Fonagy
1963). Un simple renvoi au concept de ((sensory logic)) me paralt
egalement insuffisant dans le cas des ((crossovers)) metaphoriques
de l'inanime a l'anime, de l'animal a l'etre humain. Ces transferts
representent un retour a un stade determine de l'evolution
mentale que Ferenczi (1927 [1913]) qualifie comnme ((projective)).
En general, la description du processus metaphorique pourrait
se faire d'une maniere moins metaphorique, plus precise en le
situant dans les cadres de reference de la psychanalyse (cf.
encore Jones 1961 [1916], Sharpe 1940, Kris 1950, Kubie 1953).
C'est l'absence de cet aspect dynamique que je deplore surtout
dans l'analyse de Beck. I1 n'apparait nulle part quelle est
l'importance qu'elle attribue a la composante inconsciente dans
la genese de la metaphore.
J'irai plus loin dans nvaputo ae l'importance du processus
metaphorique dans la communication verbale que l'auteur.
L'histoire de la terminologie scientifique nous montre claire-
ment que la formation d'un nouveau concept se fait I l'aide
d'une metaphore, c'est a dire, par un retour a la source qui
est l'ideation preconsciente, preverbale (Eucken 1880; cf. en-
core la bibliographie raisonnee de Shibles 1971:405). Ainsi, par
exemple, les metaphores des grammairiens grecs, latins, occi-
dentaux et orientaux, telles que voyelles ((claires)) et ((sombres
consonnes ((lisses)) et ((rugeuses)), ((fortes)) et ((faibles)) ((viriles))
et ((effemiflees)), ((mouillees)) et ((seches)), supposent une analyse
prlconsciente mais minutieuse des mouvements articulatoires,
et refletent, d'autre part, des phantasmes inconscients qui s'y
rattachent (Fonagy 1963, 1970).
 Selon l'auteur seul le paradoxe rappelle l'ideation primitive,
((sensorielle)) dans la communication verbale. Or, la structure
semantique des signes verbaux, ainsi la polysemie des lexemes
(cf. Contras 1972) reunissant des concepts distincts et dis-
parates (au niveau de la mentalite rationnelle, consciente)-
((tete)) s'applique a la partie superieure de l'epingle ou de l'arbre,
a la personne qui dirige une organisation-porte les traces de
l'ideation metaphorique. Ceci est egalement vrai des categories
grammaticales, p.e. dans le cas de la structure semantique des
constructions possessives du frangais, de l'anglais, de 1'allemand
ou du hongrois qui englobe des rapports logiques aussi dis-
parates que celui de la possession d'un objet, du rapport paren-
tal, de la partie et de l'ensemble, de l'action et de l'actant, ou
du rapport causal, congus neanmoins comme des aspects dif-
ferents de la meme entite semantique, a partir de leur identite
au niveau du processus primaire (Fonagy 1975). La concep-
tualisation d'un nouveau rapport logique exige egalement un
retour a l'ideation paleologique; la stratification semantique
des structures grammaticales conserve les traces de ces retours
m'taphoriques. Tout en opposant l'activite mentale preverbale
a l'activite verbale proprement dite, il faut voir egalement que
les procedes preverbaux s'integrent au systeme verbal et lui
pretent une souplesse, une vivacite qu'ignorent les ((langues))
artificielles.
[I fully agree with the author as to the essence of the ideas she
presents with such clarity and style. The meta-phor, as the
word indicates, is the dynamic sign par excellence (cf. Quintilian
Institutio oratoria 2.13.8ff), opposed to stable signs and sum-
marizing through its movement the ontogenesis and phylo-
genesis of conceptual thought from preverbal thought. I have
tried to outline this evolution in previous publications (1961,
1963, 1965, 1971, 1972, 1973).
I would like, however, to point out some differences between
our conceptions. I fear that the fine metaphoric terms "sensory
logic" and "sensory thought," used in the article to designate
the mental process subtending the metaphor, may be mislead-
ing. These terms seem to suggest that metaphoric thought is
closely linked to the sensory level and that its peculiarities, the
"errors" inherent in metaphoric expression, are explained by
this limitation. In my opinion, what distinguishes metaphoric
ideation from conceptual ideation is a different elaboration of
the same sensory data. The metaphor, as voluntary error, de-
pends on a more primitive analysis of these data. Freud desig-
nates the mental processes conveying dreams, myths, and tropes
by the term "primary process" and contrasts them with "sec-
ondary process," the mechanisms that constitute rational
thought, liable to give a more objective image of reality (Freud
1940-46[1900]:593-614). The primary process constitutes the
nucleus of mental structure and corresponds to a childish,
archaic world view dominated by the pleasure principle.
The author sometimes allows herself to be carried away by
her own metaphors (such as "sensory logic"). Thus, she con-
cludes, in relation to binary conceptual contrasts, that "these
contrasts cannot be made at the level of sensory input." It is
obvious that no ideation-conceptual or metaphoric-can take
place at the sensory level. I also do not believe that synaesthesia
can be explained at the physiological, sensory level. The author,
going back to a thesis from Herder (Abbhandlung iuber den
Ursprung der Sprache, 1770) that assumes the existence of a
universal sense, argues for a "profound unisensate base" in-
corporating all the senses. In fact, synaesthesia is a complex
psychological phenomenon: a concrete event, preconscious or
unconscious, with which different sensations are associated al-
lows the transition from one sensory sphere to another. The
synaesthetic metaphor is collective when it depends on a uni-
versal act, as is the case with the "light" and "dark" vowels,
pronounced by moving the tongue front (towards the light) or
back (towards the darkness); it remains idiosyncratic when it
is related to a particular event in the early years of the life of
an individual, as is the case with the attribution of a specific
Vol. 19 * No. 1 *
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
Beck: METAPHOR AS MEDIATOR BETWEEN THOUGHT MODES
colour to a vowel, a letter, or a weekday (cf. Pfister 1912, Hug-
Hellmuth 1912). In both cases, the answer must be sought at
the social and not the physiological level (Fonagy 1963). Also,
a mere reference to the concept of sensory logic seems to me
inadequate in the case of metaphoric "crossovers" from inani-
mate to animate, from animal to human being. These transfers
represent a return to a specific stage of mental evolution that
Ferenczi (1927[1913]) calls "projective." Generally, the meta-
phoric process could be described in a less metaphoric, more pre-
cise way by placing it within the framework of psychoanalysis
(cf. Jones 1961[1916], Sharpe 1940, Kris 1950, Kubie 1953). It is
the absence of this dynamic aspect that I especially lament in
Beck's analysis. Nowhere does she indicate what importance
she attributes to the unconscious component in the genesis of
the metaphor.
In the evaluation of the importance of the metaphoric process
in verbal communication, I would go farther than the author.
The history of scientific terminology shows us clearly that the
creation of a new concept is accomplished through metaphor,
that is, through a return to the source, which is preconscious,
preverbal ideation (Eucken 1800; cf. Shibles 1971:405). Thus,
for example, the metaphors of the Greek, Latin, Western, and
Eastern grammarians, such as "light" and "dark" vowels and
"smooth" and "rough," "strong" and "weak," "virile" and
"effeminate," and "wet" and "dry" consonants, presuppose a
preconscious but thorough analysis of the articulatory move-
ments and at the same time reflect unconscious mental images
associated with them (Fonagy 1963, 1970).
According to the author, only the paradox recalls primitive
and "sensory" ideation in verbal communication. Now, the
semantic structure of verbal signs, and thus the polysemy of
lexemes (cf. Contras 1972) combining distinct and dissimilar
concepts (on the level of rational, conscious mentality)-e.g.,
"head" applies to the upper end of a pin or the top of a tree,
to the person who directs an organization-bears traces of
metaphoric ideation. This is also the case with grammatical
categories, for example, the semantic structure of possessive
constructions in French, English, German, or Polish, which
include logical relations as dissimilar as possession of an object,
parent-child, part-whole, action-actor, and cause-effect, never-
theless understood as different aspects of the same semantic
entity because of their identity at the primary-process level
(Fonagy 1975). The conceptualization of a new logical relation-
ship also requires a return to paleological ideation; the semantic
stratification of grammatical structures preserves traces of these
metaphoric returns. In contrasting preverbal mental activity
with verbal activity in the strict sense, we must also see that
preverbal processes combine with the verbal system and give
it a flexibility and a liveliness that are unknown to artificial
"languages."]
by YALgIN IZBUL
Hacettepe Universitesi, Sosyal Antropoloji Biliimil, Ankara,
Turkey. 9 ix 77
Particular care has been taken in this paper, justifiably, to ex-
plain how useful the observation of commonly used metaphors
can be for an anthropological study of a given culture. I thor-
oughly agree on the points made by Beck in this connection,
with the exception that I believe there exist less laborious
means of inspecting sociocultural change. Language plays a
rather conservative role in matters relating to social and cul-
tural change. It would be a lot more satisfactory and less time-
consuming to go by the more usual methods of anthropology,
aided by the simpler language of statistics. Therefore, I would
propose that the study of the metaphor be reserved rather for
an overview of the mythical and ethical foundations of a culture.
Although we occasionally come across what is generally
judged to be a "fresh" metaphor (which is probably closer to
March 1978 91
 what Beck sees as a contemporaneous feat of the consciousness
in coming to an understanding of changing reality), metaphors
are more often than not a part of the institutionalized body of
language constructs. A further difficulty would be in determin-
ing how commonly used a particular expression is, even though
we may have heard it in the speech of a number of individuals.
Obviously, one has to guard against peculiarities of idiolect.
We must make certain that the user is trustworthy as to his
interpretative outlook upon the social scene. But in any case,
by the time an expression is part of general communication, it
is no longer "fresh," but already a point of reference constructed
in the past. Do not the old guardians of tradition sit back and
pronounce meaningful metaphors when confronted with the
vigorous young world-movers busy around them? A metaphor
is a figure of speech, then, which tests the perceived "reality"
against the accumulated body of assumptions of a people,
which is, in the final analysis, culture itself.
Congratulations are due to Beck for having provided a point-
er to another aspect of the value of linguistic alertness in carry-
ing out an anthropological study.
by STEPHEN C. LEVINSON
Department of Linguistics, University of Cambridge, Cambridge,
England. 15 ix 77
That metaphor is the key to an understanding of ritual I take
to be uncontroversial. Yet the actual mechanisms of metaphor
have received little analytical scrutiny from anthropologists
(despite their ritual attention to ritual), who rather trade on
our intuitive understanding of a universal property of human
symbolic systems. In this context Beck's essay is very welcome
indeed.
Perhaps the major thrust of Beck's comments is that the
structuralist analysis of metaphor (a la Levi-Strauss and Leach)
stresses the cognitive at the expense of the affective elements
that are also involved. (This is nicely exemplified by her discus-
sion of the phrase "to be the apple of one's eye.") She goes on
to criticize the structuralist dogmas that without contrast there
is no semantic content and that in principle ritual expressions
encode one single, central, unparadoxical message. In contrast,
she suggests that there is a level of thought which operates in
terms of affective associations, unstable and in part idiosyn-
cratic, and that metaphor plunges through the rigid categories
of verbal thought to tap riches at this other level.
In many ways I am sympathetic with this line (which, by the
way, seems quite consonant with Victor Turner's style of
analysis): loose things like emotions and associations get short
shrift in structuralist analysis especially. One immediate re-
sponse to this thesis, however, is that it may be true, but inso-
far as it is true, it is unstudyable: what we cannot catch in the
rigid categories of verbal thought we cannot catch at all.
Shouldn't we therefore stick to our analysis of the cognitive
aspects of metaphor, where at least we can make relatively pre-
cise observations? The answer, I suppose, is only if we are
sticklers for precision.
Beck's remarks seem to me weakened by a failure to make a
number of distinctions. In the first place, it is useful to dis-
tinguish different kinds of analogical expressions, as indeed is
done in traditional rhetoric. For instance, there seem to be a
number of important differences between true metaphors like
(1) "Nixon is a fish."
and proverbs like
(2) "A bird in the hand is worth two in the bush."
Both require inferential work on the part of the hearer, but
both the clue that indicates that there is an inference to be
made and the inferential mechanism itself are different in each
case. In (1) the inference-trigger is the blatant falsehood of the
proposition, while in (2) it is (presumably in context) the literal
irrelevance of the remark that leads the hearer to search beyond
the literal meaning. The mechanism in (1) is this: since the
proposition is obviously false, the only way in which I can pre-
92 CURRENT ANTHROPOLOGY
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
serve the working assumption that the speaker is conveying
something true is to presume that it is not the defining charac-
teristics, but the incidental characteristics, that are being con-
veyed (see Grice 1975); so I infer that Nixon is slippery, slimy,
scaly, or swims well. The mechanism in (2) is quite different
and much more involved: to preserve the working assumption
that speakers' utterances are relevant, I must assume that there
is some analogy between the situation being talked about and
the utterance-specifically, I must pair the subject of the
proverb with the topic of the conversation (say, my laments
over the size of my research grant) and then find a predicate
that matches the predicate in the proverb (say, "is better than
a large grant unobtained").
My point here is that the mechanisms in each case are quite
precise, and rather different, and I suspect things may be more
blurred than clarified by subsuming varied phenomena under
the single rubric of metaphor.
Another important distinction that Beck does not emphasize
is the distinction between the study of the mechanisms of meta-
phor (as above) and the study of the use to which metaphors
are put. These seem to me to be independent enterprises, and
it is the latter which is likely to bear the most anthropological
fruit. Beck's remarks about the social functions of metaphor
seem almost parenthetical, but provocative. She suggests, I
think, that they are especially used to patch up rents in the
social firmament, to aid the processes of "mystification" (as the
jargon goes), or the masking of change. But this does not jibe
well with the fact that only some cultures are full of proverbs,
and those seem to be ones associated with traditional and rela-
tively static societies. On the other hand, if one looks in recorded
conversations for the motives behind the introduction of meta-
phors in speech, one finds that, like ironies, they are typically
used to make critical remarks or points that contravene social
decorum (see Brown and Levinson 1978). What are euphe-
misms, after all? And why do parables and heresies go together?
One use, at least, of analogical allusion is not to patch up the
social structure, but to tear it down right under the censor's
eyes.
This raises a final but important point. Beck suggests that a
way to study a culture's cosmology is to study its metaphors.
But metaphors are not so easy to interpret, or perhaps too easy:
there is always a great range of possible interpretations (hence
the evasion of the censor). Take my example (1) above: was it
Nixon's slipperiness, scaliness, or swimming that I had in
mind? I'm not saying.
by FRANKLIN LOVELAND
Gettysburg College, Gettysburg, Pa. 17325, U.S.A. 18 viii 77
I thoroughly enjoyed Beck's paper on metaphor but question
whether she has demonstrated all that she set out to. It seems
to me that she attempts to integrate two major theoretical
traditions in the anthropological analysis of metaphor: the
structural school of Levi-Strauss and Leach and the movement
school of Fernandez. While she reconciles these two traditions
theoretically, I am not convinced she has reconciled them
methodologically. Granted that metaphors are mediators be-
tween "partial and abstract principles on a verbal plane and
concrete, sensual holistic images that thrive on a nonverbal
one," it is difficult to decide how to analyze them. Beck seems
to come down clearly on the movement side of the question in
her example of the "egg-to-bird" process by suggesting that
metaphors go beyond the rational bounds of experience "by
recourse to basic sensory experience." However, she seems to
overlook structural interpretations of this metaphorical process
which would give a more complete interpretation of such a
metaphor.
While she talks about showing us how metaphors help cul-
tures adapt to changing realities, she has not really given us
an example of this. In my own work among the Rama Indians
 of Nicaragua, I have come across at least one example of sym-
bols changing as a culture adapts to a changing reality. The
manatee has become a symbol for culture in the abstract sense
among the Rama, since it is capable of silence as humans are
and its large quantity of meat will provide food for an entire
community (Loveland 1976). I believe the relatively recent as-
signation of key-symbol status to the manatee is the result of
the Rama's symbolically adapting to a new coastal lagoon en-
vironment as they migrated to Rama Cay from the tropical
forest-riverine area of Punta Gorda River in Nicaragua. In the
past, jaguar and peccary were key symbols in Rama culture;
today the manatee and jaguar serve this function.
by WILLIAM C. MCCORMACK
Department of Anthropology, University of Calgary, Calgary,
Alta., Canada T2N 1N4. 13 ix 77
Beck's structuralist teleological model for cultural thought fur-
thers the attention to metaphor initiated in CA by Fernandez
(1974). In part it seems to respond to psychiatrist George Hay-
du's comment there (p. 136) that "Fernandez does not discuss
why and when a metaphor is powerful. This has something to
do with the experience form that it attempts to convey...."
Thus Fernandez's sociogenic process for metaphor has been
here extended, if not transformed and transcended, by psycho-
genic considerations. For example, riddles are explained as a
humorous use of sensory logic for verbal categories, instead of
as verbal play with a linguistic "literalization" rule, as verbal
play with discourse-orienting presuppositions related to shal-
low (sensory) memory (cf. Chafe 1974), or, sociofunctionally,
by reference to their association, as a favored communicative
style, with the state (Roberts and Forman 1971:519). Again,
it is unconventional for Beck to consider individual culture
as embodied in the sensory modality, which "remains highly
individualized throughout life in comparison with semantic
codes."
In taking Jakobson's metaphor/metonymy opposition back
to a neurophysiological source to show that "the metaphor and
the metonym are two sides of a single device [culture?] used to
bridge the gap between rational and sensory thought processes,"
Beck challenges Levi-Strauss and Leach, both of whom ac-
cepted Jakobson's formulation. Once again, it appears that
advances in anthropology depend as much on interdisciplinary
contact as on empirical hypothesis-testing (Piaget 1973). In-
deed, a crucial test-case for Beck's framework may lie in its
adequacy for explaining changes in scientific culture, in view of
arguments (Andersen 1973) that abduction and deduction,
logical models for scientific inference, explain language acquisi-
tion and language change. In this, contact with psychologist
Charles Osgood, whose work on verbal semantics was credited
in Fernandez's article and whose approach to metaphor as
synesthesia (Osgood 1959) anticipates this study by Beck,
should provide methodological clarifications as to how to make
the intuited role of metaphor in cultural process more explicit
as scientific explanation.
by ROBERT A. RANDALL
Department of Anthropology, University of California, River-
side, Calif. 92521, U.S.A. 22 viii 77
I have heard it said that generative linguists know little more
about language than anthropologists know about culture, but
linguists have a much more efficient technology for detecting
wrong ideas. Beck's review summarizes much recent work on
metaphor and contains several valuable insights, but it also
inadvertently makes obvious how little we really know about
metaphoric behavior. In my opinion, Beck does not sufficiently
stress the rudimentary nature of knowledge in this area, and
so invites criticism for drawing conclusions with too little
Vol. 19 * No. 1 * March 1978
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
Beck: METAPHOR AS MEDIATOR BETWEEN THOUGHT MODES
Where, for example, are studies that tell us in a reasonably
explicit manner what a people must know to invent and use
metaphor "to adapt to changing realities" and "to overcome
the confrontation of first principles which a radical social or
environmental change can produce"? Who has shown how meta-
phor interacts with (substitutes for?) other methods of over-
coming a conflict of first principles such as argument, inferential
problem-solving, cognitive-dissonance balancing (and other
types of conformity to peer pressure), folk pragmatism, obedi-
ence to authority, and warfare? Without such studies, claims
such as those made by Beck should be labeled hypotheses.
Is metaphoric behavior really "a fundamental aspect of the
communication process," "a fundamental data set for anthro-
pological analysis," "a primary means by which we overcome
the confrontation of first principles," or just a consequence of
something more fundamental (such as the propensity to cate-
gorize similarities and differences)? If "we must come to recog-
nize . . . metaphor as . .. a special kind of preverbal logic of
experience which substitutes for set theory," how can we simul-
taneously recognize metaphor as "a domain"? Is not a domain
some sort of set?
It seems necessary to have several systematic explanations
of metaphor invention and use before generalizations of the
sort Beck proposes can be reliably related to data and relevant
theory. Without such studies, we will continue to have diffi-
culty unravelling the cognitive, linguistic, social, situational,
and historical complexities that obviously govern metaphoric
behavior.
Because Beck has chosen to deemphasize the absence of
authoritative studies, it is difficult to evaluate her assumptions
and conclusions. Is, for example, the supposed fundamental role
of metaphor in communication a proposition in some cognitive
and/or linguistic theory (such as Chomsky's belief in a univer-
sal grammar), or is it an observation whose cross-cultural
validity is still to be determined, an empirical universal so far
ignored by orthodox psychologists and linguists, a persuasive
speculation by other scholars, or merely a hunch we should use
as a guide in further research? As a literal assertion about cogni-
tion, what, for example, does it mean to say metaphor "medi-
ates" between "levels of thought"? Before such a figurative
proposition can be used and tested, it must, it seems to me, be
given a literal cognitive interpretation.
At this stage of knowledge, then, I think we should be cau-
tious while we try to determine which problems should be
studied. Should we be investigating how metaphor is used to
ignore or even mask change? Can we discover whether the con-
tent of metaphors used to adapt to radical change is different
from the content of those used in ordinary life? Would it be
useful to study how metaphoric content is selected, or would
we expand our horizons more with studies of metaphoric form
and function, of the reasoning underlying adaptation, or of
some other aspect of communicative adaptation? Given our
gross lack of knowledge along any of these lines, it seems prema-
ture to make claims for one approach without a comprehensive
evaluation of the alternatives.
In sum, then, Beck makes several claims about metaphor
that deserve attention, but she fails to communicate what we
know, what we probably know, what we do not know, and what
we ought to know about metaphor. If she were to deal with
these matters straightforwardly, I rather think we would-to
mix metaphors-be well on the road to finding the right path.
by J. DAVID SAPIR and J. CHRISTOPHER CROCKER
Department of Anthropology, University of Virginia, Charlottes-
ville, Va. 22903, U.S.A. 21 ix 77
Beck's article evokes both admiration and many thoughts for
comment. We admire especially her lucid presentation of a
topic so often treated with mystical imprecision. We restrict
our comments to three.
1. Beck postulates the existence of a "preverbal level of hu-
man thought processes" and argues that rhetorical language
"moves" between this level and one ordered bv the logic of
verbal categories. It seems to us that such a position is highly
debatable and probably unnecessary; it asserts yet again the old
mind-body distinction which so much recent work on symbol-
ism successfully ignores if not explicitly denies. We agree that
the often implied analogy metonymy:metaphor::felt:thought is
logically unsound and ethnographically incorrect. We believe,
however, hence disagreeing with Beck, that the socio-analytical
strategies and logics of these two tropes are so very different
that one cannot yoke them as "parallel" in bridging the gap
between "felt" preverbal domains and some opposed realm of
socialized thought. Even if these tropes may and at times in-
deed do accomplish such a movement (in logically unlike ways),
this does not seem to be their most characteristic function, let
alone the phenomenological foundation for their existence.
"To kill two birds with one stone" may be concrete (all the
more if the projectile was man-made), but how many individuals
have sensorily experienced such a coup? The "proverber" may
employ this to convey the utility of some abstract economic
notion, but he may apply it as well, if, say, he is an overweight
househusband, to argue the desirability of jogging in place
while washing the dishes. In John Donne's expression "No man
is an island," the sensual thing, an island, defines man's abstract
relation to his God and his fellow men. But, at least in the on-
tology of Donne's poetry, before there was God there were
women: "O my America, my new-found-land,/My kingdome,
safeliest when with one man man'd" (so he defines his mistress
and their relationship). In using America to describe his mis-
tress, Donne is hardly defining the inchoately unknown; rather,
he seeks to make a mystery out of what is surely all too well
known, or knowable.
The delight and virtue of metaphor is just that it is open-
ended: by crisscrossing all manner of domains, each at once
sensed and semantic, it does not just "move," but focuses,
defines, and reveals. We find useful Kenneth Burke's summa-
tion: metaphor entitles complex relationships by advancing the
perspectives of incongruity (Crocker 1977). But Burke derived
his understanding of rhetoric from the tradition of Western high
literature and drama, not the (recorded) commonplaces of
ethnography. This points to a moral: before pronouncing on
the nature of figurative devices, it might be well to scrutinize
their use in our own cultural tradition. Thus we all might come
to a better appreciation of their subtlety and flexibility-in an
abstract word, their art.
2. There is a common contrast, really a continuum, that
must be kept in mind when we analyze how people perceive
metaphoric relations: that between is, implying identity, and
is like, implying similarity only. Levi-Strauss chooses to ignore
the distinction by stating that since the verb etre, with which
we translate these relations, is itself equivocal, "meme chez
nous," it would do well to settle at some mid-point between
Levy-Bruhl's participations (identity) and Durkheim's totemic
emblems (not even similarity, but only simple juxtaposition)
when formulating an analogic or metaphoric calculus (Levi-
Strauss 1962:133-34). But there are important distinctions to
be made, for example, between Catholic and most Protestant
interpretations of the transubstantiation, the former postulat-
ing a mystical identity (the bread is the body, the wine is the
blood), the latter only a similarity of a quasi-literary sort. It is
a distinction separating much of what we might want to call
religious from what is aesthetic, or poetic. A simple example:
The Kujamaat Diola of Senegal identify by way of common
soul and blood an individual with his personal totemic double,
which is a common wild animal. We may safely refer to this
representation as a feature of Kujamaat cosmology. In contrast,
these same Kujamaat reverse the terms (person to represent
94 CURRENT ANTHROPOLOGY
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
animal) when they designate a red finch by the kin term "out-
married female agnate." Here the bird is like (kankan nen) and
certainly not identical with (yo emi) a female agnate. The
grounds for the relationship are that they both supervise the
in-married wives, the female agnates jurally, the birds distally
when they congregate on the outskirts of the cooking area.
With both the doubles and the red finches, the formal relation-
ship between the terms is the same (though the direction
changed). However, the two relationships are radically different
when we place them within the context of Kujamaat culture
and beliefs. One is a profound statement about man in general,
while the other is at best a pleasant conceit and is taken by the
Kujamaat as such (Sapir 1977a: esp. n. 8). This is a difference
which it would be foolish to ignore.
3. We have found it convenient to make another distinction
(Sapir 1977b, Crocker 1977) that, unlike is/is like, has to do
with how metaphoric relationships are established rather than
how they are perceived. The distinction develops from two of
Aristotle's types of metaphor: genus substituting for genus and
analogy (chap. 21). The former, which we call internal metaphor,
establishes the relationship between terms on the grounds of
shared features. The latter, external (or analogic) metaphor, de-
fines the relationship by way of the place each term holds
vis-a-vis its respective domain. No shared features are necessari-
ly postulated. (Cf. K6ngas Maranda 1971c for a parallel state-
ment.) Generally, internal metaphors are part-icularistic, color-
ful, and unidirectional, while external metaphors are more gen-
eral, abstract, and systematic. In the extreme, external meta-
phors are entirely reversible. Needless to say, a metaphoric text
(two terms in juxtaposition) can be read either way: George is
like a lion because he shares with lions their courage, etc., or he
is like a lion because he occupies the same place among men as
a lion occupies among animals. Moreover, one type may de-
velop into the other. For example, in the metaphoric text
doves = peace and peaceful people, the two domains share
such qualities as passivity and calmness. Here we are using
doves to talk about people, not the reverse. To doves we add
hawks to get doves :hawks: :peaceful warlike people. An in-
ternal relation-peace to doves, aggression to hawks-is pres-
ent, but at the same time an overall external (analogic) meta-
phor is established based on the abstract criterion nonkiller/
killer. Adding now vultures and chickens produces a paradigm
of birds based on the criteria killer/nonkiller and independence
(hawks and doves)/dependence (vultures and chickens). Here
is an abstract set in which it becomes increasingly difficult to
establish internal relationships matching each kind of bird
with appropriate people. We are now in the realm of totemism.
But we can go further. Disregarding people and comparing
instead mammals with birds, we have a fully reversible set of
terms in which birds define mammals at the very same time
that mammals define birds-hawk :dove :vulture :chicken: :lion:
hare:hyena:mouse. From dove = peaceful people we have
moved to an abstract logical statement comparing birds with
mammals. The distinction between internal and external meta-
phors charts the course.
Reply
by BRENDA E. F. BECK
Vancouver, B.C., Canada. 19 x 77
I wish, first of all, to thank my commentators for their thought-
ful and generally positive responses to my article. Reading over
the points they have raised has provided an opportunity to
extend my own thinking on various central issues. In my answer
I shall attempt to clarify my argument in a number of ways.
Edmonson, my sharpest critic, asserts that the construction
of metaphors is impossible without language and therefore that
 my central suggestion (that human metaphors generally serve
to link semantic and nonsemantic thought processes) is wrong.
Currently available evidence on this issue clearly supports my
case and not his. For example, the chimpanzee Washoe is re-
ported to have sometimes "thought" in metaphor: One morn-
ing Washoe's human companions tried unsuccessfully to give
her an injection. Later that same morning she was observed
sticking the inside of her thigh with a long nail she had found,
using gestures similar to those earlier used by her companions
with the hypodermic needle. Rumbaugh (1977:117) writes:
"This episode is apparently a remarkable instance of a chim-
panzee engaging in an activity that would suggest both a repre-
sentation of the initial shot-giving episode and the metaphorical
use of a nail." Other recent work uniformly substantiates this
conclusion, drawing on other studies of primate and neonate
thought patterns (Ettlinger 1977:98; Meltzoff and Moore
1977:78).
Fonagy comments, in a similar vein, that metaphors must
be rooted in social learning rather than in sensate experience
per se. I would agree here, at least in the sense that the anthro-
pologist's interest in metaphor lies in its social and cultural
dimensions and not in its biological roots, but this comment
allows me to clarify what I believe to be a central issue. The
mental structures which allow for metaphor formation seem to
be present in a number of nonhuman primates. Their strength
apparently varies as a direct function of the degree to which
the species has a hominoid type of brain structure. General
primate abilities have now been shown to include symbolizing,
relating separately learned concepts, and synthesizing learned
skills for application to novel problems (Rumbaugh 1977:107).
I am well aware of the fact that, because of a great influx of
new research findings, no one definition of language finds wide
acceptance at the moment. Nonetheless, my view is that we
humans have a complex meta-code which functions to process
and communicate primary experience with great efficiency. Hu-
man languages are said to be unique primarily because of their
"openness," that is, because of the complex set of totally arbi-
trary relationships they support between lexical units and
meanings. I agree with Fonagy that as anthropologists we are
mainly interested in the social rather than the biological aspects
of metaphor, but not with Edmonson that this phenomenon
necessarily rests on predeveloped semantic structures.
Fernandez suggests that I might better contrast analogue and
digital (rather than analogic and semantic) thought processes.
I disagree. To my mind the key lies precisely in distinguishing
language-based reasoning from other, nonlinguistic mental
processes. For"example, there is evidence that these latter pro-
cesses are important in scientific engineering (Ferguson 1977),
and Einstein (quoted by Ferguson, p. 843) was quite insistent
that he thought largely in visual and muscular codes and ex-
perienced some difficulty in translating these insights into a
formal language. One of the great potential advantages of the
study of metaphor derives precisely from these links with non-
verbal processes: the concreteness of metaphoric habits makes
their translation from language to language relatively easy.
Indeed, it is particularly important that we ask, in future re-
search on metaphor, exactly to what degree culturally specific
habits of constructing concrete analogies persist when a person
moves. to new culture and becomes fluent in a new language.
It is clear, on the basis of present evidence, that motor and
spatial experience are of great importance in nonverbal reason-
ing. I cannot agree with Douglas, however, that we ought to
separate these skills from those of taste and smell. Humans
and animals alike are capable of learning to make fine discrimi-
nations within the smell and taste categories. We also follow
the smell of smoke through space with our motor movements,
just as we quickly learn if some portion of an apple is rotten by
tasting and biting simultaneously. The two types of sensation
are intimately interconnected. My point regarding synaesthesia
was precisely that all our sense inputs become merged in struc-
Vol. 19 * No. 1 *
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
Beck: METAPHOR AS MEDIATOR BETWEEN THOUGHT MODES
tures that allow for transfers independent of verbal reasoning
processes.
Fonagy suggests that we refer to the sensory as a "primary"
process and the semantic as "secondary." I am not opposed
to his terminology, but I do find "sensory" a superior term,
simply because it tells us more about the nature of such thought.
By sensory thought, to be still more specific, I mean those men-
tal activities of imitation and recombination which recreate
sensory experience in sense-based mental representations. Such
a formulation lies in a direct line with recent research findings
on neonates, for example, showing that even two-week-old in-
fants can construct abstract representations of perceptually
absent stimuli (Meltzoff and Moore 1977). (Lest it be thought
that two weeks is sufficient to "learn" such abstract and high-
order reasoning, we note that one subject of this research was
only sixty minutes old when observed!) These researchers con-
clude, then, that humans have the capacity to represent visual-
ly and proprioceptively perceived information in a form com-
mon to both modalities from the moment of birth (p. 78). Adult
metaphor construction would seem to build on this innate, non-
semantic ability, serving to link it to semantic codes.
I have been taken to task by Douglas for drawing too sharp
a line between such sensory reasoning processes, on the one
hand, and semantic thought patterns, on the other. Loveland,
in a similar vein, comments that I have not succeeded in recon-
ciling the movement and structural schools of metaphor. I con-
tend that the distinction is needed to clarify how metaphors
function. The point of analyzing reality is to abstract various
aspects of a process in order to promote an understanding of its
dynamics. Fernandez's suggestion that I use the cover term
"trope" in discussions of metaphor strikes me as inappropriate
for this very reason. "Trope" is a term used in rhetoric to de-
scribe figures of speech. Use of it would therefore link my
analysis firmly to language-based reasoning processes, precisely
the implication I wish to avoid.
I also disagree with Fernandez when he suggests (following
Levi-Strauss) that sense reasoning is logically impoverished
reasoning. The processes required to match information per-
ceived in several modalities, received at different times, and to
postulate solutions to new problems on the basis of such corre-
spondences (a general hominoid ability) surely contain as much
"logic" as word-based reasoning. Monkeys, for example,
regularly solve complex problems presented to them by re-
searchers without the help of language.
Edmonson argues that a metaphoric relation need contain no
sensory elements at all. By my definition, such analogies as
raw/cooked (for nature/culture) are not true metaphors, but
merely sets of paired abstractions. His critique is intended,
clearly, to refer to the work of Levi-Strauss. Levi-Strauss, how-
ever, insists, in all of his mythological researches, that the
reasoning processes he is studying rest on concrete images (like
rotting wood or warming fire); it is the anthropologist who has
abstracted a raw/cooked contrast from these materials. In tak-
ing such an example, then, Edmonson seems to miss Levi-
Strauss's central point. At the same time, I cannot accept a cri-
tique based on the redefinition of my subject matter. Douglas's
comment that metaphors do not deal with immediate sense ex-
perience, however, I do accept. Of course, metaphors deal, not
with immediate experience, but with sense images remembered
(and recombined) in the mind's eye.
Another group of comments concerns my failure to map out
specific research goals, on the one hand, and to link my ideas
to general theory, on the other. I agree that these are weak-
nesses. The lack of knowledge about metaphors that Randall
rightly points out led me to proceed with caution. Grand theo-
ries, on the one hand, and critical research observations, on
the other, do not spring from a vacuum. Although I am in sym-
pathy with Randall's goals, I do not think we have yet reached
March 1978 95
the point where propositions about metaphor can be stated in
a testable form. The general body of propositions with which
my essay is linked is called set-formation learning theory. This
theory stems from a substantial body of experimental findings
(mostly on monkeys) about how rapidly and accurately a set
of discriminations made in one mode is transferred to new prob-
lems presented in another mode (Medin 1977:51). Such experi-
ments have not been language-dependent.
Edmonson is right that binary discriminations can be and
are sometimes made using sense-based logic alone. Those dis-
tinctions we attend to on the sensory level, however, are dic-
tated by the arbitrary system we have learned for linking lexical
units and meanings. Experiments with chimpanzees, for ex-
ample, show that the learning of discrimination takes place
faster and more accurately when objects are named (Ettlinger
1977:81). The advertising industry has long exploited similar
findings regarding human learning. Of course we are all able
to make binary discriminations on the basis of sensory input
alone, but, given the vast weight of our human cultural codes,
it appears that semantic (and social) factors, rather than our
senses alone, dominate most of the learning we do.
Regarding further research, Fernandez has suggested the
need for participant observation. Fodor has added a request
for ethnological survey work, using specially designed question-
naires that would tap culturally (and linguistically) specific
domains such as colour metaphors. Loveland further suggests
(by implication) that we focus on "key" metaphors that have
a kind of general cosmic significance. I am not opposed to any
of these approaches, but my own thoughts have led me to design
a somewhat different strategy. It would seem that as anthro-
pologists we will most want to know (1) how cultures differ
in their use of metaphor and (2) how these differences affect
a human's adaptation to changing circumstances. I will discuss
each of these issues briefly.
Most of the research to date on metaphor has been directed
towards uncovering the particular logical processes that lie be-
hind this type of thought. The problem with this approach
is that the use of metaphor is a universal human trait. Some
limitations and variations in the verbal forms that metaphor
can take undoubtedly stem from the grammatical and semantic
differences among languages, but, in my view, these constraints
are not of great anthropological interest. Several commentators
(Fernandez, Sapir and Crocker, and Levinson) suggest that I
might have distinguished between the various logical types of
metaphor (structural/textual, internal/external, metonym/
metaphor, simile/metaphor, etc.) that can exist independent of
the particularities of a given language. But the several types of
comparison employed in these various subtypes of metaphor
are just as likely to be universal as the logic of metaphor itself.
The same authors point out that a single metaphor may some-
times have many meanings. This is not surprising, since several
logical possibilities normally exist for comparing any single set
of images. Levinson correctly asks how we are to distinguish
between such meanings. When he says "Nixon is a fish" does
he mean that Nixon is "slippery," "scaly," or what? Levinson
will not tell us, and neither would most metaphor users. There-
fore classifying and counting metaphors by logical type or by
"meaning" is not a practical strategy.
Fodor's survey approach leads us quickly towards other prob-
lems. As Izbul comments, it would be an enormous undertaking
to determine the frequency with which certain metaphors actu-
ally appear in common usage. Further, why should we try to
provide lists of culturally specific metaphors? What difference
does it make if one culture says "green with envy" and another
employs "red" to describe the same emotional state? Levinson's
concern with the contexts in which metaphors are used and
with whether or not euphemism is a critical dimension of such
thought processes only leads us deeper into the swamp.
What we want to know lies between the extremes of precise
logical structures and infinite particularity. We need to deter-
This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms
mine what categories of imagery (body parts, aquatic plants,
celestial movements, mechanical objects, etc.) provide given
cultures with popular metaphor themes. We can use existing
collections of proverbs and riddles to begin generating hypoth-
eses about such culturally specific content (Beck 1978). Such
general findings could then be further tested and refined against
individual informants, as well as by participant observation.
Following this we should ask about the creative generation of
new metaphors in novel contexts. Do new metaphors rest on
the same thematic bases as the old ones?
Izbul's comment about cliche-type metaphors, metaphors
that have become so institutionalized in a language that we
hardly notice them, is important here. Such metaphors are in-
teresting only in the statistical sense that they can help us to
establish a baseline. Only after we have such data can the inter-
esting questions about change be explored. What kinds of meta-
phors do immigrants in a new country use, for example, in de-
scribing their new situation? Do new usages tend to "translate"
traditional themes found in a "home culture" (indicating a con-
tinuity in cultural thought patterns), or do they draw upon
the metaphoric traditions of the culture most readily at hand?
Certainly both processes will occur, but thematic shifts must
take place more rapidly in some domains (perhaps on the job)
than in others (at home?). Again, at what age, etc., do people
adapt to new metaphor themes the fastest? Finally, does the
thematic distance of an original culture from a hewly adopted
one tend to speed or retard such shifts?
It is questions like these which I think are likely to lead us to-
wards a deeper understanding of how culture change occurs.
We first need baseline studies of traditional metaphoric habits.
Following this we need to do some in-depth interviewing of a
sample of persons directly involved in situations of change. I am
embarking on such a project myself and would welcome a note
from any readers who might like to collaborate in such work.
References Cited
ANDERSEN, HENNING. 1973. Abductive and deductive change.
Language 49:765-93. [WCM]
AREWA, E. O., and A. DUNDES. 1964. Proverbs and the ethnography
of speaking folklore. American Anthropologist 66(6, pt. 2) :70-85.
ARISTOTLE. 1968. The poetics. Translated by L. J. Potts. Cambridge:
Cambridge University Press. [JDS, JCC]
BARTH, F. 1975. Ritual and knowledge among the Baktaman of New
Guinea. New Haven: Yale University Press.
BASSO, KEITH H. 1976. "Wise words of the Western Apache," in
Meaning in anthropology. Edited by K. H. Basso and H. Selby.
Albuquerque: University of New Mexico Press. [JWF]
BECK, BRENDA E. F. 1977. The proverb corpus as a key to basic
metaphoric structures: The case of Tamil culture. MS.
BLACKING, J. 1961. The social value of Venda riddles. African Studies
20:1-32.
BROWN, P., and S. C. LEVINSON. 1978. "Universals in language usage:
Politeness phenomena," in Questions and politeness: Strategies in
interaction. Edited by E. Goody, pp. 56-289. Cambridge Papers in
Social Anthropology 9. [SCL]
CHAFE, WALLACE L. 1974. Language and consciousness. Language
50:111-33.
CONTRAS, E. 1972. La structure semantique du lexeme mind 'main'.
Cahiers de Linguistique Th6orique et Appliqu6e 9:71-82. [IF]
CROCKER, J. CHRISTOPHER. 1977. "The social functions of rhetorical
forms," in The social use of metaphor. Edited by J. D. Sapir and
J. C. Crocker, pp. 33-66. Philadelphia: University of Pennsylvania
Press. [JDS, JCC]
CROSSAN, J. D. 1973. Parable as religious and poetic experience.
Journal of Religion 53:330-58.
DAVENPORT, RICHARD K. 1977. "Cross-modal perception: A basis for
language?" in Language learning by a chimpanzee: The Lana project.
Edited by D. M. Rumbaugh, pp. 73-83. New York: Academic
Press.
DOUGLAS, MARY. 1966. Purity and danger. London: Routledge and
Kegan Paul. [JWF]
-. 1973. Natural symbols. New York: Vintage Books.
DURKHEIM, E., and M. MAUSS. 1963. Primitive classification. Chicago:
University of Chicago Press. [JWF]
C CURRENT ANTHROPOLOGY
EMPSON, WILLIAM. 1967. The structure of complex words. Ann Arbor: Beck: METAPHOR AS MEDIATOR BETWEEN THOUGHT MODES
University of Michigan Press. [MD]
ETTLINGER, GEORGE. 1977. "Interactions between sensory modalities
in nonhuman primates," in Behavioral primatology: Advances in LLOYD, G. E. R. 1966. -Polarity and analogy: Two types of argumenta-
tion in early Greek thought. Cambridge: Cambridge University Press.
research and theory, vol. 1. Edited by Allan M. Schrier, pp. 71-104.
LOTMAN, J. M. 1972. Die struktur literarischer texte. Translated from
New York: Lawrence Erlbaum.
the Russian by Rolf-Dietrich Keil. Munchen: Wilhelm Fink.
EUCKEN, R. 1880. Uber Bilder und Gleichnisse in der Philosophie.
LOVELAND, FRANKLIN 0. 1976 "Tapirs and manatees: Cosmological
Leipzig: Veit. [IF]
categories and social process among Rama Indians of eastern
FERENCZI, S. 1927 (1913). "Entwicklungsstufen des Wirklichkeits-
Nicaragua," in Frontier adaptations in lower Central America.
sinnes," in Bausteine zur Psychoanalyse, vol. 1, pp. 62-83. Leipzig-
Edited by Mary Helms and Franklin Loveland, pp. 67-82. Phila-
Wien-Zurich: Psychoanalytischer Verlag. [IF]
delphia: Institute for the Study of Human Issues. [FOL]
FERGUSON, EUGENE S. 1977. The mind's eye: Nonverbal thought in
MARTINDALE, COLIN. 1976. More on split-brain research and anthro-
technology. Science 197:827-36.
pology. CURRENT ANTHROPOLOGY 17:738-39.
FERNANDEZ, J. W. 1974. The mission of metaphor in expressive cul-
MEDIN, DOUGLAS L. 1977. "Information processing and discrimina-
ture. CURRENT ANTHROPOLOGY 15:119-46.
tion learning set," in Behavioral primatology: Advances in research
--. 1975. On the concept of the symbol. CURRENT ANTHROPOLOGY
and theory, vol. 1. Edited by Allan Schrier, pp. 33-69. New York:
16:652-54.
Lawrence Erlbaum.
---. 1977. Poetry in motion: Being moved by amusement,
MELTZOFF, ANDREW N., and M. K. MOORE. 1977. Imitation of facial
mockery, and mortality in the Asturian countryside. New Literary
and manual gestures by human neonates. Science 198:75-78.
History 8:459-83. [JWF]
MILNER, G. B. 1969. L'armature des locutions proverbiales: Essai de
FONAGY, I. 1961. Communication in poetry. Word 17:194-218. taxonomie semantique. L'Homme 9:49-70.
[IF] MUNZ, P. 1976. Review essay: Myth, McLuhan, Levy-Bruhl, and
--. 1963. Die Metaphern in der Phonetik. The Hague:Mouton.
Levi-Strauss. Pacific Viewpoint, pp. 61-74.
[IF]
MURDOCK, G. P. 1957. World ethnographic sample. American Anthro-
--. 1965. Form and function of poetic language. Diogenes 51:
pologist 59:664-87. [IF]
72-110. [IF]
NEEDHAM, RODNEY. 1973. Right and left: Essays on dual symbolic
-. 1971. Double coding in speech. Semiotica 3:189-222. [IF]
classification. Chicago: University of Chicago Press.
1972. Demotivation et remotivation. Poetique 11:414-31.
OGDEN, C. K., and I. A. RICHARDS. 1960. The meaning of meaning:
[IF]
A study of the influence of language upon thought and the science of
--. 1973. Languages in language. Paper prepared for the IXth
symbolism. London: Routledge and Kegan Paul.
International Congress of Anthropological and Ethnological Sci-
ONIANS, RICHARD B. 1954. The origins of European thought about the
ences, Chicago, August-September. [IF]
body, the mind, the soul, the world, time, and fate. Cambridge: Cam-
1975. "La structure semantique des constructions posses-
bridge University Press.
sives," in Langue, discours, soci6t&:Pour Emile Benveniste. Edited by
OSGOOD, CHARLES E. 1959. The cross-cultural generality of visual-
J. Kristeva et al., pp. 38-77. Paris: Seuil. [IF]
verbal synesthetic tendencies. Behavioral Science 5:146-69. [WCM]
FREUD, S. 1940-46 (1900). Gesammelte Werke. Vol. 2/3. London:
Imago. [IF]
PAREDES, J. ANTHONY, and M. J. HEPBURN. 1976. Reply [to Martin-
dale]. CURRENT ANTHROPOLOGY 17:740-42.
GRICE, H. P. 1975. "Logic and conversation," in Syntax and seman-
tics. Vol. 3. Speech acts. Edited by P. Cole and J. L. Morgan, pp.
PEACOCK, J. A. 1968. Rites of modernization. Chicago: University of
Chicago Press.
43-58. New York: Academic Press. [SCL]
GUIRAUD, PIERRE. 1967. Structures etymologiques du lexique fran~ais. PERRIN, N. 1976. Jesus and the language of the kingdom. Philadelphia:
Fortress Press.
Paris: Larousse.
HEIDEGGER, M. 1957. Der satz vom grund. Pfullingen: Neske. PESOT, J. 1975. Un probleme de semiologie linguistique: Mobilite du
HJELMSLEV, LOUIS. 1953. Prolegomena to a theory of language. Interna- signe et monosemie profonde. Canadian Journal of Research in
Semiotics 3:5-15.
tional Journal of American Linguistics Memoir 7. [IF]
PFISTER, 0. 1912. Die Ursachen der Farbenbegleitung bei akustischen
HUG-HELLMUTH, H. 1912. Uber Farbenh6ren. Imago 1:228-64.
Wahnehmungen. Imago 1:265-75. [IF]
[IF]
PIAGET, JEAN. 1973. Main trends in inter-disciplinary research. New
JAKOBSON, ROMAN. 1956. "Two aspects of language and two types of
York: Harper and Row. (First published as Chapter 7 in Main
aphasic disturbance," in Fundamentals of language, by Roman
Jakobson and Morris Halle, pp. 55-82. The Hague: Mouton.
trends of research in the social and human sciences, Pt. 1 [Paris:
Mouton/UNESCO, 1970].) [WCM]
[JWF]
PIAGET, JEAN, and BXRBEL INHELDER. 1967. The child's conception
JONES, D. 1961. (1916). "The theory of symbolism," in Papers on
psychoanalysis, pp. 87-114. Boston: Beacon. [IF]
of space. Translated from the French by F. J. Langdon and J. L.
Lunzer. New York: Norton.
KONGAS MARANDA, E. 1971a. Theory and practice of riddle analysis.
Journal of American Folklore 84:51-61.
RICOEUR, P. 1974. Philosophv and religious language. Journal of
Religion 54:71-85.
-. 197 1b. "A tree grows: Transformations of a riddle metaphor,"
--. 1975. La mgtaphore vivre. Paris: Editions du Seuil.
in Structural models in folklore and transfotrmational essays, by Elli
ROBERTS, JOHN M., and MICHAEL L. FORMAN. 1971. Riddles: Ex-
Kongas Maranda and Pierre Maranda. The Hague: Mouton.
-. 1971c. "The logic of riddles," in Structural analysis of oral pressive models of interrogation. Ethnology 10:509-33. [WCM]
tradition. Edited by E. Kongas Maranda and P. Maranda, pp. RUMBAUGH, DUANE M. 1977. "Language behavior of apes," in Be-
189-234. Philadelphia: University of Pennsylvania Press. [JWF,
havioral primatology: Advances in research and theory, vol. 1. Edited
by Allan Schrier, pp. 105-38. New York: Lawrence Erlbaum.
JDS, JCC]
KRIS, E. 1950. On preconscious mental processes. Psychoanalytic SAPIR, J. DAVID. 1977a. Fecal animals, an example of complementary
Quarterly 19:540-60. [IF] totemism. Man, n.s., 12:1-21. [JDS, JCC]
KRONASSER, HEINZ. 1952. Handbuch der Semasiologie. Heidelberg: --. 1977b. "The anatomy of metaphor," in The social use of
C. Winter. metaphor. Edited by J. D. Sapir and J. C. Crocker, pp. 3-32. Phila-
KUBIE, L. S. 1953. The distortion of the symbolic process in neurosis delphia: University of Pennsylvania Press. [JDS, JCCI
and psychosis. Journal of the American Psychoanalytic Association SAVAGE, E. S., and D. M. RUMBAUGH. 1977. "Communication, lan-
1:59-86. [IF] guage, and Lana: A perspective," in Language learning by a chim-
LEACH, EDMUND. 1964. "Anthropological aspects of language: Animal panzee: The Lana project. Edited by D. M. Rumbaugh, pp. 287-
categories and verbal abuse," in New directions in the study of 307. New York: Academic Press.
language. Edited by E. H. Lenneberg. Cambridge: M.I.T. Press. SHARPE, F. E. 1940. Psycho-physical problems revealed in language:
[JWFI An examination of metaphor. International Journal of Psycho-
--. 1976. Culture and communication. Cambridge: Cambridge analysis 21:201-13. [IF]
University Press. SHIBLES, X. W. A. 1971. Metaphor: An annotated bibliography and
LEVI-STRAUSS, CLAUDE. 1962. Le tot6misme aujourd'hui. Paris: history. Whitewater: The Language Press. [IF]
Presses Universitaires de France. [JDS, JCC] TURNER, V. 1974. Drama, fields, and metaphors: Symbolic action in
1966. The savage mind. Chicago: University of Chicago Press. human society. Ithaca: Cornell University Press.
1970. The raw and the cooked. Translated from the French by ULLMANN, STEPHEN. 1963. The principles of semantics. New York:
John and Doreen Weightman. London: Jonathan Cape. Barnes and Noble.

Vol. 19 * No. 1 * March 1978 97

This content downloaded from 181.45.90.206 on Sun, 03 Apr 2016 14:25:28 UTC
All use subject to http://about.jstor.org/terms