1. modern linguistics as a change of paradigms

Issues in the conceptual theory of metaphor
The sketch of Lakoff’s conceptual theory of metaphor in the precedingsection presents only the major premises of the model. In this section, we willexamine some issues about the conceptual theory of metaphor that have led toelaborations of the basic model.
The first issue we describe is the deceptively simple one of how best to describe
a particular metaphorical mapping. For example, the expressions in (9) illustrate a
metaphor described by Lakoff and Johnson as an argument/theory is abuilding (Lakoff and Johnson 1980:46):
(9) a. We need to construct a strong argument for that.
b. The argument collapsed.
c. We need to buttress the theory with solid arguments.
However, other expressions making reference to buildings do not participate in
the metaphor (Clausner and Croft 1997:260; Grady 1997:270):
(10) a. *Is that the basement of your theory?
b. *That line of reasoning has noplumbing.
c. *This theory has French windows.
The examples in (10) suggest that the metaphor should be formulated moreconcisely, that is, using less schematic source and target domains, in such away that the metaphorical mapping is valid for the concepts in the source andtarget domains. Clausner and Croft propose the more specific formulation the
CONVINCINGNESS OF AN ARGUMENT IS THE STRUCTURAL INTEGRITY OF A BUILDING.Insome cases, the more schematic metaphorical mapping isreplaced by two (or more) distinct specific mappings. Clausner and Croftreformulate Lakoff and Johnson’s LOVE IS A PATIENT metaphor (p. 49; examples [11a–b] and [12c] below) as two more specific metaphors,A SOCIAL RELATIONSHIP IS BODILY HEALTH (11) and A SOCIAL RELATIONSHIP IS LIFE ([12]; Clausner and Croft 1997:261–62):

(11) a. This is a sick relationship.

b. They have a strong, healthy marriage.
c. Their relationship went to the hospital.
(12) a. The marriage is dead – it can’t be revived.
b. Her selfishness killed the relationship.
c. His effort to understand her breathed new life into the marriage.
Clausner and Croft also argue that metaphors vary in productivity. Manymetaphors, such as THE CONVINCINGNESS OF AN ARGUMENT IS THE STRUCTURAL INTEGRITY OF A BUILDING, appear to be completely productive, once formulated at the appropriate level of schematicity. Other metaphorsare only partially productive, in that some expressions are acceptable and othersare not. Examples of partially productive metaphors are the revelation idioms:
(13) spill the beans, let the cat out of the bag, blow the whistle, blow the lid off, looselips
The expressions in (13) all have as a target domain TO REVEAL A SECRET. But
other similar expressions are not conventional in English:
(14) *spill the peas, *let the cat out of the house
One cannot formulate a single metaphorical mapping at any level of schematicity that would include the expressions in (13) but not the expressions in (14). Yetthe expressions in (13) are understood metaphorically. Gibbs and O’Brien (1990)found that the conventional expressions in (13) had coherent mental images associated with them by subjects, while the unacceptable expressions in (14) did not.
The fact that the idioms were ‘imageable’ (Lakoff 1993:211) indicates that there
is a metaphorical mapping present; the fact that the nonconventional yet semanti-
cally similar expressions in (14) are not consistently imageable indicates that the
metaphor is not completely productive. Finally, truly opaque idiomatic expressions
such as kick the bucket and by and large are not interpreted in terms of mapping
from a source domain. Clausner and Croft argue that the usage-based model (see
chapter 11) can be used to model degree of productivity of metaphorical/idiomatic
expressions.
Lakoff and Johnson allow for metaphors to exist at different levels of schematic-
ity, that is, in a taxonomic hierarchy. For example, they posit a schematic metaphor LOVE IS A JOURNEY, and illustrate three more specific instantiations of the
metaphor, LOVE IS A CAR TRIP/TRAIN TRIP/ SEA VOYAGE, illustrated in
(15)–(17) respectively (Lakoff and Johnson 1980:44–45):
(15) a. This relationship is a dead-end street.
b. We’re just spinning our wheels.
(16) a. We’ve gone off the tracks.
(17) a. Our marriage is on the rocks.
b. Their relationship is foundering.
Lakoff (1993:222) further adds that LOVE IS A JOURNEYcan be grouped with
A CAREER IS A JOURNEY under a more schematic metaphor A PURPOSEFUL LIFE IS A JOURNEY, which in turn is an instance of what he calls the event
structure metaphor (roughly, action is directed motion), which includesthe mappings in (18):
(18) States are locations.
Changes are movements.
Causes are forces.
Actions are self-propelled movements.
Purposes are destinations.
Means are paths to destinations.
Grady (1997, 1998) argues for a combination of decomposition of specific
metaphors and the subsumption of the parts into highly schematic metaphors
that combine with each other. For example, Grady takes the narrowed version
ofAN ARGUMENT/THEORY IS A BUILDING and analyzes it into the two
metaphorsORGANIZATION IS PHYSICAL STRUCTURE and PERSISTING IS REMAINIG ERECT (Grady 1997:273). The two metaphor parts are formulated
in a schematic fashion in order to capture other metaphors, which Grady argues
are part of the same mapping, for example other argument metaphors and other target domains for structures .
Grady’s and Lakoff’s highly schematic analyses raise the question of which
metaphors are more basic to human understanding, the more specific or the more
schematic ones? To address this question, however, we must first address another
issue, namely what conceptual structures are mapped in the metaphor.
Lakoff proposes the Invariance Hypothesis as a constraint on metaphorical mapping (Lakoff 1990:54):
(21) Invariance Hypothesis: Metaphorical mappings preserve the cognitive topology(that is, image-schematic structure) of the source domain.
Image schemas, as discussed in chapter 3, include much of the basic structuring
of experience, such as scales, causation, containment, motion and so on. Lakoff
argues in particular that reasoning in the target domain (metaphorical entailments)
is governed by the image-schematic structure of the source domain (for example, consider the epistemic correspondences for ANGER IS HEAT OF A FLUID
above).
Turner proposes an important constraint on the Invariance Hypothesis: