228 Zhiying Xin
reason relationship in lieu of intelligence. Based on introspective studies on conceptual
metaphors for both emotion and reason, the contributors focus on the metaphorical
expressions containing fear, love, mind, reason or wit, for these words are common to the two
target domains (194). They drew their data from four English language corpora, two of Early
Modern English, and two Present Day English. The data collection involves three steps. To
begin with, metaphors concerning both domains were extracted separately. Then, for each
word the metaphorical expressions with highest frequency were identified. Finally, only those
metaphors appearing in both domains were analyzed. The huge analysis section describes in
detail their findings by paying special attention to diachronic comparing. Three types of
metaphors are elaborated: ontological metaphors, personification, and metaphors involving
force. The first is subcategorized into INSTRUMENT/TOOL/WEAPON, OBSTACLE/WHIP,
VALUE COMMODITY, CONTAINER, and BODY. For each (sub) category, frequencies are
compared, diachronic changes are examined, and typical examples are provided. Their
comparative historical cognitive corpus linguistic investigations can be evaluated from four
aspects. To begin with, it is a careful attempt at integrating cognitive linguistic theories with
historical studies. It is the authors‘ contention that the two disciplines should negotiate and
cooperate so as to realize transdisciplinary operations. In addition, their findings challenge the
present cognitive concepts. Here is an example. Lakoff and Johnson regarded
UNDERSTADNING IS SEEING as the central metaphor for reason (192). However, it is ―very
rare‖ in their data. Moreover, the literature of metaphor is enriched, especially about the
similarities and differences between the two domains EMOTION and REASON. Both do share
some conceptual metaphors such as CONTAINER, and BODY. Yet there exist striking
differences as well. For example, the lexemes indicating emotion favor VALUABLE
COMMODITY while INSTRUMENT/WEAPON is much more strongly associated with the
domain REASON. Finally, the chronological approach shows that the metaphorical process
reflects gradual changes in word meaning, cultural connotation and evaluation orientation.
―A corpus-based analysis of context effects on metaphor comprehension‖ by James H.
Martin develops and verifies his notion of ―Metaphor Prediction Hypothesis‖, i.e. the
metaphor-prediction capability of different types of context determines the degree of
difficulty or easiness in metaphor interpretation. Martin begins with an introspective
discussion about how to estimate the metaphor-predictive value of context. First is to locate a
―test sentence‖ (in which a metaphor is contained) in its ―context‖ (the proceeding text).
Second is to pick up the information needed for metaphor prediction, including
source/target/ground concepts. And finally, how the context presents the information (i.e.
context types) is considered from three aspects, namely, literal source concept, literal target
concept, and metaphorical reference. The hypothesis assumed is then tested for validity with
corpus investigations and psychological experiments. In the fist testing, texts from the Wall
Street Journal corpus is tagged at the sentence level for four metaphors: NUMERICAL-VALUE-
AS-LOCATION, COMMERCIAL-ACTIVITY-AS-CONTAINER, COMMERCIAL-ACTIVITY-AS-
PATH-FOLLOWING and COMMERCIAL-ACTIVITY-AS-WAR. Here the context is
operationalized based on ―five sentence context windows‖. That is, a test sentence followed
by a four sentence window (221). The author calculated the base rates for each metaphor as
the ―ratio of the number of metaphors found to sentences examined‖ and the ―ratio of 4
sentence windows containing a metaphor to the number of windows‖. These base rates are
regarded as yardsticks to measure the metaphor-predictive capability of the three types of
contexts given in the hypothesis. The higher the metaphor occurrence frequency than the