Skip to main content

Full text of "ERIC ED357653: Language as Fluid: A Description of the Conduit Metaphor in Japanese."

See other formats


ED 357 653 

FL 021 309 





Nomura, Masuhiro 

Language as Fluid: A Description of the Conduit 

Metaphor in Japanese. 


17p.; For the serial issue in which this paper 
appears, see FL 021 304* 

Reports - Research/Technical (143) — Journal 
Articles (080) 

Kansas Working Papers in Linguistics; vl8 p75~90 


MF01/PC01 Plus Postage. 

^Communication (Thought Transfer) ; Contrastive 
Linguistics; English; ^Japanese; Language Research; 
^Linguistic Theory; Vf Metaphors; *Verbs 


The question of how 'communication' is metaphorized 
in Japanese is examined and this metaphorization is contrasted with 
Reddy's (1979) conduit metaphor. A claim is made that there is a 
strong tendency for Japanese to conceptualize 'word' as 'fluid' and 
to fuse 'word' and 'meaning.' English, which unlike Japanese, has 
overt count/mass and singular/plural distinctions, provides indirect 
support for the claim. It is suggested that one way to revitalize the 
conduit metaphor is to see how other languages metaphorize 'word' and 
'communication,' which will hopefully reveal in what respects Reddy's 
conduit metaphor is universal and in what respects it is language 
specific. (Contains 6 references.) (JL) 

Vc rt rt rt rt rt Vert rt rt Ve rt rt Vc Vc Vc rt rtrtrt Ve Vc rt rt * * Vc V: rtrt rtrt rt rtrtrt * rt rt Vc Vc Vc Vc rt Vert Vc Vert rt Vc Vc rt Vc Vc Vc rt Vc rt Vert Vc Ve Ve Vc Vc Ve rt Vc Vc Vc 

* Reproductions supplied by EDRS are the best that can be made * 

* from the original document . * 

Vc Vc Vc Vc Ve Ve rt Vc rt rt Vc rt rt rt Vc Ve Vc rt rt rt rt Vc rt Ve Ve Ve Vc Vc rt Vc Ve Vc Ve Ve Vc Vc Vc Vc Vc Vc Vc Ve Ve rt rt Vc rt Vc rt rt rt rt rt rt Vc Ve rt rt rt Ve rt Vc rt Ve Ve Vc rt rt rt Vc Ve 




A Description of The Conduit Metaphor in Japanese 

Masuhiro Nomura 


OHtce of Educational Research and Improvement 


__, ™$ document has been reproduced as 
received from the oerson or organization 
originating it 

C Minor changes have been made to improve 
reproduction quaiuy 

• Points of view or opinions slated m this docu- 
ment do not necessarily represent officia' 

OERl position or policy 

Abstract: The aim of the present paper is to see how 

COMMUNICATION is metaphorized in Japanese and to contrast 
this metaphorization with Reddy' s (1979) conduit metaphor. 
I till claim that there is a strong tendency for Japanese 
to conceptualize I0RD as FLUID and to fuse WORD and MEANING. 



lO 1. Introduction 


Q Communication is an abstract domain of experience which can be mctaphori- 

zed in terras of a more concrete domain cf experience. Reddy (1979) is the first 
detailed analysis of how our language about language is structured in terms of 
rctaphor. He argued that English expressions of COMMUNICATION arc based on what 
he calls "the conduit metaphor", which consists of the following four components 
(ibid: 290): 

(Pa. language functions like a conduit, transferring thoughts bodily from 
one person to another: 

e.g. Try to_get your thoughts across better. 

None of Mary 1 s feelings came through to me with any clarity. 

b. in writing and speaking, people insert their thoughts and feelings 
in the words: 

e. g. Try to pack more thoughts into fewer words . 

Don' t force your meanings into the wrong words . 

c. words accomplish the transfer by containing the thoughts or feelings 
and conveying them to others: 

e.g. That thought is in practically every other word . 
The sentence was filled with emotion . 

d. in listening or reading, people extract the thoughts and feelings 
once again from the words: 

e.g. Can you actually extract coherent ideas from that prose ? 
I don' t jjet any feelings of anger out of his words . 

rf\ In (la), the object of the act of transferring is "thoughts" or "feelings". 

Since words are containers for thoughts and feelings, as (lb, c, d) suggest, it is 


o * 

^ q Kansas Working Papers in linguistics, Volume 18, 1993, pp. 75-90 


possible for "words" to be objects of the act of transferring (though Reddy 
himself did not give exanples of this type): 

(2)a. accept one's word for it. 

b. He could scarcely catch the words. 

c. We exchanged a few words. 

d. He flung words at me. 

e. give him a word of greeting/advice/warnim* 

f. I hear that words passed between them. 

g. He sent word that — 

h. You should never take his words just as they are. 

i. toss a word to — 

The expressions in (2) suggest that WORD is conceptualized as an 
<individuum> that people can give and take. 

The aim of the present paper is to consider how Japanese exploits 
metaphors to talk about COMMUNICATION, and to contrast this with Reddy s 
"conduit metaphor". I will argue that there is a strong tendency for Japanese 
to conceptualize WORD as <fluid> and COMMUNICATION as a movement of ^fluid> from 
a speaker toward a hearer. 

2. M ethodological Assumptions 

1 will make the following methodological assumptions: 

(3) In some languages, there exists a set of predicates that specifically 
express the movement/state of <fluid>: e.g. 'leak', 'f low' , spill', 
'shower* , *pour' , Mouse' , 'soak' , etc. 1 

(4) If such a predicate (henceforth "fluid predicate") is used in a meta- 
phorical sense (henceforth "fluid metaphor"), its relevant argument is 
being conceptualized as <fluid> or indiscrete mass. 


English, which unlike Japanese has overt count/mass and singular/plural 
distinctions, provides indirect support for the assumption (4). 3 The following 
examples suggest that a fluid metaphor can occur with either a plural noun or a 
mass noun as its relevant argument: 

(5)a. Crowds/People flow down the street, 
b. *A boy flows down the street. 

(6)a. A lot of good ideas welled up while reading this book 

b. ?A good idea welled up while reading this book. 

c. Anger/Joy welled up. 

In light of the above assumptions, compare, as an illustration, the 
following pair of Japanese expressions which have roughly the same meaning, 
"snap at someone" : 

(7)a. hagesii kotoba-o butukeru 

bi t ing word-ACC f 1 ing 

b. hagesii kotoba-o abiseru 

biting word-ACC shower 

Since Japanese lacks the singular/plural and the count/mass distinctions as 
grammatical categories, the noun Vrtoba* has exactly the same fono in (7a) and 
(7b). The noun "kotoha" in (7a), however, can be considered to reflect 
<individuum>, because the verb "butukeru" (fling) typically takes an <individuum> 
(e.g., "isi" (stone)) as its direct object. The same noun "kotoba* in (7b), on 
the other hand, can be considered to reflect the conceptualization of <fluid>. 
because (7b) involves a fluid predicate "abisenf (shower) being used in a 
metaphorical sense. 

In the next section, based on this methodology, I will analyze Japanese 
conventional expressions of communication and demonstrate the ubiquity of fluid 
metaphors in conceptualizing COMIUNICATION in Japanese. 

3. The Conduit Metaphor in Japanese 

Reddy s conduit metaphor can be divided into two parts, (la) and (lb-d). 
The former focuses on the movement of 10RD, and the latter focuses on WORD as a 
container. In this section, I will discuss the movement aspect of the conduit 
metaphor and the container aspect of the conduit metaphor in this order. 

Movement of 108D 4 

I will examine fluid predicates one by one to sec how they are used to 
metaphorize the movement aspect of COMIUNICATION. 

(A) morasu/moreru (leak(v. t. )/leak(v. i. )) 

The transitive verb *morasu"(leak) typically takes a <fluid> direct 

(8) mizu/kuuki-o morasu 
water/air-ACC leak 
"leak water/air" 

Hence the following example indicates that WORD is conceptualized as <fluid> and 
the speaker as a container for <fluid>: 

(9) kotoba-o morasu 
word-ACC leak 

"utter words in spite of oneself" 

Furthermore, the verb "morasu" has developed a usage as a speech verb, taking a 
complementizer "to": 

(10) Taro-wa Jiro-ga gan dearu to moras ita . 
Taro-TOP Jiro-NOM cancer be COMP leaked 
"Taro confided that Jiro has cancer". 

The verb "norasu" can be combined with "kiku'(hear) and "iu"(say) to form 
a compound verb meaning "fail to catch/say some words" ("kiki" and "ii" are 
conjunctive fons of "kiku" and "iu" respectively): 5 

(11) daizina kotcro kiki -morasu 
important thing-ACC hear- leak 
"miss the important parts" 

(12) daizina koto-o ii -morasu 
important thing-ACC say-leak 

"{forget to lent ion/let out} an important thing" 

The image behind these expressions would be that 10RD as <fluid> leaks from the 
conduit and loses some portion of it when it should flow to the hearer in tota 
Interestingly, the compound verb "ii-morasu" has two seemingly incompatible 
interpretations, namely, "forget to mention" and "let out". The latter inter- 
pretation seems to be related to the fact that the verb "morasu" itself implies 
"to say something secretly", as seen in (9). 

The intransitive verb "morenf , which is morphologically related to 

"morasu", is used to express the situation where one utters words despite one- 

(13) human-no kotoba-ga kare-no kuti kara moreru 
complaint-GEN word-NOM he-GEN mouth from leak 
"lords of complaint escape his lips" 

The verb "moreru" combines with "kit?/ (hear) to make up a compound verb: 

(14) Taro-ga kekkonsuru hanasi-o more-kiku . 
Taro-NOM get married nimor-ACC leak-hea" 

"(I) hear the rumor that Taro will get married" 

The image behind this combination would presumably be that one hears I0RD as 
<fluid> leaking from some source of information. 

(B) nagasu (pour, let flow) 

The transitive verb "nagasu" (pour, let flow) typically takes a <fluid> 
argument as its direct object: 6 

(15) mizu/ti/namida-o nagasu 
water/blood/tear-ACC pour 
"pour water/bleed/shed tears" 

This verb "nagasu" combines with the receptive verbs "kiku"(hear) and 
"yomu"(read) to make up a compound verb meaning "listen/read inattentively": 

(16) Taro-wa Jiro-no kotoba-o kiki-nagasu 
Taro-TOP Jiro-SEN word-ACC hear-let flow 

"Taro lets Jiro' s words go in one ear and out the other" 

(17) hon-o yomi -nagasu 
book-ACC read-let flow 

"read a book inattentively, skim through a book" 

I surmise that the image behind these expressions is fiat the listener/reader 
lets I0RD "flow", without stopping and accepting it. 7 

then the verb "nagasu" combines with the productive verb "kaku" (write), 
the compound verb "kaki -nagasu" means "write smoothly, dash off something". 

This leaning is activated by the image that one writes smoothly and quickly as 
if pouring water. For some unknown reason, the combination "ii-nagasu" (say- 
let flow) is not comaonly used. 

(C) kobosu (spill) 

The verb 'kobosu* (spill) typically takes a <fluid> and occasionally a 
■ass-like <solid> direct object: 

(18) mizu/gohan/*enpitu-o kobosu 
water/rice/*penci 1-ACC spi 1 1 
"spill water/rice/*pencil(s) ,t 

This verb can Metaphorically be used with a noun meaning "complaint": 

(19) human-no kotoba-o kobosu 
complaint-GEN word-ACC spill 
"to complain" 

Furthermore, the verb has developed a usage as a speech verb: 

(20) Taro-wa Jiro-ga urusai to kobosu 
Taro-TOP Jiro-NOM noisy COMP spill 
"Taro complains that Jiro is noisy" 

The verb "kobosu" implies that one spills something which should have been con- 
tained. The reason that "kobosu" is normally associated with the notion of 
"complaint" might be that "complaint" is understood in Japanese as something to 
be contained and not let out. 

(D) siboru (squeeze, wring) 

The verb "si bom" (squeeze) takes as its direct object either a <fluid> or 
an object containing a <fluid>: 

(21 ) a. suponzi-o siboru 

sponge-ACC squeeze 
"squeeze the sponge" 
b. mizu-o (suponzi-kara) siboru 
water-ACC (sponge-from) squeeze 

squeese the water (out of the sponge)" 

then the verb "dasu"(let out) is added to "siboru", Baking a compound verb 
"sibori-dasu", only a <fluid> can be its direct object: 

( 22 )a. * suponzi -o sibori-dasu 

Thus, the following expression suggests that I0RD is conceptualized as a <f luid> 

(23) kotoba-o sibori-dasu 
words-ACC squeeze-out 
"force out one* s words" 

(E) abiseru/abiru (shower/be showered with) 

The verbs "abisenf (shower) and "abini"(be showered with) typically take 
a <fluid> direct object: 

(24) nizu-o abisem 
water- ACC shower 
"pour water on" 

(25) nizu-o abiru 
water-ACC be showerd with 
"pour water over oneself" 

then uttering words to the hearer, these two verbs can be used: * 

(26) hinan/syoosau-no kotoba-o abiru/abiseiu 
blaae/praise-GEN word-ACC be showered/shower 

"be showered with/shower soaeone with words of blaae/praise" 

(27) sinratuna kotoba-o abiseru 
b i t i ng word-ACC shower 
"shower someone with bitinp remarks" 

(F) haku (exhale, vo«it) 

sponge- ACC squeeze-out 
■izu-o sibori-dasu 
water-ACC squeeze-out 

'squeeze the water out" 

The verb "haku" (exhale, voait) can be said to typically take a <fluid> 
direct object: 

(28) iki/ti-o haku 
breath/Mood-ACC exhale, voait 
"exhale, voait blood" 

The following expressions show that WORD is viewed as <fluid>: 

(29) a. hituuna kotoba-o haku 

grievous word-ACC voait, exhale 
"utter grievous words" 

b. kagekina iken-o haku 
radical opinion-ACC voait, exhale 
"express a radical opinion" 

c. honne-o haku 

real intention-ACC voait, exhale 
"tell one's real intentions" 

(G) yodoau (stagnate), nigosu (nake (.water) turbid) 

The verbs "yodoau" (stagnate) and "nigosu" (nake (water) turbid) typically 
take a <fluid> argument : 

(30) mizuAuuki-ga yodoau 
uater/air-NOM stagnate 
"The water/air stagnates" 

(31) oizu/kuuki-o nigosu 
water/air-ACC Bake turbid 

"make water turbid/aake air foul" 

WORD as <fluid> *>ves fro« a speaker toward a hearer, but it is not always 
the case that 10RD *>ves smoothly: soaetiies 10RD as <fluid> can stagnate or 
get turbid, resulting in unsuccessful coaminication : 

(32) a. yodoai-naku hanasu 

stagnation-without speak 
"speak fluently" 
b. ii -yodoau 

"hesitate to say" 

(33) a. kotoba-o nigosu 

word-ACC Bake turbid 
"speak ambiguously" 
b. henzi-o nigosu 
answer-ACC aake turbid 
"give a vague answer" 

(H) siiiru (soak into), kuau (draw (water)) 

Lastly, let us consider soae expressions used fro« the hearer s viewpoint. 
The verbs "siiiru" (soak into) and "kumf(draw (water)) typically take a <fluid> 

(34) oizu-ga nuno-ni siairu 
water-NOM cloth-DAT soak 
"The water soaks the cloth" 

(35) nizu-o ido kara kugu 
water-ACC well froa draw 
"draw water froa the well" 

When WORD as <fluid> issued froa the speaker is not accepted by the hearer, the 
coopound verb "kiki-nagasu"(listep-f low) is used, as we saw in (16). When it is 
accepted, WORD as <fluid> "soaks into" the hearer: 

(36) kanozyo-no kotoba-ga kokoro-ni siniru 
she-GEN word-NOM heart-DAT soak into 
"Her words sink into ay heart" 

fhen the hearer wants to take in WORD as < fluid) of his/her own accord, 
the verb "kuau"(draw (water)) is used: 9 

(37) kotoba-o kmu 
word-ACC draw (water) 

"take someone's words into consideration" 

Container aspect of 10RD 

Compared with the variety of English expressions Reddy (1979) gave for 
the fORD AS A CONTAINER part of the conduit aetaphor (which corresponds to (lb)- 


(Id)), there are only a few corresponding expressions in Japanese that concern 
the relationship between WORD and MEANING (I assuie that MEANING corresponds to 
"thoughts/feelings" in Reddy s formulations). Here are sowie examples: 

(38) kotoba-ni iii-o koaem 
word-DAT leaning-ACC load 

(39) kotoba-ni iii-o takusu 
word-DAT leaning- ACC entrust 

(40) kotoba-ga iii~o hukuau 
word-NOM leaning- ACC contain 

(41) kotoba-no iii-o torn 
word-GEN leaning-ACC take 

It is very odd for Japanese to explicitly code the insertion/extraction aspect 
of the conduit letaphor: 

(42) ??kotoba-ni iii-o ireru / soonyuusuru 

word-DAT neaning-ACC put into 

(43) ??kotoba kara iii-o toridasu / tekisyutusuru 

word froi leaning-ACC take out 

This seeis to indicate that in Japanese, unlike in English, 10RD and 
MEANING are fused, rather than separated. Part of the reason for this light be 
that in classical Japanese there was a folk lodel where "kokoro w (heart, meaning) 
grows into *kotoba"(word)(see Ikegaii 1988. 1989). 10 In other words, heart, 
meaning, and word were considered to fon a continuua. On the other hand, if 
10RD is conceptualized as <fluid> as we saw above, then it follows, by the 
nature of <fluid>, that it is difficult to have clear-cut container/content 
separation in WORD. Thus, in this sense, it seeis natural that there is no 
clear separation of WORD and MEANING in Japcjaese. 11 

4. Soie Qualifications 

The above analysis of Japanese conventional expressions of comunication 
suggests that WORD tends to be conceptualized as <fluid> and COMMUNICATION as a 
■oveient of fluid frc* a speaker toward a hearer. 12 This contrasts with 
Reddy s conduit letaphor where WORD is conceptualized as an <individuim> and 
COMMUNICATION as its loveient. 

This does not, however, lean that Japanese never construes WORD as an 



<individuu»> or that English never construes WORD as <fluid>. Hunan beings can 
conceptualize the saae objective scene in different ways. The fact that Japanese 
prefers <fluid> and English prefers <individuw> to conceptualize 10RD is a 
natter of tendency. 

In fact, it is not uncc«on for 10RD to be conceptualized as <individuiai> 
in Japanese. The following expressions exemplify this: 13 

(44) kotoba-o okuru/kawasuAaesu/uketoru 
word-ACC send/exchange/re turn/accept 

(45) ii- kaesu / watasu 

(46) hagesii kotoba-o bu t uke ru/ nage t uke ru 
biting word-ACC fling 

"snap at (someone)" 

(47) kotoba-ga liii-ni tobikoau / todoku / hairu 
word-NOM ear-DAT juap into/reacb/enter 

That is important is that in Japanese the conceptualization of WORD as 
<fluid> is, at least, no less ccwaon than the conceptualization of WORD as 

Here are, on the other hand, so«e English exaap'cs where WORD is con- 
ceptualized as <fluid>: 

(48) a. pour out (a streaa of) words 

b. ford leaks out froa CIA. 

c. gush over(about) one' s baby 

d. a flood of words 

e. a rapid flow of speech 

f. His verse flows lusically. / Her talk flowed on. 

g. fluent ( < Latin: fluere 'to flow' ) 

It iay, however, safely be said that the conceptualization of IQRD as <fluid> is 
much less comon in English than in Japanese. 


5. Concluding Remarks 

To summarize our discussion, we have seen the following contrast between 
Japanese and English: 

Tendency: Japanese: 10RD as <fluid>. Fusion of WORD and MEANING 

English: lORD as <individuum>. Separation of WORD and MEANING. 

As Lakoff and Johnson (1980:Ch. 3) correctly point out, metaphor can high- 
light one aspect of a concept but hide other aspects of the concept. The 
conduit letaphor is so deeply rooted in English that it is virtually impossible 
to talk about language without using it. 14 One way of "relativizing" the 
conduit Metaphor is to see how other languages letaphorize 10RD and COMMUNICA- 
TION, which will hopefully reveal in what respects Reddy s conduit letaphor is 
universal and in what respects it is language-specific. The present paper is 
only a siall atteapt at this. 

Finally, I would like to mention a possibility that the distinction 
between non-fluid letaphor and fluid letaphor light parallel the distinctions 
between count noun/mass noun and perfective verb/iiperfective verbs (cf. 
Langacker 1987). These distinctions may be the different manifestations of the 
same cognitive capacity. 


iThis paper is based on my presentations at Nintigengogakukenkyuukai 
(Cognitive Linguistics Study Group) at University of Tokyo on September 5. 1992 
and at UCSD Cognitive Linguistics Workgroup on Janizary 21, 1993. I would like 
to thank A.Goldberg, Y. Ikegami, S.Iemmer. R. Langacker and an anonymous reviewer 
of KIPL for their helpful comments on earlier versions of the paper. 1 am 
indebted to R.Sheffer for checking my English. Thanks also go to my cohorts at 
UCSD, especially, Martha. Michael, lathleen. Sean, and fill for their support. 
Any remaining inadequacies are. of course, my responsibility alone. 

1. By the term <fluid>, I lean both <liquid> and <gas>. Since <liquid> 
is more basic (in terms of visibility, tangibility, usefulness, etc.) to human 
experience, I surmise those predicates prototypical ly take a <liquid> argument, 
and that this prototype is semantically extended to a <gas> argument. Fluid 
predicates are sometimes further extended to take a mass-like <solid> argument. 
Compare the following pair: 

(i) suna/lisi-o morasu 
sand/stone-ACC leak 

2. This assumption is based on Lakoff and Johnson s (1980:6) view of 
metaphor: "Metaphors as linguistic expressions are possible precisely because 
there are metaphors in a person's conceptual system." It is perfectly possible, 
however, that for some people fluid metaphors may be "dead" metaphors which do 
not evoke the conceptualization of <fluid>. 

3. For conceptual basis of the mass/count distinction, see Langacker 
(1987. 1991 :Ch. 2). 

4. The existence of the "conduit" through which WORD travels is supported 
by the following expressions: 

(i) kotoba/kangaeAimoti-ga tuuziru 
word/idea/ f eeling-NOM go through 

"make oneself understood/get one's {thoughts/feelings! across" 

(ii) tutu- nuke 
conduit-going through 
"(information) leak" 

5. It is not the case that the verb "morasu" can combine with any kind of 
action verbs to constitute a compound verb meaning "forget to do something": 

(i)a. kaki/yomi -morasu 
"forget to write/read* 
b. ?? i k i/benkyoos i/koros i -morasu 
go/s t ud y A i 1 1 - 1 eak 

It may be the case that (a) is possible, because the verbs "write* and "read" 
have something to do with language. 

6. The verb "nagasu" can take an <individuum> direct object when it means 
"to float something in the stream of water": 

(i) zaimoku-o kawa-ni nagasu 

log-ACC river-DAT float 

"float a tiiber in the river" 
This usage, however, does not lean that an <individma> object is conceptu- 
alized as <fluid>; in (i), "zaiioku-o nagasu" can never ican "pour logs". 

7. To express 'fail to hear/say", the verbs "otosu" (drop) and "nogasu" 
(let escape, miss) are used to aake a coopound verb: 

(i)a. kiki/ii -otosu 
"fail to hear/iention" 
b. kiki/ii- nogasu 
hear/say-let escape 
"fail to hear/iention" 
The verbs "otosu" and "nogasu" typically take an <individuia> direct object: 
( i i )a. enpi tu-o otosu 
pencil-ACC drop 
"drop a pencil" 
b. ookina sakana-o nogasu 
big fish-ACC let escape 
"miss a big fish" 

Thus, we nay say that WORD is conceptualized as an <individuia> in the 
expressions in (i). Interestingly, the verb "minf (see) cannot combine with 
"morasu"(leak) or "nagasu"(let flow), but it can conbine with "otosu" (drop) 
and "nogasu"(let escape), to Bean "fail to see": 
(iii)a. * a i - moras u / nagasu 
see-leak/let flow 
b. nii- otosu/nogasu 
see-drop/) et escape 
"fail to see" 

The reason fluid ictaphors like (iiia) are not used night be that we have a 
folk nodel according to which our visual field is occupied by clearly 
demarcated, discrete objects. 

8. The verb "kakeru" is known for its polysemy (hang, cover, wear, sit, 
etc.). One of its leanings is siiilar to "abiseru": "to sprinkle, throw (water)". 
It takes a <fluid> or a lass-like <solid> direct object: 

(i) lizu/sio-o kakeru 

watcr/salt-ACC sprinkle 
"pour water over/spr inkle salt on" 
The following expression light be related to the above use of "kakeru": 

(ii) atatakai kotoba-o kakeru 
want word-ACC ? 
"give (someone) kind words" 

9. The expression (37) is quite different from the English equivalent 
Keddy (1979) gives, namely, "Can you actually extract coherent ideas from that 
prose?". Tie verb "extract" can take a <fluid> direct object (e.g., extract 
juice from lemons) as veil as an <individuum> direct object (e.g., extract a 
tooth). I suspect, however, that Reddy* s intended image is that "ideas" are 
discrete objects taken out of a container. 

10. Ikega*i(1988) quotes a passage fro* the preface to Kokinwakashuu (a 
collection of waka poetry compiled in the tenth century), where "kokoro w 
(heart) is compared to a seed and "kotoba"(word) to its buds or leaves. 

11. The fusion of I0RD and MEANING in Japanese is best observed in 
examples (36) and (37), where "kotoba"(I0RD) is used to ican KEAjNING(= thoughts/ 

cf. (36)* kanozyo-no sinsetu-ga kokoro-ni simiru 

she-GEN kindness-NOM heart-DAT soak into 
"Her kindness sinks into iy heart** 
(37)' kangae/kimoti/kokoro/imi-o kuMu 

Recall, in this connection, that, in Reddy* s conduit letaphor (la), what loves 
is "thoughts/feelings", instead of "words", 
cf. ?Try to get your words across better. 

?None of Mary's words came through to ie with any clarity. 
These sentences may be acceptable only when "words" refers to actual physical 
sound. See Note 4 (i) for the contrast with Japanese. 

12. From this viewpoint, the following cliche Bakes sense: 
(i) tateita-ni mizu-o nagasu yooni hanasu 

vertical wooden board-DAT water-ACC pour as if talk 
"(He) speaks fast and fluently" 

13. In the following examples. IORD is conceptualized as FOOD/DRINK: 
(i) kotoba-ga nodo-kara dekakatteim 

word-NOM throat-from be just about to come out 
"words are on the tip of one's tongue" 

thought/feel ing/heart/meaning-ACC draw (water) 

"take into consideration someone's thought/feeling/heart/intention 

(ii) kotoba-o nomikomu 

word-ACC swallow, drink 
"swallow one's words" 

14. For the various problems entailed by the conduit Metaphor, see Lakoff 
and Johnson (1980: Ch.3) and Langacker (1991:508). 


Ikegaai, Yoshihiko 1988. *lmi no hirogari (Extension of "meaning"). Gengo, 
July issue: 28-36. 

Ikcgami, Yoshihiko 1989. Nihongo no tekusuto to komyunikeesyon (Text and 
CooMunication in Japanese). In Kazuko Inoue (ed. ) Nihonbunpoo syooziten 
(Compact Encyclopedia of Japanese Grammar), 245-266. Tokyo: Taishukan. 

Lakoff, George and Mark Johnson 1980. Metaphors le Live By . Chicago and 
London: The University of Chicago Press. 

Langacker, Ronald 1. 1987. Nouns and Verbs. Language 63:53-94. 

Langacker, Ronald 1. 1991. Foundations of Cognitive Grammar. Vol.11. Stanford: 
Stanford University Press. 

Reddy, Michael J. 1979. The Conduit Metaphor - A Case of Frame Conflict in Our 
Language about Language. In Andrew Ortony (ed. ) Metaphor and Thought , 284- 
324. Cambridge: Cambridge Dniversity Press