Professional Documents
Culture Documents
A contrastive analysis*
Dongqin Shi
Shanghai Maritime University (China)
This paper, using translation materials as data for analysis and based on Fill-
more’s (1982) Frame Semantics Theory, presents a contrastive study of com-
munication verbs in Chinese and English with respect to their representation
of semantic components related to a communication event. In Talmy’s (1991)
classification, Chinese and English fall into the same typological category as
satellite-framed languages, where large numbers of verbs characteristically
incorporate Manner with Motion, and the Path of motion is typically mapped
onto a post-verb satellite. Previous contrastive studies of Chinese and English,
however, have concluded that Chinese deviates from a typical satellite-framed
language as represented by English with respect to both the conflation of Man-
ner with Motion and the representation of the Path of motion. The present study,
which examines, specifically, how communication verbs express speaking events
in written narrative in Chinese and English, has found more intra-typological
differences between the two languages in their representation of meanings in
surface structures as well as in the narrative style in which words of a particular
category are chosen for representing events. The major original findings from
this study are that Chinese is by far more constrained than English in its encod-
ing of semantic components in communication verbs, and that a translator may
face a great challenge when doing Chinese-to-English translation with respect to
the choice of appropriate communication verbs.
1. Introduction
What and how semantic components are represented in surface structure in dif-
ferent languages has remained a focus of interest in linguistic studies. Accord-
ing to Talmy (1985:57, 2000:21), semantic-to-surface association is largely not of
cave floating”) in Spanish. In the English version, the verb ‘float’ indicates Motion
with a conflated Manner, the Path of motion being mapped onto the satellite ‘out’,
whereas in the Spanish one, the Motion verb salir incorporates the Path of motion,
with an optional gerundial element, flotando, indicating Manner.
Consequently, S-languages and V-languages have different types of Motion
verbs in their lexicon, as revealed by cross-linguistic studies. Slobin (1996b), for
example, discovers that S-languages generally have a greater repertoire, and a more
frequent use, of Motion verbs of Manner while V-languages a larger collection, and
a more frequent use, of Motion verbs of Path. With respect to Manner verbs, as
Slobin (1997:459) further observes, languages seem to have a ‘two-tiered’ lexicon
of Manner verbs: the neutral, everyday verbs like ‘walk’, ‘fly’, ‘climb’, and the more
expressive or exceptional verbs such as ‘dash’, ‘swoop’, ‘scramble’. In S-languages
the second tier of lexicon is extensive and elaborated while in V-languages it is
considerably smaller and does not play an important role.
These differences between an S-language and a V-language in the types and
subtypes of Motion verbs in their lexicon can pose difficulties in cross-linguistic
translation. Slobin (2004:200) claims that in translating into a V-language, trans-
lators often omit Manner information while in translating into an S-language,
however, translators tend to add Manner verbs. According to Slobin (1996b), in
English-to-Spanish translation, only 51% of English Manner verbs are translated
into equivalent Spanish Manner verbs, with the other half either neutralized or
omitted, while in translation from Spanish to English, translators, by contrast, add
Manner information to the Spanish original in almost a quarter of times. This is
attributed to the claim that Manner of motion is far more salient in English than
in Spanish.
Talmy’s categorization of languages into a typological dichotomy was made on
the basis of how the Path of motion is typically encoded into a surface structure
or what semantic component is characteristically encoded in the Motion verb in a
language. Thus the results obtained following his theoretical framework about the
differences between languages of two typologies in the number and use of Manner
verbs do not seem to apply to event verbs of other semantic categories, which may
also incorporate different semantic components, including Manner, with action.
Rojo and Valenzuela (2001), having analysed on the basis of the speech-act-frame
proposed by Fillmore ‘verbs of saying’ in English fiction and the ways they are
dealt with in Spanish translations, reach the conclusion that there is no big differ-
ence between English and Spanish in their use of verbs of communication, thus
contradicting, as the authors claim, Slobin’s (1996b) statement that V-languages
have a lower collection of Manner verbs than do S-languages. They have discov-
ered that, opposite to Slobin’s (1996b) finding, the Spanish translators in their cor-
pus actually add information in many cases by translating an English general verb
184 Dongqin Shi
of speaking into a Spanish specific one, which incorporates an extra semantic ele-
ment absent in the English general verb, as in the case of “ “My bloody grandson”,
said Anwar” and its Spanish translation “Un nieto, joder”, exclamó Anwar (cf. Rojo
and Valenzuela 2001:475), where the translator chooses a specific Spanish verb,
exclamar, which incorporates in it a particular Manner of saying, for the general
verb ‘say’ used in the English original.
As far as Chinese is concerned, it is classified as an S-language in Talmy
(1991:486). Chinese is thus expected to exhibit similar patterns to an S-language
such as English with respect to the number and use of Manner verbs as well as
the encoding of the Path of motion. As has been noted in the literature, however,
languages even of the same typology do not constitute a homogenous group. Chu
(2004), for example, argues that Chinese does not actually display a neat lexico-
syntactic pattern typical of an S-language in expressing Manner and Path. Accord-
ing to Chu, Chinese exhibits both a satellite-framed pattern and a verb-framed
pattern in this respect. On the one hand, the Path of motion can be realized in Chi-
nese in various structures, e.g. in a verb complement (Satellite in Talmy’s term) as
in Zhang Sheng xiang qiaoqiao liu jin-qu (“Zhang Sheng wants to sneak in”, cf. Chu
2004: 183), a prepositional phrase as in Xin niangzi xiang dongfang li zou (“The
bride went toward the nuptial-chamber”, ibid:190), or the main verb of a clause
such as shangsheng (“rise”), qianjin (“move forward”), cheli (“withdraw from”) and
so on (ibid:191). On the other hand, the conflation of Manner with Motion is
more constrained in Chinese than in other S-languages like English. For example,
the English verb ‘hop’, which encodes the Motion event ‘jump on one foot’, has
no equivalent in Chinese and has to be expressed analytically by an adverbial of
Manner as in dan jiao tiao (“one foot jump”) or yong yizhi jiao tiao (“use one foot
jump”) (cf. Chu 2004:198).
Other studies, not conducted in the framework as proposed by Talmy (1991),
have also revealed interesting differences between Chinese and English in their
representation of semantic components in surface structures. Wang and Zhou
(2004), for example, upon a careful analysis of the components conflated in the
look-type verbs in Chinese and English, have discovered that the two languages
tend to encode different components in this type of verb, thus giving rise to the
lexical “gap” between Chinese and English, a state in which lexical items in one
language find no exact equivalents in the other. For example, in Chinese direction
is often conflated with the act of looking while in English such a semantic entity
normally has to be mapped onto a post-verb satellite-like element, e.g. yangwang
vs. ‘look up’, huigu vs. ‘look back’, tiaowang vs. ‘look from above into the distance’,
etc.1
As is clear from the cross-linguistic studies introduced above, languages are
often specific in their lexico-syntactic representation of semantic structures. In
Communication verbs in Chinese and English 185
the case of Chinese and English, in spite of the fact that they are classified into
the same verb typology, they are both similar and different in their semantic-to-
surface association patterns, not only for Motion events but also for events of other
semantic categories. The present paper, following Rojo and Valenzuela’s (2001)
study of verbs of communication in English and Spanish, intends to examine verbs
of communication in Chinese and English for what and how semantic elements or
information about semantic elements related to communication are represented
in the surface structures of the two languages. Specifically we aim to achieve this
purpose by examining the rhetorical styles as demonstrated by the choice of the
writers or the translators for the type of communication verb or other surface
structure for a particular speaking event.
2. Theoretical framework
The present study examines how Chinese and English are similar to and different
from each other in their representation of speaking events in the surface structure.
It is conducted in the general framework of Fillmore’s (1982) Frame Semantics
Theory.
Frame Semantics, which has attracted in recent years a great deal of atten-
tion in computational linguistics, lexicology, lexicography, and so on, is a theo-
retical framework for studying lexical meanings in terms of cognitive frames. In
Fillmore’s (2003:267) definition, a frame refers to “a structure of knowledge or
conceptualization that underlies the meaning of a set of lexical items that in some
ways appeal to that same structure”. The basic principle of Frame Semantics is that
“word meanings must be described in relation to semantic frames…” (Fillmore et
al. 2003:235). One cannot possibly understand the meaning of a word without an
appropriate knowledge of the various participants, props, and other conceptual
roles, known as frame elements (FEs), that underlie the meaning of that word.
For example, to understand verbs like ‘buy’, ‘sell’, ‘pay’, ‘spend’, ‘cost’, ‘charge’, etc.,
one has to know something about the elements that are involved in the Transac-
tion frame, e.g. Buyer, Seller, Money, Goods and so on, as well as the relationship
between one element and another.
The Berkeley FrameNet project (Baker et al. 2003),2 an ongoing computational
lexicography project based on Frame Semantics theory and supported by the Brit-
ish National Corpus data, analyses the meanings of words within the semantic
frames from which the words originate and describes the obligatory or optional
syntactic realizations of the FEs evoked by a head word. The FEs that uniquely
characterize a frame, making it different from other frames, are referred to as core
FEs, e.g. the FEs of Avenger, Punishment, Offender, Injury, and Injured_ party in
186 Dongqin Shi
the Revenge frame, while other FEs are referred to as peripheral and mark such
notions as Time, Place, Manner, Means, Degree and so on, which can apply to any
appropriate event frame (Ruppenhofer et al. 2005:22). Core elements are identi-
fied in terms of uniqueness to a frame and are not necessarily represented overtly
in every case even if the head word allows their overt expression. For example, a
speaker could choose to say “[HeBuyer] bought [an apartment at a seaside resort-
Goods]” with the FEs of Seller and Money backgrounded though they could well be
overtly expressed. Furthermore, a head word may even restrict an overt expression
of a core FE, e.g. ‘spend’ as in “[SheBuyer] spent [a lot of moneyMoney] [on clothes-
Goods]”, which does not allow Seller to be expressed overtly, though it semantically
implies the existence of a seller. Another feature characterizing the realization of
FEs is that the same FE can have more than one syntactic form depending on the
head word. For example, the FE of Message, i.e. the content of what a Speaker
communicates to an Addressee, can be realized into a direct quote, a that-clause
or a nominal after different verbs in English, as is demonstrated by examples (1a)–
(1c):
(1) a. John said, [“I’m sorry.”Message] or John said [that he was sorryMessage].
b. John mentioned [that he was sorryMessage].
c. John expressed [his apologyMessage].
(Johnson et al. 2001:16). Examples (2a) and (2b) below, borrowed from Urban and
Ruppenhofer (2001:80) with slightly different marking of FEs from theirs, show
the FEs evoked by the communication verbs, ‘remark’ and ‘inform’.
(2) a. [In 1926Time] [Thomas HardySpeaker] remarked [forlornlyManner] [of
contemporary modernist writingTopic]: [“They’ve changed everything
now.”Message]
b. [The sellerSpeaker] informed [the buyerAddressee] [in writing Medium] [that, if
he did not pay the balance by a given date, the seller would try to re-sell
the carsMessage].
3. Data
Slobin (1996a:209) proposes that a more useful way to explore the rhetorical styles
of two languages is to compare a translation with its original, asking how each
language accommodates the demands of the other with respect to the same con-
tent. By that, he mainly refers to the comparison of the source and target language
from different verb typologies. As we believe that the same method also applies to
work on two languages of the same verb typology rich in Motion verbs with incor-
porated Manner, we decide to base our analysis on a comparison of the Chinese
original with its English translation.
We choose the Chinese classic novel Hong Lou Meng and the English transla-
tion done by Yang Hsien-yi and Gladys Yang as our data for analysis. We do that
for two reasons. First, there are numerous conversations between characters in
different settings in the original, hence a heavy use of verbs of communication
to introduce direct quotes. Second, the Yangs’ translation is deemed to be one of
the best among English translations of fiction, thus possibly more successful than
others in manifesting the differences, if there are, between Chinese and English in
188 Dongqin Shi
encoding the speaking event and its related semantic components. Though con-
temporary data might be even more convincing, the language in Hong Lou Meng
is basically the same as standard Modern Chinese, especially when it concerns the
use of communication verbs for reporting direct speech. And the fact that Wang Li
cited examples almost exclusively from the novel for his classic work A Grammar
of Modern Chinese also reassured us about our choice.
We select at random two chapters, i.e. 29 and 77, from the original and extract
from them all the communication verbs used before a direct speech. We follow
Rojo and Valenzuela’s (2001) practice and specify a classification of the speaking
verbs in our data into two categories, the general verbs and the specific ones. In
our study, a general communication verb refers to one that encodes a speaking
event alone, e.g. dao (“say”), shuo (“say”), etc. in Chinese, and ‘say’, ‘tell’, etc. in Eng-
lish, while a specific communication verb refers to one that encodes some other
semantic component related to speaking, e.g. jiao (“shout”, “say in a loud voice”)
in Chinese, and ‘exclaim’, ‘retort’, etc. in English. The results of our analysis are
presented in Section 4.
4. Results of analysis
We have extracted in total 211 instances of communication verbs from the two
chapters in the original. In each of the speaking events represented by these verbs
there is always a Speaker and a Message overtly expressed almost in the same way
in Chinese and English, and in some cases there is also an Addressee present,
realized similarly in the two languages either in a prepositional phrase or as indi-
rect object in a post-verb position. As they have no direct relevance to the main
concern of this paper, we leave aside, unless necessary for our analysis, these three
FEs, and focus our attention on whether speaking events and their related compo-
nents are represented in lexical items or in other surface forms, and in the former
case on what FEs (e.g. Manner) or what information about an FE (e.g. Message,
Speaker, Addressee, etc.) is encoded in the meaning of a lexical item.
In our data, there are 46 instances in which a communication verb in the origi-
nal is either not translated or phrased into a form other than a communication
verb, e.g. a verb or verbal constituent indicating a body movement such as nod-
ding or shaking one’s head, pointing by one’s finger, etc. Such verbs or phrases are
not communication verbs, though they do occur as introducers of direct quotes
in language. So these 46 instances have been excluded from our discussion. On
the other hand, we have included in our analysis those noise or sound verbs such
as ‘chuckle’, ‘cry’, ‘gasp’, ‘snort’, etc., which are strictly speaking not communication
verbs as far as their lexical meaning is concerned, but can acquire communicative
Communication verbs in Chinese and English 189
eral speaking verb, hence very high frequencies of occurrence of dao (“say”) and
shuo (“say”). Second, in order to better illustrate the differences in the number and
variety of verbs used in the original and in the translation, we base our counting
on form rather than on meaning. Therefore, dao (“say”) and shuo (“say”), which
basically denote the same general act of speaking, are treated as two different items
while daying (“reply”, “agree”, “promise”), a polysemous verb used in two situations
in the original for different meanings, is still regarded as one item. Similarly, in the
English translation some verbs, which are the same or similar in lexical meaning
as synonyms or near-synonyms, are listed as separate items, e.g. ‘continue’ and ‘go
on’, ‘answer’, ‘reply’ and ‘respond’, etc. In addition, in the English table the three
phrasal verbs, ‘go on’, ‘join in’ and ‘put in’, are listed as lexical items, for the two ele-
ments in each case form an inseparable whole when describing a communication
event.
As shown in Tables 1 and 2, a far greater number and variety of communica-
tion verbs are used in the translation than in the original. Of the Chinese verbs,
dao (“say”) and shuo (“say”) are classified as general, and the rest as specific. In the
English list, say and tell are regarded as general, with the rest being all specific. In
the Chinese text, 7 specific verbs are used in 16 instances only, and in the other 149
cases a general verb, dao (“say”) or shuo (“say”), is either used alone or in combina-
tion with some other morpheme, word or phrase to form an analytic structure. In
the English translation, however, specific verbs are far preferable to general ones,
with as many as 59 specific verbs being used in 134 instances as against the use of
two general verbs, ‘say’ and ‘tell’, in 31 instances.
An analysis of the specific communication verbs in our data shows that these
specific verbs incorporate the following information related to a speaking event:
Manner, Content, Response, and Aspect. As there are far fewer specific verbs in
the Chinese text, it is observed that the above-mentioned information related to
speaking is frequently realized in an analytic structure instead of a specific verb.
The following is a detailed description of how these semantic components are re-
spectively realized in the original and in the translation.
Table 4 also shows a far more frequent occurrence of Manner verbs of speech
production in the translation. This suggests the use of Manner verbs in the
translation for different ways of speech production that are not overtly described
at all in the original, and such use accounts for 23 instances in total. In 7 of these
cases there is a replacement of Manner information in the translation. For example
in (5a), the English specific verb ‘cry’ (“say in a loud voice”), which incorporates a
voice volume, is employed in place of xiao dao (“smile say”) in the original, which
describes a speaking event together with a facial expression. In 4 other instances,
while the original emotional state, facial expression or physical movement is rep-
resented in an English prepositional phrase or -ing form, a specific verb is used
additionally in the translation for a particular way of speech production as in the
case of (5b). The remaining 12 tokens demonstrate a different type of Manner ad-
dition in the translation with a bare Chinese general verb being translated into an
English specific one. In (5c), for example, by choosing the English verb ‘exclaim’
(“say in a loud voice”) for the Chinese general speaking verb dao (“say”), which
does not have a Manner modifier at all, information about the voice volume of the
speaking event is added in the translation.
Fengjie tingshuo, xiao dao: “Lao zuzong ye qu,
(5) a.
Feng sister hear smile say old ancestress also go
ganqing hao le.”
how good CRS
“ “If our Ancestress is going as well,” cried His-feng, “so much the
better.” ”
b. Wang Furen jiaozao dao: “Yong bu zhao pian-you,
Wang Lady exasperated say use not RS have
dan yong zhao le, zai zhao bu zhao.”
but use RS CRS again find not RS
“ “When we don’t want it there’s plenty; when we do there’s none,” she
exclaimed in exasperation.”
c. Wang Furen ting le dao: “Hushuo!”
Wang lady hear PFV say nonsense
“ “Nonsense!” exclaimed Lady Wang.”
It is observed that the original Manner is also either omitted or realized in forms
other than Manner verbs in the translation. There are 32 instances in which Man-
ner, overtly described in the original by a pre-verb element, is totally omitted in
the translation, including the above-mentioned 7 instances, where the original
Manner is omitted while a new Manner is added in the translation by using an
English specific Manner verb. The other forms expressing Manner of speaking in
194 Dongqin Shi
the translation include adverbs, prepositional phrases, and -ing non-finite forms.
For example:
Yingchun han lei dao: “Yi-wo-shuo, jianglai zong-you yi san,
(6) a.
Yingchun in tear say in my opinion future have a part
buru ni geren qu ba.”
might as well you each go SFP
“Ying-chun said tearfully, “Since we have to part sooner or later, you
may as well go now.” ”
b. Fangguan xiao bian dao: “bing-bugan tiaosuo shenme.”
Fangguan smile plead say not dare incite a little
“ “I’d never dare!” pleaded Fang-kuan with a smile.”
c. Jiamu huitou dao: “Hou-er hou-er, ni bu pa
Jiamu turn head say monkey monkey you not afraid
xia ge shetou diyu?”
go cut tongue hell
“ “What a monkey you are!” cried the Lady Dowager turning to His-
feng. “Aren’t you afraid of going to the Tongue-Cutting Hell?” ”
In the English translation, the use of verbs in the synthetic way often alternates
with the analytic way. This is done in some cases for the purpose of adding variety
to narration especially when both synthetic and analytic means are available, as in
the case of (6a), where a noise verb, e.g. ‘sob’, incorporating both a speaking event
and a sobbing emotion could well translate the speaking event presented in the
original. Table 5 is a summary of the English translations of the Manner expres-
sions in the Chinese original.
ily the pragmatic force Message implies, can also be incorporated in specific com-
munication verbs, and use of such verbs often guarantees a more effective convey-
ance of Message. In the case of “He said, “Don’t go alone” ”, the specific illocutionary
force intended by the speech act realized in the direct quote is context-dependent
and thus capable of different readings in isolation: it could be understood as a plead-
ing, a warning or a suggestion. Use of a specific verb such as ‘beg’, ‘plead’, ‘suggest’,
‘warn’, etc. can nevertheless specify the pragmatic force the Message intends.
In addition to the Message information, some specific verbs can also express
information about the Speaker and the Addressee. While verbs like ‘agree’, ‘ask’,
‘promise’, etc. may not involve any interpersonal relationship between the two par-
ticipants of a conversation, verbs such as ‘beg’, ‘order’, ‘plead’, ‘remonstrate’ and so
on may indicate various relationships between them. A Speaker who ‘pleads’ is
normally the one in a powerless position against or at least in solidarity with the
Addressee, whereas a Speaker who ‘orders’ someone to do something should be
the one with power over the Addressee.
In the Chinese text the illocutionary force implied in a Message and/or infor-
mation about the Speaker and the Addressee are realized in two types of surface
structures: verbs and free combinations. The two types of structure are listed in
Table 6.
As shown in Table 6, for information related to the Message, Speaker and Ad-
dressee, Chinese resorts to verbs as frequently as to free combinations. In spite of
the presence of a general speaking verb in the structure, these free combinations
are still identified as indicating specific speaking events, for the preceding ele-
ment expresses specific information about the Message and/or the Speaker and
Addressee.
The free combinations are listed in three separate groups according to their
composition. In Group a, the preceding element of each combination is a free
monosyllabic morpheme, which can be used alone as a word, in Group b it is a
disyllabic word, which consists of two bound morphemes, and in Group c it is a
bound monosyllabic morpheme, which cannot be used alone in modern Chinese
(cf. Chao 1968:143). A noteworthy point is that these Chinese structures, which
are normally free in combination, are not comparable to the English phrasal verbs
as in the data, i.e. ‘go on’, ‘join in’ and ‘put in’, in each of which the two elements
form a semantic whole.
In the English translation, however, information about the Message and/or
the Speaker and Addressee is conveyed uniformly by verbs, which are listed in
Table 7.
As revealed by the figures in Table 7, 35 different English specific verbs of Con-
tent are used in 73 instances. While the Chinese text is marked by an alternation of
verbs and free combinations for the same type of speaking events and their related
semantic components, specific verbs are always used in the translation. Compare
the Chinese examples in (7a) and (7b) with their English translations:
(7) a. Bao-yu tingshuo, mang peixiao yang dao:
Pao-yu hear hurriedly smile beg say
“hao jiejie, kuai bie dasheng…”
good sister quick not loud
“ “Hush, good sister! Not so loud!” he (Pao-yu) begged.”
b. Jia Rong…baoyuan dao: “Zao dou buzhi zuo shenme de,
Jia Rong grumble say early all not know do what DE
zhehui-zi xunchen wo.”
now blame I
Communication verbs in Chinese and English 197
“Jia Jung…grumbled, “Why was this not thought of before? Now I’m the
one to take the brunt.” ”
In each of the above examples, the Chinese general speaking verb, dao (“say”),
indicates a speaking act while the preceding element specifies the content of the
direct quote. In the English translation of each example, however, Content is en-
coded in a specific verb.
A much greater number and variety of specific verbs of Content are used in
the translation than in the original. In many cases a communication verb that
does not encode Content in the original is translated into a Content specific one in
English. (8a) and (8b) are two such examples:
Baochai xiao dao: “Ba, ba, guai re de, shenme mei kan guo de
(8) a.
Baochai smile say no no pretty hot SFP what not see EM DE
xi…”
opera
“ “It’s too hot for me,” objected Pao-chai. “Besides, there aren’t any
operas I haven’t seen…” ”
b. Jia Zhen you xiang Jia Rong dao: “Ni zhan zhe zuo shenme?”
Jia Zhen then to Jia Rong say you stand DUR do what
“ “What are you standing there for?” Chia Chen snapped at his son.”
Like any other action, a speaking event has initiation, continuation or final point,
and there are also overlappings between two speakers. We find in our data that
both Chinese and English occasionally make reference to such aspectual informa-
tion about a speaking event, but again they do it in different ways.
In the Chinese text, Aspect, instead of being incorporated into a communica-
tion verb, is realized analytically by a function word, normally you (“continuing”,
“then”) or nai (“continuing”, “then”), going before a general or specific commu-
nication verb as a pre-verb modifier. In the data there are 20 tokens of you or nai
going before a speaking verb, of which 16 instances indicate a continuing speaking
event by the same speaker, e.g. (10a), and 4 instances an inchoative speaking event
following a non-speaking event, e.g. ‘kowtow … and take her leave’ as in (10b).
Lin Daiyu yishi jie
(10) a. bu-guo zhege hua lai, Baoyu
Lin Daiyu a while understand not this speech SFP Baoyu
you dao: ‘Zuo-er hai wei zhege du le jihui zhou…”
continue say yesterday still for this take PFV several oath
“Before she could fathom his meaning he went on: “Yesterday I took an
oath because of this…” ”
b. Siqi wufa, zhide han lei yu Yingchun ketou,
Siqi have no choice have to in tear to Yingchun kowtow
he zhong zimei gaobie, you xiang Yingchun ergen shuo…
to other sister take leave then to Yingchun ear say
“Ssu-chi had no alternative but to kowtow to Ying-chun and take her
leave of the other maids. In tears she whispered…”
ing one for a situation in the original where aspectual information is not overtly
expressed. The following is one more example from the three instances:
(11) Baoyu nai dao: “Congci xiu ti-qi…”
Baoyu continue say from now on not mention
“ “From now on”, he continued, “let’s say no more about it…” ”
5. Discussion
The analysis of the data presented above has revealed some striking contrasts be-
tween Chinese and English in their representation of speaking events and their
related components in written narrative.
First, in our data, Chinese shows a remarkably high frequency of the occur-
rence of general communication verbs, either used alone or in combination with
some other morpheme, word or phrase, over specific ones while English displays
an opposite tendency (cf. Tables 1 and 2). The figures presented and the compari-
son done in the previous section have shown that the ‘second tier’, a term proposed
by Slobin (1997:459), of communication verbs, which are expressive, detailed, and
elaborate in meaning, are much greater in number and diversity in English than in
Chinese. This is especially true with those English specific verbs that incorporate
Manner, Content and Response.
As was made clear by the discussion in the previous section, many more spe-
cific communication verbs are used in the translation than specific speaking events
are described in the original, either by verbs or by free combinations. For example,
different Manner verbs describing speech production amount to 26 times in the
translation (cf. Group SP in Table 4) while Manner in terms of speech production
is only overtly represented in 3 instances in the original (cf. Group SP in Table 3).
Similarly, the English specific verbs of Content (73 instances, cf. Table 7) greatly
outnumber the verbs and free combinations of Content in the original (30 in-
stances, cf. Table 6). Obviously there is addition of information in the translation
when English specific verbs are chosen for Chinese general ones. The question
is how the translators possibly achieve such transference. A comparison of the
original with its translation reveals that the translators achieve this by assessing the
whole context in which conversations take place when looking for an appropriate
English communication verb. They mainly rely on the semantic and pragmatic
messages conveyed in a particular speech, and also on such background infor-
mation as the interpersonal relationship between Speakers, the social status of a
Speaker, etc. For example, the contents of the speech in example (8a) give the read-
ing that the speech is an objection and therefore ‘object’, an English specific verb,
Communication verbs in Chinese and English 201
is the choice for the Chinese general verb dao (“say”), which itself does not reveal
any information about which particular type of speech an utterance is. Similarly,
in example (8b), an English specific verb ‘snap’, incorporating in it both a Manner
of speaking and information about the FEs of Speaker, Addressee and Message, is
used for the general verb dao (“say”), taking into consideration the relationship
between the Speaker and the Addressee as well as the Message conveyed in the
direct quote.
It is especially specific verbs encoding Manner of speech production and Con-
tent that are used on a much more frequent basis in the translation, for these two
types of information, as is believed, may be more intrinsic than any other semantic
element to a speaking event, and use of verbs encoding such messages could help
readers understand the source text better. On the other hand, however, there is also
omission of information in the translation, which is best illustrated by the omis-
sion of facial expression as an accompanying Manner of speaking. As was shown
in Tables 3 and 5, of the 53 tokens of facial expression, as many as 30 descrip-
tions are omitted in the translation, possibly for the reason that facial expression is
not so essential to a speaking event as compared with Manner of speech produc-
tion. Addition and omission of information in our translation data well suggest
that translation is not a one-to-one correspondence of individual lexical items. A
translator often has to assess the whole context in which narration advances for
suitable surface structures for ideas expressed in the source text. The assessment
of context involves a personal judgment and/or choice of the translator over what
information is important and what is not, which naturally leads to addition and
omission of information in the translation.
Second, Chinese and English enjoy different degrees of freedom in using verbs
of different domains or communication frames for direct quotes. As was already
presented in Table 4, among those Manner verbs used in the translation, a large
proportion belong to the noise frame of the communication domain or to the
perception domain, some of them denoting human sounds while others express
non-human sounds. As Urban and Ruppenhofer (2001:80) claim, “human verbal
communication necessarily involves the production and perception of sound”, and
therefore, noise or sound verbs are often used as communication verbs (Goossens
1995, Levin et al. 1997, Urban and Ruppenhofer 2001, Zwicky 1971). However, if
not occurring together with the FE of Message in the form of a direct quote or other
syntactic structure, they do not have any communicative content, and are strictly
speaking not inherently communication verbs, as in “Hearing the news he chuck-
led”, in which ‘chuckle’ simply denotes a vocal product without any verbal message.
Goossens (1995:164) argues that uses of noise or sound verbs for communication
are of three different types: purely metaphorical, metonymic, and metaphtonymic.
In the purely metaphorical use, there is a mapping from one domain onto the other,
202 Dongqin Shi
the donor domain being clearly distinct from the target domain, as in the case of
“ “You stupid ass”, he barked”, where the noise verb ‘bark’, possibly not indicating
that a speaker produced a barking sound while speaking, merely figuratively refers
to the way of ‘saying something in a sharp loud voice’. The metonymic use, howev-
er, is only restricted to human sound verbs. In the case of “ “Oh dear”, she giggled,
“I’d quite forgotten” ”, ‘giggle’ is a metonymic use if it is understood as indicating
that the speaker produced the utterance while giggling. ‘Giggle’ is still capable of
another interpretation, i.e. the speaker produced the utterance as if giggling, hence
a metaphtonymic use, for it is not clear, as Goossens (1995:164) explains, whether
the domain of giggling and that of linguistic action are conceived separately. Ac-
cording to Goossens’ descriptions of the uses of noise or sound verbs for human
communication, the human noise verbs used in the translation are mainly met-
onymic while those non-human noise verbs are metaphorical in nature.
Similarly, the verbs or phrasal verbs used in the English translation for aspec-
tual information about a speaking event are not prototypical expressions in the
communication domain, either. Rather they describe activities in general. Their
indication of the temporal stages for a speaking event is mediated by the presence
of the FE of Message, a direct quote in this study.
Unlike the situation in English, words of different domains or frames in
Chinese could never be used so freely, either metaphorically or metonymically.
Though the verbal way of describing a facial expression (e.g. lengxiao – “coldly
laugh”) or an emotional state (e.g. wuye – “sob”) can possibly introduce a direct
quote, it is observed that such use is, however, not typical of Chinese and is thus
not prevalent. In indicating a speaking event produced along with a facial expres-
sion, an emotion or a physical movement, the Chinese way is normally, if not
always, the use of a general verb, dao (“say”) or shuo (“say”), being preceded by an
element indicating a facial expression, emotion or physical movement. Likewise,
lexicalized items are not used in the Chinese text for the aspectual information of a
speaking event as expressed by ‘continue’, ‘go on’, etc. in English. Such information
is normally conveyed analytically, by an additional element, e.g. you (“continu-
ing”, “then”), nai (“continuing”, “then”) in our data, preceding a general or specific
speaking verb.
A third difference between Chinese and English, as far as our data are con-
cerned, lies in the heavy use of free combinations for specific events of speaking
in Chinese, even when equivalent lexical items are available in the language, while
this is never the case with English. In the Chinese data, free combinations, formed
by a general verb of speaking, i.e. dao or shuo, preceded by a free or bound mono-
syllabic morpheme or by a disyllabic word, are used frequently to indicate spe-
cific speaking events and some of their related semantic components. For Content
and Response, for example, lexical items are used 17 times (cf. Tables 6 and 8, 15
Communication verbs in Chinese and English 203
tokens of Content verbs plus 2 tokens of Response verbs) while free combinations
used in 22 instances (cf. Tables 6 and 8, 15 tokens for Content plus 7 tokens for
Response). This fact might lead one to the conclusion that Chinese is lacking in
lexical items that incorporate Manner, Content or Response. A closer look at the
Chinese lexicon reveals, however, that this is far from true. On the one hand, the
free monosyllabic morpheme or the disyllabic word in such a free combination
can be used independently as a lexical item for the same specific speaking event,
e.g. quan (“persuade”) for quan dao (“persuade say”), wen (“ask”) for wen dao (“ask
say”), baoyuan (“complain”) for baoyuan dao/shuo (“complain say”), etc. On the
other other, a lexical item equivalent in meaning to the bound morpheme in a free
combination is also readily available in the Chinese lexicon for the same specific
speaking event, e.g. bianjie (“explain”) for bian dao (“explain say”), huida (“an-
swer”) for da dao (“answer say”), hui dao (“answer say”) or hui shuo (“answer say”),
yangqiu (“beg”) for yang dao (“beg say”). The verbs bianjie, huida and yangqiu are
respectively a disyllabic equivalent to the bound morpheme preceding the general
verb of speaking in each of the above examples. In spite of the fact that Chinese is
not lacking in lexical items for specific speaking events, use of such free combina-
tions as found in our data, however, is pervasive in written narrative. By contrast,
such a style of representation in which a general verb of speaking is used to specify
in repetition a speaking event is never attested in our English data.
This representation style particular to Chinese may be traced to the restric-
tions on the specific communication verb when it is followed by different FEs. Su
(2004:24) claims that the presence of a general speaking verb, dao (“say”) or shuo
(“say”), is optional because the use of mechanics like quotation marks in writing
makes it possible to drop the general speaking verb without causing any ambigu-
ity. However, our observation is that when the FE Message follows a specific com-
munication verb in the form of an indirect quote, a general verb is not necessary
or must be omitted, whereas when the Message is realized in a direct quote, never-
theless, a general speaking verb is preferably present, functioning as an introducer
of the direct quote. In contrast, English specific verbs are not constrained in this
respect. Compare:
(12) a. Ta qiu wo bie zou.
he beg I not go
“He begged me not to go.”
b. Ta qiu wo shuo, “Jiejie, ni bie zou.”
he beg I say sister you not go
“He begged, “Please don’t go, Sister.” ”
On the other hand, such a representation style found in the Chinese text is fa-
cilitated by linguistic features specific to Chinese. In the Chinese morphological
204 Dongqin Shi
6. Conclusion
Though Chinese falls into the same typological category as English, according to
Talmy’s (1985, 2000) classification, there are differences between the two languages
in the capacity of lexical items conflating or incorporating semantic elements, and
in the narrative style in which a particular type of surface structure is preferred
for expression of meaning. Our comparison of a Chinese sample with its English
translation has shown that Chinese and English display some important contrasts
in their use of communication verbs in written narrative. First, in Chinese general
verbs of communication are preferred and specific verbs are only used on a limited
basis, while in English the opposite is true, with specific verbs being used far more
frequently than general ones. Second, in many cases, specific speaking events are
represented in Chinese by free combinations made up of a general speaking verb
preceded by a morpheme or word specifying a particular speaking event, even
though a lexical item for the same specific speaking event is readily available in its
lexicon. However, this is never found to be the case in the English corpus data.
Such being the case, a translator has to assess the whole context in which com-
munication takes place when translating a Chinese text into English. Specifically,
in translating communication verbs from Chinese into English, the translator of-
ten has to add information by choosing an English specific verb for a general one
Communication verbs in Chinese and English 205
in Chinese, taking into consideration clues provided by the context with regard
to the semantic and pragmatic messages of a conversation, the interpersonal rela-
tionships and so on. And he also has to use specific lexical items for the free com-
binations used in the Chinese version for various speaking events.
While the present study shows some striking differences between English
and Chinese in reporting direct quotes, it is tentative, however, with the data for
analysis being limited in various respects. A more general and convincing conclu-
sion may be arrived at only when a comprehensive study is made by comparing
original texts produced by different writers and/or at different periods of time with
their translations, preferably produced by different translators, for it is often the
case that language use varies significantly with writers, translators, writing times,
etc. On the basis of such comparison, one needs to further examine to what extent
the personal style of a writer or translator is framed by the specific features of the
languages under analysis.
Notes
* This research was undertaken while the author was a visiting scholar in the School of Lan-
guages and Comparative Cultural Studies, University of Queensland, Australia. Generous sup-
port from the school is hereby gratefully acknowledged. The author is also grateful to Profes-
sor Ping Chen at the University of Queensland, Professor Juquan Wang at Shanghai Maritime
University, and two anonymous reviewers of Languages in Contrast for their valuable comments
and suggestions on an earlier version of this paper. She is solely responsible for possible errors
or mistakes of facts or viewpoints herein.
1. In Modern Chinese a greater proportion of the lexicon is disyllabic words, formed by two
monosyllabic morphemes, normally lexically meaningful if not both, being combined into the
same sequence. And it is often the case that the same morpheme can occur in different disyl-
labic words. As such, it might remain an open question whether Chinese disyllabic words can
be described as conflating different semantic components in their lexical form in the same way
as do single morpheme words in language.
2. http://framenet.icsi.berkeley.edu/
3. Frames in the FrameNet are organized into domains, which are very general categories of
human experience and knowledge, and only serve as groupings of semantic frames, thus slight
and indirect in theoretical significance (Johnson et al. 2001:16). The domains to be covered in
the FrameNet include a wide range of conceptual categories, varying from health care, chance,
perception, communication, transaction, time, space, body, motion, life stages, social context to
emotion and cognition.
5. We restrict the physical movement accompanying a speaking event to that of the head, and
exclude those performed by other parts of the body, which describe even less intrinsic Manner
to speaking.
6. Some verbs used in the translation conflate or incorporate more than one semantic com-
ponent with the speaking event. For example, ‘grumble’ denotes an act of complaining that is
uttered in a low and indistinct voice while ‘snap’ indicates an utterance of scolding produced in a
loud voice. Therefore, the same verbs can appear in different groups in the tables in this paper.
References
Baker, C. F., Fillmore, C. J. and Cronin, B. 2003. “The Structure of the FrameNet Database”. In-
ternational Journal of Lexicography 16(3): 281–296.
Chao, Y. R. 1968. A Grammar of Spoken Chinese. Berkeley/Los Angeles: University of California
Press.
Chu, C. Z. 2004. “Event Conceptualization and Grammatical Realization: The Case of Motion in
Mandarin Chinese”. Unpublished Ph.D. dissertation, University of Hawai’i at Manoa.
Fillmore, C. J. 1982. “Frame Semantics”. In Linguistics in the Morning Calm, The Linguistic Soci-
ety of Korea (ed.), 111–137. Seoul: Hanshin Publishing Company.
Fillmore, C. J. 2003. “Double-Decker Definitions: The Role of Frames in Meaning Explanations”.
Sign Language Studies 3(3): 263–295.
Fillmore, C. J., Johnson, C. R. and Petruck, M. R. L. 2003. “Background to FrameNet”. Interna-
tional Journal of Lexicography 16(3): 235–250.
Goossens, L. 1995. “Metaphtonymy: The Interaction of Metaphor and Metonymy in Figura-
tive Expressions for Linguistic Action”. In By Word of Mouth, L. Goossens, P. Pauwels, B.
Rudzka-Ostyn, A.M. Simon-Vandenbergen, J. Vanparys (eds.), 159–174. Amsterdam: John
Benjamins.
Johnson, C. R., Fillmore, C. J., Wood, E. J., Ruppenhofer, J., Urban, M., Petruck, M. R. L. and
Baker, C. F. 2001. “The FrameNet Project: Tools for Lexicon Building” (Technical Report).
International Computer Science Institute, Berkeley, CA. Available from: http://ccl.pku.edu.
cn/doubtfire/semantics/FrameNet/theory/FrameNet_book.pdf. [accessed July 2008]
Levin, B., Song, G. and Atkins, B. T. S. 1997. “Making Sense of Corpus Data: A Case Study of
Verbs of Sound”. International Journal of Corpus Linguistics 2(1): 23–64.
Packard, J. L. 2000[2001]. The Morphology of Chinese: A Linguistic and Cognitive Approach. Bei-
jing: Foreign Language Teaching and Research Press & Cambridge: Cambridge University
Press.
Rojo, A. and Valenzuela, J. 2001. “How to Say Things with Words: Ways of Saying in English and
Spanish”. Meta 44 (3): 467–477.
Ruppenhofer, J., Ellsworth, M., Petruck, M. R. L. and Johnson, C. R. 2005. “FrameNet: Theo-
ry and Practice” International Computer Science Institute, Berkeley, CA. Available from:
http://framenet.icsi.berkeley.edu/index.php?option=com_wrapper&Itemid=126. [accessed
July 2008]
Communication verbs in Chinese and English 207
Slobin, D. 1996a. “From ‘Thought and Language’ to ‘Thinking for Speaking’”. In Rethinking Lin-
guistic Relativity, J. J. Gumperz and S. C. Levinson (eds), 70–96. Cambridge: Cambridge
University Press.
Slobin, D. 1996b. “Two Ways to Travel: Verbs of Motion in English and Spanish”. In Grammatical
Constructions: Their Form and Meaning, M. Shibatani and S. A. Thompson (eds), 195–219.
Oxford: Clarendon.
Slobin, D. I. 1997. “Mind, Code, and Text”. In Essays on Language Function and Language Type:
Dedicated to T. Givón, J. Bybee, J. Haiman and S. A. Thompson (eds.), 437–467. Amster-
dam/Philadelphia: John Benjamins.
Slobin, D. I. 2004. “How People Move”. In Discourse Across Languages and Cultures, C. L. Moder
and A. Martinovic-Zic (eds), 195–210. Amsterdam/Philadelphia: John Benjamins.
Su, L. I. 2004. “Subjectification and the Use of the Complementizer SHUO”. Concentric: Studies
in Linguistics 30 (1): 19–40.
Talmy, L. 1985. “Lexicalization Patterns: Semantic Structure in Lexical Forms”. In Language
Typology and Semantic Description Vol. III: Grammatical Categories and the Lexicon, T.
Shopen (ed), 57–149. Cambridge: Cambridge University Press.
Talmy, L. 1991. “Path to Realization: A Typology of Event Conflation”. Proceedings of the Sev-
enteenth Annual Meeting of the Berkeley Linguistics Society. Berkeley: Berkeley Linguistics
Society. 480–519.
Talmy, L. 2000. Toward a Cognitive Semantics Vol. 2. Cambridge, MA: The MIT Press.
Urban, M. and Ruppenhofer, J. 2001. “Shouting and Screaming: Manner and Noise Verbs in
Communication”. Literary and Linguistic Computing 16(1): 77–97.
Wang, W. B. and Zhou, C. B. 2004. “A Comparison of ‘LOOK’ Type Verbs in English and Chi-
nese: Lexical Meanings and Lexicalization”. Foreign Language Teaching and Research 6:
412–419.
Zwicky, A. M. 1971. “In a Manner of Speaking”. Linguistic Inquiry 2: 223–232.
Data
Author’s address
Dongqin Shi
Foreign Languages College,
Shanghai Maritime University
1550 Pudong Avenue
Shanghai
China 200135
dqshi@cfl.shmtu.edu.cn