Semantic prosody: A new perspective on lexicography
YU Ping-fang1 ; CAI Jing-chun2
(1. Chinese-based Centre for Lexicography Directly under the National Education Ministry,
Yantai, 264025, China;
2. School of Foreign Languages, Ludong University, Yantai, 264025)
Abstract：Semantic prosody is the associative meaning resulting from its collocates and is partially recorded in English learner’s dictionaries. In spite of the proliferation of the researches on semantic prosody, most of the practice of dictionary-making still keeps it ill-marked or even untouched. The present research argues that semantic prosody is a scale, and the weaker form, which is not as easy to be perceived as the stronger one, should be also treated systematically in dictionaries, especially in those designed for EFL learners.
Key words: learner’s dictionary; semantic prosody, connotation
1. Introduction:
The term “prosody”, coined by Firth (1957), traditionally refers to the phonological coloring which spreads beyond segmental boundaries. Inspired by Firth, Sinclair, the chief editor of Collins-COBUID English Dictionary, finds that “many uses of words and phrases show a tendency to occur in a certain semantic environment”(Sinclair, 1991, p112), which means there does exist “some kind of spreading of connotational coloring beyond single word boundaries, which is called “semantic prosody”(Partington, 1998，p. 68）.
Researches on semantic prosody can be classified into two main divisions. The first focuses on the semantic prosody at textual level, exploring the perlocutionary effect caused by semantic prosodic clash. Louw (1993, p.169), for example, holds that writers who consciously diverge from the expected profiles of semantic prosody are usually with ironic and insincere intent, in which case readers might detect a semantic gap between what the writers are apparently saying and what they really mean. The second group, however, concentrates on the semantic prosody occurring at the syntactic level by analyzing the favorable or unfavorable connotation of the node words, and this kind of connotation results from the collocates of the node words and can be detected and systematically marked thanks to the establishment and utilizing of corpus in computer-readable format ( Partinton, 1996). Researches of the two kinds are mainly based on English data, and part of their findings is incorporated in English learner’s dictionaries. In China proper, the study of semantic prosody is hitherto confined to the field of second language acquisition（cf Pan Fan & Feng Yuejin, 2003; WANG Haihua & WANG Tongshun, 2005; WEI Nai-xing, 2006) and aims to explore how English semantic prosody affects Chinese EFL learners. Other researchers, including JI Yu-hua & WEI Jian-ping（2000），WANG Ze-peng（2007）etc, emphasize the need of incorporating semantic prosody into bilingual dictionaries; and there leaves large room in exploring how semantic prosody is or should be treated in Chinese learner’s dictionaries. The present research examines how semantic prosody is revealed in both English and Chinese learner’s dictionary, aiming to provide a potential and systematic way of treating semantic prosody in Chinese dictionary-making.
2. Term-differentiation: semantic prosody, connotation and complimentary/ pejorative words
Language, from a perspective of cognitive linguistics, is of subjectivity (SHEN Jia-xuan,2001); so it is with semes—the integral part of languages. The subjectivity of semes are usually shown in 3 ways: (1) an favorable or unfavorable assessment is embodied by the basic meaning of semes and is traditionally labeled as complimentary or pejorative words in dictionaries; (2) favorable or unfavorable attitudes are hinted by the associative meanings, which, in modern semantics, are called connotative meanings; (3) in a linear combination, the node words, affected by its collocates which overwhelmingly designate desirable/undesirable things, retain a positive or negative kind of semantic aura, which is called “semantic prosody”. E.g. “bent on”, which usually collocates with unfavorable items—destroying, harrying, mayhem etc, is a word imbued with an unfavorable semantic prosody.
2.1 Semantic prosody vs complimentary and pejorative words
According to the evaluation that node semes express, the semantic prosody is classified as favorable, unfavorable and mixed（Stubbs, 1996, p176）. There appears a tendency to draw parallels between favorable prosody and complimentary words, unfavorable prosody and pejorative ones, but in fact there is subtle discrepancy between them. Complimentary and pejorative words are those imbued with overt assessment, which is endowed by the speech community according to a kind of fixed social linguistic contract. That is to say, the complimentary or pejorative value is fossilized in the word meaning, constituting a kind of “feeling conceptualization”(SHI An-shi, 1993, p15). Semantic prosody, however, can hardly be perceived without its company considered. That is to say, although their collocates are apt to show a certain kind of evaluation, the node words are neutral in isolation and the neutrality can be “dissolved” when the node words are combined with their “biased” collocates and have to be harmony with its companies.
It’s clear that the complimentary and pejorative words inherently show a kind of assessment, which is a static and overt meaning, needing no support of linguistic context; while the value hinted by semantic prosody can’t be found without a linear combination, hence it is dynamic and covert, waiting to be made clear by data mining technology..
2.2 Semantic prosody vs connotation
Connotation is “the secondary implication” of a linguistic unit” (Lyons, 1977, P278); it’s “the communicative value that an expression has by virtue of what it refers to, over and above its conceptual content” (Leech,1974, p12). Typologically connotation is classified into 9 types (ZHANG Zhi-yi & ZHANG Qing-yun: 2001, p46), among which the type concerning feeling，attitudes and assessment are called expressive connotation. Semantic prosody, being the evaluation endowed by the collocates in a linear micro-context, is a subtype of expressive connotation.
Based on the analyses in 2.1and 2.2, it’s safe to say that semantic prosody, being a kind of semantic overflow occurring in the syntactic combination; is a special kind of selection restriction, in which the collocates demands a semantic harmony with which the node words have to keep.
3. The treatment of semantic prosody in English and Chinese learner’s dictionaries
3.1 The methods of labeling semantic prosody in dictionaries
The ways of treating semantic prosody in dictionaries are as follows:
(1) Label-attaching. Label attaching, by attaching an overt linguistic label to a lemma, is the most direct way to make semantic prosody revealed. E.g.
當道：② 掌握（政權）（含貶義）;圖謀：暗中策劃（多含貶義）;得勢：得到權柄和勢力（多用於貶義）。
(2) Showing the semantic prosody through definitions. E.g.
【交代】④ 把錯誤或罪行坦白出來。
(3) Marking the semantic prosody in parenthethses. E.g.
【獲得】得到（經驗、成績等抽象事物）。
(4) Hinting semantic prosody dictionary examples. E.g.
【飽嘗】長期地經受：飽嘗離別的痛苦|二十年來，他們飽嘗創業的艱辛| 只有飽嘗過生活艱難的人，才會珍惜今天的好日子。
Of the four ways of showing semantic prosody, the first one is the most obvious while the fourth one is the vaguest, which demands a high level of proficiency in language. Generally speaking, pedagogical dictionaries designed for L2 learners should be as user-friendly as possible.
The four methods can be used simultaneously in prosody-treating in dictionaries, which, in one hand, shows the complexity of semantic prosody and its inconsistent treatment in dictionaries; and on the other hand, the four-in-one way of semantic-treating slows down or even blocks the consulting process of the language learners. Therefore, a single and definite way of prosody-marking is more convenient for learners to get easy access to the linguistic information they are looking for.
3.2 Revealing semantic prosody in English and Chinese learner’s dictionaries
The large-scale research into semantic prosody is originally attributed to the broad availability of texts in electronic form and the ability to store large amounts of data economically, which has changed the practice of dictionary-making. Partington (1996) compares the treatment of the negative prosody of “set in” and “peddle”, and finds that the 3 general dictionaries designed for native speakers haven’t marked the prosody, while six of the seven pedagogical dictionaries for EFL learners have shown the semantic prosody in different depth.
In the present paper, 8 words with strong semantic prosody are selected as test data to investigate the treatment of their semantic prosody in both encoding and decoding dictionaries. Both English and Chinese dictionaries are analyzed as samples. Among the seven English dictionaries, five of them, including CIDE, LDOCE, COBUILD, OALD, MEDAL are designed for EFL learners and the other 2, viz COD and RHW, are for native speakers.
Table 1: The treatment of semantic prosody in part of English dictionaries
CIDE LDOCE COBUILD OALD MEDAL COD RHW
cause yes yes yes yes yes no no
provide yes yes yes yes yes no no
commit yes yes yes yes yes no no
happen no no yes no no no no
incur yes yes yes yes yes no no
set in yes no yes yes yes no no
Table 1 shows that there is a growing tendency that the dictionaries for EFL learners do better in the treatment of semantic prosody. Of all the pedagogical dictionaries, COBUILD is ranked number one in order of the quantity of prosody-revealing, which, as we suppose, may be attributed to its sentence definition mode.
The difference of prosody-revealing between the two groups of English dictionaries lies in the following reasons: Firstly, semantic prosody, being “invisible” to our naked eyes, can be clearly highlighted in the concordance lines extracted from large electronic corpora. And all EFL dictionaries, not like those for native learners, are compiled on the basis of large linguistic corpora, which is a great advancement in modern dictionary-making. Secondly, EFL dictionaries are designed to help with encoding tasks—the production of texts. Correspondingly, EFL dictionaries are duty-bound to help the L2 learners to achieve semantic harmony in their encoding tasks. Dictionaries for native speakers, however, are generally designed to help the comprehension of texts, which don’t need as much collocation information as text-production. Last but not the least, dictionaries for native speakers and L2 learners have adopted different approaches to define or explain their lemmas: the former ones traditionally provide a synonym or some near-synonyms, leaving the collocation information untouched because the lexicographer think that collocation rules constitute a internal part of native speakers’ linguistic institution; while the latter tend to explain the lemma from a multifarious angles, including the restriction of semantic prosody, to help the foreign learners to produce text correctly.
Next, we’ll examine how Chinese dictionaries deal with the semantic prosody. What is selected as an investigation domain is the verbs denoting “suffering”, which often co-occur with bad things and are undoubtedly imbued with negative semantic prosody. The dictionaries involved are as follows: Dictionary of Modern Chinese (《現代漢語詞典》), Dictionary of 8000 HSK Words (《HSK漢語8000詞詞典》), A Learner’s Dictionary of Contemporary Chinese(《當代漢語學習詞典》) and Commercial Press Learner’s Dictionary of Chinese (《商務館學漢語詞典》). Henceforth the afore-mentioned dictionaries are referred to as DMC, HSK 8000, LDCC and CPLDC respectively.
Table 2: The revealing of the semantic prosody of “suffering” verbs in Chinese dictionaries
DMC 8000 HSK LDCC CPLDC
挨 no yes yes yes
遭受 no yes yes yes
經受 no yes yes no
承受 no yes no yes
蒙受 no yes yes /
忍受 no yes no yes
It can be seen from Table 1 that in Chinese dictionary-making, the active dictionaries do better in making the covert semantic prosody overt, which is in line with Partington’s finding concerning the semantic-treating in English dictionary making.
From the statistical analyses of the semantic prosody in English and Chinese dictionaries, we argue that semantic prosody forms a scale. E.g, the verbs denoting sufferings in Chinese has a very strong negative prosody, which can easily be perceived by native speakers and are marked in most of EFL and CFL dictionaries. However, some semantic prosody is implicit or far beyond native speakers’ linguistic awareness, which makes it a hard nut to crack for dictionary compilers. Next, we’ll show how Chinese and English dictionaries deal with the covert semantic prosody of “build up” in English and “CHU” in Chinese.
The phrase “build up”, when used as an intransitive phrasal verb, co-occurs with “frustration”,“anger”,“resistance”,“resentment”,“pressure”etc. and correspondingly it is imbued with a negative semantic prosody. However, when it is used as a transitive phrasal verb, the objects of “build up” include “strength”, “trust”, “confidence”, etc., and it is imbued with a positive semantic prosody. Compared with“commit”and“suffer”,the semantic prosody of “build up” is more covert and harder to be perceived by native speakers and lexicographers. Table 3 is the treatment of the semantic prosody of “build up” in some English dictionaries.
Table 3: The revealing of the semantic prosody of “build up(vt & vi)” in English dictionaries
CIDE LDOCE COBUILD OALD MEDAL COD RHW
build up (vi) no no no no no no no
build up (vt) yes no yes no no no no
Table 3 shows that none of the seven English dictionaries shows the negative semantic prosody of the phrase “build up”; while the positive one of this collocation is marked in 2 dictionaries. To summarize, although English EFL dictionaries are in the leading position among the worldwide dictionary-making, there is still large room left for English dictionaries to improve their prosody-marking.
“CHU”(出) is a polysemous word with a number of definitions. Among these definitions, the sixth one is:出：產生；生長：出人才| 出成果|東北出大米|這種汽車是哪兒出的？the seventh definition is：出：發生：出麻煩| 出事故|誰能保證不出錯呢|出了問題不要互相埋怨。By consulting the concordance lines in National Chinese Corpus, we know that the sixth definition of “CHU” (出) is imbued with a neutral semantic prosody and the seventh with a negative semantic prosody. Let’s analyze how Chinese dictionaries mark the semantic prosody of the Chinese word＂CHU（出）”.
Table 4: The revealing of the semantic prosody of “出” in Chinese dictionaries
DMC HSK8000 LDCC CPLDC
出（produce） no no no no
出（happen） no no no yes
CPLDC is the only Chinese dictionaries which shows the semantic prosody of “chu” by way of providing exemplifications; and the rest three haven’t attached any labels for the semantic prosody. Overall, there is hardly any definite trace of helpful hints to keep the learners alert to their semantic prosody in Chinese dictionaries. We can safely draw a conclusion now that on the scale of semantic prosody, the seme of the highest level are tended to be recorded explicitly in dictionaries in different ways, viz, by label attaching, definitions or exemplifications; while those of lower levels on the scale, are seldom or never treated. There causes an embarrassment: On one hand, L2 learners, who have no linguistic awareness of the target language, can’t develop the knowledge of semantic prosody without explicit instruction and are more liable to make mistakes when using the words whose semantic prosody is covert, in this way, semantic prosody is one of the information categorization that learners need most; and on the other hand, because of the inherent implicitness of semantic prosody, the dictionary-compliers tend to leave these tough nuts uncracked, which makes semantic prosody difficult to ascertain. Therefore how to fill the gaps between dictionary users and dictionary compilers is top of the agenda of metalexicography.
Conclusion:
Semantic prosody, remaining hidden from our perception and inaccessible to our intuition in large measure and for thousands of years, requires their just attention from lexicographers (Louw 1993, p.173). Generally speaking, the revealing of semantic prosody in both English and Chinese dictionary-making is far from ideal. What accounts for this are as follows: Firstly, semantic prosody, which tends to remain inherently hidden to the lexicographer’s “naked eyes”, are not explicit enough to be retrieved reliably through introspection (Partington, 1996, p68) without the aid of corpus and computerized language-analyzing software; secondly, some dictionary-writers are not well equipped with up-to-date linguistic theories, which inevitably hinders their lexicographic performance. The most feasible way now to extract semantic prosody is to process the numerous concordance lines with corpus-based linguistic evidence and have it well-documented in dictionaries, esp. those complied for non-native learners.
References：
JI Yu-hua & WEI Jian-ping. 2000. Researching semantic prosody: Its objectives, methods and application. Journal of Xiamen University (3), 63-9.
Leech, G. 1974. Semantics. Harmondsworth: Peguin.
Louw, B. 1993. Irony in the text or insincerity in the writer?-The diagnostic potential of semantic prosodies. In Baker et al (eds) Corpus Linguistics and Translation Studies. Routledge.
Lyons. 1977. Semantics. Cambridge: Cambridge University Press.
Partington, A..1996. Patterns and meanings: Using corpora for English language research and teaching . Amsterdam: John Benjamins Publishing Company.
PAN Fan & FENG Yue-jin. 2003. Corpus-based analysis of semantic prosody and its applications. Comtemporary Linguistics (4), 359-66.
Sinclair, J. 1991.Corpus, concordance and collocation. Oxford: Oxford University Press.
SHEN Jia-xuan. 2001. The subjectivity and subjectivisation of language.Foreign Language Teaching and Research (4), 268-75.
SHI An-shi.1993. On semantics. Beijing: The Commercial Press.
WANG Hai-hua & WANG Tong-shun. 2005. A contrastive study on the semantic prosody of CAUSE. Modern Foreign Languages (3), 297-309.
WANG Ze-peng. 2007. A study of semantic prosody and Chinese rhetoric. Journal of Yunnan Normal University (4), 83-88.
WEI Nai-xing. 2006. A Corpus-based contrastive study of semantic prosody in learner English. Foreign Lauguage Research (5), 50－56.
ZHANG Zhi-yi & ZHANF Qing-yun. 2001. Lexical semantics. Bejing: The Commercial Press.
