A SOCIAL ACCOUNT OF TWO UNUSUAL USES OF THIRD PERSON PRONOUNS IN HONG LOU MENG*

By Cher Leng Lee (National University of Singapore) and Hongyin Tao (Cornell University)
* In Proceedings of the Joint Meeting of NACCL 7 and ICCL 4. T. Cheng, Y. Li and H. Zhang, editors. LA: GSIL, USC. 179-188.
You need a GB Chinese system to view the examples cited in the paper.

1. Introduction

Popular approaches to the issue of anaphora are either limited within the sentence (as in various formal grammatical theories) or looking at the linear properties of discourse (much as in Givón's earlier framework known as the Topic Continuity theory (Givón 1979)) . In some more recent studies, discourse structure (see Fox 1982 and Liang Tao 1993) and perspective taking (see Li and Zubin 1990) are shown to be relevant. In this paper, we will demonstrate that there are other discourse factors which are relevant for explaining the choice of anaphoric forms in text. Specifically, we will give a social account of two kinds of unusual uses of third person pronouns in the Chinese literary classic, Hong Lou Meng (HLM), and show how the factors of social status of the speaker and of the referred entity, discourse saliency of referents, and speakers' attitude towards the referent affect the use of pronouns in written literary texts as represented by HLM.

Let us first give examples to illustrate what we mean by "unusual uses of third person pronouns" in Hong Lou Meng.

The first type can be called pronouns without antecedents. In normal cases, as is well known, a pronoun is used when there is a prior mention of the referent. In Hong Lou Meng, however, we find many instances of uses of third person pronouns which do not have a clearly identifiable antecedent to which the pronoun co-refer. This is shown in example (1). 

(1)  ±¦ÓñµÀ£º¡°ÎÒÓиöÖ÷Ò⣺ϮÈËÉÏÔÂ×öÁËÒ»ÌõºÍÕâ¸öһģһÑùµÄ£¬ËûÒòÓÐТ£¬
       Èç½ñÒ²²»´©¡£¾¹ËÍÁËÄã»»ÏÂÕâ¸öÀ´£¬ÈçºÎ£¿¡±ÏãÁâЦ×Åҡͷ˵£º¡°²»ºÃ¡£
       ËûÃÇÌÈ»òÌý¼ûÁ˵¹²»ºÃ¡£¡± £¨62»Ø£©
Here, the plural third person pronoun ËûÃÇ (they) is used, but there is no textual clue as to what it is referred to by this pronominal form, although we can infer that it refers to the people associated with the protagonists in the novel.

The second type of unusual uses of third person pronouns has to do with the mismatch in number between pronominal forms and referents, mostly involving a singular third person pronoun denoting plural entities. This is shown in example (2). 

(2)  ÕÔÒÌÄïÎÞ·¨£¬Ö»µÃͬËûÈýÈ˳öÀ´£¬¿ÚÄÚÓÌ˵³¤Ëµ¶Ì¡£Ì½´º±ã˵£º¡°ÄÇЩСѾ
       Í·×ÓÃÇÔ­ÊÇЩÍæÒâ¶ù£¬Ï²»¶ÄØ£¬ºÍËû˵˵ЦЦ£»²»Ï²»¶£¬¿ÉÒÔ²»ÀíËû¡£±ãËû
      ²»ºÃÁË£¬Ò²Èçͬè¶ù¹·¶ù×¥Ò§ÁËÒ»ÏÂ×Ó£¬¿ÉË¡¾ÍË¡£»²»Ë¡Ê±Ò²Ö»¸Ã½ÐÁ˹ܼÒ
      ϱ¸¾ÃÇȥ˵¸øËûÈ¥Ô𷣡£ºÎ¿à×Ô¼º²»×ðÖØ£¬´óߺСºÈ£¬Ò²Ê§ÁËÌåͳ¡£¡±  
      £¨60»Ø£©
In this example, the pronoun is in the singular form Ëû (s/he), but the antecedents clearly involve more than one entity (СѾͷ×ÓÃÇ)

2. Hypothesis

In this study, we propose two separate hypotheses to account for the use of third person pronouns in Hong Lou Meng. Our hypothesis for the first type of unusual uses of third person pronouns is the negligibility principle, which states as follows:
Pronominalized referent(s) without prior mentions tend to be socially inferior or non-salient in the discourse context.
For the second type of unusual uses of third person pronouns, we propose the following:
The use of third person pronouns is not just another form of referring, but is an active tool to accomplish certain social interactional effect.
In both cases, a social account is needed. Let's now examine the two types in turn.

3. Pronouns Without Antecedents

In the 80 chapters we have randomly checked so far, we found 35 instances of third person pronouns without antecedents. The text we used is the 120-chapter version published by Renmin Wenxue Chubanshe (ÈËÃñÎÄѧ³ö°æÉç), Beijing, 1990. Table 1 shows the distribution of antecedent-less third person pronouns with regard to singular and plural forms. 
           Table 1. Third person pronouns without antecedents

                      ËûÃÇ (they)       30    (86%)        
                      Ëû   (s/he)        5    (14%)    
                      Total             35    (100%)
As can be seen from Table 1, the majority of the third person pronouns without an antecedent occurs with a plural form.

As we mentioned earlier, our proposal is that pronouns without antecedents are due to the negligibility principle. This principle includes two parts. In the first, a referent denoted by the pronoun is negligible due to the fact that the referent is socially inferior, or low ranking in the feudal hierarchy as represented in Daguan Yuan (´ó¹ÛÔ°). In the second, the referent is negligible because it is not salient in the discourse context, i.e., it is not necessary to keep track of the identity of the reference in discourse (see Du Bois 1980 for a discussion of this point). We have evidence to support both of our claims.

Let's first present evidence for the first part of our hypothesis, that is, the referent associated with the third person pronoun is socially inferior. Our first piece of evidence comes from the semantic content of the referent in question. A clear tendency is for the pronoun to refer to the personal attendants or entourages in ´ó¹ÛÔ°. As illustrated in (3). 

(3)  »°ÓÌδÁË£¬Ö»ÌýÍâ¼ä¡°¹¾ßË¡±Ò»Éù¡£¼±Ã¦¿´Ê±£¬Ô­À´ÊÇÒ»¸öСѾͷ×ø×Å
       ´òíһͷײµ½±ÚÉÏ£¬´ÓÃÎÖоªÐÑ£¬È´ÕýÊÇÇçö©ËµÕ⻰֮ʱ£¬ËûÕúÕúµÄÖ»
       µ±ÊÇÇçö©´òÁËËûÒ»ÏÂ×Ó£¬Ëì¿Þ×ÅÑë˵£º¡°ºÃ½ã½ã£¡ÎÒÔÙ²»¸ÒÁË£¡¡±ÖÚÈ˶¼
       ЦÆðÀ´¡£±¦ÓñæȰµÀ£º¡°ÈÄËû°Õ¡£Ô­¸Ã½ÐËûÃÇ˯ȥ¡£ÄãÃÇÒ²¸ÃÌæ»»×Å˯¡£¡±
       (73»Ø£©
Here we can infer that this antecedent-less plural pronoun refers to those personal attendants who are accompanying ±¦Óñ and whose rankings are probably even lower than Ççö©.

We found that in our data third person pronouns without an antecedent occur more than 70% of the time involving a referent of low ranking persons in the story, as indicated in Table 2. 

              Table 2. Referential types of  antecedent-less third person pronouns 

              Reference to lower ranking persons              26        74%
              Reference to higher ranking persons              1       0.3%
              Reference to both high and low ranking person    2       0.6%
              Reference to others                              6        17%
              Total                                           35       100%
Our second piece of evidence in favor of our hypothesis is that these third person pronouns are typically produced by higher ranking protagonists in the novel to refer to their personal attendants or maids (or servants). Consider example (4). 
(4) ±¦îÎЦµÀ£º¡°Âäʲô²»ÊÇÄØ£¿Ò²ÊÇÈËÖ®³£Çé¡£ÄãÓÖ²»ÔøÂô·ÅÁËÔô¡£ÒÀÎÒµÄ
    Ö÷Ò⣬Ҳ²»±ØÌíÈ˹ýÈ¥£¬¾¹°ÑÔÆѾͷÇëÁËÀ´£¬ÄãºÍËûסһÁ½ÈÕ£¬Æñ²»Ê¡Ê£¿"
    ÓÈÊϵÀ£º¡°¿ÉÊÇ£¬Ê·´óÃÃÃÃÍùÄÇÀïÈ¥ÁË£¿¡°  ±¦îεÀ£º¡¸ÎҲŴò·¢ËûÃÇÕÒÄã
    ÃÇ̽ѾͷȥÁË£¬½ÐËûͬµ½ÕâÀïÀ´£¬ÎÒÒ²Ã÷°×¸æËßËû¡£¡±Õý˵×Å£¬¹ûÈ»±¨£º
   ¡°ÔƹÃÄïºÍÈý¹ÃÄïÀ´ÁË¡£¡±£¨75»Ø£©
In this example, the underlined third person pronoun (ËûÃÇ) referring to the attendants is produced by ±¦îÎ, a member of the mandarin family and a leading character of the novel.

Of course the feudal hierarchy in ´ó¹ÛÔ° is complex. Some attendants, such as Ï®ÈË and ×Ͼê, rank higher than the other attendants whose duties are mainly of labor work. These higher ranking attendants can also use third person pronouns without clear antecedents to refer to referents lower than them. In (1), for example, we have seen a case produced by ÏãÁâ who speaks to ±¦Óñ. Next is an example from ÐåéÙ who is talking to her master Ó­´º. 

(5)  ÐåéÙ ÓÖÆøÓÖ¼±£¬Òò˵µÀ£º¡°¹ÃÄïË䲻ţ¬ÎÒÊÇ×öʲôµÄ£¿°Ñ¹ÃÄïµÄ¶«Î÷
       ¶ªÁË£¬Ëûµ¹ÀµËµ¹ÃÄïʹÁËËûÃǵÄÇ®£¬ÕâÈç½ñ¾¹Òª×¼ÕÛÆðÀ´£¡ÌÈ»òÌ«Ì«ÎÊ
       ¹ÃÄïΪʲôʹÁËÕâЩǮ£¬¸ÒÊÇÎÒÃǾÍÖÐÈ¡ÊÆ£¿Õ⻹Á˵ã¿¡±£¨73»Ø£©
An examination of the data shows that such pronouns are mostly produced by higher ranking characters in the story. This is clearly shown in Table 3, which summarizes the background of the speakers who produce such pronouns. 
              Table 3. Background of speakers producing the antecedent-less pronouns

                 By  members of the mandarin family       24      69%
                 By higher ranking attendants             10      29%
                 By low ranking attendants                 0       0%
                 No speaker  involved                      1       2%
Table 3 shows that there is a clear social dimension involved in the use of the third person pronouns: higher ranking characters are free to use pronouns to refer to lower ranking characters, while lower ranking characters are in no position to use them either to refer to higher ranking characters or to refer to characters of their equal rank, and not a single case of the latter type is found in our data.

We believe that the above two types of evidence show convincingly that the use of antecedent-less pronouns has a social motivation, having to do with ways in which higher ranking characters refer to lower class characters. We now turn to the second part of our hypothesis, that is, antecedent-less pronouns may also occur when referents are not salient in the discourse context.

The non-salientness of referents is best illustrated by the fact that some of those characters are non-referential, that is, the referent does not possibly exist in the story world. This is shown in (6). 

(6)  ¼ÖĸÏëÁËÒ»Ï룬ҲЦµÀ£º¡°¿É¼ûÎÒÈç½ñÀÏÁË£¬Ê²Ã´Ê¶¼ºýÍ¿ÁË¡£¿÷ÁËÓÐÎÒÕâ
       ·ïѾͷ£¬ ÊÇÎÒ¸ö¡®¸øÊÂÖС¯¡£¼ÈÕâô×Å£¬ºÜºÃ£¬Ëû¾Ë¾Ë¼Ò¸øËûÃǺØϲ£¬Äã¾Ë
       ¾Ë¼Ò¾Í¸øÄã×öÉúÈÕ£¬Æñ²»ºÃÄØ£¿¡±Ëµ×Å´ó¼Ò¶¼Ð¦ÆðÀ´£¬£¨85»Ø£©
In this example, ËûÃÇ does not refer to any identifiable entity; it is an empty referent which is created by ¼Öĸ to contrast with what she is going to do for ÁÖ÷ìÓñ's birthday.

Sometimes the referent may exist but is not uniquely identifiable. This is illustrated in (7). 

(7)  ÖÜÈð¼ÒµÄ·½²ÅҪȥʱ£¬±¦îÎÒòÔÚ×ø£¬ÄËЦµÀ£º¡°ÒÌÄïÇÒס¡£Èç½ñÍâÍ·È˲ζ¼
       ûÓкõġ£ËäÓÐÈ«Ö¦£¬ËûÃÇÒ²±Ø½Ø×öÁ½Èý¶Î£¬ÏâǶÉÏ«ÅÝÐëÖ¦£¬²óÔÈÁ˺Ã
       Âô£¬¿´²»µÃ´Öϸ¡£......¡±£¨77»Ø£©
Here, we can infer from the context that the plural pronoun refers to the merchant who produces the ginseng, but the exact referent cannot be specifically identified.

4. Interim Summary

In the previous discussion, we have seen that third person pronouns with an antecedent is due to the negligibility principle. That is, referents denoted by the third person pronoun may be (1) socially inferior, or (2) negligible in the discourse context.

5. Mismatch In Number Between Pronouns And Referents

We now examine the cases of mismatch in number between pronouns and actual referents. Our account of these mismatch cases is that the use of pronouns in these cases is irrelevant to anaphora per se, but rather as a means to express the speaker's degrading attitude towards the addressee in the novel or to convey a negative connotation, and sometimes also to show one's superior social status over the addressee. Although we have only 4 examples in our data, there are some interesting patterns which support o ur explanation.

The first pattern we found is that the speaker uses the third person singular pronoun Ëû to refer, with an obvious degrading color (or tune), to people of lower social status than themselves. Consider (2), reproduced as (8): 

£¨8£©   ÕÔÒÌÄïÎÞ·¨£¬Ö»µÃͬËûÈýÈ˳öÀ´£¬¿ÚÄÚÓÌ˵³¤Ëµ¶Ì¡£  ̽´º±ã˵£º¡°ÄÇЩС
          Ѿͷ×ÓÃÇÔ­ÊÇЩÍæÒâ¶ù£¬Ï²»¶ÄØ£¬ºÍËû˵˵ЦЦ£»²»Ï²»¶£¬¿ÉÒÔ²»ÀíËû¡£±ã
           Ëû²»ºÃÁË£¬Ò²Èçͬè¶ù¹·¶ù×¥Ò§ÁËÒ»ÏÂ×Ó£¬¿ÉË¡¾ÍË¡£»²»Ë¡Ê±Ò²Ö»¸Ã½ÐÁË
           ¹Ü¼Òϱ¸¾ÃÇȥ˵¸øËûÈ¥Ô𷣡£ºÎ¿à×Ô¼º²»×ðÖØ£¬´óߺСºÈ£¬Ò²Ê§ÁËÌå
           ͳ¡£¡± £¨60»Ø£©
In this example, the antecedent is ÄÇЩСѾͷ×ÓÃÇ (those minor maids), but they are referred to in subsequent text as Ëû for three times. In this context, ̽´º is trying to calm down ÕÔÒÌÄï down by emphasizing how insignificant these minor maids are and how ÕÔÒÌÄï should not lose her temper over these unimportant figures.

A similar case is found in (9), where ÍõÎõ·ï's maid ƽ¶ù refers to the minor maids ²ÊÔƺÍÓñî˶ù with the third person singular pronoun Ëû. 

(9)  ´ó¼Ò¶¼Ã¦Ëµ£º¡°¿ÉÊÇÕâ»°¡£¾¹ÊÇÎÒÃÇÕâÀïÓ¦ÁËÆðÀ´µÄΪÊÇ¡£¡±Æ½¶ùÓÖЦ
       µÀ£º¡°Ò²ÐëµÃ°Ñ²ÊÔƺÍÓñî˶ùÁ½¸öÒµÕϽÐÁËÀ´£¬ÎÊ×¼ÁËËû·½ºÃ¡£²»È»ËûÃÇ
       µÃÁËÒ棬²»ËµÎªÕâ¸ö£¬µ¹ÏóÎÒûÁ˱¾ÊÂÎʲ»³öÀ´£¬·³³öÕâÀïÀ´ÍêÊ£¬ËûÃÇÒÔºó
       Ô½·¢ÍµµÄ͵£¬²»¹ÜµÄ²»¹ÜÁË¡£¡±Ï®È˵ÈЦµÀ£º¡°ÕýÊÇ£¬Ò²ÒªÄãÁô¸öµØ²½¡£¡±
      £¨61»Ø¡³
Here, since ƽ¶ù is higher in status than the two minor maids, she is qualified to use the singular pronoun with a degrading attitude to refer to the maids and show her higher status.

From examples (8) and (9), it is evident that, as in the case of antecedent-less pronouns, relative social position plays an important role in the use of singular pronouns to refer to plural entities. What is more, in this type of use a degrading attitude is detected along with the display of the speakers' superior social status over the referred entities.

Sometimes the mismatch does not necessarily involve a superior social status, but the degrading or negative connotation is still evident. This may happen in the author's own narration of the story, showing the author's negative attitude towards the referred entity. Consider example (10). 

(10)  ÇÒ˵Ëû¹ÃÄԭƸ¸øµÄÊÇ¼Ö¼ÒµÄ "Óñ"×Ö±²µÄµÕÅÉ£¬Ãû»½¼Öè«¡£µ«Æä×åÈË
         ÄÇÀï½ÔÄÜÏñÄþÈÙ¶þ¸®µÄ¸»ÊÆ£¬Ô­²»ÓÃϸ˵¡£Õâ¼Öè«·òÆÞÊØ×ÅЩСµÄ²úÒµ£¬
         ÓÖʱ³£µ½ÄþÈÙ¶þ¸®ÀïÈ¥ÇëÇë°²£¬ÓÖ»á·î³Ð·ï½ã¶ù²¢ÓÈÊÏ£¬ËùÒÔ·ï½ã¶ùÓÈ
         ÊÏҲʱ³£×ÊÖú×ÊÖúËû£¬·½ÄÜÈç´Ë¶ÈÈÕ¡£½ñÈÕÕýÓöÌìÆøÇçÀÊ£¬ÓÖÖµ¼ÒÖÐÎÞÊ£¬
         Ëì´øÁËÒ»¸öÆÅ×Ó£¬×øÉϳµ £¬À´¼ÒÀï×ß×ߣ¬ÇÆÇƹÑÉ©²¢Ö¶¶ù¡££¨10»Ø£©
It is evident that the couple ¼Öè«·òÆÞ is not spoken of in the most favorable manner from the author's point of view, hence the use of the singular pronoun in this context.

Sometimes the singular form indicates a general negative connotation, which can be seen even when the speaker is of lower status. In example (11), ÂíµÀÆÅ -- the eqivalent of a religious medium -- uses the singular pronoun while referring to the children of the rich and the powerful. She comments that these people are often possessed by demons since birth. 

(11)  ¹ýÁËÒ»ÈÕ£¬¾ÍÓб¦Óñ¼ÄÃûµÄ¸ÉÄïÂíµÀÆŽøÈÙ¹ú¸®À´Çë°²¡£......ÓÖÏò¼ÖĸµÀ£º
         ¡°×æ×ÚÀÏÆÐÈøÄÇÀïÖªµÀ£¬ÄǾ­µä·ð·¨ÉÏ˵µÄÀûº¦£¬´ó·²ÄÇÍõ¹«ÇäÏàÈ˼ҵÄ
        ×ӵܣ¬Ö»Ò»Éú³¤ÏÂÀ´£¬°µÀï±ãÓÐÐí¶à´ÙÏÁ¹í¸ú×ÅËû£¬µÃ¿Õ±ãÅ¡Ëûһϣ¬»òÆþ
         Ëûһϣ¬»ò³Ô·¹Ê±´òÏÂËûµÄ·¹ÍëÀ´£¬»ò×ß×ÅÍÆËûÒ»õÓ£¬ËùÒÔÍùÍùµÄÄÇЩ´ó¼Ò
         ×ÓËï¶àÓг¤²»´óµÄ¡£¡±¼ÖĸÌýÈç´Ë˵£¬±ã¸Ï×ÅÎÊ£º¡°ÕâÓÐʲô·ð·¨½âÊÍ£¿¡±
        £¨25»Ø£©
In this case, the use of the singular form conveys a general negative connotation, since what is being talked about here is a negative superstitious scenario.

6. Discussion And Conclusion

We have just examined the two types of what we call "unusual uses of third person pronouns" in Hong Lou Meng. We now summarize the discussion and touch upon briefly the theoretical implications of our study.

Our study reveals that third person pronominal forms in Hong Lou Meng can be used in two unusual ways: without an antecedent or showing a mismatch in number between the pronominal form and the referred entity. We show that these unusual cases typically occur when the referent is of a lower social status or unimportant in discourse, and there can also be a degrading or negative connotation associated with the referred entity.

We believe that these findings have important implications for the understanding of anaphoric forms in general. As we mentioned at the beginning of this paper, dominant approaches to anaphora seem to have concentrated on the structural aspects of anaphora. Our study shows that anaphora can indeed be a social issue, where an understanding of the social structure of society in which referents are situated is critical for the interpretation of anaphoric forms. Furthermore, our data provide compelling evidence suggesting that pronominal forms are not merely another form of referring, but can also be a powerful tool to show social status and attitude toward what is being referred to.

Finally, the two types of unusual uses of third person pronouns in Hong Lou Meng provide new data for consideration of the historical development of the Chinese language, at least for the written genre. Whether the phenomena we discussed in this paper are exclusive of the written literary language of the late Qing dynasty or there is some historical connection involved is an interesting question which needs to be further investigated.

References

Chafe, Wallace. 1994. Discourse, Consciousness, and Time: The flow and displacement of conscious experience in speaking and writing. Chicago: University of Chicago Press.

Chen, Ping 1984. A Discourse Analysis of Third Person Zero Anaphora in Chinese. Bloomington: Indiana University Linguistics Club.

Chen, Ping. 1986. Referent Introducing and Referent Tracking in Chinese Narratives. Ph.D. dissertation, University of California at Los Angeles.

Clancy, Patricia M. 1980. Referential choice in English and Japanese narrative discourse. In Wallace Chafe, ed., The Pear Stories: Cognitive, cultural, and linguistic aspects of narrative production. Norwood, NJ: Ablex.

Du Bois, John W. 1980. "Beyond Definiteness: The trace of identity in discourse". In Wallace Chafe, ed., The Pear Stories: Cognitive, cultural, and linguistic aspects of narrative production. Norwood, NJ: Ablex. pp. 5-90.

Fox, Barbara. 1987. Discourse Structure and Anaphora. Cambridge: Cambridge University Press.

Givón, Talmy. ed., 1983. Topic Continuity in Discourse: A quantitative cross-language study. Amsterdam & Philadelphia: John Benjamins.

Li, Charles N. & Sandra A. Thompson. 1979. "Third-Person Pronoun and Zero-Anaphora in Chinese Discourse". In Talmy Givón, ed. 1979. Discourse and Syntax. Syntax and Semantics Vol. 12. New York: Academic Press. pp. 311-335.

Lee, Cher Leng. 1995. Zero Anaphora in Chinese. Taipei: Crane Publishing Co. Li, Naicong & David Zubin. 1990. "Discourse Continuity and Perspective Taking". Chicago Linguistics Society 26.253-268.

Tao, Liang. 1993. Zero Anaphora in Chinese: Cognitive Strategies in Discourse Processing. PhD Dissertation, University of Colorado, Boulder.

Acknowledgements

We are grateful to Zhang Min for presenting our paper in our stead at the Joint Meeting of the Fourth International Conference on Chinese Linguistics and the 7th North American Conference on Chinese Linguistics, June 27-30, 1995, Madison, Wisconsin. Thanks are also due to Zhang Zhengsheng for helpful comments on our conference paper. Standard disclaimers apply.