turkishenglish.com - Comparison of relative clauses in ESL students

CHAPTER V - Section 3

Other Findings

            This section is divided into two sub-sections.  The first deals with the taxonomy of discourse functions and their distribution.  The second sub-section looks at instances where students should have used an RC but did not.  In particular, I will address the claims of Schacter and Hart (1978).  They posited that Chinese students use infinitival complements as part of their attempt to avoid relative clauses.  I will discuss some examples of such constructions and offer an alternative. 

Discourse Functions:  One of the goals of the present study was to categorize RCs according to their discourse function.  A simple taxonomy of RC functions in expository text has been developed, but it undoubtedly needs refining.    Text is a complex mosaic of themes and logical relations, which is, at times, difficult to unravel.  An attempt has been made not to over-categorize the data since one would want to be sure that categories are valid and broader more general categories are easier to defend as having characteristics that make them distinct.  Nevertheless, not all of the categories are mutually exclusive.  The names of the categories themselves can be confusing because in some sense every RC is descriptive, identificatory, and characterizing (cf. Givon 1993), but the definitions given below make crucial reference to the relationship between the RC and the wider essay.  Each category will be given a brief description, and an example from the database will be provided to illustrate.

            Descriptive RCs are those that are used to provide additional new information about a referent and turns out to be the largest category for both groups.  This type of RC might provide background that lays the foundation for later assertions, or it might be description which supports earlier reasoning.  Example (57) is a description which supports the reasoning in the previous sentence.  The writer likes friendly intelligent people because they make enjoyable company.


            (57)  Another of the qualitys that I always took to choose my friends is that person have to be friendly and intelligent.  I like happy people around me wich I can enjoy and discuss any kind the conversation.  (S 3.12.4)


Another interesting example of this is (58) which provides additional description in order to emphasize a point.  The point is that family values are a thing of the past, and consequently cannot help but remind one of their grandparents. 


            (58)  ...At this time people don’t care about anyone and anything, someone could die in the street, and anybody do nothing.  The concept of family is practically something of the past that remember you your grandparent. (S 3.4.1)


            Identificatory RCs are those whose Head NP is Given and the information in the RCs is also known.  These relative clause re-establish a referent.  This is the second largest category for both groups.  But, as was noted earlier, the Spanish speakers use a higher percentage of these than the Koreans do.  In (59), the writer starts out talking about American attitudes, and then discusses the situation in Korea.  Finally, the writer returns to discuss America and its people and an RC is used to re-establish the referent.


            (59)  Most of the people in America is kind.  But some kinds of people is bad. Sometimes they ignore me.  Because they are white or black, but I’m yellow.  So I’m very sorry about that.... Just one kind of people live in Korea.  So I don’t suffer like this situation.  but many kinds of people live in America.  It’s the difference of American from Korea.  But many people that live in America are kind.  I know that.  If I’m very hard, they help me.  (K 4.2.1)


            Topical cohesion is another function of the RC.  This term is used to describe a discourse phenomenon whereby a new referent is introduced in terms of the essay or paragraph topic.  The information in the RC is known, but the Head NP is the new topic being introduced for discussion.  In (60), the writer is discussing the requirements for happiness and this sentence is the first one in an orthographic paragraph.  


            (60)  The third thing I need to do to acquire my happiness is by being a good person in all the stages of my life.  (S 5.16.8)


See also examples (11-13).

            Topic characterization has much in common with the descriptive function described above, but in this case the RC characterizes a referent in such a way as to set up a new subclass.  It establishes a topic with the characteristic that defines its importance for the following discourse.  In (61), people who are content with what they have is the first topic.  The writer then proceeds to discuss what this type of person is like. 


            (61)  first of all the most important of happiness is what he thinks about everything, that is happy are those who are contented anything.  For example, we supposed that two men are drinking in the same place.  There is a little bit of wine.  one person who is content, will say “still remain.”  (K3.21.1)


            Some RCs were used to summarize a particular portion of text - either events or propositions.  In (62), the writer discusses the characteristics of a certain type of person and then summarizes those statements with a relative clause. 


            (62)  I think all unhappiness derive from greed.  When we breathe our last,  We cannot take power and money with us.  If a rich person distributes his money           to poor person, he will be satisfied with good conduct.  Also, if a powerful who use his power and money for others is happier than the man who use his power and money for himself.  (K 4.10.4)


Here is another example, this time summarizing events. 


            (63)  In my country I enjoy the parties and go out to dinner with my friends in    the weekends.  Sometimes I went to dance, there were many places, many    times we go to rent a picture and all can see the picture...  realy I miss my friends and all the thinks that I maide with them. (S 3.1.6)


            Backreference is a discourse function in which an RC repeats the matrix verb from the previous sentence.  It functions much like a preposed adverbial clause in that it provides tail-head cohesion.  This is a common function of the RC in languages with prenominal RCs; however, the Spanish speakers use it as well. 


            (64)  Our mothers have had a different opportunity to study and to perform their career.  This new opportunity that our mothers have had lately, has influenced on our familiar environments.  (S 5.17.1)


            Some RCs seemed to encode the essay or paragraph theme.  In (65), the essay is about the requirements for happiness and the last sentence in the first two paragraphs states the theme of those paragraphs. 


            (65)      I think that the happiness start from at home.  Because, our life is at home from start to finish.  We was born and grow up at home.  After we be an adult, we make another home.  Then I think the man who has a happy home is happier than any other person. 

                        I think to meet with other people is important too.  We must be happy if we have a good parents, and we can be happy wife or husband if we meet a good match.  It is same in any other case.  Then I think the man who has good person around him is happy.  (K 4.24.1)


            Exemplification is the name given to all of the Korean examples of a set-some-member relationship.  I chose this term because the purpose of the RC is to give examples of a more generic term. 


            (66)  In simility, I don’t get foreign goods which are Japanese, England, or Jerman because our economic blows if people but foreign products. 


            Relative clauses were also used to restate an idea or theme.  In example (67) the last sentences of the first paragraph and of the concluding paragraph demonstrate how the student is restating the main point of the essay. 


            (67)  Learning English as a second language is something that open you mind, that help you to meet better other people’s idea and it can stop the afraid that some people has when they have to travel to other countries. 

                        In conclusion, knowing English as a second language is something that   has been helping me to increase my knowledge about other people’s cultures, to meet different friends from others countries and also it is something that makes easy traveling to others places of this world without feeling afraid.  (S 5.7)


This example includes elements of summary as well, but it is very clearly restating the thesis statement of the essay.

            The last function of the RC is similar to that of topic characterization above except that instead of introducing a local topic, this RC frames the entire discussion.  This is often done by using abstract nouns with RCs, such as aspects, qualities, etc.  In example (68) the writer uses the generic noun aspects and an RC to constrain the essay to three things (communication, spending time together, and love) in terms of how they relate to maintaining good family relationships.   


            (68)  There are three indispensables aspects that must exist in a family to have a good relationship with each member of the family:  communication, spend time together, and to love and respect each other.  (S 6.2.1)


            Table 30 gives the number of RCs in each category. 


Table 30.  Number of discourse functions by category and language background.


       Language background

Discourse Function









Topical cohesion



Topic characterization

























            There is some overlap between the categories, summary, theme, backreference,

restatement, and the category identificatory because of the nature of these functions.  For example, an RC that is functioning as backreference is definitely identificatory because it is encoding known information on a known Head Noun.

            The discourse function of RCs have been shown to play an important role in the production of RCs.  Spanish students use significantly more RCs for identification and topical cohesion than the Korean students.  We should not, however, overlook the similarities between the two groups.  Description and identification were the two largest categories for both groups.   

            More work needs to be done on the function of RCs in discourse and finer distinctions need to be drawn.  For example, had not the Korean students mistakenly used RCs to encode a set-some-member relationship, this study would not have discovered or known that RCs do not function in this fashion in English.  It seems that functional errors might be useful in determining what the boundaries of a particular form are in the functional domain.  Edmondson (class notes) frequently made the point that deviant utterances are much more interesting for linguists than correct ones.  


Infinitival Noun Complements:  Though the main focus of this paper was not the

incorrect use of infinitival complements, an effort was made to note any instances of  sentences that should have had RCs but did not.  Schacter and Hart (1978) had specifically claimed that Chinese students used infinitives as a way of avoiding RCs.  The evidence from this study, however, suggests that this phenomenon may be better  explained as a case of students not knowing the boundaries of a particular construct; in other words, over-generalization.  In English it is possible to modify a noun with an infinitive as in (69). 


            (69)  In my country, Columbia, there are many beautiful, interesting places to visit.  (S 3.7)


This same sentence could just as easily been written with an O relative, as in (70), where the subject of the RC is unnecessary information. 


            (70)  In my country, Columbia, there are many beautiful, interesting places that you can visit.


This is not always possible, however.  If the noun being modified functions as the subject of the verb in the infinitival complement, then an RC is often necessary.  In (71), from a level 6 Korean essay, the student has used an existential as in (69) above, but the infinitive is wrong because women is the subject of the infinitive.  It should have been an RC.  Examples (72) and (73) are similar.  The first one is taken from a Spanish essay and the second from a Korean essay.   


            (71)  In the case of Korea, there are a lot of women to give up their career to get      married.  (K 6.7)

            (72)  There are many things to cause happiness.  (S 4.4)

            (73)  There are many things to make me happy.  (K4.23)



            The Korean essays had four instances of infinitives that should have been RCs.

The Spanish essays had two.  The number of tokens is so low that no firm conclusions can be drawn, but since both groups make the mistake and there are conditions in English under which infinitives can modify nouns, it is very likely that these students simply have not learned the conditions for usage.  With so few mistakes, it is impossible to suggest that the Korean students are using infinitives as a strategy for avoiding RCs.

Next: Conclusions           




            1The level five Korean students did not follow the same pattern as the other Korean students.  In almost every category, the level five students were different.  They used more Object relatives, more Indefinite Pronouns, fewer RCs, fewer Subject Heads.  All of these were in stark contrast to the normal pattern.  This particular group was the smallest, however, producing only 17 RCs, so these numbers may not reflect the IL competency of these students. 


            2Unlike Korean, NRRCs in English are sometimes used to encode a subsequent event.  If the students had used NRRCs in this way, it would have meant that they had acquirred a TL function, but none of the NRRCs in the Korean corpus were of this type.

            3Rutherford (1983) has noted that Korean and Japanese students overproduce

existential constructions.  This might explain why the Spanish students decrease their usage while the Korean students do not. 


            4This tendency to frame NPs in terms of the student’s own activity is related to another difference in distribution between the two groups.  Spanish students often ground a Head NP by referring either to themselves or to the reader by using 1st of 2nd person pronouns as the subject of an O or X relative.  They used 65 such pronouns; the Koreans used only 18.  Further study is needed to determine exactly why Spanish students choose to link referents to themselves or to their listeners and Koreans do not.  This may also account for the large production of O and X relatives by the Spanish students.


            5Language they speak is also important.  (s 4.12.6)


            6One difference that I really like between living here and in Durango is that at home I’m under the eye of my parents.  (S 5.1.6)  This RC does not frame the referent since this is done by the preposition.

            7This is similar to the cleft and pseudo-cleft constructions which are also used

to focus information, e.g., “It was John who broke his leg.”


            8As we noted in Chapter 2, languages like Korean and Japanese have the same construction for RCs and CCs.  Because they have one form in the NL which corresponds to two forms in the TL, one might expect them to have some difficulty distinguishing between the two, especially since the target language, in this case English, also has a construction whereby they may use an RC with a generic Head to substitute for the complement clause.  However, these students rarely used this construction.  Why they do not is an area for further study. 


            9In this example, Mr. Park may have been attempting to maintain a topic-

comment structure by keeping house in the subject position.