1

THE SYNTACTIC DIFFICULTY OF RELATIVE CLAUSES

FOR KOREAN STUDENTS:

AN ANALYSIS OF ERRORS AND AVOIDANCE STRATEGIES

by

Chanyoung Park

An Applied Project Presented in Partial Fulfillment

of the Requirements for the Degree

Master of Teaching English as a Second Language

ARIZONA STATE UNIVERSITY

DECEMBER 2000

THE SYNTACTIC DIFFICULTY OF RELATIVE CLAUSES

FOR KOREAN STUDENTS:

AN ANALYSIS OF ERRORS AND AVOIDANCE STRATEGIES

by

Chanyoung Park

has been approved

December 2000

APPROVED:

______, Chair

______

______

Supervisory Committee

ACCEPTED:

______Department Chair

______Dean, Graduate College

ABSTRACT

This paper reports on the syntactic difficulty order of relative clauses (RCs), and analyzes the errors and avoidance strategies used by 10 Korean learners of English. Three hypotheses from previous studies are examined in the present study, that is, the Perceptual Difficulty Hypothesis (PDH), Noun Phrase Accessibility Hypothesis (NPAH), and Parallel Function Hypothesis (PFH), which indicate the difficulty of center embedding, pronoun fronting, and non-parallel function, respectively.

A one-page questionnaire, which contains 12 sets of sentence combining tasks, was constructed and used. The questionnaire was designed to force the students to write each 3 of 4 different types of RCs: OS, OO, SS, SO types ( O for Object and S for subject). The first symbol shows the function of the noun phrase (NP) in the main clause, and the second, that of the relativized clause.

The present study supports NPAH and PDH, but not PFH. The data show the order of usage as SS/SO, OS, and OO from the greatest to the least. Many errors were made in SS and SO types; this indicates the difficulty of center embedding and thus supports the PDH. A novel finding of this paper is that L1 transfer influences the perceptual difficulty of RC, and is shown in access and avoidance strategies. Another contribution of this paper is the introduction of a 4-step teaching method for RCs.

ABBREVIATIONS

The following abbreviations are used to label the linguistic terms employed in this volume.

* Ungrammatical (when placed before an example)

AdjPAdjective Phrase

CComplementizer

DDeterminer

ESLEnglish as a Second Language

EFLEnglish as a Foreign Language

L1First language

L2Second language

NPNoun Phrase

NPAHNoun Phrase Accessibility Hierarchy Hypothesis (NPAHH)

PDHPerceptual Difficulty Hypothesis

PFH Parallel Function Hypothesis

RC Relative Clauses

VVerb

VPVerb Phrase

TABLE OF CONTENTS

ABSTRACT ……………………………………………………………………… i

ABBREVIATIONS……………………………………………………………… ii

TABLE OF CONTENTS ………………………………………………………….. iii

CHAPTER 1: INTRODUCTION……………………………………………… 1

CHAPTER 2:RELATIVE CLAUSES ……………………………………………. 4

2.1. Definition and the Structure …………………………………………… 4

2.2. Different Types of Relative Pronouns and Relative Clauses ………….. 6

2.2.1. The Relativization of the Subject ………………………………... 6

2.2.1.1. Subject-subject Relatives ………………………………. 6

2.2.1.2. Object-subject Relatives ……………………………….. 7

2.2.2. The Relativization of the Object ……………………..…………… 8

2.2.2.1. Subject-object Relatives ………………………………… 8

2.2.2.2. Object-object Relatives ………………………………….. 8

2.2.3.Relativization of the Possessive ………………………………….. 9

2.3. Relative Pronoun Deletion Rules ………………………………………… 10

CHAPTER 3: STUDIES IN THE ACQUISITION OF RELATIVE CLAUSES …….. 12

3.1. The Complexity of Relative Clauses …………………………………….. 12

3.2. The Order of Difficulty and Three Hypotheses on the Difficulty of Relative Clauses …………………………………………………………………… 13

3.3. Native Language Transfer ……………………………………………….. 17

3.4. Corpus-Based Approach ………………………………………………… 18

CHAPTER 4: HYPOTHESES AND METHODS …………………………………... 20

4.1. Research questions ………………………………………………………. 21

4.2. Hypotheses ………………………………………………………………. 21

4.3. Methodology …………………………………………………………….. 22

4.3.1. Subjects ……………………………………………………………. 22

4.3.2. Materials …………………………………………………………… 23

4.3.3. Procedure …………………………………………………………... 24

CHAPTER 5: RESULTS AND DISCUSSION ……………………………………… 25

5.1. The Order of Difficulty …………………………………………………... 25

5.2. Data Analysis …………………………………………………………….. 27

5.2.1. Error Analysis ……………………………………………………… 28

5.2.1.1. Error Type 1. Pronoun Retention …………………………….. 29

5.2.1.2. Error Type 2. Relative Clause Marker Selection …………….. 29

5.2.1.3. Error Type 3. Adjacency ……………………………………... 30

5.2.1.4. Error Type 4. Be-verb Omission ……………………………... 31

5.2.1.5. Error Type 5. Misplacing or Omitting Subject in Objective RCs.31

5.2.1.6. Error Type 6. Modifying Wrong Antecedent ………………… 32

5.2.2. Avoidance Strategy Analysis ………………………………………. 33

5.3. Discussion ………………………………………………………………... 36

CHAPTER 6: TEACHING SUGGESTIONS AND CONCLUSION ……………….. 40

6.1. Conclusion ……………………………………………………………….. 40

6.2. Teaching Suggestions: 4 steps …………………………………………… 41

6.3. Implicit Activities …………………………………………………….…. 43

REFERENCES …………………………………………………………………….… 46

APPENDIX …………………………………………………………………………... 48

CHAPTER 1 INTRODUCTION

From my own experience of learning English as a foreign language, I remember that a friend of mine one day came to me, pointing to another person, and said “She must be very good at English. She even uses relative clauses when she talks in English.” Relative clauses (RCs) have been considered as the symbol of complex English structure and even some extent, the indicator of mastery in English.

The reason can be found in the process of producing RCs. In order to acquire relative clauses, students have to learn the identification of identical NPs, relative pronoun substitution, word order rearrangement, and the process of embedding.

In addition, English is syntactically very different from Korean. Korean is an SOV language, while English is an SVO language. Korean RCs are prenominal modifiers, but English RCs are post-nominal modifiers. Korean has an invariable relative pronoun, whereas, English has variable relative pronouns.

Schachter (1974) and Gass (1980) explored acquisition of relative clauses in the perspective of native language transfer. Chiang (1980)’s replication of Schachter’s study reports that overall target language proficiency is the best predictor of relative clauses production.

Several SLA researchers have studied on the order of natural difficulty among four types of relative clauses, that is, SS, SO, OS, and OO. The first letter stands for the function of the head noun in the main clause and the second letter stands for the function of the relative pronoun of the relative clause. Kuno (1975) argued that center embedding is perceptually more difficult and therefore, OS/ OO clauses are easier than SS/ SO clauses. On the other hand, Keenan (1975), in his noun phrase accessibility hierarchy, focused on embedded relative clauses rather than the main sentences. He hypothesized that relativized subjects are more accessible than relativized objects, which means SS/ OS types are easier than SO/ OO types. The hierarchy is as follows, from most “accessible” to the least “accessible”: subject NP, direct object NP, indirect object NP, oblique object NP, genitive object NP, and object NP of a comparison.

Ioup and Kruse (1977)’s data supported Kuno’s study and a later study by Schumann (1980) also supported Kuno’s hypothesis. The present study questions an assumption in Schumann’s study, namely that the more frequently produced type is the easiest form. It is plausible that easier structures are used more often because his subjects were non-native speakers of English. However, there can be a variable that subjects produced OS, OO more often, just because those are occurring more frequently in a daily conversation.

Stauble (1978)’s research provided the same frequency of these four types of relative clauses with native English speaker subjects; this weakens the logic of Schumann’s study, which assummed that if one type of RC is used more often, it can be considered as easier . However, certain types of RCs can be used more often by the non-native speakers just because they are more often used in a real language. Therefore, in this study, errors and avoidance rate were also considered to in measuring the difficulty of each relative clauses type in addition to the frequency of production.

In this paper, I investigate the syntactic difficulty that lies in the formation of relative clauses and I analyze the errors and avoidance strategies of 10 Korean learners of English. The methodology used was a questionnaire with 12 sets of combining sentence tasks, which force the subjects to write 3 of 4 different types of RCs.

Chapter 2 contains a general description of relative clauses, including definitions and different types of relative clauses. Chapter 3 reviews the theories and studies regarding L2 acquisition of relative clauses. It discusses three different hypotheses on the difficulty of relative clauses. The Parallel Function Hypothesis, the Perceptual Difficulty Hypothesis, and the Noun Phrase Accessibility Hierarchy Hypothesis. It also includes the studies from the perspective of native language transfer, and corpus-based approach. Chapter 4 explains the research questions, hypothesis, and methodology of data collection, the subjects and the procedure. Chapter 5 investigates the result obtained by the questionnaire. It also provides an error analysis and an avoidance strategy analysis. Chapter 6 draws the conclusion from the study and gives some teaching suggestions.

CHAPTER 2.RELATIVE CLAUSES

2.1. Definition and Structure

Clauses that function inside the phrase as modifiers are called relative clauses (RCs). For example, the clause in brackets, who crossed Antarctica in (1), is a relative clause.

(1) The man [who crossed Antarctica] was happy.

English relative clauses are introduced by a relative pronoun and modify their NP antecedents or head noun. In (1), man is the antecedent that is modified by the RC in brackets. The relative pronoun who indicates or replaces the antecedent man and also serves as a complementizer. The relative clause always contains a gap t, which is indicated by the trace of the relative pronoun as in (2).

(2)S

NP VP

NPS’VAdjP D N C S

NP VP

tVNP

The man who crossed Antarctica was happy.

The basic structural relationship in RCs is formed by a process called embedding, which is the generation of one clause within another higher-order or superordinate clause such that the embedded clause becomes a part of the superordinate main clause. For example:

(3)The fans [who were attending the rock concert] had to wait in line for three hours.

(NP[S] )

a. The fans had to wait in line for three hours.

b. The fans were attending the rock concert. (Celce-Murcia & Larsen-Freeman, 1999, p572)

Originally, sentence (3) is derived from (3a) and (3b). (3a) is a main clause and (3b) was embedded in (3a). The fans is the noun that occurs in both sentences. When embedded, this noun will be substituted by the relative pronoun, which is who in the sentence (3). In this process, the relative pronoun will take the same case as the original embedded sentence. In other words, the fans in the second clause, (3b) is the nominative case, therefore, it will be replaced by the nominative relative pronoun, who. In the main sentence, the fans will be an antecedent that will be modified by the relative clause.

There are four processes in producing RCs. First, the identical NP or modified antecedent should be identified. Second, proper relative pronoun should be chosen and substitute the identical NP in the relativized clause. Third, relative pronoun should be fronted when the function of the identical NP is an object in the relativized clause. Fourth, the relativized clause should be placed after the modified antecedent. This process, as I mentioned above, is called embedding. When the relative clause modifies the subject of the main clause, the relative clause is embedded in the middle of the main clause; this is called center embedding.

Identifying identical NPs, relative pronoun substitution, and embedding applies to all types of relative clauses. On the other hand, relative pronoun fronting and center embedding applies to certain types. The following section introduces the different types of relative clauses and explains the process involved in each types.

2.2. Different Types of Relative Pronouns and Relative Clauses

Relative clause structures can be broadly categorized in terms of function of the head noun in the main clause and of the function of the identical noun in the relative clause. In the main clause, head nouns can function as subject, direct object, indirect object, object of the preposition, or predicate noun. In the relative clause, heads of the NP can function as subject, direct object, indirect object, or object of a preposition. Moreover, the possessive determiner whose can relativize any noun functioning as a subject, direct object, object of a preposition, or predicate noun. I will go into these in the following sections.

2.2.1. The Relativization of the Subject

2.2.1.1. Subject-subject Relatives (SS type)

When the subject of the embedded sentence is relativized, it will be replaced by a nominative relative pronoun, who, which, or that. Consider the following sentence:

(4)The man who speaks Korean is my brother.

This sentence is formed through embedding The man speaks Korean into The man is my brother. Who is the relative pronoun that can be used to refer to persons of either gender in singular or plural. In this sentence, who replaces the man.The man in the embedded sentence functions as subject, and the man in the main clauses also functions as subject. In order to function as a relative clause, the embedded clause must contain an NP that is identical in form and reference to the NP in the main clause. In the above example, NP is the man which refers to the same thing in both sentences and it is in identical form. The relative clause is functioning as an adjective, modifying the noun preceding it. The relative pronoun which replaces a non-human thing, and that replaces both which and who. The processes involved in the producing of SS type RCs are relative pronoun substitution and center embedding.

2.2.1.2. Object-subject Relatives (OS type)

The subject of the sentence can be relativized and embedded into the main clause modifying the object of the main clause. In this case also, the identical NP, subject of the embedded sentence and the object of the main sentence, is needed. For example:

(5)I know the man who speaks Korean.

This sentence is composed of two base sentences. I know the man and The man speaks Korean. In the former sentence, which is the main sentence, the man is the object. And in the latter sentence, the man is a subject. A relative pronoun is replacing the subject of the embedded sentence; therefore, it adopts subject case. In the above, sentence, the man was the direct object of the sentence. Indirect objects and the object of the main sentence can be modified by the subject relative clauses as well. The process involved in OS type of RCs is relative pronoun substitution.

2.2.2. The Relativization of the Object

2.2.2.1. Subject-object relatives (SO type)

In this type of relative clause, the object NP of the embedded sentence will be relativized to modify the subject of the main sentence. For example:

(6)The girl whom you met is my sister.

In this sentence, You met the girl is embedded into The girl is my sister. The girl is an object in the embedded sentence and the subject of the main sentence. Since the object of the embedded sentence is relativized, it will take the relative pronoun with the objective case, as whom in the above sentence. The objective relative pronouns are who, whom, which, and that. The relative pronoun who is used for both nominative and accusative cases, and whom is used for formal accusative cases and after prepositions.

The two identical NPs are not adjacent as they were in the two previous sentences discussed above. Therefore, after replacing the object with the relative pronoun, the relative pronoun has to move to the front of the relative clause to be adjacent to the noun modified. This rule is referred to as “relative pronoun fronting” (Celce-Murcia and Larsen-Freeman, 1999, p576). The processes involved in SO type of RCs are relative pronoun substitution and relative pronoun fronting.

2.2.2.2. Object-object relatives (OO type)

In this type of relative clauses, the object of the embedded sentence will be relativized to modify the object of the main sentence. For example:

(7)I read the book that you wrote.

The base sentences are I read the book and you wrote the book. The book is the identical NP in both sentences. In the embedded sentence, the book is replaced by the relative pronoun that, which takes the object case of the original sentence.

In summary, the process of identical NP identification, relative pronoun substitution, and embedding applies to every type of RC. Objective relative clauses, such as SO and OO types, require relative pronoun fronting. When the subject of the main sentence is an antecedent such as SS and SO types, center embedding is required (see table1).

When a target language construction has several processes involved, it is more complicated and more difficult for second language learners than simple structures. Therefore, from the table 1, we can infer that SO type may be the hardest and OS type may be the easiest.

Table 1. The Process in Each Type of RC.

Identical NP identification / Relative pronoun substitution / Embedding / Relative pronoun fronting / Center embedding
SS / Yes / Yes / Yes / No / Yes
SO / Yes / Yes / Yes / Yes / Yes
OS / Yes / Yes / Yes / No / No
OO / Yes / Yes / Yes / Yes / No

2.2.3. The Relativization of the possessive

In sentence (8), whose replaces his in (8-b). Whose generally refers to human head noun but sometimes it refers to inanimate noun too as in (9).

(8) I met the man whose name is John.

  1. I met the man.
  1. His name is John.

(9)I found an old coin whose date has become worn and illegible.

Table 2. Relative Pronoun

HUMAN / NON-HUMAN
SUBJECTIVE / Who / which
That / that
OBJECTIVE / Whom / which
Who
That / that
POSSESSIVE / Whose / whose

Table 2 shows a summary of the types of relative pronouns. Subjective relative pronouns are who, which, and that. Who is used for indicating humans and which for non-humans. That can be used for both. Objective relative pronouns are who, whom, which, and that. Who, whom, and that are used for humans, and which and that for non-humans. Who is used as both subjective and objective. Whom is more formal and used mostly after prepositions. The possessive relative pronoun is whose and can be used for both human and non-human antecedents.

2.3.Relative Pronoun Deletion Rules

A relative pronoun in a relativized object can be deleted unless it follows a preposition, which is fronted with the pronoun. The relative pronoun +be verb deletion rule applies when a restrictive relative clause containing 1) a relativized subject and a verb in the progressive aspect , 2) a relativized subject and a verb in the passive voice, and 3) a relativized subject followed by the be copula and a complex adjective phrase. Following is the example of each cases:

(10)I saw a man (who was) crossing the street.

(11)I like cookies (which are) made by mom.

(12)I stayed in a hotel (which is) the most luxurious in the world.

CHAPTER 3.

STUDIES IN THE ACQUISITION OF RELATIVE CLAUSES

3.1.The Complexity of Relative Clauses

Relative clauses are considered as complex postnominal modifiers in English. Therefore, they appear in the latter part of second-language English textbooks, and are taught at an intermediate level in an English class. For example, the Grammar Book (Azar, 1995), which is used for beginning classes at the American English and Culture Program, the language institute in Arizona State University, doesn’t include relative clauses at all.