23
Taiwanese argumentation skills:
Contrastive rhetoric perspective
Fei-Wen Cheng, Yueh-Miao Chen
National Chiayi University, National Chung-Cheng University
Abstract
Argumentative writing is a fundamental writing style across various EAP and ESP writing tasks.
Despite its importance, no second language writing research has examined how Taiwanese students
compose this genre based upon a well-established theoretical model of Western argumentation. The
goal of this study is to bridge this gap by building upon Stephen Toulmin’s (2003) model of argument
to examine the use of English argumentation features. Participants included 40 Taiwanese and 39 US
college freshmen writers. Writing samples collected included 119 essays, 80 of them were composed
by Taiwanese participants (40 Chinese and 40 English texts) and 39 by American students. The
findings indicate that Taiwanese student arguments are less extended and complex, and display a
limited range and quantity of argumentative structure in comparison to American arguments. Yet,
both Taiwanese and American students are weak at handling oppositional structures, an essential trait
differentiating Chinese and English rhetoric. Equally important, Taiwanese students, when composing
Chinese texts are able to construct certain argument features in a way similar to American students.
This illustrates that culture may not necessarily account fully for the argument features manifested in
Taiwanese writing of English. Other factors, such as L2 language proficiency and developmental
factors, also play a mediating role in the use of argument structures.
Keywords: argumentative writing, contrastive rhetoric, Toulmin’s model, second language writing,
genre
1. Introduction
English argumentative writing, based on Aristotelian rhetoric, has the goal of
convincing an audience (Connor, 1996) in a situation where there exists a conflict
Taiwan International ESP Journal, Vol. 1: 1, 23-50, 2009
24
between the beliefs and attitudes of the writer/speaker and the reader/audience (Hinkel,
2002). This rhetorical tradition emphasizes articulation of one’s stance, justification of
one’s position and ideas, a logical progression of one’s ideas, and refutation of opposing
arguments to defend one’s claim (Connor, 1996; Hinkel, 1994 & 2002; Matalene, 1985;
Oliver, 1971). These discourse practices are embedded within a larger socio-cultural
context that encourages individual self-expression, creativity and critical thinking.
In contrast, the goals of Chinese rhetoric under the influence of Confucian philosophy
and classical Chinese rhetoric are to achieve general harmony, to express the views of
the group, and to promote social cohesion (Becker, 1986; Hinkel, 2002; Kaplan, 1966
Matalene, 1985). To achieve these purposes, Confucian persuasion includes employing
various indirect modes of expression to suggest one’s claim, supporting one’s ideas
with appeals to history, to tradition, and to authority rather than one’s own individual
opinions or beliefs, relying on accepted patterns of expression, and avoiding any
contentious forms of argument (Hinkel, 2002; Matalene, 1985; Oliver, 1971; Scollon,
1991). Furthermore, since the social values and practices under the influences of
Confucian culture emphasize group harmony and conformity, Asian students in general
are portrayed as deficient in critical thinking abilities (Atkinson, 1997; Fox, 1994),
skills essential to composing effective English argumentative writing. Qualities such as
conformity run counter to the spirit of critical thinking since critical thinking involves
individualistic and adversarial practices(Atkinson, 1997).
Given these socio-cultural differences, previous L2 writing research has characterized
several writing features manifested in Chinese-speaking students’ English/Chinese
writing and assumed these features are transferred from Chinese rhetorical traditions
(Hinkel, 2002; Kaplan, 1966; Matalene, 1985; Scollon, 1991; Shen, 1989). This
presumed rhetorical transfer often leads to ineffective English writing due to its
incongruence with Anglo-American rhetorical traditions. Often, these researchers
examine features of general essay framework: location of thesis statement (Shen, 1989;
Wu & Rubin, 2000), lack of self-expression (Gregg, 1986; Bloch & Chi, 1995), or the
general macro-level Chinese organizational pattern qi-cheng-zhuan-ho (Cai, 1993;
Kaplan, 1966; Matalene, 1985). These features are generalized as Chinese rhetorical
Fei-Wen Cheng, Yueh-Miao Chen
25
practices across all genres.
Although the above studies do provide insight into how Chinese-speaking students
may transfer these general rhetorical strategies to their English argumentative writing,
several important rhetorical features associated with Western argumentation have never
been examined, such as the negotiation process. Negotiation process in argumentative
discourse refers to the way one deals with opposing views (Golder & Coirier, 1994). By
addressing opposition, a writer can continue to present a claim as being valid and
acceptable even under circumstances that might refute or undermine its strength
(Crammond, 1998; Golder & Coirier, 1994; Ramage, Bean, & Johnson, 2001).
Research in L2 writing has been hindered by the lack of utilizing a well-established
theoretical model of Western argumentation to examine how Taiwanese students
compose this genre. Thus, there is no accurate picture of how Taiwanese students
compose English argumentative writing and how effective their English argumentation
is. The purpose of the current project is to take a genre approach, drawing upon Stephen
Toulmin’s (1958 & 2003) model of argument (as elaborated in Crammond, 1998), to
examine closely how Taiwanese students employ diverse structures to defend the
validity of their thesis in their English writing, and to what extent culture exerts influence
over a Taiwanese student’s ability to compose effective argumentative writing.
2. Toulmin’s model
Toulmin (1958 & 2003) analyzed arguments in terms of six functional elements
and their relationships: (a) claim, (b) data, (c) warrant, (d) backing, (e) qualifier, and (f)
rebuttal (Toulmin, 2003, pp. 90-94). A claim refers to the conclusion to be argued for
and data denotes the facts or the premises drawn upon as the basis for the claim (p. 90).
Toulmin introduced the concept of warrant, which serves as the bridge to justify how
the claim is derived from the data (p. 91). Backing refers to facts, authorities, or
explanations used to strengthen or support the warrant (p. 91). Qualifiers refer to
modals, such as probably, possibly, perhaps (p. 92). By qualifying or narrowing the
TIESPJ, Vol. 1: 1, 2009
26
claim, a qualifier serves as an indication that the claim is not absolute or universal. A
rebuttal specifies the conditions which might defeat the major claim (p. 94).
Toulmin’s model is suited to analyzing structural variations in argumentative
writing because of its broad acceptance and wide application. First of all, his model is
developmentally appropriate for the analysis of general L2 argumentative essays at
ESL/EFL college levels since standards for this type of writing are general enough to
cover the basic structures underlying various types of argumentative discourse.
Although several theories of argumentation, such as Walton’s (1996) argumentation
schemes, and van Eemeren & Grootendorst’s (1992) pragma-dialectics have been
proposed in contemporary history of Western rhetoric, these alternatives to Toulmin’s
model all assume a level of sophistication beyond the ability of L2 college students and
lack broad acceptance in pedagogical and research contexts. In addition, the rhetorical
elements specified in Toulmin’s model can predict and explain the textual quality of
students’ argumentative/persuasive writing (McCann, 1989; Connor, 1990; Knudson,
1992; Ferris, 1994; Crammond, 1998). Furthermore, several researchers (Knudson,
1992; Yeh, 1998; Cheng, in press) have adapted Toulmin’s model to design instructional
programs for native-English or ESL/EFL learners and have reported improvements in
student writing in terms of an increase in the use of Toulmin’s rhetorical elements.
3. Purpose of the study
The goal of this study is to describe Taiwanese students’ ability to compose
English argumentative writing. To gain a comprehensive understanding of cultural
effects on rhetorical features, a cross-language and cross-nationality comparison design
has been adopted. By comparing Taiwanese students’ English writing not only with
their own Chinese writing but also with the argumentative texts of their American
counterparts, we should be able to ascertain whether certain rhetorical features are
culture-typical styles, patterns induced by language proficiency, or other potential
factors. Specifically, this study addresses the following concerns:
Fei-Wen Cheng, Yueh-Miao Chen
27
1. How do Taiwanese students employ diverse argument structures to justify their claim
in their English writing as compared to their American counterparts?
2. How do Taiwanese students employ diverse argument structures to justify their claim
in their Chinese writing as compared to their American counterparts?
3. How do Taiwanese students employ diverse argument structures to justify their claim
in their English writing as compared to their Chinese writing?
4. Method
4.1 Participants
Participants included 40 Taiwanese and 39 US college students. The Taiwanese
participants were freshman English majors in Taiwan, who have received one-semester
of instruction on English expository writing at a college level, but have yet to receive
any instruction on English argumentation. An equal number of US freshmen enrolled in
a required composition class in a large comprehensive university participated in the
present study. The number of participants recruited for this study was restricted because
of the heavy workload involved in the textual analysis.
4.2 Instrument: Writing prompts
Given that the Taiwanese students were required to compose two essays, in order
to avoid collecting contaminated data, writing prompts for two different topics were
designed so that students would not respond to the same topic twice. The Taiwanese
students and US students were assigned the following prompts.
In the first prompt (Topic 1), participants were given a hypothetical rhetorical
situation describing a foreign student studying in the US. The student had to go back to
his hometown in Africa during the school year since his father was ill and his relatives
insisted he should stay and take over the family business. However, this student was torn
in the conflict between family obligations and his personal pursuits. Participants were
TIESPJ, Vol. 1: 1, 2009
28
asked to give their opinions about whether they agreed or disagreed that this African
student should give up his studies and go back to help his family. This writing prompt
was adapted from an ESL textbook, Identity by Shaules, Tsujioka & Lida (2003, p.6).
The second prompt (Topic 2) presented a controversial issue involving the famous
long-necked women on the Thai/Myanmar border. Participants were given background
information about the Padaung tradition, which requires women to stretch their necks
by wearing brass coils, and about the impact of tourism on this tradition and the
lifestyles of these women. Participants were asked to argue whether tourists should or
should not visit these long-necked women. This writing prompt was adapted from an
ESL textbook, NorthStar Building Skills for the TOEFL iBT, Intermediate by Beaumont
(2005, pp. 96-97).
4.3 Data collection
In total, the writing samples collected included 119 essays, 80 of them composed
by the Taiwanese participants and 39 by American ones (See Table 1). Taiwanese
participants composed in-class essays on both topics, with order of topic and language
(Chinese or English) counterbalanced. American participants wrote only on one of the
topics, half randomly assigned.
The American participants composed only one 50-minute English essay in their
freshmen composition class, half randomly assigned one of the topics. To prevent topic
and treatment order effects, half of the Taiwanese participants composed Chinese essays
first and the English essays the following week, and vice versa. Also, half of the
participants wrote the Chinese essays on the first topic and the English essays on the
second topic, and vice versa. Their essays were hand-written in class, generally within
a 50-minute class period. The writing samples were collected at the beginning of the
second semester of their freshmen year, before they received any instruction on English
argumentation. Thus, the participants have general knowledge about how to compose
an English essay ensuring researchers that their Chinese and English texts were analyzable
using specific criteria based on English argumentation. Also, their lack of knowledge
Fei-Wen Cheng, Yueh-Miao Chen
29
about English argumentation prevents their transference of these rhetorical elements to
their Chinese writing so that the data is not contaminated for the purpose of exploring
cultural effects on their English argumentative writing.
Table 1 Student written texts by topic and language
Taiwanese students American students
N= 20 N= 20 N= 20 N= 19
Topic 1:
20 Chinese
essays
Topic 2:
20 English
essays
Topic 1:
20 English
essays
Topic 2:
20 Chinese
essays
Topic 1:
20 English
essays
Topic 2:
19 English
essays
4.4 Data analysis
All participant texts were subjected to the following analyses before conducting
any statistical analysis.
Step one: Identifying argument substructures. To examine student texts, the six
major elements in Toulmin’s model— claim, data, warrant, backing, qualifier, and
rebuttals— are further elaborated in the present study. Note that Toulmin’s model has
been criticized for its weakness as an analysis model since its applicability to analyzing
full-length arguments is problematic (Fulkerson, 1996; Inch & Warnick, 2002). Thus,
Toulmin’s model was modified in this study for the purpose of developing a model that
can deal with the range of argument structures that may be encountered in English
argumentation. Following Crammond’s study (1998), four structural elaborations are
added to Toulmin’s model: data backing, constraint, countered-rebuttal and alternative
solution. First of all, the backing element is further elaborated in the present study
to include not only backing for the warrant but also backing for the data. Second,
qualifier is expanded to include not only modality but also constraint, which refers to
conditions under which claim will apply. Thirdly, in addition to reservation (refers to
Toulmin’s rebuttal), two more oppositional structures are added: countered-rebuttal and
alternative solution. This elaboration of Toulmin’s model is represented in Table 2.
TIESPJ, Vol. 1: 1, 2009
30
Table 2 Comparison of rhetorical elements in Toulmin’s model and in the present study
Elements in
Toulmin’s model
Elaborated
elements in
present study
Operational definition/
Participants’ examples (Italics)
Claim claim
The conclusion to be argued for.
If I were the African student, I would quit school and go back to
take care of my father and take over my family business like him.
Data data
Facts or premises drawn upon as the basis for the claim.
At this period, going back assisting his father’s business is more
essential than finishing his study.
Warrant warrant
General assumption used to connect the claim and the data, such
as moral principles and cultural values.
Family is more important than personal pursuit.
Backing
backing
(including warrant
backing and data
backing)
Explanations, the background or context, examples and
authorities used to support the warrant or data.
At this period, going back assisting his father’s business is more
essential than finishing his study. *He still has the opportunity to
finish his study.* (data backing)
Qualifier
modality
constraints
Modal terms
I strongly recommended him to go back.
Conditions under which the claim will apply
He only need to go back to help his father if he were the only son
in his family.
Rebuttal
reservation
countered-
rebuttal
Conditions under which the claim will not apply.
He should go back but not under the pressure of his family.
Opposing views that challenge the writer’s claim and the writer’s
refutation to the opposing views
Some people may claim that we should stick to our dream instead
of sacrificing it for our family. Although pursuing goal is
important, a person’s biggest responsibility is not to himself but
to his family.
alternative
solution
A possible solution to the problem statement that is under
consideration other than the claim advanced by the writer
Tourists can help this tribe in another way. They can donate
money to set up school.
Among these argument substructures, claim and data are considered to be the
backbone of every argument while other elements are regarded as optional structures.
All writing samples (See Appendix A for a sample of argument diagram drawn by the
Fei-Wen Cheng, Yueh-Miao Chen
31
researchers based upon a Taiwanese participant’s English essay) were subjected to an
analysis of argument substructures and argument chain (consisting minimally of a
claim-data complex). Identification of arguments was based on semantic structures and
linguistic elements that typically signal the presence of supporting data as suggested in
Crammond (1998). Moreover, when the connection between claim and data is not
explicitly linked by conjunctive devices, the present researchers made judgments about
the writer’s intent by inferring arguments based on knowledge of reasoning structures.
For further analysis, all argument structures were subjected to double-coding, assigning
two labels to one substructure. For example, data was also double-coded as claim when
it is followed by backing. This double-coding can facilitate the analysis of argument
chains. Two independent raters established inter-rater reliability with the first author in
terms of these argument structures. The three raters coded the same 30 essays (38%) of
a randomly selected sample comprised of the written texts in English. On the analysis
of various argument structures, the inter-rater reliability ranged from 0.73 to 0.98.
Step two: Analyzing structural complexity. To explore the extent to which
students develop and elaborate their arguments, the following features were analyzed
for each text (See Table 3). The quantitative measure of argument size is the total
number of argument chains (Each argument chain consists of claim and data) to be
found in a text. Depth measure includes the longest argument chain in an argument
structure presented in a text. Elaboration is measured on two accounts: the maximum
variety (how many different types) of optional substructures a participant incorporates
into an argument structure and the total number of optional structures.
Table 3 Quantitative measures for overall textual variables
Argument features Quantitative measure
Size Total number of an argument (a Claim-Data Complex, labeled as
“Argument 1, Argument 2...” in the Appendix A) per text
Substructures The number of each argument substructure (i.e. claim, data, warrant, reservation,
etc.) used in each text
Depth The number of a Claim-Data Complex presented in the longest argument chain per
text
Elaborations 1) Maximum variety (how many different types) of optional substructures2) Total number of optional substructures
TIESPJ, Vol. 1: 1, 2009
32
After completing the textual analyses as illustrated above, two types of MANOVA
and one repeated measures ANOVA were undertaken for the present design. Significant
MANOVA effects were followed up by separate univariate ANOVAs for and of the
dependent variables. All textual features, acting as depen