Review of Collaborating towards Coherence
AUTHOR(S): Tanskanen, Sanna-Kaisa
TITLE: Collaborating towards Coherence
SUBTITLE: Lexical cohesion in English discourse
SERIES: Pragmatics & Beyond New Series 146
PUBLISHER: John Benjamins
Angkana Tongpoon, Department of English, Northern Arizona University, Ph.D.
''The concept of the lexical item ... is not totally clearcut; ... although
clearly defined in the ideal, it presents many indeterminacies in
application to actual instances ... it [the lexical item] is an essential
concept for the understanding of text'' (p. 202). This remark by Halliday
and Hasan (1976) reveals the challenge in the exploration of lexical
relation in discourse, but at the same time indicates its vital role in the
understanding of textual characteristics. Based on the challenge of lexical
cohesion, Tanskanen conducted a study on lexical relations, presented in
this nine-chapter book.
The first chapter provides the introduction and the rationale of the study.
Due to the absence of a model of lexical cohesion to study texts across
discourse types, Tansakanen states that ''for complementing our
understanding of the functioning of cohesion in discourse, comparisons of
the operation of cohesion in texts produced under different conditions
would be essential'' (p. 3). Also, she asserts that coherence is best
studied under the collaboration framework, which takes into account the
producer and the receiver's cooperation in creating understanding of text.
The goals of the study are to: 1) develop a model of analysis capable of
capturing all cohesively meaningful lexical relations in texts, and 2)
examine if and how their use varies depending on the conditions under which
the selected texts have been produced.
The second chapter discusses how cohesion and coherence are independent but
also interconnected. Cohesion is the property of the text, and coherence
depends on communicators' evaluation of that text. Both elements are
interrelated since cohesive devices help facilitate coherence. In
collaborative processes, cohesive devices serve as signals of collaboration
in which the communicators attempt to successfully interact with their
participants; these processes can be found in both dialogic and monologic
texts even though it is more obvious in dialogues.
Chapter 3 presents approaches used to analyze lexical cohesion in previous
studies and in this present study. Through the review of previous research
studies, Tansakanen compares the methods adopted in those studies, combines
them, and proposes a new model for the analysis of lexical cohesion
relations. Similar to Halliday and Hasan (1976), Tansakanen's model
consists of two main categories: reiteration and collocation. However,
Tansakanen's model differs from Halliday and Hasan's in terms of lexical
devices included in each category. In Tansakanen's model, reiteration is
the use of general nouns as cohesive agents; it pertains to the repetition
of a lexical item either directly or through the use of a generally related
word. Reiteration comprises eight devices to create lexical cohesion: 1)
simple repetition defined as repetition of same words (e.g., student -
student, play - play), 2) complex repetition meaning repetition of same
words but different functions (e.g., grade (v.) - grade (n.), cultural
determinism - a cultural determinist), 3) substitution (e.g., any racist
employer or group - they, Sylvia - her), 4) equivalence defined as synonym
in Halliday and Hasan (1976) (e.g., Nazi slaughter - Nazi extermination of
the gypsies, carbon dioxide - C O two), 5) generalization or the relation
between an item and a more general term (e.g., energy products - imported
oil, political party - Labour Party), 6) specification meaning the
relation between an item and a more specific item (e.g., Ebonican -
dialects, health education - the other social services), 7)
co-specification more specially the relation between two items which have a
common general item (e.g., RP speakers - Standard English speakers), and 8)
contrast or antonym (e.g., general - particular, out of fashion - up to
date). The second category of the model is collocation. Collocation is the
co-occurrence of words which regularly appear together in the language.
This category consists of ordered sets, activity-related collocation, and
elaborative collocation. Order sets are sets of lexical items such as
colors, months, days of the week. Activity-related collocation is defined
as relation between items based on an activity (e.g., eat - meals, drive -
car). Elaborative collocation is lexical items elaborating or expanding
other items (e.g., Cambridge - Cambridge's Mill Lane lecture room).
Chapter 4 is devoted to the rationale for the selection of discourse types.
Four types of discourse are included in this study: face to face
conversation, academic writing, mailing lists, and prepared speeches. These
discourse types are chosen based on Biber's (1988) principles of text
comparability. That is, spoken and written materials that can be compared
should be of the same typicality or markedness from each mode (i.e., spoken
and written). On one end of the spoken-written continuum are face to face
conversations, and on the other end are academic articles. The other two
text types (i.e., mailing lists and prepared speeches) are in the middle of
the continuum. Also, the four texts are compared adopting Biber's (1988)
multi-dimensional analysis by using estimated means, showing different
situational characteristics of the four texts.
Chapter 5 reports the results from the analysis of four face to face
conversations. Reiteration relations, particularly simple repetition pairs,
are most commonly found in all four face to face conversations and in both
two-party and three-party conversations. Both reiteration and collocation
relations are most often produced by the same speakers and typically within
a turn. Chain formation -- lexical units forming chain of previous related
units -- followed similar patterns in all four conversations. That is, the
beginning and ending of each chain identifies a topic of conversation.
The results from the analysis of the discussion mailing-list texts in
Chapter 6 show that the cohesive profile of the mailing-list texts is
similar to that of the conversations, particularly two-party conversations.
Reiteration relations, particularly simple repetition, are most frequent,
and collocation relations are rarely found. Different lengths of cohesive
chains seem to carry different roles in mailing-list discourse. The long
chains reflect the general topics of the discussions; the shorter ones are
related to topical segments.
The results from Chapter 7 concerning academic writing reveal that written
monologue is different from conversation and mailing-list texts in certain
respects. The total number of lexical devices is less than those of
conversation and mailing-list texts due to different communicative
conditions. The occurrences of reiteration pairs are lower even though they
are still a dominating device in this type of discourse. Moreover, the
proportion of collocation is higher. However, the three text types are
similar in chain formation, which apparently marks textual segments.
Especially in the written monologue, chains identified different sections
of the articles. In addition, roles of cohesion in the text are evidenced
in the construction of a summary of a text through the combining of
sentences with an average or a higher number of cohesive units.
Chapter 8 reports on the analysis of prepared speeches. The cohesive
profile of the prepared speech shows a combination between conversation and
academic writing. That is, the number of simple repetitions is close to
that of conversation; the use of other devices (i.e., generalization,
specification, and co-specification) is similar to academic writing.
Similar to other text types, prepared speeches contain both long and short
chains. Long chains reflect the general topics of the speeches, and shorter
chains indicate topical segments. Also, a summary of spoken monologue
through the use of sentences with an average or a higher number of cohesive
units can be effectively created as it is in a written monologue.
Chapter 9 is the concluding chapter, which summarizes and discusses the
results from the previous four result chapters. Based on the results from
the four discourse types, the author concludes that the created model under
the collaborative framework reveals useful findings across discourse types.
Cohesive profiles are affected by cognitive contexts (i.e., cognitive load
required by different contexts) and social context features (i.e., the
absence or presence of visual contact has an impact on the use of lexical
cohesive devices, while the channel (i.e., spoken versus written) plays
little role. Finally, cohesive strategies are sensitive to communicative
conditions such as numbers of participants in the conversation.
I found this book easy to read and to understand. Due to the fact that the
book follows a conventional style of a research report, the reader's
expectations are met. Also, the writer's clear writing style and
appropriate use of technical terms appeal to the reader. Most technical and
new terms are defined, redefined, and accompanied with clear samples. This
writing style helps enhance the reader's understanding of novel concepts
and new information. In addition, the result chapters are well organized in
a consistent manner in terms of sections and tables presented. This leads
to a clear presentation of the results even though the content is dense in
nature. Concerning the reconstructed model, the author employs a successful
effort of reconstructing a model based on previous approaches. Including
features from other approaches into this new model (e.g., trigger for the
analysis of chain formation), Tansakanen's model is likely to capture
collocation relations in a concrete way, which is not often found in those
kinds of studies. As the results of the study shows, the model can be used
to effectively analyze lexical cohesion relations, particularly collocation
which is often suggested to be left out in the lexical cohesion relations
due to its difficulty and subjectivity (Hasan, 1984). In addition, the
study contributes to the reader's understanding of the use of lexical
cohesion in different text types, which is still lacking in the field.
The major problem of this book seems to be concerned with the
interpretation of the results, particularly the explanation of the
differences between frequency figures. In Table 5.3 on page 100, for
instance, the normed counts per 1,000 words of collocations are just two in
both Conversations 1 and 2 (i.e., three-party conversations) and five and
four in Conversations 3 and 4 (i.e., two-party conversations). In prose,
the author reports that ''in a two-party conversation collocation is more
frequently used'' (p. 100). This is reported again on page 111: ''there were
differences in the use of collocation as well: collocation relations
produced by different speakers were much less frequent in the three-party
conversations than in the two-party ones.'' I would argue that the
occurrences of collocations in the two-party and the three-party
conversations were very close, and it seems the difference between the
two-party conversation and the three-party conversation was marginal,
between two and five cases. This information, instead, seems to suggest
that the two types of conversations follow a similar pattern of collocation
use. As a consequence, the conclusion of the sensitivity of cohesive
strategies to communicative condition in multiple-party conversations may
be questionable. Accordingly, when small differences are found, the
interpretation of the results should be made with caution. However, this
study is not the only study struggling in the comparison of occurrences due
to subjectivity in determining differences. This has been evidenced in
other cohesion studies (e.g., Bolton, Nelson, & Hung, 2002; Chen, 2006).
How differences can be determined as strongly being differences is a
question needing an urgent and clear explanation from the field.
It was a surprise that while it seems to be the main focus of the study,
collaborating (i.e., the process that the producer and the receiver
actively cooperate to achieve coherence) accounts for only a small
proportion of the book, about ten pages. Even when the results are
discussed under collaboration, it seems the discussion relates more to the
varied contextual features of different text types than the collaboration
between the producer and the receiver. This may be due to the fact that
collaboration is not clearly defined and operationalized in terms of how it
varies in different communicative conditions of the four text types
analyzed, except in a general overview of spoken and written texts. In
relation to this, while the method of text selection is relatively well
grounded in its link to previous literature (i.e., the use of Biber's
multidimensional analysis to select text types), it is not clearly
understood how the multidimensional information of the texts reveals the
varied collaborative features of the four texts (i.e., cognitive levels, p.
27); the information from the multidimensional analysis again identifies
different communicative conditions (or situational characteristics) (Biber,
Conrad, & Reppen, 1998). As such, it is hard to understand how the
collaboration affects or describes the use of lexical devices in different
texts, as the study is trying to point out and indicates in the title of
Finally, the book does not report raw frequency counts of the occurrences
of cases found in cohesive devices (e.g., equivalence, ordered set,
activity-related collocation, and elaborative collocation) or the total
number of words used in the analysis. The normalization method seems to be
appropriately used, since this method provides a way to compare texts with
different length more accurately (Biber et al., 1998). This study, however,
does not report the total words of each discourse and raw frequency counts.
Without this information, it is hard for the reader to make an appropriate
judgment about the use of the normed count per 1,000 words as well as the
real picture of the occurrences, which is crucial for understanding and
interpreting of the data.
In spite of these weak points, this book is useful. The style of writing is
remarkably clear, and technical terms are well defined. The main
contribution of the book is the reconstructed model of lexical cohesion
relations. It offers an insightful way to analyze lexical cohesion,
particularly collocations, which are perceived as a challenge to deal with.
In addition, it sheds light on the characteristics of lexical cohesion in
different text types from varied contextual features. In regards to its
audience, this book may be of interest not only to experienced academics,
but also to graduate and post graduate students wishing to learn more about
lexical cohesion analysis and a new model for such analysis.
Biber, D. (1988). Variation across speech and writing. Cambridge: Cambridge
Biber, D., Conrad, S. & Reppen, R. (1998). Corpus linguistics:
Investigating language structure and use. Cambridge: Cambridge University
Bolton, K., Nelson, G., & Hung, J. (2002). A corpus-based study of
connectors in student writing: Research from the International Corpus of
English in Hong Kong (ICE-HK). International Journal of Corpus Linguistics,
Chen, C. (2006). The use of linking adverbials in the academic papers of
advanced Taiwanese EFL learners. International Journal of Corpus
Linguistics, 11, 113-130.
Halliday, M.A.K. & Hasan, R. (1976). Cohesion in English. New York: Longman.
Hasan, R. (1984). Coherence and cohesive harmony. In J. Flood (Ed.),
Understanding reading comprehension: Cognition, language, and the structure
of prose (pp. 181-219). Newark, DE: International Reading Association.
ABOUT THE REVIEWER:
ABOUT THE REVIVEWER
Angkana Tongpoon is a Ph.D. student in applied linguistics at Northern
Arizona University. Her research interests include discourse analysis,
second language vocabulary acquisition, and Computer Assisted Language
Learning (CALL). She is an assistant professor at the Department of
Foreign Languages, Khon Kaen University, Thailand.