Publishing Partner: Cambridge University Press CUP Extra Publisher Login
amazon logo
More Info

New from Oxford University Press!


Sorry About That

By Edwin L. Battistella

Sorry About That "explores why we apologize or don't and how our apologies succeed or fail."

New from Cambridge University Press!


Sociolinguistics from the Periphery

By Sari Pietikäinen, Alexandra Jaffe, Helen Kelly-Holmes, Nik Coupland

Sociolinguistics from the Periphery "presents a fascinating book about change: shifting political, economic and cultural conditions; ephemeral, sometimes even seasonal, multilingualism; and altered imaginaries for minority and indigenous languages and their users"

Review of  Patterns of Text

Reviewer: John Hammink
Book Title: Patterns of Text
Book Author: Michael R. Scott Geoffrey Thompson
Publisher: John Benjamins
Linguistic Field(s): Computational Linguistics
Text/Corpus Linguistics
Issue Number: 13.1733

Discuss this Review
Help on Posting

Date: Tue, 18 Jun 2002 18:08:59 +0300
From: "Hammink, John"
Subject: Scott & Thompson (2001) Patterns of Text: In Honour of Michael Hoey

Scott, Mike and Geoff Thompson, eds. (2001) Patterns of Text: In Honour of Michael Hoey. John Benjamins Publishing Company, vii+323pp, hardback ISBN 90-272-2572-9 (Eur) / 1-55619-792-6(US).

John Hammink, F-Secure Corporation, Helsinki

'Patterns of Text' contains an introduction and 12 papers which explore text patterns above the sentence level, as the previous reviewer for this list has noted . The subtitle: 'In Honor of Michael Hoey' appeared at first glance to be a reference to Michael Hoey's 1991 'Patterns of Lexis in Text' although after viewing Hoey's bibliography, I found there to be much more inquiry into this relatively underexplored, yet long-inhabited terrain on his part. Nonetheless, the editors have assembled a star-studded cast of contributors bound by the written-discourse analysis thread, and this collection is a decisive contribution to the field.

The book opens with an introduction by the editors themselves. 'Why', do they ask 'should there be a Patterns of Text?' The editors acknowledge a shift in linguistics concerns over the past three decades, from language as a set of isolated syntactic structures to language 'as a set of functional resources in use', or more succinctly, from the clause to the text. In short, the focus has turned from micro to macro. That doesn't only mean a shift purely to semantics, but the study of texts as objects themselves. What are the texts made of? How do their constituents 'hang together'? What can be said about their function in and of 'the society that produces them?' As language events, the editors ask us to study texts as 'complete' events even though the completion is relative. In speech, is the complete event the punchline, the joke, or the conversation in which the thing takes place? In text, is it the encyclopedia article, Volume A, or the whole edition? How do we know we are dealing with the same 'story' even though it appears on different TV channels, radio stations and newspapers on different days? On this note, the editors point out some of the limits of general linguistics when assessing completion as well as considering text events themselves.

Since participants and users can recognize completion in a language event, and therefore elicit a complete text, the language event must therefore be 'patterned in a distinct way'. The editors propose the concepts of 'Conjunction' and 'Repetition'. Conjunction is seen as being concerned with how different parts of a text fit together; Repetition is an aspect of continuity or 'how the...writer signals that...she is still talking about the same thing.' All of the articles in the collection would appear to be tied to one or the other or both of these concepts.

'Colligation, lexis, pattern and text' by Susan Hunston derives its founding ideas from the Repetition concept. Colligation 'depends on repetition between the text and other texts'. The author presents it as a 'wider phenomena' of collocation, that is, the tendency of certain words to co-occur. Hunston sheds light on colligation and even provides some corpal and syntactic evidence of the phenomenon in relation to lexis, pattern and text.

Antoinette Renouf, in her article, 'Lexical signals of word relations' used CL (Corpus Linguistics) techniques to study and identify semantic-relation signals, which, as the previous review has noted, include sense relations like hyponomy, antonomy, and synonomy. In the process, she also studied correlation between those lexical signals themselves and meaning. Automated extraction of sense-related word-pairs from text is applicable in text retrieval as well as linguistics itself--word pairs can provide an 'abstract' of a larger 'textual thesaurus'.In 'Patterns of cohesion in spoken text', Susan & Geoff Thompson attempt to answer the question 'How do chunks relate to each other?' Their texts are a selection of television advertisements where viewers must simultaneously handle information at various levels. 'What makes the advertisement a single audiovisual unit', in other words, a text event, 'as opposed to a set of elements?' The authors view the text events through the 4-band spectrum of repetition, replacement, conjunction and theme. Repetition build a framework with 'slots', which 'signals continuity of aboutness.' Replacement switches lexical and other items into the familiar slots. Conjunction works alongside theme, signalling connectedness. Both words and sounds are considered as cohesive devices.

Peter Fries uses Martin's (1992) terminology to distinguish between 'Presenting' and 'Presuming' reference. In 'Issues in modelling the textual metafunction', Fries explains how the writer or speaker may signal the reader about the status of chunks of information in a flowing text, that is the 'newness' or the contextuality of a given chunk. An early distinction is made between participant entry and information structure.

Mike Scott's paper dealt again with repetition in 'Mapping key words to problem and solution.' A word can possess 'keyness' in a text if it occurs more frequently in that text than in a large corpus. If this is the case, than does that key word reflect in some way what the text is about? Scott explores the Problem/Solution pattern. As there appears to be no direct match between the pattern in the signals (the texts might show the pattern, but not the keywords themselves which are supposed to be the signals.). That is, the key words appeared outside the part of the text constructed as Problem-Solution, or the keywords only had a local scope. One of the many questions raised: How can the user distinguish between an incidental mention and a pattern signal?

In 'Negotiation of evaluation in written text', Adriana Bolivar discusses the role of evaluation shaping text structures and determining types and genres of text. While aiming to understand text as social interaction she starts from the key role that evaluation plays in that interaction. 'Lead, Follow and Valuate' are three parts of this interaction. She uses editorials as the test data. In the last section of her article, she touches on the implications of this for teaching reading and writing.

As many of the articles focus more on repetition, 'Some discourse patterns and signalling of the assessment-basis relation' focuses strictly on conjunction. Assessment and Basis in text can also be evaluated in terms of the conjunctive devices at macro- and micro-levels. As author Michael Jordan writes: 'binary relations of meaning exist between (usually adjoining) stretches of text', it is possible to also compare Assessment-Basis conjunctive devices to those used for Topic-Appraisal, Cause-Effect, and Purpose-Means. Also worthy of note, are the issues of Multiple Bases, Assessments as Basis, and the complications which arise when combinations of several such relationships exist.

Repetition in context comes into view in Ann Darnton's chapter 'Repeat after me'. By repeating contexts rather than raw lexical items (in ever-changing contexts), beginning readers may be more easily able to recognize a word each time it reappears throughout the text, rather than re-decipher it. Moreover, a tagmemic analysis, where a text may be analysed hierarchically in terms of its grammatical constituency; phonological constituency; and its referential or lexical applied to the stories included in many holistic reading schemes. Several parallinear 'vectors' can be drawn through the narratives, which create a framework. The framework supports the reader's expectations in story development and what words/meanings to expect when the words reoccur.

In 'Lexical Segments in Text', Tony Berber Sardinha 'proposes a detailed and ingenious method for getting at the boundaries within a text,' by looking at changes in aboutness, again, based on repetition (in this instance, by analyzing lexical links across texts). The author asks, 'Can a computer program do this?' While Michael Hoey's model applies to connectivity rather than segmentation, and if groups of sentences with a degree of lexical similarity can be considered 'similar', then presumably the opposite can also hold true. However, the volume editors themselves point out that the world is too big for any limited approach that a computer system solve-it is through the continual process of 'model making, model implementation and model testing' that insights are found and progress made.

The Repetition-Conjunction dichotomy reappears in 'Patterns of Lexis on the surface of texts', a sly acknowledgement of Michael Hoey's best-known work. Author Malcolm Coulthard is concerned with textual plagiarism, and as plagiarists copy not only lexical items, but even entire sentences, a Matching relation where lexical repetition between texts can be compared, is a useful step toward this end. Interestingly, Coulthard uses literary texts, but also student essays and police records of interviews as his 'corpus'. Since an important issue in plagiarism is directionality, (who wrote what first?), Coulthard even proposes a means for determining this, by comparing cohesive devices between the texts. Admittedly, while this is still an emerging methodology, the possibilities for 'identifying written voices', as a previous reviewer pointed out, nonetheless exist.

'Patterns of text in teacher education' asserts the need for studying the macro, as opposed to the micro level of texts, where the lion's share of educational concerns seems to rest on grammatical and syntactic concerns. Unlike the rest, authors Julian Edge and Sue Wharnton evoke a narrative and use it to explain Hoey's Situation-Problem-Response Evaluation (SPRE) model. Where non-native speakers may complain of, say, problems with grammar and lexicon, L1 language speakers may only be able to say that 'writing is tricky.' The goal of such macro awareness, therefore, appears to be to 'obtain the mental tools to describe and solve...problems'.

HMI (Human Machine Interaction)offers a type of discourse whereby the user must interact with a machine according the machine's agenda, which allows for, perhaps, a range of prompts at a command line or some clicks on a specified part of screen. In 'The deification of information', one of the founding fathers of corpus linguistics, John Sinclair, describes such one-way model so typical of HMI, as opposed to the two-way communication model which humans use amongst themselves. As HMI uses a one-way type of communication, the human user is unable to revert to her own agenda to ascertain the current state of the machine and adjust the communication strategy accordingly. This is a dangerous state of affairs when humans are increasingly ill-equipped to handle 'the vast amount of information available by computers', as a previous reviewer put it.

As I am not only familiar with Michael Hoey's work from material related to this review, I was at first a bit confused by the 'In Honor of' part of the title, as it implied to me that there would be a keener sense of 'aboutness' or connectedness between the works represented here. Nonetheless, this volume brings to the surface a number of separate issues which appear to have been invoked by Michael Hoey's work (see the references below). And the articles themselves illustrate the range of application of his ideas.

In any case, one doesn't have too much trouble finding a binding thread among the articles when one considers the notions of Repetition and Conjunction as cohesive devices.

For added readability, the volume's index is broken into a 'Name Index' and a 'Subject Index', which those perusing the book for specific knowledge or reference purposes may find useful.

Hoey, M.P. 1983. On the Surface of Discourse. London: George Allen & Unwin.

Hoey, M.P. 1991. Patterns of Lexis in Text. Oxford: Oxford University Press.

Hoey, M.P. 2001. Textual Interaction. London: Routledge.

ABOUT THE REVIEDWER John Hammink holds a B.Sc. in Linguistics from Eastern Michigan University. He has been working for the last five years in the areas of data-security software localization and validation, where his work focus ranges from system test automation using trainable language corpora to applying corpus linguistic research in software document inspections. His other interests include Linguistic Evolution and the. Finno-Ugric Language group.