LINGUIST List 13.2054

Fri Aug 9 2002

Review: Lexicography/Pragmatics: Wray (2001)

Editor for this issue: Naomi Ogasawara <>

What follows is a review or discussion note contributed to our Book Discussion Forum. We expect discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in. If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for review." Then contact Simin Karimi at


  1. Hurriyet Gokdayi, Wray (2001) Formulaic Language and the Lexicon

Message 1: Wray (2001) Formulaic Language and the Lexicon

Date: Tue, 06 Aug 2002 22:21:39 +0000
From: Hurriyet Gokdayi <>
Subject: Wray (2001) Formulaic Language and the Lexicon

Wray, Alison (2001) Formulaic Language and the Lexicon.
Cambridge University Press, 352pp, hardback ISBN 0-521-77309-1, $64.95.

Book Announcement on Linguist: 

Hurriyet Gokdayi, Department of Turkish Language and Literature, 
Mersin University


Our everyday language significantly contains routine elements.
Speakers (of a language) easily recognize their form and keep them in
their memory as prefabricated units. Though they are commonly used,
their structural boundaries have not been clearly stated and what role
they play in the production and comprehension of linguistic messages
has not been firmly established. With her book "Formulaic Language
and the Lexicon," Alison Wray attempts to explore the nature and
purposes of formulaic language and gradually build up a "unified
description and explanation of formulaic language and its status
relative to the lexicon" (p.5). Wray also aims at giving us a full
account of the vast literature in various fields that somehow includes
formulaic language and providing an inclusive approach to explain what
language is and how we use it.


Wray's book is divided in 6 parts and 14 chapters, which are followed
by notes (pp. 283-300), references (pp. 301-326) and index
(pp. 327-332). It also includes 21 figures and 5 tables (listed in
pp. vii-viii). The book is organized as follows:

Part I. What Formulaic Sequences Are

Chapter 1. The Whole and the Parts (pp. 3-18) 

In this chapter, Wray states the focus of her book as formulaic
elements of the language because we need to include them in our
explanation of how language works and certain modern theories of
linguistics (e.g. generative syntax) ignore them. She asserts that
linguists should recognize "the role of formulaicity" that
fundamentally affects our "understanding of the freedoms and
constraints of language" (p.5). Having reviewed the majority of
findings in this field, Wray puts forward a hypothesis to explain what
formulaic language is that "it is a dynamic response to the demands of
language use and, as such, will manifest differently as those demands
vary from moment to moment and speaker to speaker" (p.5). She also
offers replacing "formulaic language" with "formulaic sequence"
(defined in p. 9) because it has not been used in the literature and
it can capture all characteristics of formulaicity. Wray also
proposes "a dual-processing systems" (p. 14) to explain how speakers
treat linguistic material. This proposition contains two processing
ways, namely holistic and analytic. She favors holistic processing to
utter and decode formulaic elements because it simply reduces
processing effort for the speaker and the hearer.

Chapter 2. Detecting Formulaicity (pp. 19-43)

In this chapter, the author discusses what kinds of tools could be
used to identify formulaic sequences (FS). Wray reviews the potential
roles of the following features associated with FS in her pursuit of
identifying the subject: Intuition of individuals; shared knowledge of
the same speech community; frequency counts in texts; structure,
internal composition, fixedness, and phonological form of FS; fluency
and pause, stress and articulation, and pronunciation while speaking;
liaison in French; and code-switching in bilingual speech. The author
concludes that there is no single criterion that helps to distinguish
FS from similar structures, e.g. idioms. Therefore, FS should be
governed by some unified criteria.

Chapter 3. Pinning Down Formulaicity (pp. 44-66)

This chapter includes a detailed review of form, function, meaning,
and provenance, which are the main themes used to describe FS in the
literature. Form-based approach enables researchers to determine
descriptive characteristics of FS but do not explain the wide variety
of fixed and grammatical sequences and invariable structures.
Function-based approach associates FS with situations but some do not
require a situation for their utterance. When it comes to
meaning-based account, Wray thinks that semantics and pragmatics are
easily mixed in this method. Provenance is another theme adopted to
determine the boundaries of FS. Wray cites Peter (1983) saying that
'the speech formula' [FS] is fabricated either though social
negotiation or individual evaluation (p. 59). However, this model
also cannot solve all the problems related to the identification of
FS. For example, when listeners hear a new sequence, it is difficult
for them to describe the utterance as formulaic, as in the example of
"to take a rain check" in British English. In addition, Wray
considers using continuum model to determine precisely what
formulaicity is. Nonetheless, the inherent problem with continuum
models that how to decide the base of any given model still remains.
Wray's discussion in this chapter suggests that attempts to define FS
based on a single theme or characteristics are eventually destined to
fail because none of them covers all aspects of formulaicity.
Consequently, there might be a need to utilize all discussed features
together to pin down FS.

Part II. A Reference Point

Chapter 4. Patterns of Formulaicity in Normal Adult Language (pp. 69-92)

Having focused on the language of adult native speakers in this
chapter, Wray aims at determining some reference points about the use
and functions of FS to make comparisons with the language of children,
non-natives and aphasics in later chapters. She evaluates probable
roles of FS in traditional oral texts (oral storytelling and poetry),
sports commentaries, auctions, and weather forecasts. While composing
those texts, speakers purposely utilize FS to lessen processing effort
both for hearers and for themselves, and eventually become very fluent
in their delivery of the texts. It seems that reducing the processing
effort is the most visible function of formulaicity. Besides, Wray
identifies additional functions that FS serve such as signaling
identity, manipulating the situation, and elevating the interlocutor
or someone else. The evidences presented in this chapter imply that
there seem to be several roles for FS relating to the speaker and the

Chapter 5. The Function of Formulaic Sequences: A Model (pp. 93-102)

Based on the discussions in previous chapter, this relatively short
chapter proposes that all of the functions of FS actually serve the
promotion of the speaker's interests (listed in p. 95-96). Wray
presents a model in which various functions of FS are reduced to three
(p.101): The reduction of the speaker's processing effort, the
manipulation of the hearer, and the marking of discourse structure.
Then, they are further reduced to one single function that overrides
all others. That is the speaker's promotion of self. Even though
that is a non-linguistic problem, speakers solve it through a
linguistic tool (FS).

Part III. Formulaic Sequences in First Language Acquisition

Chapter 6. Patterns of Formulaicity in Child Language (pp. 105-127)

The author shows that early stages of child language also feature
various kinds of FS such as rhymes, songs, and simple greetings.
Based on the literature review, Wray argues that children use both
analytic and holistic processing while acquiring their first language.
In the process of learning and making use of their first language,
children basically have four main goals, that are to get things done,
express individuality, gain control of language, and feel part of the
group (Figure 6.3 in p. 125). In order to achieve these goals,
children tend to employ FS mostly that are either created by fusing
strings of words or formed by reproducing complex utterances of care
givers. The reason for this tendency is that FS lessen the processing
effort for both parts of the communicative exchange.

Chapter 7. Formulaic Sequences in the First Language Acquisition
Process: A Model (pp.128-139)

This chapter discusses how children know not to analyze FS and add
them in their memory as a whole when acquiring the first language.
There are identificational and definitional cues for FS that were
presented in Chapter 2 and 3. Wray maintains that those cues are not
necessarily available to children who do not learn language for the
sake of learning grammar but do acquire language for the purpose of
communicating with others. Children rarely apply the analytic
processing during first language acquisition and "operate with the
possible largest unit" (p. 138). As Wray points out, children do use
analytic processing only when they need to do (e.g., acquiring
literacy, learning how to read and write). To explain the relevant
balance of holistic and analytic processing during first language
acquisition, Wray proposes a developmental account (Figure 7.1 in
p. 133). This account suggests that children pass through four phases
concerning analytic and holistic processing during first language
acquisition. Children start with holistic processing and end up
having a balance between holistic and analytic processing in late
teenage years.

Part IV. Formulaic Sequences in a Second Language

Chapter 8. Non-native Language: Overview (pp. 143-149)

In this another short chapter, Wray discusses how to make use of the
findings of second language learning research in her exploration of FS
and how to make judgements about them. She reviews the evidence from
second language acquisition research to explore "the ways in which
formulaic sequences seem to be used, or not used, to support the
individual's promotion of self, by promoting fluency and ensuring
hearer comprehension" (p. 145). The evidence suggests that the use of
FS by second language learners depends mainly on their priorities in
socio-interaction and processing. Then she presents a summary of the
main patterns found in this data in pp. 147-149.

Chapter 9. Patterns of Formulaicity in Children Using a Second
Language (150-171)

In this chapter, Wray evaluates the results of 14 case studies, which
specifically focused on the role of FS in children's learning a second
language. Those studies were conducted with 21 children (12 girls, 9
boys), aged approximately 2 to 10. The subjects in these experiments
established their first language before they were exposed to a second
language. Based on her review, Wray maintains that children also
express themselves holistically in a second language by employing FS
in most cases. They primarily have four goals as in the case of first
language acquisition (presented in Figure 6.3 in p.125). FS serve as
a means of accomplishing one or more of those goals for children in
the second language learning process.

Chapter 10. Patterns of Formulaicity in Adults and Teenagers Using a
Second Language (pp. 172-198)

Wray reviews the available data on second language learners concerning
the roles FS in the language acquisition process. This data consists
of 7 studies examining FS in the process of adults' learning a second
language naturally and 20 studies carried on classroom setting. Based
on the data, Wray indicates that teenagers and adults learning a
second language are unable to balance between formulaicity (holistic
processing) and creativity (analytic processing). The evidence
suggests, as Wray argues, that they seem to use either too much or too
less FS and behave too creatively in the process of acquiring a second
language. This approach might hinder the success in the acquisition
of a second language.

Chapter 11. Formulaic Sequences in the Second Language Acquisition
Process: A Model (pp. 199-213)

Having reviewed the relevant literature, Wray proposes "a second
language version of the first language model" (p. 199) to explain the
complex roles of FS in the process of learning another language.
According to the model, there are three main points: 1. Adult second
language learners make mistakes when they retrieve and reconstruct FS
from their memory because they take apart FS to get the lexical
constituents, store them separately, and do not keep information about
how word strings stay together. 2. They are not able to acquire and
employ FS in the same way children do because they perceive the word
as the main unit of the language structure. 3. Contrary to the native
speakers, they understand collocations as separate items, which become
paired. This model is able to account many differences between child
and adult second language learners in relation to the acquisition and
use of FS. Children could easily learn, store, retrieve and use FS as
a whole with holistic processing because their goal is to control
the environment and have a place within it. However, adult second
language learners usually prefer the analytic approach towards FS,
which indicates a firm control to the language units. Adults
frequently fail to recognize that FS are units that belong, not go,
together and there is no need to separate them.

Part V. Formulaic Sequences in Language Loss

Chapter 12. Patterns of Formulaicity in Aphasic Language (pp. 217-246)

This chapter presents the role of FS in one type of language loss,
aphasia, in which the left hemisphere of the brain is damaged. It has
been discovered that FS could survive the most types of aphasia
because people store at least some, if not all, FS in the right
hemisphere and/or subcortex. Wray examines the data to find answers
to following questions: 1. How do aphasics treat FS in their
processing? There is plenty of evidence, Wray presented, that
aphasics treat formulaic word strings like words rather than phrases
or sentences. 2. What kind of role do FS play in language processing
of the aphasic speaker? Aphasic people have much more difficulty in
the promotion of self because their language processing ability is
somewhat impaired. Therefore, as studies suggest, aphasic speakers
depend on FS more than normal speakers do in order to get their
message across for it reduces the processing effort for the speaker
and the hearer. In addition, they use FS for commenting and giving
answers, which is different from normal speakers, who use them "for
questioning and expressing personal beliefs, attitudes, feelings and
emotion" (p. 236).

Chapter 13. Formulaic Sequences in Aphasia: A Model (pp. 247-258)

In this chapter, Wray draws another model to explain aphasic language
with reference to normal adult language. With her model (Figure 13.1
in p. 249), she divides the lexicon of a normal person (adult native
speakers and in this case non-aphasic) in five sections in relation to
the function each serves. First section includes the lexical units
used to grammatically construct novel utterances. Second stores
referential expressions. Third contains elements used in routine
interactional situations. Fourth retains memorized texts. Fifth
consists of fully reflexive expressions. In this model, first and
second sections are located in the left hemisphere while the rest is
located in the right. These five sections all holistically store
elements with different sizes (morphemes, words, word strings). In
addition, Wray's model includes multiple representation of elements in
different sections as exemplified by the author in pp. 252-253.
According to Wray, it is possible with this model to explain that
aphasic people whose left hemisphere damaged are not able to produce
fully grammatical novel utterances because their lexicon have lost
first and second sections. It is also possible to explain why
aphasics are able to utter and use FS because third, fourth and fifth
sections of their lexicon located in the right hemisphere are still

Part VI. An Integrated Model

Chapter 14. The Heteromorphic Distributed Model (pp. 261-281)

In this final chapter, Wray proposes a "macro" model of lexicon in
which all the models presented in different chapters were drawn into a
unified model (Figure 14.1 in p. 263) called The Heteromorphic
Distributed Lexicon. The model divides lexicon in five subgroups each
featuring the distribution of three types of formulaic elements (the
morpheme, the formulaic word and the formulaic word string). This
model is "the repository of all linguistic units which are not subject
to further segmentation and which are therefore handled as holistic
units" (p. 264). Wray asserts that individuals' priorities determine
the nature of the lexicon. She reaffirms that a single system
processing cannot be adequate to explain the handling of all
linguistic material by speakers and hearers. Instead, a dual-system
model, in which language is processed in both holistic and analytic
ways, possibly provides a better explanation of processing linguistic
input and output. Therefore, we can clearly see how grammar works
and "how language use determines patterns of distribution and
frequency" (p. 278). Wray concludes declaring that researchers must
examine linguistic behavior "as one manifestation of more fundamental
... socio-interactional priorities such as the promotion of the self"
(p. 281).


I have been stimulated a great deal by Wray's book 'Formulaic Language
and the Lexicon.' The author gives an extensive review of the
literature concerning FS in the fields of discourse analysis, language
acquisition, and language pathology. Having surveyed a vast amount of
research findings, Wray provides different kinds of tools that could
help us to identify and define FS. She then presents four models to
explain how the speakers and the hearers handle FS in adult native
language, first and second language acquisition processes, and
aphasics. In the conclusion, Wray proposes a final model, which
unifies all previous models and presents the distribution of all
linguistic units, formulaic and non-formulaic, in the lexicon of a
speaker. This model enables us to explain how we use, process and
store formulaic and other elements of the language.

Wray makes three important points: 

1. Speakers possess a dual-processing system, in which analytic
processing is used to produce and comprehend grammatical novel
utterances, and holistic processing system is used to utter and
comprehend formulaic elements of the language. The processing effort
required for the holistic processing is lesser than that of the
analytic process. 2. Formulaicity is commonly used to promote the
self. 3. Formulaicity is placed at the center of the language.

While the book is both interesting and inspiring, it's worth to
consider some minor weaknesses. First, though the title contains the
term "Formulaic language," it has been replaced with "formulaic
sequence" in the rest of the book. Wray offers this term for the
place of 'formulaic language' and 58 other terms used to refer to
formulaic parts of the language. She thinks that formulaic language
"is too commonly used to be free of such associations [the same term
used to used for different things]" (p. 9) and it cannot be defined
clearly. Wray gives a definition of formulaic sequence which is "a
sequence, continuous or discontinuous, of words or other elements,
which is, or appears to be, prefabricated: that is stored and
retrieved whole from memory at the time of use, rather than being
subject generation or analysis by the language grammar" (p. 9). This
definition is inclusive but seems to be incomplete and not clear
enough. Wray uses the term 'other elements' in the definition but
do not explain what the 'other elements' are.

In addition, a sequence implies units or morphemes that are attached
to each other. In this case, it is hard to assume that such words
like 'yes, no, huh' are formulaic sequences. Wray later
differentiates between types of FS and uses the terms 'formulaic word
strings, formulaic word and morpheme. Wray also admits that "the term
[formulaic sequence] will have to be used fairly loosely and a
coverall" (p. 9).

Second, in Chapter 2, Wray considers liaison in French as a criterion
to detect formulaicity (p. 39). I think this is a little confusing
because liaison might be used to identify formulaic elements in French
but not in English.

Third, in chapter 2 and 3, Wray reviews the relevant literature and
evaluates the criteria for the identification and description of FS
which "are not a single and unified phenomenon" (p. 66). Therefore I
assumed that Wray would provide us a framework in which we could
utilize a bunch of criteria to determine and clarify FS. Instead, she
presents models that explain how FS are treated by native adult
speakers, children, second language learners, and aphasics.

Fourth, I am concerned with the contradictory methods in the
discussions of FS in first and second language acquisition processes.
Having considered teenage years as a stage in children's first
language acquisition process in chapters 6 and 7, Wray discusses and
evaluates the research findings from this perspective. However, she
regards teenage years as a period in adults' second language learning
in chapters 10 and 11, and examines the literature based on this
categorization. I think this approach do not give a reliable account
of the research findings.

Fifth, I don't accept the term 'polyglot' for the place of bilingual
in Chapter 13 (p. 298). Wray uses 'polyglot' "simply to avoid the
much more loaded term bilingual. Normally a polyglot speaks at least
three languages, but we shall take the epithet here to refer to anyone
who has knowledge of at least two" (note 10 p. 298). I think this
replacement is not necessary because these terms have been widely
employed in linguistics and any researcher could easily recognize and
define them.

Sixth, in page 97, the subject and the object of a sentence is the
same: "When the speaker wishes to manipulate the speaker..." The later
'the speaker' should be replaced with 'the hearer.'

Finally, Wray cites Nickels (1997) saying, "a single Turkish noun can
have more than four million different forms" (p. 268). I wonder how
Nickels came up with this huge number. As a native speaker and
researcher of Turkish, I would argue that this statement is an
exaggeration. While its true that Turkish is an agglutinative
language which has a productive morphologic system, four million
different forms for a single noun is inconceivable.

Despite these minor criticisms, the book covers a wide variety of
topics related to the nature and purpose of FS. It will, for sure,
inspire a great deal of future research. I would recommend the book
to ethnographers of communication, researchers of language
acquisition, language pathologists, cognitive scientists and anyone
who has a desire to understand how we use and comprehend formulaic
elements of the language.


Hurriyet Gokdayi teaches grammar, syntax, and linguistic courses at
the Department of Turkish Language and Literature, Mersin University.
His doctoral dissertation (University of Washington, Seattle) was
about the ways of speaking and ethnography of communication in Turkey.
He is currently doing research on formulaic language in Turkish. His
other interests are semantics, language and culture relationship, and
language policy.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue