·:'.

:·~~W':-~
.

'" \

THE CAMBRlDGE APPLlED L1NGUISTlCS SERIES

Series edirors: Michael H. Long and Jack C. Richards Genre Analysis

This new serics presents rhe findings of recent work in applied linguistics which are ';~
of direcr relevance to language reaching and learning and of particular inrerest to
applied linguistics, researchers, language teachers, and reacher rrainers, ':'~
. ;-~
English in academic and
1" tbis series: ::;'~
Interactive Approaches 10 Second Language Reading edited by Patricia L. Carrell,
;:i.~
research settings
[oanne Deuine, and Daoid E. Eskey
~~.
Second Language Classrooms - Research on teaching and learning
by Craig Chaudron

Language Leaming and Deafness edited by Michael Strong John M. Swales

The Learner-Centrcd Curriculum by David Nunan
The University of Michigan, Ann Arbor
The Second Language Curriculum edited by Robert Keith [ohnson
Language Transfer - Cross-linguistic influence in languagc learni~g il
by Terence Odlin \ ~
t~\
Linguistic Perspccrives on Second Language Acquisirion edited by'Susan M. Gass ;t~:.
.',t-."
~:;.
and jacq_uelyn Schachter'
':~
Learning Strategies in Second Language Acquisition byJ. Micbael O'Malley ] .. J'
and Anna Uhl Chamot
.:~
~'
The Development of Sccond Language Proficiency edited by Birgit Harley, ~
Patrick Allen, [im Cummtns and Merrill Swain ~
Second Language Writing - Research insights for rhe classroorn
edited by Barbara Kroll
'~i

Genre Analysis - English in academic and research sertings by john M. Swales

Evaluating Second Language Education edited by J. Cbaries Alderson and

:;i"
A Ia" Beretta

,
:'"
Perspectives on Pedagogical Grammar edited by Terence Odlin .;;~

Academic Listening edited by John F/owerdew

Power and Inequality in Language Education edited by james W Tollefson
~
Languagc Program Evaluation - Theory and practice by Brian K. Lynch

Sociolinguisrics and Language Teaching edited by Sandra Lee McKay and

Nancy H. Hornberger

Contrastive Rhetoric - Cross-cultural aspects of second language writing by

CAMBRIDGE
UNIVERSITY PRESS
Ulla Comlor

Teacher Cognirion in Language Teaching - Beliefs, decision-making, and

c1assroom practicc by Devon Woods

Second Language Vocabulary Acquisition - A rationale for pedagogy edited by

james Coady and Thomas Huckin

L..
 :'~~'~;
.:"(j
.
.,
.''.>c

PUBLlSHED BY THE PRESS SYNDICATE OF THE UNIVERSITY or CA.'vIBRIDGE

The Pirt Building, Trumpington Streer, Cambridge CB21RP, United Kingdom Contents
CAMBRIDGE UNIVERSITY PRESS
The Edinburgh Building, Cambridge CB2 2RU, United Kingdorn
40 West 20th Streer, New York, NY 10011-4211, USA
.~. .,;

10 Stamford Road, Oakleigh, Melbourne 3166, Australia .,

';1
© Cambridge Univcrsiry Press 1990 ,~l.
,~~
;;;:;
This book is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
,~ Series Editors' Preface VII
no reproduction of any par! may take place wirhout
rhe written permission of Cambridge University Press.
" Thanks IX

Acknowledgements XI

First published 1990

Fifth printing ] 998 ~O;),o:ooi,~ PARTI PRELIMINARIES

Printed in rhe United Kingdom

at the University Press, Cambridge
59i-l~ 1 Genre analysis - setting the scene
].1 Aims and purposes 1
1.2 Ourline of the book 8
Library of Congress cataloguing in publication data 1.3 Origins and influences 13

Swales, John.
Genre analysis.
PART 11 KEY CONCEPTS 2]

(Cambridge applied linguistics series). 2· The concept of discourse community 21

1. English language - Study and teaching (Higher) -
': ~ 2.1 A need for clarificarion 21
,t5:
Foreign speaker, 2. Interdisciplinary approach in ',':\ 2.2 Speech communities and discourse communities 23
:',
education. I. Title. n. Serics. 2.3 A conceptualization of discourse cornrnunity 24
·~":i
PEl128.A2S931990 428'.0071'] 90-2464 .'I":~ 2.4 An example of a discourse community 27
:"i
2.5 Remaining issues 29
British Library

Swalcs,John
cataloguing

M. (John Malcolm),
in publication

1938-
data

, .. ':;-
~----
3~fgenrV 33
3.1 Genre in ~e srudies 34
Genre analysis: English in academic and research settings.-
(The Cambridge applied linguistics series). 3.2 Genre in literary studies 36
1. English language. Analysis 3.3 Geme in linguistics 38
L Title 3.4 Geme in rhetoric 42
·~:.'I

428 3.5 A working definition of genre 45

./:1
3.6 Pre-genres 58
.::;,:
ISBN O 521328691 hardback 3.7 Differences among genres 61
ISBN O 52133813 ] paperback

~i 4 The conceptof task 68

I
4.1 Task and academic English programs 68
r-"-'- . '.-~-."...,...,_..._... _ 4.2 Towards a concept of task 73

' i
. Universidade (',~ i.";i'i:'a 43 A pedagogical illustration 77

1) "iU~~""'''-'i .~\~

Genres, schemata and acquisition

bg-12- .~
5 83
.--r:;_ .._.---_....
"'--'1.,
~ j EU mo> i
li'
~.1
M v
L~9.DE~7!1 7J! ~~'~j
I .~~

'~!:
..,-,
.#,1'
 ""
32 The concept af discourse community
where professor and srudent interacted together in a joint problem-
solving environment.
The next issue to be addressed in this section is whether certain
groupings, including academic classes, constitute discourse communities.
Given the six criteria, it would seem clear that shareholders of General
Motors, members of the Book of the Month Club, voters for a particular
political party, clienteles of restaurants and bars (except perhaps in
soap-operas), employees of a university, and inhabitants of an apartment
block ali fail to qualify. But what about academic classes? Except in
exceptional cases of well-knit groups of advanced students already
familiar with much of the material, an academic class is unlikely to be a
discourse community at the outset. However, the hoped-for outcome is
that it will form a discourse cornrnuniry (McKenna, 1987). Somewhere
down the line, broad agreement on goals will be established, a fulI range
of participatory mechanisms will be created, information exchange and
feedback will flourish by peer-review and instructor comrnentary, under-
standing the rationale of and facility with appropriate gemes will
develop, control of the technical vocabulary in both oral and written
/
contexrs will emerge, and a level of expertise that perrnits critical
thinking be made manifesto Thus it turns out rhat providing a relatively
constrained operational set of criteria for defining discourse communities
also provides a coign of vantage, if from the applied linguist's comer,
for assessing educational processes and for reviewing what needs to be
dane to assist non-native speakers and others to engage fully in them.
Finally, ir is necessary to concede that the account I have provided of
discourse community, for alI its attempts to offer a set of pragmatic and
operational criteria, remains in at least one sense somewhat removed
frorn reality. Ir is utopian and 'oddly free of many of the tensions,
disconrinuities and conflicts in the sorts of talk and writing that go on
everyday in the classrooms and departments of an actual university'
(Harris, 1989: 14). Bizzell (1987) toa has claimed that discourse cornrnu-
nities can be healthy and yet contain contradictions; and Herrington
(1989) continues to describe composition researchers as a 'community'
while unveiling the tensions and divisions within the group. The precise
status of conflictive discourse communities is doubtless a matter for
future study, but here it can at least be accepted that discourse cornmuni-
ties can, over a period of time, lose as we1l as gain consensus, and ar some
critical juncture, be so divided as to be on the point of splintering.
3 The concept o] genre
, i''Y
~~
yJ
"lI.
ô'J
Genre i, a term which, as Preston "Y', one approaches with ':
trepidation (Preston, 1986). The word is highly attractive - eve~t~C the
Parisian timbre of its normal pronunciation - but extremely ~.As
a first step in the arduous process of pinning it down, I sha1l di~nt ali
uses of the term to refer to non-verbal objects. These include the original
meaning of the term (in English) to refer to a type of smaIl picture
representing a scene from everyday domestic Iife and its growing employ-
ment as a fancy way of referring to classes of real world entities. The
latter is illustrated in Webster's Third New International Dictionary by
'large floppy rag doIls, a geme favored by rwo-year olds'.
The use of geme relevant to this study is glossed by Webster's Third as
'a distinctive type or category of literary composition'; however, the
dictionary's citation - from The New Yorker - usefully expands the
context of literary ta include 'such unpromising genres as lndian T reaties,
colonial promotional tracts and theological works'. ln~
l~~~~.!iP5~~cou!E-ofan:y-
tyP~1_spok~.9J2.uy.ntten,WI lull:-with2l!tlitrr~spirati~. 50 when we
i1ów hear or read o t e geme of the Presidential Press Conference', 'the
new geme of the music video' or 'the survival of game-show gemes', we
do so, I believe, without feeling that a term proper to rhetorical ar literary
studies has been maladroitly usurped.
Even so, geme remains a fuzzy concept, a somewhat loose term of art.
Worse, especially in the US, geme has in recent years become associated
with a QigeputablyJorm~laic wa of constructin -
stru~ion of) p~icular t~~!~ - akiºsL9bIT.itJn~QLSpeaking.ll}: numbe!s.
This assocíation ~racterizes ge9~~I as~echanism, and hence
is inimical to thé'éntíghitiihi and eií1lghtemng concept that language i~
ult.illJately a matter of choic~. The issue then \ i~~her-g~E~~
~g device for language ~..adllngjul~9.me~
unthinki~ati2!l...Qiformulas, or:Ylhether such.."ªRQ!ltcome is rather
~mplific~º-rought_ª-Qº.!!!J2y,_~§i@ik<ll. convenlence: An
initial way of tackling the issue is to examine what scholârsn'ã"veãdually
said about gemes in a number of fields. For this purpose, the following
four sections briefly consider uses of the term indolklore, JitêfaFJ
33
1n
..
34 The concept o] genre
studies, linguistics and rhetoric. (Another possible area would have been
film studies, e.g. Neale, 1980.)
3.1 Genre in folkJore studies
The concept of geme has maintained a central position in folklore studies
ever since the pioneering work in the early nineteenth century on
German myths, legends and folktaIes by the Brothers Grimm. And yet as
a major figure in folklore studies has rernarked, 'thus far in the illustrious
history of the discipline, not so much as one geme has been completely
defined' (Dundes, 1980:21). Ben-Amos (1976), whose valuable survey I
have relied on, comments that this failure is partly ascribable to high
standards of rigor and clarity expected in scientific definitions, and partly
to continuous changes in theoretical perspective. As he pertinendy
observes 'the adequacy of generic descriptions depends entirely on the
theoretical view they are designed to satisfy' (1976:xiii).
I; l}.en~
A~....gue~.Q.1l..!.<.?.C?~~!~.!.!l_t:l~JJlberof.
rhese .R~~~~~Q~
y ~~ followlog Llllnaeu~3La_ cla_ssl.fi2.toi cat<:gory; 19J:.-
b e.Kª.mpk~~my.th,..kg_e.~d o.!:!al~The value of
B' c1assification is seen to lie in its uSé1-sa research roolfor Ca.tegorizing and
ç /' filing individual texts, rhat is, as an effective sf8Ç~é andretrieval systern.
This, in rum, can lead, as mi-gh~~~pS':.çted,..to __ th~_deyjsin~e~
_mapsJhat.Rlace particular genres along various kinds of planes such as
tne···prosaiclpoetic and the secular/religious. However, it is apparently
i common in this classificatory work t~ emes as 'ideal es'
I r~~ge.Ld~~_~~_s.t\t~L:.~.~~~..:Actual texts will deviate from t ,.e ideal in
\ varicus.kinds.of wajs: . ~A.' v-
I --~no}he!m:ljor..gr.~!-p_2L~pr!"~?:~E~_.~~s_.g.eAr..e.L<! /ftm:!t-!, 9J1<}·est@.-
/ lish:eã-t!:ª-dition__t~ÍJlg..Jhe.s_~.Jorl!1Lilli_p.enna.Rem. Th us.•.Jeg.ep~cL
~~t~h~nge~_0_ejr-çhatact~r-o.\!.~u- ord:..~JlÍ.story~~hey
nave an mdependentllterary mtegnty, whlc~oclal vananons
and technological developments' (1976:xx). T~'l~
co&~i~~yedeeR.s~~~çt~Pl~.9.-hy_.th.c..!el<uio.ns-ªillgng.the·<li~<:-º:\!!sal
co-!Dponentsofthe.t(fxts.tb..emselves. What does change, of course, IS the
roleof-such texts in society: Vlêíous political satires become innocuous
nursery rhymes; incantations to prevent the soul from leaving rhe body
reduce to formulaic 'God B1ess You' responses to a sneeze; and proverbs
no longer play as central a role in popular education as they once used to.
A strong motive for the concept of an underlying permanent form
apparently derives frorn the long-standing interest among folklorisrs in
using the classic exemplars of rnyth and legend to trace beliefs back imo
pre-history, For that motive, the assumption of an enduring substrate is
Genre in [olklore studies 35
c1early useful, perhaps even necessary, but closely tied to a field-specific
research agenda.
The functionalists in folklore would rather stress socioculturaI value.
Em:..M!t~.,.1%.Q)Tlor_~_!!!.rle, folklore ~l!!§.g>..Dtrib~te to the
~-ª-!1f..e...a.l.MLsu.cri.)(.aLºL~sial_grouI2Th-:§l!:.l!.~J.he.y-.se.r'y~_s-ºsj..al..arrd
<sJ2i.ili.Y-ª-lne~d..s..
Perhaps inevitably, to assign cultural value also requires
me investigator to pay attention to how a community views and itself
c1assifies gemes. Thus, for many folklorists major narrative gemes such
.:1,
~
..
as myth, legend and tale are not so labeled according to the [orm of the
I:~
narra tive itself but according to how the narrative is received by the
community.
Do rhe people regard lhe narrative as sacred? If so, rhen it would
seem a myth. Do they enterrain the narrative as a potentiaIly
:·1 accurate recounting of anual events? Then it is a legend. Do they
regard the narrative as a total fiction with a requisite suspension of
}
belief? Then ir is one form of tale. The central poinr is that the
:}i
folklorist is primarily concerned wirh the folk narrative in some
larger context of belief and behavior. The folklorist recognizes that
)
folk narratives are the production of individuais, produced during
social inreractions and informed by surrounding cultural
traditions. The entire sense of folktale is not sandwiched in
between 'Once upon a time' and 'they lived happily ever after'.
., A tale is much larger than rhar, The folklorist must arternpr to
?I
undersrand why people reli stories in rhe first place, why listeners
appreciate thern, and why they favor some stories over others. The
problem is not only to undersrand how a text 'hangs together', bur
,,:'\
also to understand why a particular individual or group of people
would find such a text meaningful, worthy of artention, and
deserving of repetition.
~~!,
.:'~
(Oring, 1986: 134-5)
.':~
A final observation in this brief survey is thar not ali folklorists accept the
permanence af [orm concept. Some are more interesred in the euolution
of the genres themselves as a necessary response to a changing world.
.. This.is.parricularly.rrue. ai those wbo study relatively recent gemes in
ckY.elopedcountries, such as 'The Blues' in the USA, or have watched t~
ezolution ..(and.atm.pbr.)..ofJolklor.ist..genr.es-in.-u.adiÜõnaLcoiliI'nun:Jti€s---
a ffeúcd hy mndernizing influences./
~..k.~~I}!j ..Jm_m_th~lklorists_ for a genre-based approach to
academic English are, Ibelieve, severa!. First, the c1asslfyingorgenresTs
seen as having some Iimited use, but as an archival or typological
convenience rather than as a discovery procedure (a point we have
already seen Geertz make at the end of Part I). Second, a community,
whether social or discoursal, wi~1oiten view genres as ~ans to ends.
I.·g
 ~~,"oct,

36 The concept af genre Genre in literary studies 37

Third, a community's perceptions of how a rext is generically inrerpreted
is of considerable importance to the analyst,
3.2 Genre in literary studies
We have already seen thar folklorists may have special historicist reasons
for holding onto rhe permanence of formo In contrast, literary critics and
theorists may have special reasons for de-ernphasizing stabiliry, since
their scholarly activiry is typically designed to show how the chosen
author breaks the mould of convention and so establishes significance
and originality. Moreover, actual literary practice in rhis century would
seern, on the surface, to have so thrown away convention - in form, in
content and in authorial role - as to render obsolete the very term genre
itself. As Todorov remarks:
To persist in discussing
t' \y.ri;'u
genres roday mighr seem Jike an idltif ~or
These processes make it possible to c1aim that the whole issue of geme
conventions and their realignment is central to the evolution of rhe
creative ares - in film, in music, in art and in lirerature. On rhe last,
Hepburn (1983) has rhis to say:
How a competem reader approaches a work of lirerature, his
attitudes and expectations, depend irnportantly upon rhe genre he
sees it as exemplifying. A work thar rebels against gerire-
conventions equally relies on rhe reader's recognition of the
conventions being rejected. Aesthetically relevam features of a
work may srand out only if its reader has a ~!:Q!!nd awareness
~he hislorical development of rhe ~.ru:e...QLºf the s~::tIiãi.-:tlli"~o_"
w.prk is transforming il}..it.s
..distinl;Jive v.:.a..LiI!1..q
..puh(!Jl~..wi~~2~t
dj~~gJi3Y!!.lillnh~~,$xt

(Hepburn, 1983:496)
. ,c,v-I"v(
itst;!['fhe "Y.q~kmaydemandJ.çtpe
~~~ ~~r~fiTr_r~~~TÜ°~;Çtb;~~~~~:~~~~.fr,9}P.:rYb.iSh ...iLg~ms,.
,--"",-,---,~"."""_.,,,,
_...o.-.-..Y.........• __.... Y .. 0...0·-··
..0·..0.....· .... l'
",'

obviously anachronistic pasrime. Everybody knows thar rhey Th us afi I··

alm 15 a dv~~E...aPl?re~..!.~~
~À .d. , .. o f gen:~1.~~..neç.e~soª-rv_...
zenre i ií , J'
exisred in
tragedies,
the good old
and comedies
days of the classics - ballads, odes, sonnets,
- bur roday? J
. ~gglLconditiQnJ:QJ: Ieis. neCe!!gry /
a!!..~.pp.~.~.fi.ª-tlQlLQETit~I.iiJ.lJJe. l,
~~~~jLn-ºU>Eh:oprOyide~.ª.IJ. ..interpretarive.andevaluative ..frame.for.a Íí'

(Todorov,1976:159) work of a!:'U1llt,more to.the pointçthat frame.is.as much textual.ss.u.is .. ~

,
cUItuêãí,historical, socioeconorniç or politieal..
However, the above quotation comprises the opening sentences of Towfer"(19-S2Y, Ti1"tfiemost exhaustive contemporary study known to
~ Todorov's paper and our geme knowledge of such papers leads us to me of.fuerary genres, additionally stresses rhe value olgenre ~Q.!h.e.uLl:itgr:\p
< expect, in this case quite correctly, that the author is indeed about to
o

~." persisto He argues that the fact that works 'disobey' their genres does not Far from inhibiting lhe author, ~..a-positi.v.e..suppo.rt. They 11 J"
'\ mean that rhose genres necessarily disappear. For one thing, trans- ~1}1, O.!l~IJ.!ig!Hsay,f.Qcbim_loowite.in'-"""ª...habitarion.of.; I(

r~. gression, in order to exist, requires regulations to be transgressed. ~

o
another,---.!.~~.norm~----º!!!Y..i~tainvisibility and vitality by being trans-
.gr~~~~~Ibi~J~h~_Q!Qcesh.~.~~ºr_<hl!got-ºI-ºº-qrov,--ºLgwre genera~ionJ
m~djar~çlddi!}i~~~
i.ií.JI..!!ü'-h::~~:hich
Ins~ead of adau~void,
ª-Jlt.~raJ-Y__
a. PJ9RQ.!JjonaLlCloÇ.l!!r:::iÜRª.çe;
to .order.h iux:p~ti~ºç~o-ºº.P.ng.compQ~iti.o.n.::
they e~~~.If~y~!;~ocatively definit~
IQXJtatIon.The writer IS invired ro ~!!.?xpenence and form JO a
.. ..)
i ':.1.'

r
!I·
\1
I' I
:11
. 1

--:~:b_~~~~'oK~E.F~~~bY~sthe _~~an~formatiQ!1---ºLQ!J~_ or sJ~y.eralold genres.bz, specific yet undetermined way, Accepting rhe invitation does nor 'i
~-..inversi9-'1'_0ºvhsp-,-ªocel1}.m!,J1)~Sgmm.llª tion: (1976: 16,1). He then turns - solve his problems of expression ... Bur ir gives him access to
:,'
-l-'S tQ..theis~f wha~.1~~_~~-!:.d rej~~~~0':'-'i90ely~~~I~~iew;especially formal ideas as to how a variery of consrituents might suitably be ~I:
common.in.lirerary ci r:aes.:iliãi.genres ..are .classes of text~~Iders- combined. Genre also offers a chaJlenge by provoking a free spirit
to transcend rhe lirnitations of previous examples.
i'
i'
.instead.ro.argnei.. :Ii:fll
(Fowler, 1982:31)
In a society, lhe recurrence of cerram discursive properties is "I
\!i
institutionalized, and individual texrs are produced and perceived !!l
.i:
Although Fowler discusses geme classifieation wirh great erudition, he
I, in relation 10 the norm constitured by rhat codification. A genre,
concludes that alI such constructions have relatively lirtle value when li
lirerary or otherwise, is norhing but this codification of discursive
properries. seen against the inescapable evidence of continuous geme evolution. At illl:Ii
(Todorov,1976:162) the end ofthe day, geme analysis is valuable because it is clarificarory, not 'I,·
because it is classificatory. Ir provides 'a cornmunication sysrern, for the li:
t •
II
Further, since ideological changes affect what a society chooses to codify, use of writers in writing, and readers and crirics in reading and inter- 1

11
'I 50 change may come about from institutional sources as well as from preting' (1982:286). In taking this srance Fowler is able to lay at rest a "1
lil
!i
"

li
individual experimentation with discursive (or discoursal) properties. number of 'ancient misapprehensions': P
li
I
,.
't
(
,I,
 p

38 The concept o] genre Genre in linguistics 39

1. Geme theory is of little relevance because it corresponds ill with actual tber~.is.,Isl!.gg~~.t.~i!1g~.li!.t.!~.!lI!~<l!j.~f~~.?!Y ~~~.El!Y~i..~!?.??~.~_
.....
works of literature. ;~' l~.$par<\tmg.genr.e...and_*eech-eventanal;:ti.calfy. ~ .
2. Geme theory leads to heavy prescriprion and slavish imitation. t:rties o~(~umQIQ..\!S.-.eHecrTs--clear\t<-J:L~g..lls
3. It sets up highly conserva tive value hierarchies ('no great novels since -.·;j'1 delivering a sermon, if only because rhey have very different communica-
Joyce ar Lawrence'). .~ tive purposes. If, on the orher hand, some o~J2f.9-pert.ics-'are...
4. It is inevitably obsolescent in its arrernpts to characterize a present i~y~?.!~~~J~~.~~Ht;.g, .then.this...maY...(.o~~~u~tht;
period by then gone. rp~efkcliYeness_9J.J!!lg.thel ..genre-sU.Ç!Las a !2OJiti.calsp.e~.Çb. It is
J" still nor a sermon, however 'sermonizing' it may be. Whereas if ali the
This very brief excursio~ into literary views of gem.es has singled ~t a
characreristics are transposed, then we can recognize rhe fact that we are
few authors who have glven genre particular artennon. In consequence
lisrening to a sermon occurringin an arypicallocation. As Preston (1989)
they may be atypical, but in fact none represents a view as extreme as rhat
notes, it is not rhar speech euents and genres need to be kept apart, but
of Hawkes, who contends that 'a world without a theory of geme is
rather that situatians and genres need to be.
unthinkable, and untrue to experience' (1977:101). Ih~w authors
A position much doser to that adopred in rhis book is rhat of
~p~r t05~tinl~ations gE..~~~~ç,h~ª-n~e.as_iJêsult Õf
~1982). Like most other ethnographers, she takes~
, ~~QfQ!!RqJ.!.qJce, das~ilkiUOrysche~esar€:~
!"~~llu)LCQmmunicativ.e-ev:enLand_olf.ers th.~~iQllowing..as_
J '13S0ndary ~utcom~~f~~!y~is. As SC'fiãuber and Spolsky (19S.~) observe,
g.amRks: i~s, stories, lectures, gr.~ti!}&S-i!ng,Çnill'..e.rsatiuns. Like some
Ij . gemes form an open-ended set. N~.eLTodorov.nor Fowlera ...s pt.thaj;
of the folklorists, there is interest in discovering in a communiry which
~\( g.enres.are .....
simJJ!Y_~~m~l!~~.o .. f. m.~..~:::gr-Ies..s_.s.irlilar. t..e.xtual o~:&:'E.~_t, communications are generically ryped and what labels are used, as these
, "<o i.Dsteadf are. CQ.ckd.~~cr key~cr evef!ts_ ser within social cQmmuniçari.v~_~
will reveal elements of verbal behavior which the community considers
.~~ pJ:Q.CeSS.CS. Recogniziog:!I:i§'~:ewdés and keyscan be a powerful fa.~.i.~~at.or
sociolinguistically saliento In addition, the ethnographers give consider ••..·
-
of..-b.oth_c.omprehensionan<i..<:omp_osition.
............•. __ ...----... ... r:;
,
able arrention to how best to interpret and utilize the elicited meta-
language. Saville-Troike is quite dear on the matter:

3.3 Geme in linguistics Since we cannot expect any language to have a perfect

1
metalanguage, the elicitation of labels for categories of talk 15
c1early not adequare to assure a full inventory and must be
Linguists as a group have been more partial in the attention they have supplemented by other discovery procedures, but it is basic to
given to the term genre. This may be partly due to traditional tendencies ethnography thar the units used for segmenting, ordering and

l
, to de ai with aspects of language below the levei of texts and partly due to describing data should be rhose of the group, and not a priori
a reluctance to employ a 'terrn of art' (Levinson, 1979) so dosely caregories of the investigaror.
associated with Iiterary studies. In any event, rhe term is only found with (Saville-Troike, 1982:34)
any frequency among linguists of eirher ethnographic ar systemic per-
suasions. Ir is not, of course, difficult to recognize the danger of basing units on the
For the ethnographer Hymes: 'a priori categories of the investigator', and indeed text-linguistics and
Genres often coincide with speech evenrs, but must be treated as certain rext-rypologies are somewhat prone to this very danger. ln rhat
analytically independent of thern. They may occur in (or as) respect, the ethnographic position as represented by Saville-Troike is
different events. The sermon as a genre is typically identified with a both salutary and admirable. However, what we might call 'folk'
certain place in a church service, but irs properties may be invoked, categorization and the investigaror's a prioricategorization are not
for serious ar burnorous effect, in orher situations. necessarily in exclusive opposition. lndeed, ir can be argued that the
(Hymes, 1974:61) investigator's role in geme analysis is neither to follow slavishly the
nomenclatures of groups, nor is it to provi de his ar her own deductive
As for speech euent itself, it 'will be restricred to activities, or aspects of and introspective categorial system. Rather, the procedure should be to
activities, that are directly governed by rules ar norms for the use of develop sets of a posteriori categories, ones based on empirical investi-
I speech' (1974:52). Leaving aside the restriction to only rhe oral mode, gation and observation, within which e1iciting the community's category-
\ ,.- (
I,; p","""_fl,· •.·,·,,·r_(

J~L-,~'-r
.
I::
/'
I, . ,;~~;
I '.' I
"'--
"
 '"
40 The concept af geme
labels plays a central role. Indeed, this seems to be what Saville-Troike is
getting ar when she observes that languages do not have 'perfeet'
metalanguages and 50 need supplementation and refinement.
~oncepsgf..g.:n,:J:!,!s_aJs() in rec.ent years b«.t:I1diSClI~~dJ).j1-rhe
s~~ ~:_ 'H~~_~i~_~~'linguistsi:f. Hallida~, 1978). However, ._the
rcla.b.onslÜp_1K~eeDgenre ãn(h~1Qngeustabl~s.hed concep!of.regl~er..
is not always very cleãr-=see-Ventola (l984) for ã-dtscússion of rhis
unêert;iínry:-~;·-or::.functiQn._illª-,]gu_ª~iatiºD, is~acon@uaJ
~':
c~_gory correlating groupings of linguistic_J(!.'!WJeLwitb. __ J'.!~ÇUn:ent
r\: " si!uatiõiwrteãt~Gregory àrl"aCãITolT,-f978:4). This category has
typicãlly-been-amrtyzed in terms of three variables labeled field, tenor and
0,} '; mode. Field indicares rhe type of activiry in which the discourse operates,
\: \) i its contem, ideas and 'institutional focus' (Benson and Greaves, 1981).
;l~J
I
'--c.:,," ..\::
Tenor handles the status and role relationships of the participants, while
mode is concerned with the channel of communication (prototypically
·~~~I speech or writing). "The field, tenor and mo de act collectively as
determinants of the text through their specifieation of the register; at the
same time they are systematically associated with rhe linguistic system
through the functional cornponents of rhe semantics' (Halliday, 1978:
122). Thus, field is associated with the management of the ideas, tenor
with the management of personal relations, and mode with the manage-
ment of discourse itself. The categories provi de a conceptual framework
for analysis; they are not themselves kinds of language use.
Ir is only cornparatively recently in the systemic school that geme has
become disentangled frorn regíster: Frow (1980:78), for instance, refers
to 'discourse genre, or registcr'. On the other hand, ~~
Ihticllowin~êvfutill(;tiQIl;..genr.e.s.are.J.'.~ªlized..thr.o_ugh.r~isters,
and registr:~~..i!l-.turn .. ~reureaIj~~dhthrough Ianguage. As fOLgrnre...s
~ems.d'{s;s.:- - Since the rwo sides of the scale are independent, a writer could
\ Gemes are how things get done, when language is used to
! accomplish rhern. They range from literary to far fram literary
!) forms: poems, narrarives, exposirions, lectures, seminars, recipes,
manuais, appointrnenr making, service cncounters, news
I broadcasts and so on. The term geme is used here to embrace each
of rhe linguisticallyrealized activity types which comprise 50 much
of our culrure.
(
(Martin, 1985 :250)
-,
Martin gives rwo kinds of reasons for establishing geme as a system
l' underlying register. One revolves around rhe fact rhat gemes constrain
the ways in which regisrer variables of field, tenor and mo de can be
eombined in a particular society. Some topics will be more or less suitable
for lectures than others; orhers will be more or less suitable for informal
conversation berween unequals. Recognizing the gaps is not only valu-
..•.l!
Genre in linguistics 41
able in itself, but can have important consequences for cross-cultural
awareness and training.
The second reason for recognizing rhat genres comprise a systern for
accomplishing social purposes by verbal means is that this recognition
leads to an analysis of discourse structure. Genres have beginnings,
middles and ends of various kinds. Verbal strategies 'can be thought of in
terms of stares through which one moves in order to realize a geme'
(Martin, 1985 :251). Genre 'refers to the staged purposeful social pro-
cesses through which a culture is realized in a language' (Martin and
Rothery, 1986:243).
Couture (1986) provides unusual clarificarion of the use of register and
genTe within systemic linguistics. Registers impose constraints at the
linguistic leve\s of vocabulary and syntax, whereas geme constraints
operate at the levei of discourse structure. Further, 'Unlike register, genre
can only be realized in completed texts or texrs that can be projected as
complete, for a genre does more than specify kinds of codes extant in a
group of re\ated texts; it specifies conditions for beginning, continuing
and ending a texr' (1986:82). For Couture then the two concepts need to
f·
be kept apart: genres (research report, explanation, business report) are
completable structured texts, while registers (language of scientific
reporting, language of newspaper reporting, bureaucratic language)
represent more generalizable stylistic choices. Genres have 'complement-
ary' regisrers, and communicative success with texts may require 'an
appropriate re\ationship to systems of genre and register' (1986:86).
In a detailed application of how genres and registers could relate
differentially to a scale which runs from the highly explicit to the highly
elliptical, Couture gives the following illustration:
select a genre that implies a high levelof explicitness (like a
business reporr) and at the same time select a regisrer rhat demands
less explicitness (such as bureaucratic language). ln doing 50, the
writer must decide which critera for explicimess he or she wishes
to dictate linguistic choice (clear hierarchical development of
rnessageand support demanded by the report genre or implicit
expression of the cultural values of irnpartialiry, power and
prestige associated with bureaucratic sryle).
(Courure, 1986:87)
Aside from scholars such as Martin, Rothery and Couture, linguistics as a
whole has tended to find geme indigestible. The difficulty seems to derive
from the fact that register is a well-established and central concept in
linguistics, while gente is a recent appendage found to be necessary as a
result of important studies of text structure. Although genre is now seen
as valuably fundamental to the realizarion of goals, and rhus acts as a
\,
 ,?
42 The concept af genre
deterrninant of linguistic choices, there has been an understandable
~to demo te register to a secondary P?sition, anqIDWi1tíllip)
ness ~n the one hand, by large-scale mvestrnent 10 analysis
of language varieties (for lexicographic among other purposes) and
underpinned, on the other, by relatively lirtle interest in seeing how rexts
are perceived, categorized and used by members of a community.
Despite these equivocations, linguistic contributions to the evolving
study of geme lie in the emphasis given to: (a) gemes as types of
goal-directed communicative events; (b) gemes as having schematic
structures; and most srrikingly (c) gemes as disassociated fro~~g!gtls
.~
3.4 Genre in rhetoric
I
Ever since Aristotle, rhetorical inquiry and criticism has been interested
in classifying discourse. One common approach has been 'to proceed
deductively, in a top-down manner, and construct a closed systern of
categories, A prominent modern example - and one of many - is
Kinneavy's A Theory af Discourse: The Aims af Discourse (1971).
Kinneavy c1assifies discourse into four main types: expressive, per-
suasive, literary and referential. ~disçj}.llrse-w.ill-·be-cl~i6ed jnto a
!?J!l~~~ne.ntin_.the.-communicatimL.,_
.llli>..Ç.e.Ss-~tb~l]).aLJl-focus ..lf th.e fO.Ç!!s..Qr.aimis
on.rhe.sender ,..the-.
di~<:Q!ill~lLb~1í,pressUre.; if on the receiver, persuasive; if on the
linguistic form or code, it will be lirerary; and if the aim is to represent the
realities of rhe world, ir will be referenrial, Although s~h classifications
have irnpressive intellectual credentials and considerãble organizing
power, the propensity for early categorization can lead to a failure to
undersrarid particular discourses in their own terrns. For example, the
scientific paper appears, in Kinneavy's system, to be a classic instance of
referential discourse bur, as we shall see in Part IlI, rhere may be very
good reasons for not coming to quick conclusions about its predornin-
antly referential nature.
In contrast, rhetorical scholars who have taken a more inductive
approach have tended to take context more into account anel to give
geme a more central place. This is perhaps particularly so among those
who study the historical development of discourses in recurrent settings,
as has been done by Jamieson (1975). She outlines her position as
follows:
Three bodies of discourse may serve as evidence for the thesis rhat
ir is sometirnes rherorical genres and nor rhetorical situations that
are decisivelyformative. These bodies of discourse are the papal
encyclical,the early state of the union addresses, and their
L
Genre in rbetoric 43
congressional replies. I will argue rhar rhese discourses bear rhe
chromosomal imprint of ancestral genres. Specifically,I propose to
track essential e1ementsof the contemporary papal encyclical to
Roman imperial documents and the apostolic epistles, essential
elernents of the early stare of lhe union addresses to the 'King's
Speech' from rhe throne, and essential elemems of the early
congressional replies to rhe parliamentary replies to rhe king.
Uamieson,1975:406)
amieson is able to show, in these cases anyway, haw anrecedcnt genres
o~ul constraining models. As she observes, without such a
concept, ir would be difficult to reconcile the fact, on the one hand, that
rhe firsr leaders of the United States incorporated monarchical forms into
key early public starcments and rhe fact, on the other, that one of their
prime purposes was to reject the tyranny anel trappings of a monarchical
system. p,
Jamieson is careful not to assert that established rhetoric will necessa-
rily be a prevailing influence on a particular rhetorical response. Whether
,,1
,J ir is siruation, audience expectarions or genre itself is, she advocates, a 4"
~L
)
~ rnatter of inquiry. Even 50, it will come as litrle surprise to find that many )
·;K~ '" rhetorical scholars with an inductive and/or hisrorical orientation stress {J
':
~ the recurrence of similar forms in geme creation: ~.
A genre is a group of acts unified by a constellation of forms that,~ /
~. recurs in each of its memb.ers.These.forms, in isolation, appear in ''X\
-'2_ other discourses. What is distinctive about the acts in a genre is a \)
~ recurrence of the forms together in consrellation.
. (Carnpbell and jamieson, 1978:20)
This kind of generic analysis, as in most others we have seen, aims to
illuminate rather than classify. It offers, amongst other things, a way of
studying discoursal development over time that is detachable from an
analysis of an individual event or an individual author; it also suggesrs,
by way of comparing rhetorical similarities and differences, a porential
method of establishing rhe genre-membership or otherwise of a particular
texto
Miller (1984), in a seminal paper, shares Campbell and Jamieson's
view that analysis of actual gemes can c1arify certain social and historical
aspecrs of rhetoric that might orherwise be missed. She is also like (hem
an anri-taxonornist, because gemes are unstable entities: 'rhe number of
gemes in any sociery is indeterminate and depends upon the complexity
anel diversiry of society' (1984:163).
However, Miller also advances the discussion in a number of im-
portant ways. First, she has principled reasons for extending the scope of
geme analysis to types of discourse usually disregarded by rhetorical
scholars:
 44 The concept af genre
To consider as potential genres such homely discourse as lhe letter
of recommendation, the user manual, lhe progress report, the
ransom note, the lecture, and lhe white paper, as well as lhe
eulogy, the apologia, the inaugural, rhe public proceeding, and lhe
sermon, is not to trivialize the study of gemes: it is to take
seriously the rhetoric in which we are immersed and lhe situations
in which we find ourselves.
(Miller, 1984: 155)
Secondly, she argues that 'a rhetorically sound delinition of geme must be
eentered not on the substance or form of diseourse but on the aetion it is
used to aecomplish' (1984:151).
fL/'~gives serious attention to how gemes fit inro the wider
ij scale of human affairs. She suggests that:
\ What we learn when we learn a geme is not just a partem of forms
\\,or even a merhod of achieving our own ends. We learn, more
imporrantly, whar ends we may have ., .
(Miller, 1984:165)
As students and strugglin&, scholars, we may learn that we may create a
research space for ourselves, we may promote the interests of our
discourse eommunity, we may fight either for or against its expansion, we
may uneouple the ehronological order of researeh aetion from the spatial
order of its description and justification, we may approaeh unexpeeted
sources for funding, or we may negotiate aeademic or editorial decisions.
~enre analysts a,mong the rhetoricians thus ~aJs.e a subst~ntial eontri-
bution to an evolving concept of genre~ fô'f the applied purposes
of this study. They provide a valuable historieal context for the study of
genre movements and they finally destroy the myth - ar so I hope - rhar
genre analysis necessarily has sornething to do with constructing a
)
classification of genres. ~QnaLwork-reinJQills...th~ÇDllCep.L,
of geme as a means of social acnon, one situated in a wider sociorhetori-
cal context and operating not only as a mechanism for reaehing commu-
nicative goals but also of clarifying what rhose goals might be.
Overview
The foregoing brief survey of how gemes are perceived in four different
disciplines indicates something of a common stance. Its eomponents can
be summarized as follows:
, (Levinson, 1979:368)
1. a distrust of classification and of faeile or premature prescriptivism;
\ 2. a sense that genres are important for integrating past and present;
~ 3. a recognition rhat genres are situated within discourse eommuni-
A ioorking detinition af genre 45
ties, wherein the beliefs and naming practices of members have
relevance;
4. an empbasis on communicative purpose and social action;
5. an interest in generic structure (and its rationale);
6. an understanding of the double generative capacity of genres - to
establish rhetorical goals and to further their accomplishment.
This stance suggests that it is indeed possible to use genres for teaching
purposes without reducing courses to narrow prescriptivism or formal-
ism and without denying students opportunities for reflecting upon
rhetorical or linguistic choices.
3.5 A working definition of geme
This section offers a characterization of genres that I believe to be
appropriate for the applied purposes that I have in mind, although
detailed consideration of links to language-teachi!1g.. activity and
l~ theory wilI be heW over to Chapters 4 and 5
respectively. I shaIl proceed by making a series of short criterial observa-
tions, which will be followed in each case by commentary. Sometimes the
commentaries are short and direcdy to the point; at other times they are
more extensive as they explore wider discoursal or procedural issues. I
hope in this way - as the section title indicates - to create a sufficiently
adequate characrerizarion for others to be able to use, madify or reject as
they think ~c--d.~d;c.Q
r
1. A genre is a cfass of çommunicative events.
./" I will assume that a comrnunicative event is one in which language
(andlor paralanguage) plays both a signilicant and an indispensable role.
J .
Of course, there are a number af situations where it may be difficult to
say whether verbal communication is an integral part of the activity or
noto Levinson neatly illustrates the possibilities for speech contexts:
,,/ ,
On the one hand we have activities consrirured entirely by talk (a
telephone conversation, a lecture for example), on the other '
l-
activities where talk. is non-oc:uring or if it does occur is inci.d.ental
j/ (a game of football for instance). Sol'newhere in berween we have
,-.l~s, or a Bingo session, or a visit to rhe grocer's.
----- And rhere are sornerimes rather special relation~
s.tli<iancLwhatis...done,-as-irr a sports commentary, a slide show, a
cookery'demonstrarion, a conjurer's show, and the like.
Activities in whichjalk is incidental e- as~~ng in physicalexercise,
~hold cli:~driving, will nQt.liê:-considered' as
----------------------,-,_._._-,,---

46 The concept af genre
i:~,Y1!ica.!i.v.~.~yents.;!.tQr will activities that involve the eyes and ears in
I non-verbal ways such as looking at pictures ar listening to music.
I Secondly, communicative events of a particular class wi\l vary in their
occurrence from the exrrernely common (service encounters, news items
in newspapers) to the relarively rare (Papal Encyclicals, Presidential Press
Conferences). B.J' and large, classes.. wi.th few insJançe§. need to Qave
~ce._:yi.t~ln-the"ie1~v;iit~I!ure. tQ.exisU_La.ge-;;-r~s:· 11' a
coriimunicative evenfõf a particular kind only oecurs once a year it needs
to be ~or
n~Finally,
class formation: a TV advert using a talking dog wiU
and tO repeat an earlier claim, a communicative event is
ere conceived of as comprising not only the discourse itself and its
participants, but also the role of that discourse and the environment of its
production and reception, including its historical and cultural assoei-
ations.
2. The principal criterial [eature that turns a collection af communica-
tive events into a genre is some shared set of communicative purposes.
Placing the primary determinant of genre-membership on shared purpose
rarher than on sirnilãrities of form or some other eriterion is to take a
i; position that accords with that of Miller (1984) or Martin (1985). lhe
decision is based on the assumption that, except for a'few interesting and
exce~~~,()nalca~es,B~E.~~~5Qmmunicatiy.e-vehiçl§.fo.r:Jb~.a.fh~~v~.m~ot
I of goa1s. At this juncture, it may be objected that purpose IS a somewhat
l~ss 'o~ht and demonstrable feature than, say, form and therefore serves
li: less well as a primary criterion. However, the faet that purposes of some
genres may be hard to get at is itself of considerable heuristie value.
i;: Stressing the primacy of purpose may require the analyst to undertake a
fair amount of independent and open-minded investigation, thus offering
protection against a ~on--l:>as.cl.Sill stylistiç features and
i~~~e~~.~~I:~_~E.~fyi!1K.!~~.:.~~_.~!.ricle~.~~_s9--;rej)Ol1.S
expenrnents.
"Qt
In some cases, of course, identifying purpose!~?nJle relatively easy.
Recipes, for example, would appear to be~gfíi'.@'~fti..!ll~tmctiQ.llal
texts designed to ensure that if a series of activiriesris earried out
aecording to the prescriptions offered, a successful gastronomic outcome
will be achieved. In others it may not be so easy. For instance, we might
suppose that lhe examination and cross-examination of witnesses and
parties carried out by lawyers under an adversarial sysrem of justice are
designed and struetured to elicit 'rhe facrs of the case'. However,
i;'
independent investigation shows this not to be so (Atkinson and Drew,
1979; Danet et al., 1980). The elaborate sequences of closed 'yes-no'
questions are designed to control how much the hostile or friendly
witnesses will be allowed to reveal of what, in fact, rhey do know.
Or, to take another example, we might suppose that the purposes of
11
.,
~ A working definitian of genre 47
parry political speeches are to present party policies in as convincing a
way as possible, to ridicule the policies and personalities of opposition
parties, and to rally the faithful. However, especially in these days of
massive television covefage, party political speeches may now be being
written, structured and delivered in order to generare the maximum
amount of applause (Atkinson, 1984). And certainly there are signs in
Britain that the 'applause factor' is becoming raised in consciousness, as
it were, not only as a result of the interest in Atkinson's work, but also
because of the recently established journalistic practice of measuring the
length of ovations following major speeches at conventions..
The immedi~e.~g..example..suggestwhm..iLi~J...uncQInmon.
1-
~-nd genr~ tb.aLhaye sets of cornmunic-ati-ve..p.w:poses. While news I
broadcasts are doubtless designed to keep rheir audiences up to date with
events in rhe world (including verbal events), they may also have
purposes of moulding public opinion, organizing public behavior (as in
an emergency), or presenting the controllers and payrnastérs of the
broadcasting organization in a favorable light. When purposive elernents
come into conflict with eaeh other, as in the early Environmental Impact
} Staternents srudied by Miller (1984), the ,cillecttvefleS5-_0Lthe_gefH:e...a~
v
sociorherorical a.J:.tiQ.!!.l>~çome~questi.Q!lable. In th~-ª,Ç.ade1llÍc...context,..a_
'., g~J?0temi.<!LÍQLçonflLc:_t~l}g_p~.:.p~~jvhatof the student- .
, ., ...wntt€fI..examm,atlon (Searle, 1969; Horowirz, 1986a). ,b"f/_''''''.r ")
There remain, of course, some genres1o!..~h.12.lJJp-bSe iUlilsuited.as ..
a primary criterioE,' ~genres are ao obvious e~e. Although
there may be overt political, religious or patriotic tracts put out in the
form of verse, the poetry that is taught, remembered, known and loved is
rarely of that kind and inevitably makes an appeal to the reader or
Iistener so complex as to a\low no easy or useful categorization of
purpose. P"OeJIlb-aIlclQ.l.b.cr.g.gnFes....w~p-p-~ay·-lie--in-the-veWê.
r pkas.w:eJ@_~~_~i1E..-.jW~,;PJ.~ately
v" they de ~fcó-mtnumcanve
char.i!f.t.e.Jizedby the fact that
purpose.
j e nced, in ali but exceptional cases, to ascribe privileged status to
~ purpose derives not only fram a general recognition of the power it has to
:! shape our affairs, but aiso because it provides a way of .separating 'the-.,
'" ~. The Oxford Dictionary defines parody as 'A
composition in which rhe characteristic turns of thought and phrase of an
author are mimicked and made to appear ridiculous, especial\y by
applying thern to ludicrously inappropriate subjects'. However, Mac-
Donald (1960:557) is surely right when he complains that the final dause
does not sufficiently distinguish parody from its poor relations, trauesty
and burlesque. GJ20dparody is often appliW-t()..Subject.matJ..eJ..1b.a.t.is..o.nlY--
s1i~htly or subtly in<lpp~ A.s-.a..result,content.andJQ1'm.I!!ªy-.QQ.L
rmaLthe fact tbat pat~~~_~t':!ng ~~~_I!lP!e4,~.~,J~_ÇYl'ilConnolly's
p_~.ody.,oí-Aldous-Huxlqin~Tol~,i!1q,!!~'_.C>~ .Henry Reed's ce\ebrated
 48 The concept of genre
"
A working definition of genre 49

iJ Eliot-esque 'Chard Whitlow'. Consider, for instance, the opening rwo find contextual e1ues that help us to separare the..-sP-urious from the
paragraphs from a paper by Michael Swan and Catherine Walter ~, we need to rely on the privileged property of identifiable
published in the English Language Teaching Journal entitled 'The use of comn1unicative purpose to disentangle rhe e1ever parody from 'the real
li!

t t
I sensory deprivation in foreign language teaching':
The terrn 'sensory deprivation' is probably familiar to rnosr of us
rhing'. t

..
" from recent reports of interrogation procedures, but ir may seem 3. Exemplars or instances of genTes vary in their prototypicality .
strange to find the expression used in a discussion of language 50 Iar I~..!b.~~L~.JIKmbershilüs...b~~mmUIJ..is:-ªtiY.~

! teaching, especially since, ar first sight, ir is difficult to imagine how

deprivation of sensory input could conrribute to learning.
However, Tecem experiments iri this field (carried out principally
by the Chilron Research Association at Didcot, near Oxford) have
suggested that sensory deprivation (SD) could well become a
";"i;'>'
.... 10.,
'•.. ,?
.i.

/..
~.

. '.
What e~se is it based on? What additional

_1.. and •he c.: '11.. __ .J~e-'-'~.L,=,--

' appooa~~-1~m.wmu:e h
features will be
reqUlre<i to establish such rnernbership? There are, as far as I can see, two
possible ways of trying to find answers to such questions: the de{initional
. .ap.j.UmJ..Q::.
powerful pedagogic tool in the not toa distam future. The purpose r:;;E:. . The .definitional .view is ~ch the. b~tter .establishe~ and, inde~d,
of this article is simply to provi de a résumé of current research in f \:f">. '\!!!deq:!IDS the o:e.a.t!..0n and w:ill:'f1i.of dictionaries, glossanes and specia-
SD; readers who would like more complete inforrnation are .1·> .. ' lized technologies. It asserts that, in theory at least, ir is possible to
referred to the very detailed accounr by Groboshenko and produce a small set of simple properties that are individually necessary
Rubashov (1980).
and cumulatively sufficient to identify ali the members and only the
Interest in the use of SD in language teaching arose initially as a members of a particular category from everything else in rhe world. Thus,
natural extension of the work of such researchers as Gattegno, a bird can be defined in terms of being an animal, having wings and
Rand Morton, Lozanov and Watanabe. Gattegno's refusal (in rhe feathers, and laying eggs, or some such Iist of~tIe§)As long as rhe
'Silent Way') to allow learners more than minimum access to rhe
object has the stipulated features, it is a mernber of the category; it
second lariguage (L2) model; Rand Morton's insistence on
elimioating meaning entirely from rhe early 'phonetic matters not whether rhe bird is a 'normal' one like a sparrow or a
programming' stages of language learning, Lozanov's ccncern to 'far-out' member of the category such as an ostrich or a penguin - they
i purge rhe student of his former idenriry and to build a new, are ali equally birds. The definitional view has had some success in certain
autonornous L2 personaliry through 'Suggesropaedia', and finally areas. Kinship terms have been extensively analyzed in this way; a
Watanabe's controversial but impressive use of 'hostile bachelor is 'an adult unrnarried male' (Katz and Fodor, 1963); and other
environment' as a conditioning factor - all these elernenrs are areas where ir seems to work with relatively litrle problem are numbers
c1early recognizable in current SD pracrice, But SD goes a great
deal furrher. (ordinal, cardinal, real, rational etc.) and physical and chemical e1ements.
However, in practice, great difficulty has been experienced in drawing up
(Swan and Walter, 1982:183) lists or defining characteristics of such everyday categories as [ruit,
vegetables, [urniture and vehicles. ,~hat. is so, thm..iliere would ~
Most of the regular readers of ELTJ with whom I have discussed this ~p~dittl.U1O-p~ol.idenüly-ing....the._ªll:9!::.!1on~ningkatw:.es-.cl:V
paper stated that they read it with increasing incredulity. However, they
~s, _staL"J.!.eti'2gs, r~e!!!ch PEP(!rs, iº~(}s or _c.?.'!.~l!lt!!!~ons.A)..!;
also admitted rhat they were by no means sure ir was a 'spcof' until they
funher difficulry is created by the easily-attested phenomenon that we
reached the end and saw the words 'Received 1 ApriI1982'. After ali, the
still recognize category membership even when many of the suggested
content is conceivable (just), and certainly not 'ludicrously inappro-
defining characteristics are missing; the roast chicken emerging frorn rhe
priate'. Further, the Swan and Walter paper is of an appropriate length,
cooker is still identified as a bird. As Armstrong, Gleitman and Gleitrnan
I. uses standard style, has the expected information-structure and is appro-
observe, 'It's not ar ali hard to convince the man in the street that there
L! priately referenced, some of the references being genuine. Although the
are rhree-legged, ta me, toothless albino tigers, that are tigers ali the same'
publication of rhis fake paper may have been an exceptional event in the
(1983:296).
world of language teaching publications, other academic groups, par-
It might therefore be the case rhat what holds shared membership
ticularly scientists, have an established tradition of parodying both their
together is not a shared list of defining fearures, but inter-relationships of
research methods and their publication forrnars. For instance, there exist
a somewhat looser kind. This indeed would seem to be the view taken by
'specialized' periodicals like the [ournal of lrreproducible Results and the
Wittgêhstein in a justly famous passage in the Philosopbical Investi-
[ournal af lnsignificant Research (see Gilbert and Mulkay, 1984,
gations rhat is worth quoring in full:
Chapter 8 for an excellent discussion). In the end, although we may welI

,~!'.
r! '"
,
50 The concept af genre
66. Consider for example the proceedings we call 'games'.
I mean board games, card games, ball games, Olympic games, and
so on. What is common to them ali? - Don't say: 'There must be
something common, or they would not be called "garnes" - but
/ook and see whether there is anything common to alI. - For if you
look at thern you will not see something that is common to aI!, but
sirnilariries, relationships, and a whole series of them at that. To
repeat: don't think, but look! - Look for example at board games,
with their multifarious relationships. Now pass to card games;
here you find many correspondences wirh the first group, but many
common features drop out, and orhers appear. When we pass next
to ball games, much that is common is retained, but much is lost. _
Are they ali 'amusing'? Compare chess with noughts and crosses.
Or is rhere always winning and losing, or competition between
players? Think of patience. In ball games there is winning and
losing, bur when a child throws his ball ar the wall and catches it
again, this feature has disappeared, Look at the parts played by
skill and luck; and at rhe :!Hference between skill in chess and skill
in tennis. Think now of pmes like ring-a·ring-a·roses; here is the
elernent of arnusementç.but how many other characteristic features
have disappeared! And' we can go through many, many orher
groups of games in the sarne way; can see how sirnilariries crop up
and disappear.
And the result of this examination is; we see a complicated
networkof sirnilarities overlapping and criss-crossing. sometimes
overall sirnilarities, sometimes similarities of detail.
67. I can think of no better expression to characterize these
sirnilarities than 'family resemblances'; for rhe various
resemblances between members of a family: build, features, colour
of eyes, gait, temperament, ete., etc, overlap and criss-cross in the
sarne way. - And I shall say: 'garnes' form a family.
(Wittgenstein, 1958:31-2)
Thus, we could perhaps argue thar in, say, the case of lectures 'we see a
complicated network of similarities overlapping and criss-crossing:
sometimes overall similarities, somtimes similarities of detail'. Thus some
lectures may be like others in terms of some of the following: the
arrangement of speaking roles, seating arrangernents, the levei of forma-
lity in language, the use of visual support, the number and positioning of
examples, the ernployrnent of metadiscoursal features of recapitulation
and advanced signaling, and so on. Others will be like others in
equivalent bur different ways.
However, 'farnily resemblance' has not been without its critics. To
start with, we would do well to bear in mind Ladge's observation that 'no
choice of a text for illustrative purposes is innocent' (Lodge, 1981:23).
Just as kinship-terms suit Katz and Fodor, so games may particularly
favor Wittgenstein's observations. In fact, we can note that nearly ali
A working definition o] genre 51
"g:Jmes affer a contest or ~ challenge. The Oxford Dictionary proposes
,;ihisdefinltIon of a game: A diversion of the nature of a contest, played
.'àccording to rules, and decided by superior strength, skill or good
.fortune' - and, of course, this conrest can be against the game itself as in
patience, solitaire, or in a jig-saw puzzle. Admittedly, we are left with an
uriaccounted-for residue as represented by such children's games as
'ring-a-ring-a-roses'. -:
Rather more seriously, it can be objected that a family resemblance
theory can make anything resemble anything. Consider, for instance, a
set-up like thar shown below.
A'· -B· "C. -D
same mouth same eyes same nose
Thus whilst B and C share a common feature, A and O have nothing in
common themselves except that they share a different feature with B and
C. 50 a knife is like a spoon because they are borh eating instruments, and
a spoon is like a teapot because rhey are both used to contain Iiquids, and
a teapot is like a suitcase because they both have handles, 50 a knife is like
a suitcase. Indeed it was precisely this kind of undisciplined chaining that
Vygotsky (1962) characterized as being typical of the young child, to be
replaced in maturity by a more orderly systern of categorization.
However, as Bloor (1983) has argued, we need to remember thar
Wittgensrein Wa5 concerned with fam..ibLJ:esemhlances-ª!l~
cohere º-y..r.eason_Q(.Q.theF,Jhing'5J.le"side--pnysi€al,charac.tt;;risIiS:
such
2 as
~~.=Ües-and-shared'experien{:es; therefore it would be inappropriateto'
leave the domain of eating activities (knives and spoons) for that of
traveling activities (suitcases).
Wiugen~n's discussion of faIlJilyJe.s~~ and su~~~quem
comment have given rͧe.to.-a-'proJOtype'or c!uster tneofl_<!eSigned-w
áç~nuõ!.'C.Qurcapãêity to recogniu:.ins.tances-.oLcmg9Iies.
The prototype approach to categories is particularly associated with
the work of Eleanor Rosch (Rosch, 1975; 1978; Mervis and Rosch,
1981; Armstrong, Gleitman and Gleitman, 1983; for a useful intro-
duction see Clark and Clark, 1977: 464-8; for a ful! discussion of the
'issues see Smith and Medin, 1981). Rosch and her co-workers begin with
rhe observation that alrhough by~clinitiQ!! __ !:,~.~bins,
eagles, swallows,
~triches and~ns are ali birds, we somehQ.wJêclj.fiit t~y do llõtãIT--
 52 The concept of geme
have the same status. Some are 'birdier' than others. Rosch then
c'ônauCteefanumbcr-àf experiments to establish this; she was able to
show, for example, that the time subjects took to verify the correctness of
a statement depended on whether the subject was a 'typical' member of
its class or noto Hence, in the United States, verification times for 'a robin
is a bird' were faster rhan for 'an ostrich is a bird'. Further, Rosch (1975)
was able to establish that when subjecrs were asked to rank examples in
order from most typical to least typical they did so with a large degree of
agreement. Thus, in US culrure chairs and tables were the most typical
members of items of furniture and lamps and ashtrays least typical;
similarly, apples and plums were typical fruits and olives and coconuts
atypical.
The most typical category members are prototypes; a chair is what is
likely to come to mind when we think of an item of furniture and an apple
similarly comes to mind in the case of fruit. A robin is a prototype bird
within US culture because its body and legs are average size, and it flies,
perches in trees and sings. According to Rosch (1975) a category has its
own internal structure, which will assign features or properties a certain
probability for being included in category membership. An ostrich is a
marginal member because it fails to meet the high probability expecta-
tions of flight and relatively small size. Organisms like bats and whales
are problematic because they carry properties that meet highprobability
expectations of categories to which they do not technically belong.
Armstrong et aI. bring rhe definitional and cluster approaches together
in the following generalized way:
There are privileged properties, manifest in most or even ali
examples of the category; these could even be necessary properties.
Even so, these privileged properties are insufficient for picking out
all and only the class members, and hence a family resemblance
~, description is still required.
(~ r>; (Arrnstrong er aI., 1983:270)
, (
~'
'"
..
This integration has consi~erable í!ÉEe~> lt allo",:s the geme. ~nalyst to
find a course between trymg to prodtíce unassailable definitions of a
J'
~ I particular geme and relaxing into the irresponsibility of family resem-"
~ iJ dj blances. As we have seen, communicative purpose has been nominated as
I the privileged property of a geme. Other properties, such as form,
I structure and audience expectations operate to identify the extent to
~hich an exemplar is prototypicai of a particular genre.
4. The rationale behind a genre establishes constratnts on allowabie
ri contributions in terms af their content, positioning and formo
I!), Established members of discourse communities employ gemes to realize
1" communicatively the goals of their communities. The shared set of
A warking definition af genre 53
purposes of a genre are thus recognized - at some levei of consciousness-
by the established members of the pa~~l~tdiscourse cornmunity; they may
be only partly recognized by appi'M1ice'members; and they may be either
recognized or unrecognized by non-members. Recognition of purposes
provides the rationale, while the rationale gives rise to constraining
conventions. The conventions, of course, are constantly evolving and
~ir!fleedcan be directly ch~Dllt~:c9DJIí1üe
.-._~. -""-.-.-'
to~el9~rt.
..... _-
I will i!lustrate these observations by taking two simple examples: one
from administrative correspondence and one from professional inter-
views. Cor~denc~y_}!2t.-yet açlm!ni.~tr.ative corre.s2.()n_c!~.n.S_e,
j t~e~f
does not constiture a genre as it does not represent a coherent set of
shared purposes. Rather it represents, as a convenient label, a supra-
generic assembly of discourse. W!!hin-.adn:ú.tllii!!".~_tiy~x~smden.ce
thcre-ª-re,.howe'ler,.aJll!rrLb..fu)f.establish.ªºl~.~er:~.e~J '!YO-dQs.el)éI~
ones are the indi'!i®-ªJly~9ir~Ç!e_d:g.QQclns:.w:s:.Jett.erAlJd:badne",,-~'Jetter
(MUrpl1y3iiC! Hildebrandt, 1984). These gemes are formal responses to
applicatiófis, or sometimes cornplaints. Classic instances are responses to
applications for jobs, scholarships or grants. At one levei, it might be
argued that both kinds of letter constitute a s.ingle genre of responses to
applications, but a little reflection will show that, while the textual
environment and the register may be the same, the rationale is sufficiently
different to require a separate geme for each.
The rationale for the 'good news' letter is based firstly on the
assumption that the information transmitted is welcome. It therefore is
conveyed early and enthusiastically, while the rest of the letter is set out in
such a way as to remove any remaining obstacles and engender a rapid
and positive response. Part of the rationale behind a 'good news' letter is
that communications will continue. In contrast, the 'bad news' letter is
based on the assumption that the information is unwelcome. Ir therefore
y is conveyed after a 'buffer' has prepared the recipient for its receipt and
couched in language that is regretful and non-judgmental. Part of the
t rationale of the 'bad news' letter is that it minimizes personal resentment
so that no long-term disaffiliation from the institution occurs; another
part is to signal that communications have ended. For that reason, in 'bad
news' letters the nega tive decision is usually represented as having been
taken by some impersonal body, such as a committee, over which the
writer gives the impression of having little influence, the purpose being to
insinuate that complaint, petition or recrimination will be of no avail.
The rationale thus determines what Martin (1985) r~~
s~ctI.II:e.....oi....!h~,jis~.aOO--alsº_~ons!Lains-.lexical an4.
. syntactic choice.
~lllustration is taken from medical consultations and
is designed to' highlight differences in ~nak.J'~I_ç.eprion...be~~_n

/54 The concept o'[ geme
established and non-esrablished members of discourse commurunes.
Apparently many medica I doctors trained in Britain use rhe systern called
SOAP to structure their consultations (jones, 1982):
I. S = Subjective (whar rhe patient says is wrong; whar rhe patient
perceives as his or her symptoms)
2. O = Objecrive (results of tests; symptoms perceived by the docror)
J. A = Analysis (of the symptoms so as to lead tO a diagnosis)
4. P = Prescription (pharrnaceutical and/or giving advice or trearrnenr)
However, patients rarely have any conscious recognirion that rhe doctor
imposes order on rhe consultation by the use of a structuring system such
as SOAP. Part of the reason may be that orher things going on, such as
, greetings and leave takings and various types of urterance designed to
l setrle and reassure patients and to effect transitions between stages
I'
(Candlin, Bruton and Leather, 1976; Frankel, 1984), could appear_more
salient to patienrs. Equally, there may be things apparently not going on:
the doctor's carrying out of stages O and A rnay welJ be a largely silent
(
and privare matter.
Undersranding of rationale is privileged knowledge, but is neither the
whole story nor any guarantee of cornmunicative success. Erickson and
Schultz (1982) in rhcir remarkable study of academic caunseling sessions
i-i; make the folJowing observation:
;1'
There is a similar sequential order of discourse topics across
i'!! interviews- an order which manifesrsan underlying logic of
---T--
\
garekeepingdecision making. Bur ir is not the underlying logic, the
interactional deep structure, that is essential, for much more is
manifested in performance - in cornmunicarional surface structure
\\ - than an underlying abstract logic of garekeeping. Distinctive
packages of social meanings and social identities are also
manifested comrnunicativelyin each interview.
(Erickson and Schultz, 1982:12)
{
i The point to note here is that even when we granr that surface features
I and local dccisionsare highly contributory to the performance outcorne,
' it is still very much rhe case that for a participant to have a sense of rhe
'underlying logic' or rationale is facilirarive in borh reception and
,\ production.
5. A discourse community's nomenclature for genres is an important
source o] insight.
As we have seen, knowledge af the conventions of a genre (and their
rationale) is likely to be much greater in those who routinely or
professionally operate within that genre rarher rhan in those who become
involved in it only occasionally. In consequence, active discourse com-
l
;
A warking deiinition af genre 55
munity members tend to have the grearest genre-specific expertise - as we
often see in interactions between rnernbers of a profession and their dient
publico O$-COns.eqlteng:jSJ:hauh~active membgr~giYeg~ill~names to
slasses-4tommuniC1tixe.ev.:ents_th'lt~iiç:QgDg~_'!~_RmYLQi!!~IT5J![:
~h~Q.t:iC1Lacti01l. These names may be increasingly adopted first by
overlapping or dose discourse cornrnunities and t\::lenby farther and
b~ifíes:-P~ârtí~i~~~_tt~~-t-i~n the~~f<?r~.l]i~4~.tob~g;-ven_tQ_
~em:latlJ[es-created_by~tho~~ijQ:are_rp()stfamiliar witha~d_
~professionalJ}!-jnvQly~_d_in..those.geru:e.~~
As far as acadernic genres are concerned, many, if not most, are terms
that incorporate a pre-modifying nominal of purpose: introducto(}'
~
lecture, qualifying exam, survey article.. review session, writin~()fk-
shop, Others reverse -thé~~der by-~sing a purposivéh~;d-noun: grant
application, reprint request and course description. Still others indicate
the occasion rather than communicativepurpose, such as final exarnin-
ation, plenary lecture, fesrschrift, faculty meeting or graduation address.
However, members of rhe discourse community typically recognize that
particular occasional gemes have particular roles to play within rhe
I.
academic environrnent and rhat, in consequence, the sets of purposes are,
on the one hand, evidenr and, on the orher, constrained.
In the previous section that dealt with linguistic contributions to
genre analysis (2.3), it was argued that insider metalanguages should
certainly be considered seriously, but also viewed wirh circumspection.
lndeed, it was suggested that an appropriate approach for the analyst
would be to establish genres based on investigations inro actual cornrnuni-
cative behavior, two aspects of which, among severa I, would be partici-
pants' naming procedures and elicited categorizations. There are a
number of reasons for caution.
One ~D.is.l:halihe nalJli!:uLC2LCOl:!Ln;LUnicative~ents
that occur and
rec--u-rtnpost-secondary educati9_Q-ªL~t!i!J~ - tq.Jf~JJlçt_discuSsi()ll-tothe
rilaín.fõCliSõl]Jili::siüdy~~;-t~~d to be institutional labels rather than
desénptÍveõ;;-es. I mean by rhis that the timetable or course handbook
will idenrify group activity A in setting X as a lecture, and group activity
B in setting Y as a tutoria/o However, as cvery student in higher education
knows one member of staff's 'tutoria!' can be identical as a communica-
tive event to another member of staff's 'lecture' and vice versa. Of course,
instructors may modify their approach depending on whether rhey are
supposed to be giving /ectures or tutorials, but the fact thar a cornmunica-
tive event is labeled by the institution as being an event of such-and-such
a kind does not necessarily mean rhat it wiIl be so.
Secondly, names tend to persevere against a background of substantial
change in activity, Lectures rnay no longer be rhe monologic recitations
they once were, but actively invite inrercalared discussion and srnall-
group tasks. Tutorials today may consist of student interaction with a
 6>

56 The concept af genre A uiorking definition of genre 57

cornputer program or a tape recorder and no longer involve a 'tutor' in what they expect others will want to hear. This aIlows the speaker to
the traditional sense of the termo We inherit genre categories that get pitch his talk right'). However, I now know whar a haiku session or
passed down from one generation to another. screening panel lecture might be like, even though I have never experi-
In direct contrast, genre-narning can equaJly be generative. While the enced eirher of thern; and I dare say having rcad about such possibilities,
coining and deliberare usage of new labels for event categories can at my interest is raised and so my parricipation is encouraged. Thus it is rhat
times create substance and structure out of an amorphous background, ar the naming and description of new sub-genres can have pre-ernptive force.
others the names may reflect empty categories with no claim to geme Oscar Wilde had an inimitable ability to stand the world on its head, and
status. A pertinent instance of these processes can be seen in the advance when he observecl that 'life imitates art' rather rhan the commonly-held
information for the Nineteenth International IA TEFL Conference converse that 'art imitates lifc', he may have been closer to the truth than
(lATEFL Neu/sletter no. 84, August 1984:54). The section entitled his witticism is generally given credit for. Certainly here we seem to have
'Contributions' quotes ar length from The Working Party Report on been discussing porential cases where 'conferencelife imitares formar'
Conferences, ApriI1984. rarher than the converse.pn the other hand relativel few of these enre
;3ggestions seem to .h.:!.ye
b~~ . Documentation rom su sequent
The range of ways in which presentarions and workshops could
conferences fails to make mention of the 'haiku' or the 'screening panel'
operare might be broadened considerably. If contriburors were
offered a range of possible formars to choose from, rhere would be forrnars, even if others such as 'resource rooms' and 'specific interest
scope for many members who are currently inhibired by the groups' have made some headway.
formaliry of presentation. Ar the same time many presentarions If there are genre narnes with no genres attached to thern, so must there
would continue in rhe well-trieJ formars of rhe pasto be genres without a name. I believe there is at leasr one of rhese that
The advance-information then lists and glosses eleven possible sug- occurs quite commonly in my main professional discourse community
gestions: and which 1 arn sure many readers will recognize. This is a type of
presentation given to colleagues and graduare students which is built
1. Basic prescnra rions
around a number of episodes in which participants, often working in
2. Haiku sessions pairs or small groups, are asked to reach and then share conclusions on
3. Resource rooms short texrs distributed among them. The rasks might involve ranking
4. Traditional talks/lectures texts in order of evolution or quality, re-assernbling textual fragments
into their original order, or using interna I evidence to guess a text's
5. Experimental workshops
provenance. While I have twice experienced the use of such informed
6. Creative workshops
guessing episodes in other disciplines (in geology and art history slide-
7. The buzz-group lecture supported presenrarions), interestingly in both these cases the presenter
8. Curran-style lecture prefaced his remarks with the same phrase 'Now ler's play a party game'.
-' In my own discourse community, I believe that involving others in
9. Screening panel lecture
context-srripped and task-oriented text analysis is viewed as toa central
10. The traditional debate
and too valuable an activiry to be dismissed as 'playing a party garne'.
11. Specific interest groups Anel as far as I am aware, presenrations of this distinctive anel relatively
] think it reasonable to suppose that 11 different formats is decidedly prevalent type have no name.
more than the average conference-goer is familiar with, and I would guess This section opened with the promise that it would produce an
that there are very few people in the English-reaching world who could adequate characterization of geme. The working definition thar follows
confidently explain what is expected to happen in ali 11. Certainly, I had may in fact not be fuIly adequate, but ir has I believe benefited from the
not heard of haiku sessions ('People who have one very good idea to discussion of the term in allied fields and does represent some advance on
present that can really be properly got across in 10 rninutes or one my earlier formularions (e.g. Swales, 1981a). Although there remain
minute') or screening panellectures ('Before the lecture begins three to several loose ends, some to be discussed in rhe next two sections, my
five people from the audience come to the frant and spend five minutes present understanding is summarized below.
discussing what they expect to and want tO hear from the speaker and

!1ri.""
;;f,ifW
.'t.;

L
 '1::1
,9
c.c"frJ p
58 The concept af geme Pre-genres 59
?
Genre defined ~. seem to refer to such a ubiquitous levei of speech performance that one
A geme cQffipriss:s-a....das~Q.!!!.l!.l.unigtive e.vent&;-th~-ID~~f would sense a difference between it and anything else one might wish to
~bar~ some seLOL~~w Q.s~. These purposes are call a genre'.
recognized by the expert members of the parent discourseu.CítmmuniQ> Casual conversarion presumably occurred early in the evolution af the
\ human race, as it does in a child's acquisition of first language. Ir takes
and therebL const~.!~ the ratio~~l.~.JC?!j1e geme. Tl1íS-~ationale_~hapes
the schematic structure of the discourse and influences and constrains up, for rnost of us, a fair part of our days; indeed involvement in
ch:~;~~:~T~Qnt~.nLa~dsUl~-çQrrlt!1qniÚfi~E~EP.9.Se iSJ10ti:l a privileged conversation can be quite hard to avoid, Further, our sense of the
criterion. and ..one that Qpen.teS!9.Jeep thÇ5.f-ºPj':_.2L,!_gçI1_r~ ãSh~Ee enveloping nature of conversarion is broughr horne when we consider its
conceiveçln~rrowlyJ9qJSegQIJ.J;ºI11RM~.pk.rlltlºÜÇ.aÜs_rjQ!l. In addition absence. Therein, after ali, lay many of the trials and tribularians of
to purpose, exemplars of a geme exhibit various patterns of similarity in Robinson Crusoe. Ir is often said that the severity of placing a prisoner in
terms of structure, sryle, content and intended audience. If ali high 'solitary canfinement' resides as much as anyrhing in the denial of verbal
probability expectations are realized, the exemplar will be viewed as interaction, and a 'vow of silence' is no light undertaking.
protorypical by rhe parent discourse community. The geme names Additionally, there would appear to be attesrable individual discrep-
inherited and produced by discourse communities and irnported by ancies berween conversatianall ano non-conversarional skills. Probably
i
others consritute valuable ethnographic communication, but typically ali of us have known people who may be highly effecrive cornrnunicators :f':;.
need further validation. in certain roles (as teachers, salespeople, joke-tellers, armchair critics and ,,!

so on) yet who are adjudged to be lacking in the skills of ordinary

;:f
conversation and thus are thoughr of as individuais who are 'difficulr or . '~'
uncomfortable to talk wirh'. Conversely, we probably know people who
3.6 Pre-genres I'
seem to have a remarkable facility to sustain casual conversation,
who are the first to announce, for instance, that they couldn't srand up
but ;[
lone of the basic assurnptions underlying much of rhe prcccding discuss- 1.(,
and give a vore of rhanks to save their lives. These observarions ali seem
I ion is thar human beings organize their comrnunicative behavior partiy to point to the fact that general conversational ability and genre-specific
1:1
I rhrough repertoires of gemes. Thus, it is not the case that ali cornmunica- verbal skills may be phenomena of a somewhat different kind. ;!
I. tive events are considered instances of gemes. ln fact, there are at least If these observations have substance, it would secm that ordinary !~
two arcas of verbal activity that I believe are best considered to lie outside conversation is too persuasive and roa fundamental to be usefully
genres: casual conversation or 'chat' and 'ordinary' narrative.
The nature and role of conversation will be considere~Ffirst, and
considered as a geme. Rather, it is a pre-generic 'form af life', a basis ,i~
from which more specific types af interaction have presumably either
Levinsori's opening posirion will serve perfectly well: "''''!W'i'!
evolved ar broken away. The interesting question for rhe geme analyst is
Definition will emerge below, but for the present conuersation may not so much wherher conversation is a genre; instead, the interest lies in
be taken to be thar predominam kind of talk in which rwo or more ..~,, exploring the kind of relarionship that might exist between general ,\
participants freely alternate in speaking, which generally occurs conversational patterns, procedures and 'rules' and those rhat can be
outside specific institutiona! settings like religious services, lato discovered in (to give rhree examples) legal cross-exarninations, medical
courses, classroom.and the like.
consultations and classroom discourse. In those three cases, are the
(Levinson, 1983 :284, my emphases) unfolding interactions best seen as mere extensions and modifications of
common conversational practice and thus ultimarely parasitic on such
This kind of talk has, of course, been massively studied and discussed, practice? Or, alternatively, would we gain a greatcr understanding of
particularly since the advent 01
the tape recorder (e.g. Grice, 1975; what is happening by considering them as existing independently in
Goffrnan, 1981; Levinson, 1983; Richards and Schmidt, 1983; Gardner, separa te universes of discourse? Are Unequal Encounters (CandJin,
1984); and Atkinson (1982) gives the ethnomethodological arguments 1981) such as normally occur between doctor and patient, lawyer and
for the cenrraliry and significance of conversation, As he and many witness, and teacher and pupil, af a different kind to the more equal and
people have observed, 'ordinary' conversation is a fundamental kind of less goal-directed encounters rhat take place in casual conversation?
language use: for example Preston (1989:225-6) comments: 'Since ~es!iE1L_~_~pec~~ the. PU!~EÜ'_~._(f:.Iarionslljp~._~<:~(:eJUh.~
':;"
conversation in some sense is basic to ali face-to-face interactionçit may ~re=ge~!J.d_~~~.ff.urs j_I!~~e.r.e_:Qr..dinary'~g;~!2~fa~.

/ "

r

60 The concept of genre
conversation is replaced by telecommunication. Schegloff (1979) has
shown that telephone conversations actually open with the ringing of the
telephone and that rhe person lifting the receiver and speaking is
responding to a summons. He has also analyzed and described the limited
range of procedures that Americans use to identify and recognize each
other on the telephone (much less of a problem, of course, if you can see
to whom you are abour to talk). Owen (1981) has written interestingly
on the use of 'well' and 'anyway' as signals given by British telephone
speakers to indicate a wish to dose a topic or a call. However, to establish
that a particular kind of communicative event has specific, situation-
bound opening and dosing procedures is not, in fact, to establish very
_/
.>-
"mnch'lbecause specificity may well be concentrated at initiation and
ternhnation (Richards and Schmidt, 1983: 132-3). For example, open-
ingsllike 'Merry Christrnas', 'Cood morning, Sir', 'Oh, we are srnart
today', 'Come here often, do you?' refIect particular circumstances that
are Iikely to be of rapidly diminishing importance as the conversation
proceeds. Thereforeçon p~es~I1t~"igeps.~ljU~Q!!Iª.s_e.e~el1sible
exclude peE~!l!lIJ~1~p.li6iI~....COJJ.Y~.llil..tions.Jro!!!....J;.eru:e..status-'~
~~t~~rp.,·,~e.~P~~
---p~nr",.
rheir relativ~ly~.h()E!.. his.tor.y,-,as.pan..91-!he'
..-.~.- ... , ~... ,
In contrast, we can imrnediately recognize the unusual nature of
radio-telephony. Robertson (1985; 1988), for exarnple, outlines the
purposes of plane-ground radio-telephony as to:
i) prevent eollisions in rhe air;
ii) prevent collisions berween aireraft and berween aircraft and
obstructions on the manoeuvering area;
iii) expedire and maintain orderly flow of air-traffic;
iv) provide advice and information useful for the safe and efficient
conducr of flights,
(Robertson, 1985:295)
Civen these aims it is not surprising that there have evolved especiaIly
rigid rules for turn-taking (Sacks, SchegloH and ]eHerson, 1974) and
special conventions for clarifying both rhetorical function and identity.
These conventions have to be learnt by native speakers as well as
non-native speakers, as the following fragment illustrates:
Control: Sierra Fox 132, correction, Sierra Fox 123,
whar is your flight levei?
Pilot: Flight levei 150, Sierra Fox 123.
Control: Say again flight levei, Sierra Fox.
Pilor. Flighr levei 150, Sierra Fox 123.
(Robertson, 1935:303)
( Radio-telephonic Air Traffic Control meets the criteria for geme status.
\
\./
v.
Differences among genres 61
If casual conversation is a pre-generic dialogic acnviry, is there a
comparable pre-genre for monologue? The obvious candidate is nar-
ration (if viewed as a process) or narrative (if viewed as product).
Narrative, Iike conversation, is a fast expanding research field (Van Dijk,
1972; Crimes, 1975; Longacre, 1983) and has developed its own
disciplinary name, Narratology (Prince, 1982). For present purposes I
will simply follow Longacre and suggest that narration (spoken or
written) operates through a framework of temporal succession in which
at least some of the events are reactions to the previous events. Further
characteristics of narrative are that such discourses tend to be strongly
oriented towards the agents of the events being described, rather than to
the events themselves, and that the structure is typically that of 'a plot'.
These pre-generic long tums commonly occur in letters and also arise as
responses to such prompts as 'How was the vacation?' or 'How did the
meeting go?' and so on. "
In a way analogous to that described for conversation, specific types of .~;;
:1 :~'
to narra tive diverge frorn the pre-generic norm and thus begin to acquire f,i
genre status. Thus in news stories the temporal succession is disturbed by
~
putting 'the freshest on the top'. ln reports of various kinds, such as those + )-;
describing scientific work, events rather than agents predomina te. ]okes .t;
have temporal sequences, agem orientation and plot, but the resolution
of the plot is specific: the moment of resolution needs to be overt!y
signaled (the onset of the punch line) whilst the manner of resolution
needs to be unpredictable.
A final point perhaps worrh making at this juncture is that thê
English-speaking world (as one of many) uses names to describe classes
of communications that quite appropriately operate as higher-order
categories than gemes. One very common example is the letter. This
useful term, of course, makes reference to the means of communication,
but lacks as a dass sufficient indication of purpose for geme starus. The
same observation holds for subsets of the dass that refer to fields of
activiry such as business letters or officialletters. lt is only when purpose
becomes ascribable that the issue of geme ariscs, as in bcgging letters ar
letters of condolence. Category labels like letters do not therefore refer to
pre-gemes in rhe scnse used here, but operate as convenient multigeneric
gencralizations.
3.7 Differences among genres
If there were only minor differences among gemes there would be little
need for genre analysis as a theoretical activity separable from discourse
analysis, and probably no need at ali for an analysis driven by applied
concerns. But, of course, it turns out that genres vary significantly along
 I!
62 The concept af genre
quite a number of different parameters. We have already seen that they
vary according to complexity of rhetorical purpose - from the ostensibly
simple recipe to the ostensibly complex politica! speecb, They also vary
greatly in the degree to which exernplars of the geme are prepared or
constructed in advance of their communicative instantiation (Nystrand,
1986). Typical prepared genres might include research papers, letters of
personal reference, poems, recipes, news broadcasts and so on, while at
the other exrrerne arguments and rows typicalJy flare up wirhout malice
aforethoughr. Ç,emes also Ya.r.}'jll~~~lJlS
oL~h<o._mode oU!!eIDUJUJhrough..
w~ich they~(Ç.sscl;.in<k~guu~~,~ configurati~~...9..L~'pç~_ch verSl!~_
w!illitg can become gt!ite complex (Grégory~T9bl). F,º!J.n.~.t;mce,of the_
pr ' of re ared en es.,JIlQSllf~..E!edominant!x~
However, research papers can be presenred at conf(:renCéS·'j'i1·'manuscript
delivery' (Dubois, 1985) or as 'aloud reading' (Goffman, 1981), while
:1 references and recipes can in an emergency be communicated by the
: 'I
telephone. Poems in western cultures have in rnodern times been a
predominantly written form, although 'aloud reading' of them is an
ongoing tradition and one thought of as requiring uncommon skill in
modulated performance (in the case of actors) or in interpretation (in the
case of poets reading their own work). In other cultures the converse mar.
; ~~:
':i
apply with poetry as an essentialJy oral medium, written forms operating
,; as archival repositories. News broadcasts are scripred and then read
~i
!:
aloud.
Prepared-rext gemes like those we have been considering vary also in
the extent to which their producers are conventionally expected to
consider their anticipated audiences and readerships. However, this
variability is somewhat ar odds with much current thinking on and
research into writing processes. An influential and representa tive advo-
cate of the interactional orientation to reading and writing processes is
wiádowson (Widdowson, 1979; 1983; 1984). He expresses his 1979
position, which has little changed (d. 1984:220) in this way:
As I write, [ rnake judgements about lhe reader's possible
reacrions, antieipate any difficulties rhat I rhink he might have in
understanding and following my directions, conduct, in short,
covert dialogue with my supposed interlocutor.
(Widdowson, 1979: 176)
According to this view, writers, at least cornperent ones, are trying to
second-guess both their readers' general state of background knowledge
,o,,)
and their potential immediate processing problems. At rhe same time
;i
~ (competem) readers are imerrogating authors on their present positions
'\ as weU as trying to predict where rhe aurhors' lines of rhought or
: 1 description willlead. There is, as it were, a reciprocity of semantic eHort
h
I>; . ~
: .~
li.
lti~
....
.
) ,.
d>
Differences among gemes 63
to be engaged in by both sides; a contract binding writer and reader
together in reaction and counter-reaction.
Investigations imo various gemes would, however, suggest that this
supposed sociocognitive activity is over-generalized, since a producer's
contract with a receiver is not general, but subject to quite sharp genre
fluctuations. Of course, rhe interactional view is obviously both appro-
priate and useful in certain contexts such as the processing of recipes and
news broadcasts. Indeed, Hugh L'Estrange (personal communication)
has pointed to the fact that recipe-mongers who fail to be considerate of
the reader can conrribute to gastronomical shipwreck, as in 'T ransfer
immediately to a pre-greased tin'. And news braadcasts go to quite
considerable lengths to ensure that they are comprehensible both by
repetition ('Here are the main points again') and by providing back-
ground information ('President Kyprianou of Cyprus', 'Faya-Largau, a
strategic town in Northern Chad .. .') (Al-Shabbab, 1986), While recipes
and news broadcasts may be marginal to the purposes of this book, we
can also affirm that a unifying characterisric of instructional-process
gemes will be consideration for the reader or listener.
However, it remains the case that in certain gemes, usually written
ones, the writer has the right to withdraw from the contract to consider
the reader because of an overriding impera tive to be 'true' to the
complexity of subject matter or to rhe subtlety of thought and imagin-
ation (Elbow, 1988). Thus we find that in a significant number of geme
texts, in laws and other regulatory writings, in original works in
philosophy, theology anel mathernatics (and arguably rheoretical linguis-
tics), in many poems, and in certain novels of which Joyce's Finnegan's
Wake would be an extreme example, there is a diminished consideration
for the reader. Joyce, after ali, is reported to have commented on
Finnegan's Wake to the effect that as rhe book took him 18 years to write
he didn't see why the reader shouldn't take as long to read it.
There is in fact a standard defense of the legal draughtsman's practice
of using very long sentences containing numeraus and e1aborate quali-
fications (ali those elements beginning notunthstanding, in accordance
with, without prejudice to etc.). This defense would c1aim that it is
ultimately more satisfactory for a legal text to reveal clarity after detailed
and expert study than to be a text that, however immediately accessible
to an educated lay audienee, falls into ambiguity upon rnultiple reading
(Bhatia, 1983).
Thus ir turns out that certain legal, academic and literary texts ali point
to another kind of contract that can exist between writer and reader. This
is one not based 011 'consideration' but on respect. If we use Widdowson's
device of imagining the rhoughts of the writer, it might come out
something like this:
II,
'"
64 The concept ar genre Differences among genres 65
As Iwrite, Iam aware that, whatever I do, what I write will be ... a c1early defined topic, introduetion, body which explicates ali
difficult for most readers. Because of what I arn trying to achieve,
but nothing more than the stated to pie, paragraphs which chain
this is unavoidable. This is why Imusr convince the readers that
from one to the next, and a conclusion which rells, the reader what
their efforts wiU be rewarded; I need to keep their faith that I arn
has been discussed ... no digression, no matter how interesting, is
not making my text unnecessarily difficult.
permitted on the grounds that it would violate unity.
While Flower (1979) and her co-workers may be generalIy right in their (Kaplan and Ostler, 1982:14)
theory that the immature writer produces 'writer-based prose' and the
mature writer 'reader-based prose', ir would seem equalIy cIear that in They then contrast this pattern with the elaborate paralIel structures
certain genres mature writers also produce 'writer-based prose'. found in Arabic prose, with the more digressive patterns of writing in
GhrU:f§--<llso_Y.a-!:L~xtcnLto..wbklLll19:': are likely .!9~hit Romance languages which permit 'tangential' material to be introduced
~~~W~cific~. On the one hand, it would in the discourse, and so on. Clyne (1987), in a particularly careful study,
appear that the diplomatic press communiqué has developed a global if has examined the Exkurs or 'digression' in contemporary academic
devious set of conventions whereby, for instance, 'a fulI and frank German and is able to show, among other things, that the Exkurs is not
exchange of views' is interpreted by discourse community members only institutionalized in certain German genres but has no easy trans-
throughout the capitais of the world as signifying that the parties failed to lation equivalent in English.
agree. On the orher hand, one might reasonably assume that marriage Comparison of languages is notoriously difficult, especialIy at the
proposals wilI differ widely from one language community to anorher discoursal level (see Houghton and Hoey, 1983, for a specification of
because they are deeply embedded in particular socíoeconomic cultural caveats). Among such caveats it is irnportanr to compare texts of the same
matrices. genre in two languages. Ostler (1987), for example, can be criticized for
The sociolinguisric literatureon rhe form, structure and rationale of comparing student placement essays with extracts from published texts.
specific communicative events is vast and faIls largely outside the scope of In general terms the existence today of 'invisible colIeges' and of
this book (see Saville-Troike, 1982; Downes, 1984; and Preston, 1989 transnational discourse communities is likely to lead to universalist
for overviews). However, there is one investiga tive area that is directly tendencies in research gemes. A strong form of the universalist hypo-
relevam to a pedagogically-oriented study of academic English, one thesis is offered by Widdowson:
known as Contrastive Rhetoric. Scientific exposition is structured according to certain patterns of
The concept of Contrastive Rhetoric was originaIly elaborated by rhetorical organization whieh, with some rolerance for individual
Robert Kaplan in a 1966 articIe entitled 'Cultural thought patterns in stylistic variation, imposes a conformity on members of the
intercultural education' (Kaplan, 1966). Kaplan, who has remained scientific community no matter what language they happen to use.
active in this area, more recently summarized the concept as foIlows: (Widdowson,1979:61)

There are, it seems to me, important differences berween languages

Najjar's 1988 study of research articles in English and Arabic dealing
in the way in which discourse topic is identified in a text and in rhe
way in which discourse topie is developed in terms of with agricultural science shows sufficiently few and sufficiently unimpor-
exemplifieation, definition, and so on. tant differences to provide some support for the universalist argumento
However, as we have seen, Clyne (1987) provides some counter-evidence
(Kaplan,1987:10)
from German as does Peng (personal communication) from Mandarin.
The jury is stilI out.
The notion that the rhetorical structure of languages differs is not orily
Although universalist tendencies may be apparent in research activi-
relevam in itself, but more particularly because much of the work to date
ties, those who have taught in different higher education institutions
has been based on the study of expository prose (Connor and Kaplan,
around the world have typicalIy been struck by the peculiarities of study
1987). Kaplan and Ostler (1982), in a review of the literature, conclude,
modes, teaching styles and of general educational expectations within
despi te a minority of studies to the contrary, that different languages
particular institutions (james, 1980). If we examine, say, the first years of
have different preferences for certain kinds of discourse patterns. For
undergraduate study in Faculties of Science, I believe it would be difficult
instance, they argue that English expository prose has an essentially
to argue that what goes on in those faculries is part of a universal
linear rhetorical pattern which consists of:
scientific culture. Rather, we tend to find in this are a of scientific activity
 '.,ih
!;

'"
66 The concept of geme Differences among gentes 67

powerful local influences of many kinds: national, social, cultural, (1987) has shown the existence of two rhetorical sryles in contemporary
technical and religious. The ways in which such inl1uences form par- academic Korean, one deriving from tradicional rhetoric and rhe other
ticular 'educational cultures' have been described for various parrs of the much influenced by English. Although Eggington does not put it in these
world: Thailand (Hawkey and Nakomchai, 1980); Iran (Houghton, terms, we can see here the existence of twodiscourse communities: an
1980); the Arab World (Dudley-Evans and Swales, 1980; Holliday, elite group of US-educated scholars who are members of the inrernational
1984); Asia (Ballard and Clanchy, ]984). There have also been some community of researchers in their specialization, and a larger national
inreresting studies of the 'rhetorical gaps' that apprentice researchers comrnunity using traditional Korean rhetoric. Indeed, rhe discourse
from overseas have to cross when leaming English academic style: a cornmuniry concept, as a sociorhetorical construct, offers some general
Yemeni Arab student (Holes, 1984); a Brazilian (james, 1984a); a Thai illumination 00, the difficult and irnportant question of academic lan-

+ aOd a japanese (Ballard, 1984). Ali in ali, ir looks as though the relativist
hypothesis has some substance in teacher-student
books, lectures and turQrialsYevertheless,
interpretation. We canôrhcrfean
or we can prefer an cxplanation
genres such as text-
we face a difficulry in
towards intrinsic cultural differences,
rhar woüldgo no further than stress the
guagevariation across cultures and generations.

(.
relevance of recenr history. For instance, are thedifferences between
western and Arab educarional genres a reflecrion of differences in :;
rhetorical and ideological cedes, ar do theysignify litrle more than
different stages in an educational cycle? More precisely, would we do ~:
better to interpret such differences as dcriving principally from, on the
one hand, an Islamicized verbalistic tradition and, on lhe other, a
.' i
secularized pragmatic European or Norrh American tradition? Or should
we conclude that modes of study and modes ·oi expression commonly
accepted and practiced in the Arab World. today are in surprising
numbers of ways similar to those existing in the West 50years ago (the
teacher qua reacher as respected authority, a stress on rote-learning, a
style of wriring in the tradition of belles-lettresexc.vi An educarional
ethos which may, of course, yet revive in the West. ..
At present, our perspectives on the forrnarive influence of the edu-
cational environment rest largely on anecdóte, incidental observation
and the single-subject case study. Mohw..~..u 985·S] S~are certairyly
CQ.ITect in lhciLcÚtique..oLCOlurastive Rhetori~.w.henJ.he.)!...I2.9inL~
need~rearer
- ..
--' -.--..'
awareness of students'
---.----.-.-
narive
-_
literacy andeducatio!l<!!
_--~ -'- T- ..-~--·-
f~$J:ienc.e.<lS.factoF& influeru:ingJhe_Q~ITloJ1m~~Eot acaaen:jç,Y.-:.r.!J,ingJn..-
~~ond I-ª'Q~~g~ Ir is hoped that the concept of ge~re-ae;eIoped in this
booGspecially with regard to features of text-role and text-
environment, wilI conrribute to a less narrowly Iinguistic orientation in
Conrrastive Rhetoric studies. And indeed, independently, rhere are signs
that rhis is already happening. Hinds (1987:143-4) has suggested that
English-language culrures tend to charge the writer 'with the responsi-
bility to make clear and well-organized staternents', whereas in japanese
culture 'ir is the responsibility of the listcner (or reader) to understand
what it is that the speaker or author had intended to say'. Hinds' typology
can thus be related, in a cross-cutting way, to the previous discussion on
genre-specific differences in the writer's responsibility. Finally, Eggingron