ELIA 2, 2001
213
A PRAGMATIC ANALYSIS FRAMEWORK FOR THE
DESCRIPTION OF MODALITY USAGE IN ACADEMIC ENGLISH
CONTEXTS
Jo di Piqué (Uni e si a de València)
San iago Pos eguillo (Uni e si a Jaume I, Cas elló)
J.Vicen And eu-Besó (Uni e si a de València)
In ou p agma ic app oach o esea ch a icles (RAs), we ake a look a one
o he main cha ac e is ics o academic w i ing: he use o modalized
s a emen s. Modaliza ion, in he o m o modal e bs, has been ex ensi ely
s udied by ESP esea che s and p ac i ione s. Mos o hei s udies,
howe e , ha e sligh ly ouched upon he dis inc ion be ween epis emic
modali y, which ques ions he ce ain y o p obabili y o a s a emen , and
deon ic modali y, which lays obliga ions o gi es pe mission o he audience.
This dicho omy is an impo an ool o desc ibe disciplina y a ia ions in
academic and p o essional w i ing. We con end ha di e en disciplines
a o di e en ypes o modali y. Resul s in his s udy indica e ha heal h
science RAs mos ly use epis emic modali y, whe eas li e a y c i icism RAs
combine he use o bo h epis emic and deon ic modali y.
1. In oduc ion
Wi hin he p agma ic linguis ic pa adigm, gen e analysis, as de ined
by Swales (1990) o Bha ia (1993), is a majo app oach in he cu en s a e-
o - he-a s udy o English o Speci ic Pu poses (ESP). In his line o
esea ch, much a en ion has been gi en o he de ini ion o some specially
ele an gen es, such as abs ac s (C emmins, 1982; Lancas e , 1991;
Pos eguillo, 1996b). A numbe o s udies ha e ocused on RA s uc u e
(Hu chins, 1977; S anley, 1984; Swales, 1990; Nwogu, 1997; Bha ia, 1993;
Pos eguillo, 1996a, 1999; Piqué & And eu-Besó, 2000), on indi idual
sec ions o he RA, such as he in oduc ion (Swales, 1981; Dudley-E ans &
Hende son, 1990; Piqué & And eu-Besó, 1998); he esul s sec ion (B e ,
1994); he discussion sec ion (Hopkins & Dudley-E ans, 1988; Holmes,
1997) and including o he discou se unc ions wi hin he RA, such as he use
o ci a ions and e e ences in RAs (Dubois, 1988; Lynch & McG a h, 1991),
on RA a gumen a ion (Hyland, 1990; Thompson, 1993), o name bu a ew.
J. Piqué, S. Pos eguillo y J. V. And eu-Besó
214
Resea ch a icles --as he mo e deeply analyzed academic gen e--
ha e been al eady been s udied o un old some o he mos salien
disciplina y a ia ion o be de ec ed. Howe e , mos o hese s udies ha e
ocused on whe he he IMRD pa e n is applicable o all disciplines as he
ep esen a i e mac os uc u e o he gen e. We ha e also pu sued his line o
esea ch in p e ious s udies when, o ins ance, desc ibing he semi-
applicabili y o he in oduc ion-me hod- esul s-discussion (IMRD) pa e n
o compu e science esea ch a icles (Pos eguillo, 1999), o he
app op ia eness o Swales’ (1990) CARS in oduc ion pa e n o medical
and nu sing esea ch a icles (Piqué & And eu-Besó, 1998, 2000). I is ue
ha hese mac os uc u al desc ip ions ep esen a majo ea u e o RAs as a
gen e, bu he e a e o he linguis ic and communica i e phenomena which
di e en ia e he esea ch a icle om o he gen es which ha e no been
analyzed in such de ail, especially ha ing in mind disciplina y a ia ion.
Ou aim in his pape is o analyze he language o h ee di e en
academic and p o essional con ex s (heal h sciences, jou nalese and li e a y
c i icism) in o de o de ec possible a ia ions in he use o epis emic o
deon ic modali y and o compa e ou esul s wi h hose ob ained by Simpson
(1990). Ou ini ial hypo hesis is ha di e en disciplines o p o essional
se ings a o di e en ypes o modali y.
2. Co pus and analysis
Mood and modali y exp ess he speake ’s a i ude o opinion
ega ding he con en s o he sen ence o wha he sen ence p oposi ion
en ails (Palme , 1986: 21). He de ines mood as ealized by he e bal
mo phology, whe eas modali y appea s as a linguis ic ea u e gene a ed by a
a ie y o linguis ic phenomena, as desc ibed by Downing and Locke (1992:
383-384), among which modal e bs play a special ole. Bo h Palme (1986:
18-19) and Downing and Locke (1992: 382) dis inguish wo main ypes o
modali y: epis emic and deon ic.
Epis emic modali y, on he one hand, implies ha he speake
assesses “ he p obabili y ha he p oposi ion is ue in e ms o he modal
ce ain y, p obabili y o possibili y” (Downing & Locke, 1992: 382), as in I
may be he case ha ..., Resul s migh change i ce ain condi ions ..., o The
ELIA 2, 2001
215
conce mus be o e . On he o he hand, deon ic modali y means ha he
speake “in e ene[s] in he speech e en by laying obliga ions o gi ing
pe mission” (Downing & Locke, 1992: 382), as in One mus look in o his
ma e in de ail..., o Shall we nego ia e peace now?
In ou esea ch we ga he ed h ee dis inc co po a: i s a co pus o
heal h science RAs (wi h 30,086 wo ds) made up o di e en RAs dealing
wi h AIDS, cance and su ge y; a second co pus o newspape a icles (wi h
30,481 wo ds) dealing wi h all ypes o opics downloaded om The
Gua dian; and a hi d co pus o li e a y c i icism RAs (wi h 30,042 wo ds)
dealing wi h Ame ican and B i ish li e a u e om a selec ion o essays. The
names o he au ho s, he i les o he a icles --o headlines--, as well as he
ables, pic u es o igu es ha e been elimina ed.
We a e awa e ha co pus 2 is made o a gen e (i.e. newspape
a icles) o he han he RA, bu we belie e in he ele ance o compa ing he
use o epis emic and deon ic modali y ac oss h ee dis inc academic and
p o essional se ings --heal h sciences, jou nalese and li e a y c i icism.
Ne e heless, he di e ences --should he e be any-- be ween co pus 1
(heal h sciences) and 3 (li e a y c i icism) can only be explained in e ms o
disciplina y a ia ions in he use o modali y.
In each co pus we ha e sys ema ically conside ed he modal e bs
can, could, may, migh , mus , ough , shall, should, will, would, and he wo
semimodal e bs da e, need, as de ined by Downing and Locke (1992: 315-
316).
3. Resul s
Tables 1, 2 and 3 below o e he esul s ob ained in ou
in es iga ion. The da a om he heal h sciences co pus in p esen ed in Table
1. On he le -hand column one may see he modal and semimodal e bs
analyzed; on he second column we ha e he numbe o imes which a
pa icula e b has been used in ha co pus; column h ee shows he
equency ( ) o use o he modal e b wi h epis emic meaning; column ou
gi es he ela i e pe cen age (%) o use o epis emic modali y; column i e
shows he numbe o ins ances ( ) in which he e b was used wi h deon ic
J. Piqué, S. Pos eguillo y J. V. And eu-Besó
216
meaning, and, inally, column six gi es he co esponding ela i e
pe cen age (%) o deon ic usage. Tables 2 and 3 below a e o ganized in he
same way.
Table 1. Epis emic and Deon ic Modali y in Co pus 1.
1-Heal h Science-RAs
Epis emic Deon ic
To al
Modals
%
%
can 124 123 99.19 1 1.01
could 31 31 100 0 0
da e 0 0 0
may 125 125 100 0 0
migh 20 20 100 0 0
mus 17 15 88.23 2 2.27
need 30 28 93.33 2 2.14
ough 0 0 0
shall 0 0 0
should 49 41 83.67 8 9.56
will 30 30 100 0 0
would 24 24 100 0 0
To als 450 437 97.11 13 2.97
Acco ding o hese esul s, he use o deon ic modali y is almos
non-exis en in heal h sciences RAs ep esen ing only 2.97% o he o al
modal and semimodal e bs. I is also no iceable ha many o he modals
and semimodals show a 100% o epis emic usage. Such is he case o could,
may, migh , will and would, as illus a ed in he ollowing examples:
(1) I all cells ha ha bo ed HIV could quickly be des oyed in his
manne , he i us migh be clea ed om he body.
(2) Wha 's mo e, he immune sys em, in killing o in ec ed cells, may
play a signi ican ole in i s own des uc ion.
(3) On he basis o hese indings, esea che s ha e de eloped
heo ies on how HIV migh indi ec ly des oy ne e issue.
ELIA 2, 2001
217
(4) This p ocess will allow o he iden i ica ion o hose measu es
which hold he mos p omise o mo ing he ield o wa d and o
c i ical gaps ha pe sis in he HIV HRQoL measu emen a ena.
(5) Mos clinicians would p obably ag ee ha his amoun o
di e ence would be a clinically signi ican di e ence.
The esul s ob ained in ex s om The Gua dian a e shown in Table
2. Epis emic modali y con inues being he main ype o e b used by
jou nalis s (91.20%), al hough a sligh inc ease is p esen in ega d o e bal
exp essions wi h a deon ic o ien a ion (8.80%).
Table 2. Epis emic and Deon ic Modali y in Co pus 2.
2-Jou nalism—The Gua dian
EPISTEMIC DEONTIC
To al
Modals
%
%
can 64 52 81.25 12 18.75
could 41 39 95.12 2 4.88
da e 00 0
may 22 21 95.45 1 4.55
migh 10 10 100 0 0
mus 22 14 63.64 8 36.36
need 19 19 100 0 0
ough 1 1 100 0 0
shall 00000
should 39 35 89.74 4 10.26
will 112 111 99.11 1 .89
would 79 71 89.87 8 10.13
To als 409 373 91.20 36 8.80
The modal e b mus , howe e , p esen s a pa icula ly in e es ing
change. While in heal h sciences his e b was used as an epis emic
modalizing de ice (c . Table 1), in jou nalism i combines epis emic
(63.64%) wi h deon ic (36.36%) meaning. We wan o illus a e his use in
he ollowing wo examples om co pus 2 in which he modal e b mus is

J. Piqué, S. Pos eguillo y J. V. And eu-Besó
218
p esen ed in his double modalizing unc ion: epis emic in example (6) and
deon ic in example (7).
(6) As i is, he showe o G ammies and BPI awa ds ha Telboy
ea ned om his 1987 debu album now seems a long ime ago, and
he inds himsel in wha mus be a sligh ly uncom o able limbo.
(7) 'Wha you did ha e ening in M s Longden's la was a disg ace ul
exhibi ion o uncon olled bad empe which spli o e in o he
commission o a c ime o which you mus be punished.'
Co pus 3, om li e a y c i icism essays, p esen he mos s iking
di e ences. As shown in Table 3, he pa e n o epis emic/deon ic usage is
e e sed wi h modals mus and shall, bo h o which a e mo e equen ly used
in deon ic modalizing exp essions han in epis emic ones. I should be
poin ed ou , howe e , ha al hough he modal e b shall may no be
pa icula ly ele an , since i only appea s in 7 ins ances, he modal e b
mus , wi h 58.62%, con i ms he double unc ion o his modal.
ELIA 2, 2001
219
Table 3. Epis emic and Deon ic Modali y in Co pus 3.
3-Li e a y C i icism Essays
EPISTEMIC DEONTIC
To al
Modals
%
%
can 61 37 60.66 24 39.34
could 23 17 73.91 6 26.09
da e 00 0
may 26 20 76.92 6 23.08
migh 29 27 93.10 2 6.89
mus 29 12 41.38 17 58.62
need 7 5 71.43 2 28.57
ough 5 5 100 0 0
shall 7 1 14.29 6 85.71
should 10 8 80 2 20
will 55 37 67.27 18 32.73
would 36 31 86.11 5 13.89
To als 288 200 69.44 88 30.56
The ollowing examples illus a e once mo e his double unc ion o
mus , in his case in he ield o li e a y c i icism: epis emic, in example (8),
while deon ic in (9).
(8) The e o o accu a e ansc ip ion in ol ed mus ha e been
he oic. Bu exac ly wha he u ili y o he exe cise was I, o one,
ha e ne e unde s ood.
(9) I we wan o unde s and Ame ica, you mus ealize ha he
g ea es lo e is his homoe o ic passion, which is in some ways as
s ong, hough also as sublima ed, as Dan e's lo e o Bea ice.
I should also be poin ed ou u he di e ences in he use o modal
e bs, especially in e e ence o co pus 3 (jou nalism) as compa ed o co pus
1 (heal h sciences). Fi e modal e bs –could, may, migh , will, would– in
co pus 1 showed a pe manen use o epis emic modali y. In co pus 3,
howe e , hese same i e modals p esen se e al ins ances o deon ic
modaliza ion, as can be seen in he ollowing examples om jou nalism
J. Piqué, S. Pos eguillo y J. V. And eu-Besó
220
ex s:
(10) The e's a amous essay by Ma k Twain, which is called 'The Li e a y
O enses o Fenimo e Coope ", in which he poin s ou ha Coope couldn'
build a no el. His s uc u es a e pi i ul; he couldn’ w i e dialogue.
(11) I is, one may hink, a elie ha mo e eco ds do no emain, since i
hey did, and i McKenzie ansc ibed hem as assiduously as he does hose
o 1696-1712, his his o y o he p ess would un o 120 qua o olumes.
(12) In a supe icial and simpli ied way (which eaches, ne e heless, as a
as we migh wish o ex end i , and wi h all o he nuances which a e
conside ed necessa y), li e a u e is a p oduc o human ac i i y ...
(13) And his akes me back o my beginning which will also be my end.
(14) Bu I would like o ake his as a as possible.
This p og essi e bu e iden inc ease in he use o deon ic modali y
is shown in Figu e 1 below which g aphically summa izes he esul s shown
in Tables 1, 2 and 3. In u n, epis emic modali y use is close o 100% in
co pus 1, i sligh ly diminishes in co pus 2, bu he e is a signi ican dec ease
in co pus 3.
Figu e 1. Epis emic s Deon ic Modali y in he Th ee Co po a (in %).
0
20
40
60
80
100
120
Deon ic 2.97 8.8 30.56
Epis emic 97.11 91.2 69.44
1-HS-RAs 3-Jou nalism 2-Li C i
ELIA 2, 2001
221
4. Discussion
The esul s ha ha e jus been p esen ed in his esea ch con i m ou
i s hypo hesis in which we en u ed ha di e en disciplines a o
di e en ypes o modali y. This is especially so in he case o heal h
sciences and li e a y c i icism esea ch a icles. These wo specialisms, as
exempli ied in co po a 1 and 3, qui e simila in size and gen e, ep esen wo
di e en uses o modal e bs in e ms o epis emic and deon ic modali y.
While in heal h sciences (co pus 1) seems o be p ac ically he only
accep able op ion o he w i e , in li e a y c i icism bo h op ions seem o be
equally eso ed o by hei au ho s. Thei communica i e pu pose seems o
dic a e he use o a gi en modal e b h ough which hey exp ess hei
epis emic o deon ic meaning. These indings open a whole new a ea o
linguis ic and sociolinguis ic esea ch and makes us pose he ques ion
whe he he communica i e pu pose o an essay in li e a y c i icism di e
om he communica i e pu pose o a medical esea che /w i e . This may
e y well be he case.
Simpson (1990) no iced he endency o li e a y c i ics o be highly
asse i e in hei s a emen s. He desc ibed ins ances o unmodalised
asse ions in he ex he s udied --The G ea T adi ion, by F. R. Lea is.
Simpson quo es Lea is’ (1950: 1) opening s a emen as a ypical example o
unmodalized asse ion:
(15) The g ea English no elis s a e Jane Aus en, Geo ge Elio , Hen y
James and Joseph Con ad— o s op o he momen a ha
compa a i ely sa e poin in his o y.
Simpson’s s udy, howe e , has he sho coming o ha ing analyzed
only one ex and, acco dingly, he w i ing o only one li e a y c i ic. In
addi ion, he does no p o ide speci ic quan i a i e da a. Ou analysis o he
w i ing o di e en li e a y c i ics con i ms ha li e a y c i icism educes he
use o modals in compa ison o o he academic disciplines. I also con i ms
ha when li e a y c i ics do use modals hey may do so in ei he epis emic o
deon ic asse ions. This allows us o sugges ha he combina ion o deon ic