An Introduction To Linguistics Stuart C Poole 1999 205p

iK (1.
3*•
An Introduction to
Linguistics
cAJTAri
Stuart C. Poole
An Introduction to
Linguistics
:1 L4Hr
Stuart C Poole
(WZ)VR41' 155 -14
: 01 - - 0169
.11kWfi(cuP)itt
i71,A1'1/(*)*(Pooles. c_ )—M40'1* LIEFE
f f i ±. .7
ISBN 7 — 5600 — 1869 — 6
.Cibr---1ff -N. H319, 4
" 1=1
PP WM*CIP tom4-(2000M
English edition © Stuart C. Poole 1999
This edition © Foreign Language Teaching arid Research Press 2000

All rights reserved. No part of this publication may be reproduced, stored or transmitted by any
means without the prior permission of the publishers.
This dual language edition is published by exclusive arrangement -with Macmillan Publishers
Ltd, United Kingdom. It is for sale only within the mainland of China and may not be
bought for export from the mainland.
*-11EkAA* Jtklii31WFiffri:*k3-filfluAtPiti_06
aMIAtivIIZA
its*An
Stuart e. Pi3ole*
gi*X
N'irl..#*4410-E kX4±
L;(11-,Th-gENNLIEff 19 q.• (100089)

httpi //www. atrp. corn. al
650 x 980 i'16
15.5
)1** 6 )1* 1 &EPA

A)fez ; 1-5000 fiff
: ISBN 7 - 5600 - 6/G. 789
1111` : .90 3-E
tInUrP)01 AMMAN:1M MAMMA
WIff (Wt- )
RVIm
AqA4 *AUL
KM* #1i+f2 #14-104 -AM #4*
wE1M A44
M .174 (NAVIatry-
)
t4k ii _Et1:=
iiA
_L*)FAI t-4:4
.A t *4-1
frr A MMA M 4*_
VE_41TA
.t44 t Acik *5
A*
ttt TjL
K42t
AAA 4A17;
Yitrc A•.],t
, +sz ft-Ltq. A
FQ
*mi jI A4A-
A
) J1i it ;VOA
t
Preface
This book introduces the nature of language. It deals with the sounds that we
make when we speak, with the way in which we construct sentences, with the ways in
which our speech varies between social situations, and so on. It does so by the
commonly adopted method of having chapters with such headings as phonetics, syntax,
social variation_ Less conventionally, it has a major chapter which, against the
background of these aspects of linguistics, presents the principal features of the
languages of western Europe. You have at least one remarkable skill; you can speak a
language. For at least one language you generally know what can and what cannot be
designated by the everyday words,. you know what is and what is not an acceptable
way of combining words to form a sentence, you have a good idea of whether or not
a particular statement would be acceptable in a particular social situation. As a skilled
user, then, you already have a substantial foundation for the study of language. That
foundation and an interest in language are
'all you need; given those, this introductory book can lead you to an
understanding of what language consists of, of how it works. It serves as an
introductory book for students of linguistics and as important background material for
students of modern languages. Reflecting my work in continuing education, I have
attempted to write in a readable style that will also make the book attractive to the
many people who want to explore the fascinating world of language without entering
full-time education.
I would like to thank my wife Beryl and a student, Arthur IsilcIvor, for taking
time to read and comment on my draft. I also thank Beryl for her support and
tolerance while I was working on the book I also thank Arthur as a representative of
those students whose interest and enthusiasm help to inspire and reward my work.
The
facilities of the University of Edinburgh such as the library and word-processing
facilities have been of great assistance.
F9
Preface by Halliday
Foreign Language Teaching & Research Press is to be congratulated on its

initiative in making these publications in linguistics available to
foreign language teachers and postgraduate students of linguistics in China.
The books are a representative selection of up-to-date writings on the most

important branches of linguistic studies, by scholars who are
recognized as leading authorities in their fields.
The availability of such a broad range of materials in linguistics will greatly help
individual teachers and students to build up their own
knowledge and understanding of the subject. At the same time, it will also
contribute to the development of linguistics as a discipline in Chinese universities and
colleges, helping to overcome the divisions into "English linguistics'', "Chinese
linguistics" and so on which hinder the progress of linguistics as a unified science.
The series is to be highly commended for what it offers to all those wanting to
gain insight into the nature of language, whether from a theoretical point of view or in
application to their professional activities
as language teachers. It is being launched at a time when there are increasing
opportunities in China for pursuing linguistic studies, and I am confident that it will
succeed in meeting these new requirements.
M. A. K . Halliday
Emeritus Professor University of Sydney
F 10
wartc_lifvwTyi
01/.4
-kEJ ‘V-7t.9“4.--01-
X*1.1i4 W4f W4:14
'Z*LIJ,V
Pt R-M-E],1100, tr40--?g41
a4z*Lti4.41-40$4— tnt14*---\fr Z-,1a,
"444-4§ 41:41-,V4-**
—T*
Jr
VV '*rgirto
p4c—*W*
TA,1- F g*-W
-kv_g414W41)(*-/
VM
°1 zt1-
4411-
EJTI
'*g/ *•Ell-Ifil7c .P1f4*(
W4 *1 T tJY
'4J6W'W1.
rOY !at kl '1 T-1 Oct
fig ‘[ J.,WVT ci 4- +
'4F-4-
roni
±W1Ar ‘2(1 C-4 V4.0T
fY4-tv* 4E-
qL1
Y §1-1,V41 kk \#14f,-k, 'V - Y5'S
lc 4.41_4
‘4174. TIL,c“film
/4,kher —a=
W:11P
kt:.46JA- fN/tlki,_) 11?
AAPitit.-14*-4-7#14/03*If1,
__Fi—g46k74
kfriitii A-4, &JEl ith0-iff.#4415-1A
altiff-t,*F0EkrA44$1YJ,
Ei1AffTir
kÀii:ELN4lit1;1Jtk _,
it-161--
14,*kritA*414-,
t t-i_th1I4Tg4-
A3C80
t, -401*111-Pfi-gl ,T
• -jiH;.0 *-Aikr, -it---CWttRAKAA- ifr-
11-- s- t
*,- Kt 44* kr -k*A.iirts X t14
chornsky;
Agoh-iff,(44, l It,4..4.1#3-
t7t4fAT,i1A.1-4t-
VFtikJ.
A-Avit - '6`;-1A -t*fAl (`
,4**F‘VIHki,
i*,- Ft:tiff. A1r51,
igktA45(AA, t4*-Ato-1-
"J-441Ttisr-#- , -thi#4,-
AtZjic 4591,to
T Afnli
T44.4-Mi-fttiksIC ittISIktt4ofdtlki0 PRAtitt
, A4t-5-Mf1±7-**Jko
aintmt-f-F75bAci
*41/. t 144 TIT t ifT- it- itirt*ki

Vr344-14,-A filtAT-t*;/13'AA44,
F12
AiffittUdT*AAAC-*to
Alfri .71-Jr vtlt -I4-k±-4-1/T‘

i L LTrtiktli- Amo
.—t ift-TiAk#0 itYN
449g-rtkirdi-A-4-4,
***AO
AffIt rktlirf -AttAA11*+4-ao

Ailla-if4- A*1t;
*f_M'gAit'f-tfP1-1-4AftiA; *il
AIFF—'K, .3015-
1VV3.*;
Ait*AliMIN-4-141:11411*N-
V404/A1/0J0
tt if*Pk CIDY1 A, Xt
ifr-
ttl-it,11±4113X1-flititio —
ItilV;fY41261}E*1 +;)711n7
FliftEJ&76 zt1J, ih*PiA0
rfi
±t-OriJ
Mt*Ig41411/1* (Akt11) MAA'qM, 1J-ig
f x4ififit',1kIrArhiff*.i19/41.ttiVN'00
.g.tt01_r%M,
idt1111,*44
.qA4--1%-nAiQiklf6411,
titgaigtriAC- -)14,t1*
414T-ko
i#; U HJLAN=t11
JAA-*15•0_'44:4fA1-N;Vfl, Pt-
,&NifiWf*gtAII
iff ll111.tRE**, retilifiVVEtfuR,6„
JIZT
-1121,1AWIT, Ma-.-
tn,041MTIrtt i§ii*
Ti,
fPL >fficAtaiq
ii
't) ; ±OttuT -ISMitc, ntiq, itZAMAT-
TRA5KM*ifr-,
itA,ZKOHOirMfir7.
Aa5_14- MLE-IMA
,1,031E4,
—#M0114]*.%; 111-74-TAR1110/1*.?c*MPA—#1111A-1.9,Azo
AlfiffiCr*,
gffIft, iffrovtiEtt, 5A-0111443I4AX{M fA0

TALMW#IC AAVOVS3S)11J,RAA#'&-fti3tt, %ffniff
M'AffrLk qfit034010iW W?
EiP1 P1-1, Atf1-*IY,iPf4-4AJA,
A.2ilf*If*Sic6TMEIL
iitgAt Mf*, RA-JfZ; A-4\*.ft. (A), tL3A-
)Q4411. *)0 R-q,AA
F14
tftW(Mit) A- ± if X
X;
•Xo F -p.A.F3-61,ifftr2
A., tivr„i liff ;
S4.
._1016AX*,, qiCf-
EMX*,
_t_ort-
ifftwqx*,
YLit'rrAt f%•?,
PkAint4E0111-
A.:
i.t,;-
W:5113frfuitrimit.
-ORY1,1#-
k,
qJX*0
EiyA-
TLESS
1'zifiT4
itt A-A3Cit3Z ,,
iTc
I .AI ,
igrMtrAnz.,
imc*, iwx*,
ti-&-iff7A-t OKA,
AC\ ),ii_suffrrto
trAill'EVT
SitAltlil-q-AA.-* -
1111-MiC.,‘Riff
111.0
tICOLM "MtiZigg"
*FM A11-
4.6"t45111--
ftT4N„
t 1r- t .fFtRAA
Eiy3trt:
*3111i4ikgW*10,1XA-it
lUE, ra.*
*fftIU f t IJ k011EA,
21“WPkr
aptt-t?g, )41-F-
ILAFYit ir-ta-t-?ft,
Al'4.P,OE14A-4
vtA itteu...tiai., a It
izILTIMUM A-
EtJ fl1 Tc
igt-flAiji#11 .AW-4011,
iMttAffihEAVXV.;
PJA43*A-
4*±JAT-Xlt-
41t.,
qMiA,
Aftifx*XVOT7'**M**3Mito
V.,L g1A "Off", h4J1:3
RIMIWriEit
ri f[14d±_hA1t
-LT:11-i*ffiTLikt, criA4E-
Plevos-Aa,
.1917•NAO lgi'L;
itgi/PNVWthifil*ift.,
Af4T
F15
PA)Vil4A-1Aif-A*'10,*0.14A“t, SfkrJL ilT.TM4t—t

f1-9 T—R3C--4PALLS:hE3.t, t73* T:giAjcAl
Sociolinguistics is a branch of linguislics which concentrates on language in

society, in other words, it tries to examine how and why people use language as they
interact with other members of their society, Sociolinguistics examines variety in
language and has shown that language is not merely used to communicate ideas but also
to communicate our opinion of others and of ourselves. Even the simplest utterance such
as Hello! can reveal that the speaker wishes to be friendly and informal, and that he or
she is probably British ( many Americans would prefer Hi! ) . In considering any
spoken
communication, therefore, a student will notice that a speaker's language
reveals information on his sex, appropriate age, regional and perhaps ethnic origins,
education and attitude to his listeners.
Pq_PL a t AO" 1-*%,

iF: Rit? VFL
"Hello, it s me." That may be all that we need to hear to recognize somebody
who rings us up if we know them fairly well, for no two people speak exactly alike.
His or her speech profile has been shaped by such factors as where he or she grew up,
by the kind of education that he or she received, by his or her occupation and by the
same range of factors in respect 01 his or her parents. His or her speech profile is, of
course, affected by whether it is his or hers that is appropriate. It is
affected by age. ft is affected by many other physical differences that influence the
individual s acoustics. That makes up the individual's voice quality. The unique nature
of a person's speech is his idiolect
Tf(f 1r F,
T "f4-*" Pita.M11431EM--ik.4-E, ' C14-0A3KMM, t
APNINt EA, -5!P9f6, WrEil *-MX-ORA
itOlig*Afitrrictlirt'
F16
Li
Y*TI %ANT „
V*1041-
Y:K44.0.1.14
Vir0-44- EJ,V
'4&11ift-
--
774.,RU
LiSi:Wg41- °M-f1l-
401±1061W11 41)
„RgAlv„
/ 14;17-if1-=1-4W-
4c7EY1
t(41MY:Wfat11-
11[1WAigDkYfistNN
Eil))KVT.IT,ciPPLYAlEc.44Y
"01-
YIVTr ‘$'?4 ta-rilvopo m gv.
L ti4-1d-
Wk
r4q- owq-r*AT
‘wg‘ovwmtuitovo
lisLUANia'PrOhn
V-
r41440 - .i0(
'609ita(W-
L4filM411-
‘TA,W441340
°W111 tY1rfliMbi4W
tinfloW4ftrO.
'AM:iiitaf5114titP4-
4fia
‘fkmaromowit glistiw gaii.ro-
twr ‘„gh
-4(—wto ig* WZii--*H4f4tcit_t
c}rAltil
WC, V-14- *fikici*.*4eV/n°2/1±1
OfkitirOti*V
“riLp .-
WkZ'f.WISIKAZtkithiM4kiJiar-.4.1-A1141.411?ftMS
L'1141-%.-t±iffi,*-1c4i--41431T1f---W-?ir*V
araza"lt
tIUj V-11
f*V3iftt-4:iigNWY4M41
'fM.410 '4P4WINW-kft4VVZ-41144Y-[Wtil ‘)V2N
V- ' —D
W1-*CW*TVYAM-0L tfAl•WaWYW1441 °_11-111,4481rx
.0.VAT‘19A014441)1411 'V-I-J4VITEU4-**
‘7;.-14.7ETR-31- 4-W
E4i14P;t4;3-W*
'(5=V .661WWWW) WV},'114 A5:44.0.4
4fg[t141J5=
-IgW41)5;) "n11-0Thyig1UM 4-
MVIIMI4t 'Aga,M1141:tfAW-
14 ‘10)-
tMV-E-V 1-ig YNc (90Z — 86T
'cid) * -42-4-Ea-r).11-elitilii-PYIA
ffjf— '(Axessom)
ViWPU
*V4-W :t4 LI '6*[4 -f4 VI.n11,9.-q_JZ
cirsqmiY
,t11411-Y ‘iv4T.441---itc*ruktwris* ‘Yiilft.&3kYrffUV
$0 f-t XE*4.111f litAifitityjnito

An-hiCE, X4iiltEM
iACY.:illitTALI,
ifT:54:?\
rfifiNk'?ff.91014?tialatibio
ttir, TRAKM.Aftifilaqifi-g
ATH, T11114±-kgq_knAAft)44*,iflli,
*17,ifffinwf-tIti±m5 PIT.Z4
Mr.*1
AA ifi * M - -6
spick and
span 4
.4t6c D
i1
/j\, kick the
bucket -4T die
at--0.10
Tiq tirn
unhealthy 11304)-un
(morpheme) c, tiM J1.5!4
frA 'NA
ill* On unthinkable
'II
goes, went, gone, T—
going,
(1exeme)
a ifl give in,
make up,
look after
gi—
tiAlls:i441
ifRA, T1
fireman fl J
blackbird
ittArlUf fk
R43 J
/AfFiTNWN/},
ifi-M I shall
nk,Z10-
1:411=ifl
Je donnerai, f
—
tiCIMT:
clareo
iiA4M41,3Kfl'
kili4igr4f
ZAVIVA
*t6ZIMA.fgr
VA?
mitAiii-
mromg-0
F 18
XfA•g-TZADAM.2, —1-1±1AVOYr.A.A6913AAA,
L'EfTL *IGActtqiftlAN
iferVUz -t'iMi..-4; a-
OiTtAPAgti'
i•Niff si-gj**ff lf
jillcr-
VAR*4A-
"ff)qfeAdil*",
trn crash,
crush „, OA,
ilplq
thittkAR4TEPRU4P4., --
R.11
N, A
KfciA,
MIrujtrA-tei,
R-FM-
iraliZIEMM-
F*M,
-,,AjtItZ -44QM - ,1tP1 *A
Eci-
q4PV}W413141±,
4r-LT: fit Ji
lffgl5firt fff
paimelo (il\*
-91Pri (CitiV)11MMT);
ArAiateif- +
P14%
handkerchief
ZtrififfACWIX?
OPZ
91A
A' Mk (conversion), gjir4
fE
M=fIllt_riLW
tri—t-iyAL tzu
discover .5EOZ
undiscovered 4„
&thi Efi1j4 editor fil t
ffi3KA:
IE Tik IT edit )Thlit
j "'w",
"MEI" i "
1E1 "
ci
ift3ti EMji\-14 MIA
tU shoemaker, A`lit
fire-fighter,
fireman
lf, ti Ji T, breakfast
brunch, H ± u came
camcorder. +1414 --- '3K

floaftirli
b‘IYA0
Ri6rfq "Nt" P1A—fittita;

31E0 "Va", "fq
*nf- F-13A0 413M11#1.41111 t#thiliiti4/41, A4§.(44,
AuXit(iq honeymoon n -tiFA lune de naiel.
F19
iMykid N.JO, 51- it Niqt" 0
=.-A.zicu kgEirx*,
nAE0 - Pfi4"6 *tt,

thfffUlAit
-,1-1:R
IPIVOIStt%0
R4c"efift6 A fi if
f .oItff4trtliffA-q-
J1A9MtR44lii Mfr-Tg.
klt[i.X*J0:1:1ii&A.111
-1-1L-titt tiWo M ,k F tn
m
TiltfiftWW-'15 V-4%, fA:ffi., ij j
AVL, i & Wk
(semantic
range) ,
tirMiff fit
uncle ric'jiTir-51.. bta "
UR" Jd4 "J.t.tt rf) -
PHA tiMIA —
ifr34A,-41 "NW, "IRV" SO
Wlif4,1AL
(semantic field); iiijejko
"gi.", "M", "0"
fltilik
.4n, "4"
ig4A-
NallTh?
""
"f§" W ? 4Mtt-11-4-ittlflIETM T.
"4" i _Ellt/fd•JA (h
(IR-ARYL "A
i-FVÌci‘i ( )c a-
,V3p
Mt3H-J'.4flt,
glIfttirr—i--
imosoviE, un
"oc"
acK4—tbAvEci (s
ilraoifise.,RwAlrltisu fu
a3(- (antonym) *MN
A
fElfIr
TOVETtIM-
Lf113t.W.LP.KNI
P4T,
-M!i1A --:VVOcj ( "PE" j"th"),
t ( t" "4\").0.
*WO iff Y., DJ if (it bl
hair itfil aiffiY3
"AA/3\A" till
AA for the
sake of
appearances,
Vrii
'Fait
"tTitAlEA,4"0
iwiniqiff54.V.AtNk --tifir,N
,X6
( poi ysemy ) ( homonym) fij *Frit.4
SZAlr
chair )1E; "*T", " w
A3E; " A
; meet meat f1r. wind
mt ±j'iMlt
wind i
F20
Mizg* Air*
ifi*T-fFtife ni3W1 6-0 TiffIZWii9rEd-,
WFAAA K610310t1A-fflYIk'Ac. *404VOikiltAftftiVifire-M-aiM, VAIR

(IVWFIN),
(-N ) V,
&Òfn (AJAsputxvymt,
ift),
pr-Line.)
M-1-
741t4tiiiA44-
4k,)
*ARA. ..-30{t, NTICI;
WWII'
11 fk1
tl rfriII.Ttflo AlOTIE#M., ANK;
tI -Ohk&-
601M1,11.1**,
ft-NAM XQN
WiltM -**0
t4 Mt4)%, t,1 tA1frff,
ht ULI (L ) (plosive).
NAtitAligbl i;#
ft, LA..41.1t-ft .,i49t.kITIA,
(fricative)
WctStlf, (affricative). -t-
,N31111M31.
ftI.PEMIN4\,
4-
;`)Elf'111g*IV,
gt36/4M-K
(glide), * [11
*11
[wl
R{MPRAZ7—
ilni4
fti1F+Rad
ifit&
TtAlziMfri z#
f.=, awvia
LPi *11 [bk
57)q7fS ft fir - W
writ tOUiri A-
MYEE,Ji Ni.Ftf? +-rt6 , 41.h-rt. g; Irk03F,
NAlliff2a4- V_TtiiTRIE*JER
1181Mfff,
&tTlec%
(!!U fire 111VG-6 )0

A*TAitiALLAIK,-=0141',ttr
F21
STFTCkiRitii [kJ p
AV4NARIETPito
ME* ig,,f3y—*Es
]-,
4
:0d( spr-fl st), ttrrai#
Akiivituiw/MMT4-aiibLZIMI
M%14M0
ft-fTiTif4-444, P.,(
(phoneme)0 , bit /b/411
*11
pit
itg
4-;
MA,
it>/itn`-t4ito TINVI.5fit /z/filisiTa)Aft54,±Ah_Nij,
lowttciav-4K. /pi /p/Ok.1210
*ializritikco/p/411
[11-4CAA E/1/ ini OVVAA-f2k
.
—# CUM IYJ *032R1,
it?4ft--ATA-VA o stape
zciapeo iA A-EfTMWsi, /z/trliizi,
airot*Twoz*%c
31itill*0 Ai6A-AiPAYMIJERI, kiti*TIRA-A-NteLOW,Z9.10

WV* tititl f-01.1T WI Mft--'k.-A
O ARTi431EM schola, t%**111StftliEricii*W1 T

f f i Mill*11 5115finOtWtEFIti*tii-dt-TG • 3-)11_1
AT. ecole ll escuelai ifri=13)t,T -

kouitio AAR, ifrM
j'f-F.A
girlfriend T 111
1)1T garufurendo?
ORM R OM- irr -f ilk trI
t
t O
IMita7rfR3J ki#4
Eteiwvittio
YETIMriitgNA .
M IA* Z5-*
--I-A V; -fr - ;
AAA - A-
t- iAr
F 22
EZ
‘-(01frAi4r4(1144-(crILICAJ,
tiMW M *4141r4hrog, °01.1.1r1-r) -

1W‘E'(
'(11191)
-F it 4 Vg Vtf °
4
W a-4 R
VE1:*
VqgiafftiFifiNtliM c'-
kNIRMII.-&-
81.r4/611A111164TEEll
41031f
Ps+WA-
Mht014311Wiraiti3&!--
4/-V W °V1144-k411-
VWfil
kY,--- iffn ‘470N40-V- ,
1444raiLi1i 4171AW- Y;I
M4,—
kotiAaT kNW
•3;arrli461V_
ot VEr_if
.10411-1Y. [61- ‘ri4
(4WTta*-4
'WE) 1k1REA*1- (
UikiliNtV4
sW45+5134YRINI-7-i4
*YL4)5( iiltRTWTs gfWV1-•4.i_

glY1K014114
°A:C-4t1W-
MO:kifg-
i5rIkalfftrt7LW
omrfos-w or aal..li -
"TWITA. 0
u
'fit 4gTir#TV-VTWi °Q1VI ,
VU3kEilra.k 'rig
T(14 TItrikiqi-3--Yiffli Y141.1T, EPA.
E11 '2'
L W °'4Lkil3tlqi- IN 14-
3 KaMTI.T,1-
74TV
W54 'XV[q.-144)44(ITITAg E
°(13Dim5Pgici n4) 11,
tu niq WL41N4.-4-NP-H-W
owwamo ‘rAita(4'-
w*Nitzli:px/xfiavvigu±
Nt !TA -4.71frkorm von
hilry9 v*4417.s-
x
(21-193qpio
IRIM
TIci-Eq-}3 ENT °(uReios!) (
4
2
°(211Pull!inik122-4 E
=i1-- *12f6V.
k145151.5MViltii- cHU_k.r1)1
X UJI_kg4i:L1411.
'-f;3Vin
LCIPI1(E014W cla- r114-.4-1 T
tfrAI-1APIPVM111Ortticia/V6-9",k*J$31E,
Ii
II
FF VIET-
tWt4i —
fAilri-t*, ft
Xft T
M,ckifflar410-11NWM.
*MMtig, giiffMTli=rz144Fiff*g
tL..IiitRXft
IA_L-
Miffiri=94.11444*TAIVAT
J-if u ;AA
kigolmgri.nitritx:c
XA,
EttVilc
A4h. 121
EM÷A.Iffia
( idiolect) 0 M 101$ t uitti *
Afn
TVITHM. -go
if V1141Alitfli "It'f*".
KMMIFAR--- AAAP.i_h#1VN-
iNMOMMO Rft
HA/ +114Palgfclt —
irfit
*ER 4141PA*414 01*gAtf TIME
AF-R04111 0
t I titt
—a,
AtAITA_R*0 -0,41E4--fati6r*
a Ott tE bus/ ) it E • ml -
bus
(isogloss); W- , )14 .-
Eitiltr0 11(14JVI.N.o
--iff-tftrliTIMY7- TR—
144.1f1Dtti3i0 M-
4IMM
--f4rEATit2A*71, — -L. tf*
fr11175
EnAtOANT- V±A. Witt
ffc, i;101A
fiElfCiiiiMAgf4rIff71
tfM tL A P * - Wit
it it
OX2AZA: A
ltlEiff4attf2Eiri
't AVE& 1,4A1 t I‘r E
aliMA.R0 WlYi.T.Prio
NiCt
rffi13131F4MMIAA-AR 0
F24
XA,* ±-A144
iY3 iq!a—#:1.... Aft .*15). *.fft i#it FYi
Af414#:'kIll*railUEriq
""ret" nii4±=k- f*ci
il?%1Attniv
.ZE4 401/4 E5M, ft
friMiWN4tlifn
ftctIMMI. IMrlDJ:11
iM 4#0
Cgiffri.210E-ft.RM— T'" -
i*M4, Aldi.•4P:ti'VEETIt-
#4MgMA,V , iC Yr-.Z,
TA N.
ri2AVA,
)111-MR4
*45.0 )8-AVV149TA,
i o ?MF,
3Zfe.tilibt5MC
r-ci JAIZO44 {Ctti-i?fMT-
t.,fr-
WWtS911010
A-±Wtt.-
SM-
J34IRN4E-
11-0 A-
A.At91±S
EMINAMVIS iLit)j2.. k, S,
EiYJI"± ffit.±0P_VA-
3?
ifffftN±WttliTkirFINc. i M CM74,
*1110ifig
WiNit E0Mri iEff? M o 95±IVIT- Mfit
At,
WRE4-TATT4—#.
NI t A Mt A-4 *;
I:Mft
ZAirt-
PMSAA±T.91
f4,
*.12,11 .*; TAif JL
—1-
AVTI1PWW-
FWeAT
Z'IN,Tteit* *FY141 A,
A.0,
* ( register -t- 1,014M riLl (
) code-
switching)
AÌlliriAiliA/V;ATA W FfiMO,
fttInttifi
(taboo).,
r.r."
MY1-
144Zffift.t, /AA
T
( euphemism)
.3 k ig Elk
T ittIA flAQ
TA,
CfiTWkOrf iiMPRiittl,
T- AfiX4-
4411EIM
P`11 f flifi (slang),
"Mr&IHRII
t IM± -A-11=rit fll VI Mit
F25
M-1-*
ilittEltfElf$A$
frEiRA,, -1\-- s,f$J, #

fr4:Ziffki *ZHICETIPìtY111111- 1
T (synchronic linguistics);
liFFAiiiiTtE/FRTIEitPAII.:1tO,
ifrgr* (diachronic linguistics), ,INI--Th._sUflcg1:A
ifiglfitrSiffiZTIMPIL.C? (h
bird), trt-54-* (a naddre At an addre) W
Wg-ft,
mwiAr*4E)AnEli1til*,
Rif/i4E(W4, iftZtlAi0 f4
ii4gtf(--,/ Ltzt-4t? h" ,e.t.M_LO111*1
NP MIA-114,
rAfETC#N*
(redundancy)
(frigi) t il
V
Fht, itialtift6tTFA(41,.
Aftiflif*r3Cift4010ill
f11 0[14, W iE1114.
frl-tiffVFVtA91-fR, AfiANAifr-tgi140#13Z-c,
tfliff*MtATiii, xiiifriaiff*oEL i§(}moRAgel
4iTi-5641/t/friik/OT
4)AT/coil/x1. atgatIt"ftA4liti ri$W1W,Y1A5 MIELPLIM

MIMIC atb.At4fiF4*I1
iAVZEPV6i4 INtiltili-CiITI5.1410frOtirtiWITT -FAO, tA

--EM5PUillrlfA-3-}gihZ4N
113A)fi_stiC4* Mii*IVIT)
MO, AMEVF151.—ftift-Vivial3)145X Tigirtlilioi,t,
fMTVIIif-trAAni tliRtN0 iS-14 131-.4•IYIR CI;

*i-541WPA,
'10W irttRPOITiA- :?7:111-tEHVETI0
A-16
JAi illA11-6- --Zflitiz —fil.A_T.Aiflr (creole),
t'EAE[- htEYJo P4fliZttfviff-
IXIMilirCrOkiWill 11141.1141*
Ut,h- —fr1/ AiA- ( pidgin),
ft VREVR,
Tr• itONWIZT3K, iCESt- E,
F26
T-1-`0144 T AVAiff.
*AIN] Vitt Tk fireUgiL trig, w
iqff
aTit-ititkiu eitEikrh7c]
*iff _hk.11 L11,30 H j
- i F, E1 7
AAAIFVRAIY04M141;
tillift4 if c ir-gititUEfr )
TiffikA(fot-A
ift- i, ‘'*-111.)A3)- -,
iiMIit,?149.7.111 Vp54;11V71111E
sMiltilN#C•fia-MAXiff,
VLLif4“tift.
NUL M ifefriff E
--- Mri
fpli31.1EM*APJX*L'
E,
wm-Atum-Aa.A,
Wi1YlifighZT*iff0
E
ift; - i-N; .Zit7t
Lf*JAARG.tril ift*
—APNA-MMI:ft, =A441119[2
IL JT
WSW.-iffX3-Mjit* P4M3-)3
ffflii(01(gfirit-fri&VZ- fff03
mtiffit-h1 X t)i,
itr ,15$iN Ltik141', t *at,
+111ARANOMPWOFV4111
xMlIsIN.*g ifitftgifi-W
tfitit-W-
01.4ExciVrgigtERM,
EWIik *OW* 0
ff*Z-
ft.11Y6.
v4-T.*Patfl •A.*tO, fthri.fttATLF I
ff*fi5 Eli
***I*, WA)tififiNc,
iffflAVIX.4, uT WL
, j JPj
.5-Y)A7- ±, NiirfuNiff;
frITATAirfo
Milc14iW7IsziaRRUM,
I-3K1k F1 t%Afu
-
F27
q*.)2jft-, iWti ±ifrfah.,11±tUtifii

=At/AIL
" , igiffcriR)Q14A4441 rf; fT
VI*
iii5q14k.*A1PRIVSJ#W.,
tvfm04YA**no -&iMYIffILAT,
V.AJM-fi
MqVifriVT:T*it, -4T-A-Ltt:_elA.Xf
A*PMMT*:
Ot, .7)\ì~rffn/D-WARY2ii ,
RAIATXTEPRIcifv*Aianrr, AWERifffiltn
M+—* 3Z*1**
±11 ,SC ZIl'-A•17
Itte, TirkEIgilt iff4A4$1[

(morphemic)
C#?.., NUL
A-61.(1 .4rf-9-
o4Trf-ir-
, VP-MAJVIE
:tEtOrri*,
Ntoll fit
(phonetic)
(syllabic)
aiff4bidC4fC iffq,
AVAYIJAIX-
*5TIA, AAti*
Mit4Pci A-AUF (.A FIN
A4);
A*fillitMA- ,—::*1tEgO
(API, T.)
( V±Tlf1);
Etivezrf-9-
om (tam,
ig).
HtAjYRM*-- **, trk)11301
ifirttM*3C
Y't±
Tg -is*THMI C fc4fRA'iEf-
fi4f413Z *_SC
iztA{,
bk.*114
AIffiflAbki*A1-
Alt1LR- -4, lEriRAAiff
At
it4:P*IrEf
AVE1- A*filcfMitfrk*ific
M-t-
AIMV±4EA-
INILIP-**
(M*iiffit Kr11-1- -Llf,
011RNifr\t
iMq3M ill
RM. /1.
F28
Contents
Preface F9
Preface by Halliday FlU
A4 Ft 1
F14
What Is Language?
.1 The Significance of Language
.2 What Is Language?
.3 The Functions of Language
.4 What Is a Language?
.5 The Elements of Language
Summary
Exercises
Lexis
.1 What Is a Word?
.2 Where Do Words Come From?
Summary
Exercises
Semantics
.1 The Function of Lexernes
.2 The Meaning of Meaning
Semantic Range
A The Definition of Semantic Range
Collocation and Idiom
.6 Homonymy and Polysemy
.7 The Human Element of Meaning

.8 Pragmatics
.9 Discourse Analysis
Summary
Exercises
F5
Contents
Phonetics
.1 The Organs of Speech
.2 Consonants
Phonemic Notation
.4 Vowels
Sounds in Sequence
Summary
Exercises
Phonology
.1. Sound Systems of Languages
.2 The Phoneme
.3 Phonological Rules
.4 The Phonology of English
.5 The Phonology of Other Languages
.6 Suprasegmental Features
Summary
Exercises
Morphology
.1 The Composition of Words
.2 Morphemes
.3 Derivation and Inflection
.4 Productivity and Word Formation
.5 Problems of Morphological Analysis
Summary
Exercises
Syntax
.1 Syntax as Opposed to Morphology
.2 Word Classes
.3 Constituent Structure
.4 Noam Choinsky
.5 Syntactic Forms
Summary
Exercises
Regional Variation
.1 Variations of Variations
.2 Horizontal Definition of Dialect
.3 Vertical Definition of Dialect
.4 The Nature of Variation
Summary
Exercises
Contents
Social Variation
.1The Social Dimension
.2The Standard Language
.3The Urban Vanguard
,4Men and Women
.5Power and Solidarity
,6Registers and Diglossia
.7Taboo and Political Correctness
,8Slang
Summary
Exercises
Historical Linguistics
10.1 The Diachronic Dimension
,2 How Language Changes
,3 Why Language Changes
.4 When Language Changes
.5 Divergence
.6 Convergence
.7 Pidgins and Creoles
Summary
Exercises
The Languages of Western Europe
.1 The Indo-European Family of Languages
.2 The Germanic Languages
.3 The Romance Languages
.4 The Celtic Languages
.5 Finnish
Summary
Texts
Exercises
Writing Systems
.1 Communication across Time and Space
.2 Morphemic and Phonetic Script
.3 Chinese Script - A Morphemic System
.4 Roman Script - An Alphabetic System
.5 Allographs
Summary
Exercises
Glossary
Guide to Exercises
F7
Contents
Bibliography
Index
215
3Ci -hZ i
F8
An Introduction to Linguistics
and now; this becomes clear as soon as you try to arrange to meet a friend
outside the town hail at seven o'clock without saying or writing anything. Language
allows human beings to learn and adapt to changing circumstances far more quickly
than would be achieved by evolution; a word of warning passed from father to son is a
much more direct and efficient way of adapting the species to its environment than is a
process of natural selection whereby those not inclined to do the right thing are
eventually weeded out. Human language is infinitely versatile; honey-bees have a
remarkable way of letting others know the direction in which a source of nectar lies
and how far away it is by means of a 'dance', but they could not begin to discuss the
merits of other foodstuffs or alternative ways of constructing their homes.
The power that language gives us is, indeed, so great, the Bible tells us, that God
felt obliged to restrict it; in Genesis, chapter li, we
read that our plans to build a tower, that of Babel, up to Heaven were countered
by God putting an end to our common language because it gave such power:
And the Lord said, Behold, the people is one, and they have all one language; and this
they begin to do: and now nothing will be restrained from them, which they have
imagined to do.
Some would even maintain that the very contemplation of building a tower
reaching up to Heaven would be impossible without speech, arguing that even our
thoughts are impossible without speech, that we cannot conceptualise without the
framework that our language provides for perceiving the world around us, In short, the
nature and the dominance of the way of life that the human race leads is, despite any
obstacles that God may have put in its path, due in very large part to the fact that
human beings can
speak.
L2 What Is Language?
We have just seen that even chimpanzees lack the physiological equipment required
for speech as we know it. That is, of course, not
to say that animals cannot communicate, Animals have developed
ways of letting others know that they have found a supply of food,
of warning others of danger, of attracting a mate, and so on. Such communication
may be by means of sight, smell or sound. Given the huge range of forms of
communication, it is important that before we embark on our study of language we have
a clear understanding of what we mean by the term language. Our first response might
be that language relates to communication between human beings and not to
communication between animals, and that is certainly a useful first step towards a
definition. But are applause in a theatre, an expression of friendship by means of a
smite or attracting somebody's attention by means of a 'cough' any more a part of
language than are the alarm calls of vervet monkeys which distinguish between snakes,
leopards and eagles?
One attempt to define human language was made by the American linguist
Charles F. Hockett (Hockett, 19.58). He enumerated a number of features which, he
argued, constitute human language. Other communication systems might exhibit one or
more of these features but only human language has them all. The 'dance' of the honey-
bee which informs other bees about the location of a source of nectar meets many of
the criteria. It meets, for example, that of interchangeability: any creature that can
transmit the information can also receive such information and vice versa. It meets that
of productivity, the ability to vary a message to reflect differences in the circumstances
concerned; this is clearly necessary in a case where the source of the nectar may be
constantly changing. The dance does not, however, meet the criterion of cultural
transmission for the bees are acting instinctively, not behaving in a
way that they have learnt from others. This last criterion is particularly associated with
human language for the one stimulates the other; we acquire our native tongue by
cultural transmission and it is by means of our native tongue that we receive cultural
transmissions, that we learn and adapt. This is the spiral that has driven human
development
So how, then, might we define the term language? An earlier American linguist,
Edward Sapir, gave a definition in a book published in 1921 (Sapir, 1921, p. 8). He
supported the hypothesis that language relates to communication between human beings.
Just as Hockett was to associate human language with cultural transmission, so too Sapir
considered that it is 'non-instinctive' and voluntarily produced'. Thus for him language
does not include
such instinctive forms of communication as smiling and cries of pain. His

definition is as follows:
Language is a purely human and non-instinctive method of communicating ideas,

emotions, and desires by means of a system of voluntarily produced symbols.
He goes on to say that these symbols are, in the first instance, auditory; thus language
is primarily a matter of speech as opposed to, say, sign language.
The element 'symbols' reflects the tact that there is rarely an inherent association
between a word and the object or concept that it denotes. Any sequence of sounds can
serve to denote an object as long as the speakers of the language concerned make the
same association; we could just as well denote a dog using the word perro, the word
alien or the word hand, as the Spaniards, the French and the Dutch have shown. The
element 'system' reflects the fact that language provides us with the framework for
generating appro
priate utterances rather than providing us with an infinite store of
ready-made utterances. We can create utterances never uttered before; this may
possibly be the first time that anyone has ever written the following: 'An elderly
mariner leading a monkey by a chain staggered into a bank and asked a teller for a
glass of whisky and a banana'!
For comparison we may look at a definition given by a modern British linguist,

David Crystal, who wrote the following (Crystal, 1989, p. 251):
The discussion may be summarized by referring to language as human vocal noise (or
the graphic representation of this noise in writing) used systematically and conventionally
by a community for purposes of communication,
Thus this definition also proposes communication as the principal function of

language_ What it does not do is attempt to specify what is communicated; as the
British linguist John Lyons points out (Lyons, 1992, p. 3), Sapir was too restrictive in
this.
Nor is there any element corresponding to 'non-instinctive'; while any particular

language is culturally transmitted — an infant acquires
What Is Language?
the language of the society in which it grows up, irrespective of the language of
its parents — it is now generally accepted that humans inherit a predisposition towards
acquiring language. Indeed the modern linguist and cognitive scientist Steven Pinker
uses the word instinct to embody the essence of human language (Pinker, 1994, p. 18).
Whether one considers language to be instinctive or not depends on precisely what one
is talking about. Language is instinctive in so far as we are all born with a
predisposition to speak, we all acquire a language without tuition and when we speak
we do
not consciously convert our thoughts into speech. Language is,
however, non-instinctive in that we can choose what to say or whether to say
anything at all; it is not instinctive in the way that removing one's hand from a very
hot plate is, done before we are
even aware of the situation.
Both definitions refer to the element of system and both allude to the fact that the
association between the words used and the things
that they denote is not inherent, Sapir by using the word symbols and Crystal by
referring to the fact that the association is the result of
convention. Crystal, in referring to vocal noise, is more specific about the
principal way in which the message is physically transmitted.
While, as R. H. Robins suggests (Robins, 1990, p. 12), there is a danger of

definitions of language being simplistic, it might help us to focus our study of language
if we try to distil a definition. Such a
definition might be something like the following: 'Language is a
form of human communication by means of a system of symbols principally
transmitted by vocal sounds.'
.3 The Functions of Language
The term communication, then, can be used to cover most of the function of
language. But the function of language is varied. I've got
a knife could imply that it is now only necessary to find a fork before
one can start eating or it could be a warning. Do you have a knife? could be an
offer to lend a knife or a request to borrow one. If the person we are talking to has
been ill we probably want an honest
answer to the question How are you?; if we ask it simply as part of a greeting we
may not want an honest answer. Linguists have different terms for the different
functions of language. In the case of How are you? used just to be sociable, for
example, they use the term
phatic communion, that being the use of speech with the aim of establishing or
maintaining social relations. In such cases the
important thing may be simply that one says something, as saying nothing
might be taken as a sign of displeasure. An expression of emotion such as That's
fantastic! or Shit! may be called an emotive
utterance. When an utterance is an act in itself, the utterance being spoken by
somebody with relevant authority, it may be called a performative utterance; a
bridge, for example, may be officially opened by some dignitary saying I declare this
bridge open. Such matters will be developed further in section 3.8.
IA What Is a Language?
Having considered what language is, let us briefly contrast what we have
hitherto been considering with what we are concerned with
when we talk about the Russian language or the Arabic language.
The use of the word language in both cases clouds a very significant
difference.
Language, the faculty for communication by speech sounds, is a universal

characteristic of the human race. But we do not share one medium of
communication; Russians and Arabs speak different languages. A language, then, is a
medium of communication specific to a society; it forms part of the culture of that
society. Being a feature of the human race, 'language' is inherited genetically,
whereas we acquire 'a language' from the society in which we spend our first
years. A child born in Russia of Russian parents will acquire Arabic if it is taken
away from its parents and spends its earliest years in an Arabic environment in, say,
Jordan or Algeria.
.5 The Elements of Language
If a young child sees a dog he may draw it to his mother's attention by pointing to
it and saying 'dog'.
Even such a simple utterance involves a number of facets of language. The speaker
has to recognise which category of the world around him the animal concerned
belongs to and he has to know the label that attaches to that category. He has then
to transmit the sequence of sounds that convey that label to the hearer, thereby
generating the thought of a dog in the mind of that person. The study of words is
lexis and that of meaning, of the relationship
Language is generally considered to be a form of communication between

human beings by means of a system of symbols which are principally transmitted
by vocal sounds. While the faculty of speech is considered to be inherited, the
system that we use, the specific language, is determined by the society in which we
grow up, it
being culturally transmitted.
Communicating by speech requires symbols to be transmitted orally in an

order that shows the relationship between them. The form of utterance will be
affected by geography and social factors. Our utterances have a variety of functions
in addition to commu
nicating facts; we may speak to express our emotions, for example,
or to reinforce a relationship with somebody.
Exercises
Write what you consider to be a good definition of the term

language. Justify the choice of the elements that you have incorporated in
your definition.
1.2 What do you consider to be the principal benefits to the human

race of language?
1.3 How well, in your opinion, does the word communication

represent the function of human language?
Lexis
or
Are Slithy Toves Smoothly Wet?
.1 What Is a Word?
if we stripped language down to its barest essentials we would, we might

intuitively feel, have to retain words. When a child shouts Dog' we have little material
for a discussion of syntax or social variation. What we do have even in. this simplest
of utterances is a word and, as the nature of a word implies, an association with
something in the real world and a sequence of sounds that conveys that something to
other people. If we set aside exclamations of joy, fear, and so on we can communicate
little verbally without words. It is words which, expressed as sounds, convey the thought
of a thing, of a concept, from the mind of one person to the mind of another person,
providing, of course, that these two people speak the same language. For speakers of
English the word dog, conveyed by means of the sound sequence idpgi or, in the case
of many Americans, /dug/, transmits the concept of a particular kind of animal. The
word large, conveyed by the sounds /1a:d3i, refines the image. But what idea is
represented by the word spick? Very little. Only when that word is accompanied by
two more words, and span, do we have a symbol that represents a concept, a concept
similar to that represented by the word clean or the word spotless. The meaning of the
phrase kick the bucket can, in the literal sense of striking one's foot
against a pail, be arrived at by a summation of the meaning of the constituent words,
but in the idiomatic sense of dying it cannot. The
verb is variable for subject and for tense as is the verb die, but otherwise the
phrase is set; we do not have the leeway that we have in the literal sense to say 'kick
the pail', 'kick this bucket', and so on.
Lexis
morphemes, morphemes that cannot stand independently as words, such as un-, from
what might be allowed as a word. This definition does not seem to admit words that
are compounds of other words; it would admit fire and man but not fireman. But
elsewhere Bloomfield says that a free form which is not a phrase is a word, which
would admit fireman.
We may find the sound system helpful in determining what is a word. In both
black bird and blackbird the principal stress indicates where the noun starts. In Italian,
morphemes that are word-final almost always end in a vowel; this is not so of other
morphemes. Hockett turns to the sound system when he suggests that a word is 'any
segment of a sentence bounded by successive points at which
pausing is possible' (Hockett, 1958, p. 167).
In the end, however, we may have to accept that — setting aside the convention
of leaving spaces between sequences of characters in the written language — it is
difficult to define satisfactorily what a word is. The range of meaning or function
encompassed by a word varies from language to language, The French equivalent of I
shall give is represented by two words: je donnerai, the marker for the future tense
being incorporated in the verb. In Spanish only one word is required, dare, the element
-e being considered sufficient to indicate who is going to do the giving. The Swedish
equivalent of the house is a single word: huset. Where we say in my house a Swede
would also use three words, but a Finn, assuming that he was not a Swedish-speaking
Finn, would use only one; in the agglutinative Finnish language the equivalent of in my
house is talossani. How do we treat elided forms? How many words are there in the
French
facheterai or the English equivalent buy? The sound system of
French might suggest that fa- constitutes one syllable; syntax, on the other hand,
would argue for two separate words as an object pronoun can be inserted to give je
racheterai. How do we treat separable verbs in German? Aufstehen, meaning to get up,
would appear to be a single word, but then we find that in the equivalent of, say, I get
up at seven o'clock it is split into two parts: ich stehe um sleben Uhr auf. In the
Spanish equivalent there is a reflexive pronoun which is suffixed in the infinitive,
levantarse, but freestanding in the
present tense: me levant° a la s side.
This variability is reflected in Edward Sapir's description of a word as 'merely a form, a

definitely molded entity that takes in as much or as little of the conceptual material of
the whole thought as
12An Introduction to Linguistics
the genius of the language cares to allow' (Sapir, 1921, p. 32). This American
linguist tells us that in. some of the highly synthetic languages of the Native Americans
it is not always easy to say whether a particular element of language is to be
interpreted as an independent word or as part of a larger word (Sapir, 1921, p. 33).
So, too, in the analytic or isolating language Chinese characters
represent morphemes rather than lAtords.
Perhaps, after all, it is not of such great significance how languages combine
morphemes, units of meaning, into words. Perhaps the concept of the word is not as
important as we may have thought,
.2 Where Do Words Come From?

We speakers of English refer to a dog by means of the word dog. Why? What is
the connection between the sequence of sounds /dg/ or /dug/ on the one hand and a
particular kind of animal on the other? The spoken word does not sound like a dog, the
written
word does not look like a dog. There is no reason why a person who
does not speak English should associate the word with a dog rather than with, say,
a ship or the action of cooking
There is only an association between word and object because the English-speaking
community accepts that this is the case. The association is a matter of convention. Any
sequence of sounds could serve just as well to denote a dog; as we have seen, the
sounds /sjel, the sounds produced by the word chien, do the job just as well for the
French and the sounds fhonti serve just as well for the Dutch. As
Juliet comments in Shakespeare's play Romeo and Juliet, 'that which
we call a rose,/ By any other name would smell as sweet'. The nature of the
association between word and object can be likened to
that of the association between a red cross arid medical services, the association
between a red light and the requirement to stop.
So if words have no inherent relationship with objects in the real world, what
determines their form?
Firstly, of course, some words do have something of an inherent relationship with the
thing that they denote. This is clearly the case with onomatopoeic words, words which
represent sounds or things which make a sound. The words bang and crash are clearly
imitative of the sounds that they denote. While the word dog is arbitrary, the young
child's equivalent bow-wow is basically imitative, although
Lexis
here, too, a conventional form has developed; a Danish child imitates a dog with
vovuov. Indeed, the theory that human speech arose as the result of humans imitating
animals has been given the
light-hearted name of 'the bow-wow theory', as opposed to, say,
the 'yo-he-ho' theory, the theory that speech arose out of man's co-operation with
his fellows.
Other words, though not directly imitative, may also exhibit sound symbolism,
may also reflect the nature of the object or concept that they denote. Which would you
expect to be larger, a
*geek or a *gock? Which would you expect to move more quick/y, a *gish or a
*gunip? What is *Moviating? (The use of an asterisk in conjunction with a word
indicates that it is a hypothetical word, that it does not exist.)
It has been argued, for example, that high front vowels, those produced with the
tongue raised towards the front of the mouth, tend to be indicative of smallness; cf.
Scots wee, Swedish liten, French
and Catalan petit, the diminutive suffixes of Italian (-mno), Spanish
(-ito) and Portuguese (-inho), This may be imitative, the mouth being rounded,
more voluminous, for the back vowels such as the /3/ of lot. Thus, if you thought that
a *gock should be bigger than a *geek
you help to substantiate this view.
The initial cluster si- of English, it has been suggested, is indicative of the
condition that Leonard Bloomfield called 'smoothly wet' (Bloomfield, 1933, ID_ 24,5);
cf. slide, slippery, slimy. This is one of a large number of associations between sound
and characteristic listed
by Bloomfield. The cluster kr- suggests noisy impact as in crash and crunch, the cluster
-amp suggests clumsiness as in clump and thump,
and so on. Thus he would surely agree if you felt that a *g-ump would be slower
than a *gish.
So does Lewis Carroll's nonsense phrase slithy toves suggest something smoothly
wet? If it does, is that indicative of a universal association between word and
characteristic or are we simply drawing an analogy with other words that sound similar?
To put it another way, is the association between word and characteristic or between
word and word? After all, Carroll's word slithy is supposed to have been derived by
fusing the words tithe arid slimy. The initial cluster hi-, it has been suggested, may
imitate inflation, as in, for example, blow, blast, bladder. The equivalents of Mow in
the
Germanic languages generally begin with the same cluster, as in German blasen, Dutch
blazen, Swedish Wicisa. By way of Latin flare we
the verb discover by the addition of a morpheme implying removal. This removal
can be negated by the prefix un-: undiscovered. The verb discover becomes discoverer
by the addition of a morpheme implying the person who does the action. This process
of word formation is known as derivation.
Clearly, different languages form derivations in different ways. In the Romance

languages, for example, it is common to derive a noun related to an activity from the
feminine form of the past participle of the verb denoting that activity. In this way, for
example, French has derived its words for an entrance, an exit and a view: entree,
sortie
and vue. The German equivalents are a prefixed nominal form of the
verb: Eingang, Aussgang and ,42,45,5icht.
Occasionally a word is coined on the false assumption that it was presupposed by

an apparently derived form. In this way edit
appeared on the assumption that an editor must be a person who edits. The words
pea and cherry have evolved from a form that was assumed to be plural; cf. pease
pudding and the French word cerise.
This process is known as backformation.
Our label for some things is the result of analogy with something else; thus we
talk of the foot of a mountain and the mouth of a river. These are metaphors. We are
dealing here with semantic extension, a new concept being included within the semantic
range of an existing word. In this way the word coach, originally denoting a horse-
drawn vehicle, now denotes a long-distance bus or a railway vehicle. The range of the
Italian word carrozza has been similarly extended, as has that of the German word
Wager'. The German equivalent of sky, Himmel, was extended to cover the new concept
of heaven when Christianity came along. The Swedish word for a telephone receiver, lur,
had long before denoted a viking horn. Semantic extension may reflect the society
concerned. The Swedish word for a book, bok, is the same as the word for a beech
tree, this reflecting the practice of inscribing the bark of the beech tree. A general term
may be applied to a more specific item where that item is of particular significance to
the society. In The Italian Language (1984), the Italian linguist Bruno Migliorini gave
as an
example the development of the Latin word machina into the Italian
word macina with the more specific sense of a millstone because The machine par
excellence was the miller's grindstone'. One might
Lew
similarly explain the Italian leN,ford rnacchina in the sense of a car in terms of the
car being a malor machine of our age.
Alternatively, a new word may result from the compounding of existing

elements in the language. In this way we acquired numerous words like houseboat
and boathouse. Icelandic has generally done this. While other languages have
derived their word for a kitchen from the Latin coquina (English kitchen, French
cuisine, German Kikhe, and so on), Icelandic coined the word eidlnis by
compounding its equivalents of fire and house. Similarly, the
Icelandic word for a volcano, eldfiall, is constructed from the words for fire
and a mountain. Chinese does the same thing in this case; there, too, the word for
a volcano, huoshan, is constructed from the
equivalents of fire and mountain.
Like other Germanic languages, English makes extensive use of compounding.

It is, however, rather arbitrary about whether the elements should be written as
one word, as two words or as a
hyphenated word; the New Shorter Oxford English Dictionary (1993)
has fireman, fire station and firefighter.
This all helps to justify the concept of the lexeme, but one is still left with a
grey area: is leather bag a compound noun or do we simply have here a modified
simple noun as we do with big bag?
The elements of a compound may be merged such that at least one element is
abbreviated. In this way, for example, the phrase motor cavalcade became
motorcade, camera recorder became camcorder. Such forms are blend words. Slithy
would seem to be one of the many blend words coined by Lewis Carroll. The
Swedish word for a sauna, bastu, is a blend of the compound word bad5tuga„
literally bathing hut. When talking of a stationmaster the Swedes may use the full
form stationsinspektor or the blend form sting.
As an alternative to using native elements to produce a new label, one can, of

course, turn to other languages for material. A language can form new compounds
using elements taken from the classical languages. The word television is composed
from a Greek element indicating distance and a Latin element indicating sight. In
the field of medicine, terminology is generally based on classical elements; hepatitis,
for example, is based on the Greek equivalent of liver. Such words are called nee-
classical compounds_ Like blend words, they are composed of elements that cannot
exist
by themselves,
A language can also, of course, acquire a word ready-made from
An introduction to Linguistics
another language. An object or concept acquired from another society may be giver
the label used by that society. Thus the word sauna - which is riot, as we now know,
Swedish - came to us from Finnish. The words chocolate and tomato came from a
native language of what is now Mexico. The people of the Andes gave Europe the word
llama and received from the Spaniards the word kawallu, the horse being previously
unknown to them. Societies that are pioneers in a particular field often provide the
labels for other societies. In this way Italy has exported musical terminology, France has
provided terminology in the field of cooking - or cuisine! - and the German-speaking
world has provided terminology in the field of psychology. Thus, from the Latin verb
crescere, meaning to grow, we have by way of Italian the musical term crescendo and
by way of French the word for a bakery item: croissant. In the seventeenth century the
Dutch were a major maritime power and Dutch maritime terminology was adopted
widely. The Dutch word for a sailor, matroos, for example, went into German as
Matrose, into Russian as maTipoc; the Dutch word jacht went into German as facht,
into Russian as mxTa , into English as yacht. When a word is adopted by one language
from another it is referred to as a loanword. Loanwords may retain a pronunciation
similar to that which they had in their language of origin or the pronunciation may be
adapted to the phonology of the adoptive language. Some people retain the French
pronunciation of the word garage with the sound /3/ but others replace this rather alien
sound with /c13/. So, too, the word may or may not retain its native morphology;
spaghetti retains
its Italian plural form, possibly because we do not recognize it as such - the word
means small pieces of string - but musicians play cellos, not *celli.
One might, of course, argue whether or not a foreign word has been adopted by the
English language. The word spaghetti is so widely used in the English-speaking world
that most people would accept that it has been adopted by the English language, but
might not its retention of an Italian plural morpheme and the fact that it denotes a food
so closely associated with Italy lead one to argue that it is an Italian word that may be
inserted into an English sentence? Perhaps less likely to be deemed an English word, if
only because its use is largely restricted to the people that it denotes, is cognoscenti.
Also counting against its adoption is the fact that much of its
Lexis
function is already performed by a cognate of French origin, the word connoisseur
.
How long can a word be in a language and still be considered a loanword? The
word connoisseur has not come from modern French;
the diphthong /oil was a feature of Old French; the modem French equivalent is
con naisseur. The word chef would generally be accepted to be a loanword, but what
about the cognate chief which, as the spelling and the original sound /tf / show, was
transferred from Old
French? Is take to be considered a loanword because it was of Norse origin and ousted
the Old English niman? We must conclude that such concepts as a loanword are useful
guides to the processes
involved but that there is a danger of oversimplifying what happens in the real
world.
A Swedish book on the subject of the influence of English on Swedish by

Magnus Ljung is entitled Skinheads, Hackers & Li:1771C1 Ankor (1988). The first two
words of the title are loanwords that have passed from English to Swedish. But what of
lama ankor, a translation of lame ducks? Instead of borrowing, to use the usual term, or
adopting, to use a more appropriate term, a word from another language, a society may
take the principle of a foreign word but translate its constituent elements, In this way
the literal sense of the English word honeymoon was imitated as Tune de miei in
French. The English term skyscraper gave rise to the literal equivalents gratte-ciel in
French, rasocielos in Spanish, Wolkolkratzer in German, and so on. On the model of
the French term for a railway, chemin defer, literally iron road, German has Eisenbahn,
Swedish has jeirnveig and Finnish has rautatie. Terms acquired in this way are called
loan translations or calques. Here, too, we do not have a category with distinct
boundaries. It is relatively easy to identify a loan translation when the term is as
idiosyncratic as honeymoon. But when the term is more prosaic, like mousetrap, it is
quite feasible that different communities happen to
have independently produced a term from the equivalent semantic elements. It is hardly
likely that the Icelanders modelled their word
for a volcano on that used by the Chinese!
A word may derive from the name of a place or person. The word coach may be
derived from the name of a Hungarian town. The word copper derives from the name
Cyprus. Items may be named after the person who invented or discovered them; thus
the Italian Alessandro Volta gave his name to the volt and the Italian Luigi
Galvani gave his name to the process of galvanization.
A new word or phrase in a language may be adopted in a modified
form which makes it more acceptable, which makes it appear more meaningful.
The word crayfish is derived from the Old French word crevisse; the element fish was
considered to be more appropriate for the name of a creature that lives in water.
The Spanish word for a bolt was once berrojo (cf. French verrou). It later became
cerrofo by association with the function of a bolt, cerrar being the equiv alent of to
close. The Portuguese wordferrolho, on the other hand, was influenced by the
material of which a bolt is made, namely iron. The alteration of words in this way is
called popular etymology.
Other ways in which lexemes change once they are in a language will be looked
at in chapter 10.
Summary
We commonly consider the word to be a fundamental building block of language,

but as units of meaning may be expressed by more than one word, linguistics has
the concept of lexeme.
Some lexemes arguably have some intrinsic association with the things that
they denote, but the origins of most words are lost in the mists of time. What we can
establish is how these elements subsequently provide the lexemes that we use. New
lexemes can be
produced from native elements by such means as derivation,
semantic extension and compounding or from foreign elements by such means
as borrowing or translating foreign words.
Exercises
.1 Write what you consider to be the best definition of the term

word. Justify the choice of the elements that you incorporate in your
definition.
2.2 Suggest alternatives to the word towel to denote a towel.

Account for your suggestions.
2.3 A certain object is generally denoted in the other Germanic

languages either by a word that literally means *cowpot or *cowleg, these
alluding to its form, or by a word that literally means *breakiron, this alluding to
its function. What do we call this object?
Semantics
or
Why Don't Spaniards Watch Blue Movies?
.1 The Function of Lexemes
Chapter 2 dealt with the word, with its origins. But it said nothing of the function
of the word, of the lexeme. Without a function a word is a mere sequence of sounds,
just as a red light would be little more than decoration if it did not indicate to drivers
that they should stop (or indicate whatever a red light in a 'red light district' indicates!).
So what is the function of a word, of a lexerne? In the sentence The daughter of the
terrorist has been caught we can identify two distinct types of word. While we can tell
somebody what a daughter or a terrorist is, we cannot tell them what a 'the' or an 'of'
is, for while the words daughter and terrorist denote something in the real world, the
words the and of do not. What these latter words do is serve the others in some way,
the by specifying, of by indicating a relationship. Daughter and terrorist are content
words; the and of are function words. One can draw a parallel with an algebraic
expression like 3y 2z; the <y> and the <z> can be likened to content words in that
they are symbols that represent something substantial, while the symbol <-1-> can
be likened to a function word - it represents the function of addition in much the same
way as the function word and indicates summation. The use of content words changes
as society changes but the use of function words is much more stable; nouns and verbs
may come and go for various reasons but we rarely
have reason to change articles and prepositions.
In this chapter our attention will be focused on content words. Function words
belong rather to the field of syntax.
Smartt ics23
and spaniels but not foxes and wolves. We have to reach a com
promise between having on the one hand an unmanageable array of categories and
on the other insufficient precision.
The trade-off between manageability and precision clearly depends on how

precise one has to be. Arabs may need to be able to make finer distinctions between
different types of camels and different types of horses than we do. It has often been
claimed that Eskimos have many more words than we do to identify different types of
snow, although it is now widely accepted that this claim has been greatly exaggerated. I
am in no position to judge as I do not speak an Eskimo language. I do, however,
speak Danish fairly well and in the Danish novel Froken Smillas .fornemmelse for sne
by Peter Hoeg (1992) the eponymous narrator, an expert on ice who was born in.
Greenland, refers, for example, to ganik, defining it as fine-grained powdery snow, and
to pirhuk, defining it as light snow Similarly, those who specialise in a particular sphere
of activity need to make finer distinctions in that sphere than others do. For most of us
the word horse and a few others like mare and font are all that we need to be able to
talk about horses, but those involved in horse racing need such terms as filly, yearling
and gelding. We could, of course, refer to anything, if only by using what seems more
like a dictionary
definition than a lexeme; we could denote a filly using the phrase young mare. But the
more significant an object or concept is to a community, the greater the tendency to
texicalise the label used to
denote it, to have a more succinct term.
The divisions that we draw in order to define our categories are very haphazard. In
English we use the same verb, play, to denote the very different activities of a child
amusing himself with his toys and a pianist performing at a concert. And yet the person
who supervises a cricket match is given a different name to that given to the supervisor
of a football match, the one being an umpire and the other a referee. When we eat
lambs we call the meat lamb but when we eat calves we use a different word for the
meat: veal. We distinguish between roofs and ceilings but Spaniards do not; they use
techo for both, On the other hand Spaniards distinguish between an internal corner
(rinco'n) and an external corner (esquina) while we do not. The arbitrariness is clearly
likely to be all the greater where there is little natural distinctiveness in the subject area
concerned. It is fairly clear cut what is and what is not an egg. But how big does a
hill have to be before it is a mountain? How big does a village have
to be before it is a town? How big does a stream have to be before it is a river?

The more arbitrary the division, the less reason there is for
different languages to make divisions at the same points; thus, for
example, the Spanish word pueblo equates to a large village or a small town.
What we call a river can be denoted by either flame or
riviere in French, the general view being that the former flows into the sea while
the latter flows into another river. Sweden also has two
types of river, an Ilk, and an 4, the former being larger than the latter. Moreover,
Swedes use a third word,flod, to denote a river outside of Scandinavia.
The colour spectrum would appear to be a prime example of arbitrary division.

Russian has two equivalents of blue: ronyoorA and cvn-nul The Welsh word glas equates
to blue but also overlaps with
green, as it also denotes the colour of grass. According to Berlin and
Kay, however, colour terminology is not random for a study by them suggested a
fairly universal hierarchy of terminology If a language only has two colour terms they
will, they claimed, be based on black and white; if it has three they will be based on
black, white and redL
The next distinction will be green or yellow, and so on (Berlin and
Kay, 1969).
The set of items that we identify by means of a word or lexerne is the semantic
range of that .word or lexerne. Such sets may be grouped with others with which they
share a common feature to form a semantic field. Thus the semantic range of the
English word red is that part of the colour spectrum that we denote with this word, a
range considerably more restricted than that of the corresponding term in a language
that distinguishes only three colours. The range
of red together with that of the other colour terms can be referred to as the semantic
field of colour.
.4 The Definition of Semantic Range
Presented with a colour chart or a box of paints we can, barring colour blindness,
select a colour that most people would accept as
being red. But there may be disagreement about whether something
is red or orange. Similarly, few of us would deny that a crow is a bird for it is
and does what we expect a bird to be and do: it has two legs and feathers, it flies, it
builds a nest and it lays eggs. But there are
birds that do not fly and there are creatures such as bats that do fly
Semantics
animal
mammal
cat dog wolf
poodleterrier spaniel
Figure .1
but are not birds, creatures such as snakes that lay eggs but are not birds. Thus we need a
way of determining the boundaries of the
semantic range of the words red, bird or any other word if we are to be able to
judge when it is appropriate to use it.
We can define the English county of East Sussex either by saying which other
counties surround it (West Sussex, Kent, and so on) or by saying which second-tier
authorities it encompasses (Eastbourne, Wealden, and so on). Similarly, we can define a
word in terms of
what it is not or in terms of what it subsumes_ Thus we can define
the range of the word dog either by saying that it is any animal that is not a
wolf, a cat, a goat, and so on or by saying that it is any animal that is either a
poodle, a terrier, a spaniel, and so on. We can be
assisted in this by a hierarchical diagram like Figure 3.1. Such a
relationship between words whereby more specific terms are arranged under their
more general superordinate terms is known as hyponymy. The word poodle is a
hyponyrn of the word dog, which in turn is a hyponym of the word mammal, and so
on. One can compare this arrangement to the natural history classifications by class,
order,
family, genus and species; the dog and the wolf both belong to the genus Canis,
the dog being the Canis familiaris and the wolf the Canis lupus. The hyponyms of a
word define its semantic range. A superordinate like mammal or animal can serve to
designate a semantic field.
Complications can, however, arise. Firstly, as Figure 32 shows, we may ascribe different
meanings to a word with the result that that word appears at different points on the
diagram; we use the word
mammal
,
catdog
doghitch
Figure 3.2
dog to indicate gender as well as to denote the species in general. Secondly, we

cannot put dog and bitch alongside poodle, ierrier arid spaniel as hyponyms of dog, for
while being a poodle precludes being a spaniel it does not preclude being a bitch; we
are dealing with another dimension. Love can be presented as a hyponym of emotion.
But is it an immediate hyponym or is there something else in
between? Roget's Thesaurus, which adopts a hierarchical arrange
ment like hyponymy, has love as a hyponym of sympathetic affections, which in
turn is a hyponym of affections. How does love relate to loyalty or desire? Are they related?
If so, are they all on the same level
or is one a hyponym of another?
Clearly some fields lend themselves to arrangement by hyponymy more than

others do. Similarly, componential analysis, which defines the range of a word in terms
of the presence or absence of particular components, is more easily applied to some
fields than to others. Words denoting family relationship lend themselves to such a
binary approach; words like daughter, on and mother denote either a male person or a
female person, either an earlier generation or a later generation), and so on, In Figure
3..3 the ranges of the words aunt, mother, son and daughter are defined by giving each
one a positive sign or a negative sign in respect of maleness, previous generation and
relationship by birth. This is
enough to give each word a unique range.
What components would one identify to define the range of the word terrorist? There is
no predetermined system of categories; one keeps identifying distinguishing components
until one has a set unique to each word unless one believes that one is dealing with true
synonyms or that the difference is one of style, formality, attitude, and so on rather
than denotation. Terrorists are people who
Sernantics
MalePrevious Related by
generation birth
Aunt -+
Mother - ++
Son +- +
Daughter +
Figure 3-3
use force to achieve a certain aim. But then the same could be said of soldiers
and bank robbers. So what other components can we adduce to distinguish between
them? Perhaps +authorized for soldiers and -authorized for terrorists in so far as soldiers
operate within a framework established by their government whereas terrorists do not.
Perhaps +political for terrorists and -political for
bank robbers, thereby reflecting the aims of their actions.
The semantic range of a word can be defined with the assistance of another word
that means the same thing, a synonym, or a word
that means the opposite, an antonym. Thus it can help us to use the word large if we
know that it means much the same as the word big or that it means the opposite of
the word small.
Due in large part to the overlay of Norman French onto Old English, the English
language has an extensive vocabulary. From Old English, for example, we have the
word hide and from Old French we have the word conceal. We can say He was
determined to
hide the truth and lie was determined to conceal the truth. Thus, as we
can use either in this sentence, we might call the words hide and conceal
synonyms. But if we try to replace hide by conceal in the sentence He was determined
to hide we find that we do not get a
satisfactory sentence, the reason being that, unlike hide, conceal
cannot serve as an intransitive verb, cannot, that is, be used without an
accompanying object. Thus the two words are not complete synonyms as they cannot
substitute for each other in all circumstances. In
this case the difference was one of syntax. In the case of the words high and tall
there is a difference of reference; both may be used to
qualify buildings but only tall can be used to refer to the height (1) of people.
In the case of antonyms we have to consider different types of relationship. The

words tall and short are opposites, antonyms, as
are the words alive and dead_ There is, however, a fundamental difference between
these two pairs, for while one person can be shorter than another, one person cannot be
more dead than another;
thus tall and short are termed gradable antonyms and alive and dead are termed
complementary antonyms_
Different again are pairs like buy and Al/ which denote, for example, two facets
of an action; these are converse terms. When looking at antonyms we should note that
a word can have different antonyms in different contexts; while the opposite of a short
person is a tall person, the opposite of a short walk is not *a tall walk. Context,
indeed, plays a very important part in the definition of the meaning of words. If
somebody tells us that a chair is yellow we think of a particular colour; if they tell as
that a man is yellow we may think of cowardice. If they tell us that the chair is
comfortable we understand that we would feel relaxed if we sat on it, but if we are told
that a man is comfortable we are unlikely to conclude that it would be relaxing to sit
on him. The meaning of the word bank becomes clear when it is preceded by either
savings or gravel_ The context, then, can do much to determine the meaning of a word.
It can, indeed, result in a different word being used; is there any
substantial difference between what a referee and an umpire do or
is the choice of word solely dependent on the sport that he or she supervises?
Some linguists have concluded that one knows a word by the company that it
keeps (J. R. Firth) or that the meaning of a word is its use in the language (Ludwig
Wittgenstein).
.5 Collocation and Idiom
When a word becomes closely associated with a particular context to the exclusion
of other words with a similar meaning such that they form what is almost a set phrase,
we have what linguists call collocation, From a logical point of view we could perhaps
refer to a *complete moon but we don't, we refer to a full moon. White coffee is not
white and black coffee is not black; it would seem that white and
black are being used to indicate polarity rather than to give an accurate indication of
colour, in this way Catalans refer to white wine (vi blanc) and black wine (vi negre —
not to be confused with
,_iinagre!). If something is not far away a Catalan might refer to the
espagnole). If a Spaniard thinks that a person is acting dumb he may say that he is
making himself the Swede (Sc hace et sueco).
.6 Homonymy and Polysemy
As we have seen, for convenience we force a uniformity on the world around us.
The word dog can denote quite different animals. The word lamb can denote an animal
or a dish, A person's body can similarly be alive or dead. The word play can denote
the activity of a pianist and that of a child. Thus the question arises of how extensive
the semantic range of a word or lexeme has to be before we feel that we are dealing
with two separate words that have the same form rather than with one word that
denotes a variety of things. In the case of lamb we can consider the animal and the
dish to be different referents of the one lexeme. In the case of sound in the sense of
something that one hears and sound in the sense of a narrow stretch of water, on the
other hand, we are likely to consider that we are dealing with two distinct lexemes that
happen to have the same form. In the case of one lexeme with a variety of referents
we have an example of polysemy. Two or more lexemes with the same form are
homonyms. Lexemes may be alike to the ear but not to the eye, as with meet and
meat, in which case they may be referred to as homophones. Conversely they may look
alike but sound different in which case they may be referred to as homographs;
examples are wind denoting a current of air and wind denoting tortuous movement.
Unlike polysernic items, lexicographers will treat homonyms
as different items, giving each one the status of a separate headword.
As so often, however, the real world does not fall neatly into our categories. Introducing
language in the medium of language, the linguist encounters the problem of defining
semantic range when talking of semantics! There can be little doubt that in the case of
the two referents of the word sound that we referred to we are dealing with different
lexemes; one is of Latin origin and the other is of Germanic origin. So, too, the word
bill denoting an account or invoice and that denoting the beak of a bird must be
deemed to be different lexernes, the former, a cognate of the French word billet, being
of Latin origin and the latter being of Germanic origin. But we a1so have a word bill
denoting a kind of axe, which may be related
we come to distinguish between a terrorist and a freedom fighter our distinctions

are likely to be founded to a much greater extent on our attitudes towards the cause for
which the people concerned are
fighting; the same person may well be called a terrorist by an opponent and a
freedom fighter by a sympathizer.
Somebody's tenacious behaviour might be called stubborn or obstinate by those

who are obstructed by this behaviour, while sympathizers might call it determined or
resolute. As we considered with the word bionti, do we attempt to account for the
difference as a componential feature, in this case perhaps -good and 4-good, or
does it constitute a different dimension?
Our attitudes towards some things may be so strong that we are reluctant to refer
to them directly. A word that we are loathe to use may be called a taboo word and its
more acceptable substitute may be called a euphemism. In this way Americans are
reluctant to refer to a farmyard bird as a cock because of its association with the
male
genitals (note the inoffensive phrase used by the author!) and they
use the word rooster instead.
We can differentiate between kill, murder and assassinate by reference to such

components as +intentional and +political, but what components can we adduce to
differentiate between kill and do in? Here again the difference might be seen to lie in
another dimension, in this case on an axis of formality. When in the play Pygmalion by
George Bernard Shaw the flowergirl Liza Doolittle, her speech not yet perfected by the
phonetician Henry Higgins, says 'it's my belief they done the old woman in' it is not
only what she says that arouses some consternation but also how she says it, the
register that she uses.
Words, then, do more than identify things in the world around us. Ogden and
Richards gave what they referred to as a representative
list of the main definitions of meaning, that list ranging from 'An Intrinsic
property' to 'That to which the Interpreter of a symbol (a) Refers (b) Believes himself
to be referring (c) Believes the User to be referring' (Ogden and Richards, , pp.
186-7).
Geoffrey Leech presents a simpler breakdown of meaning into seven aspects (Leech,
1974, pp. 10-23). The first is the fundamental denotative - or, as Leech refers to it, the
conceptual - meaning, that which defines the meaning of a lexerne in terms of its
constituent features. Thus, to use the componential analysis introduced in
Semantics
section 3.4, the conceptual meaning of the lexeme woman can be defined in terms
of such properties as +human, -male and +adult. Here, then, we are dealing with a
direct link between word and thing. But, as we observed in section 3.2, the intermediary
of the human mind often affects the nature of the semantic range of a lexeme. Different
people, having been subjected to different experiences in life, have different mental
images when they hear a lexeme. When different people hear the lexeme woman such
qualities as beauty, compassion, practicality and emotion will feature with differing
relative significance. Moreover, the relative significance of features will change as
society changes; the association of housewife is much less prevalent in our society than
it was fifty years ago. Such association is called connotative meaning by Leech. A
lexeme may be more appropriate in a particular style. To draw from Leech's examples
again, cast is associated with literary style while chuck is colloquial. This is stylistic
meaning. Shut up in the sense of being quiet similarly has stylistic meaning, being
colloquial, but in so far as it is indicative of a disrespectful attitude on the part of the
speaker it also has affective meaning. Our response to one sense of a lexeme may be
affected by another sense of that lexerne. We can, for example, scarcely use the word
gay in its older sense of merry as it now invokes homosexuality. This is reflected
meaning. Lexemes may have a collocative meaning, requiring that they are used together
with certain other lexernes but not with others. As we saw in section .5, collocation
may be so restrictive that we can guess which word will follow a given word, an
example being blond. Finally Leech refers to thematic meaning, to the emphasis that
attaches to a lexeme as the result of the speaker's choice of grammatical structure or his
use of stress; if in saying the sentence
John broke the vase the speaker stresses the name John we understand
that the question to be resolved was who broke the vase, whereas if he stresses
the word vase we understand that the question was what
it was that John broke.
.8 Pragmatics
In section 2.1 we saw that a unit of meaning can be as small as a

morpheme and as large as an idiom. Here we extend the range to include whole
sentences. The speaker of the sentence The daughter f)./C
the terrorist has been caught might, for example, be simply conveying information,
but he might also be suggesting that there is now an opportunity to exert some pressure
on the terrorist, Like an idiom, then, a sentence may have a meaning that is only
conveyed by the whole. What did you ay? may simply mean that the speaker did not
hear, but it could also be a challenge to repeat something that will not find favour. Do
you hazy a spare biro? could constitute an offer of a biro, a request for a biro or
simply a request for information. Here, too, context clearly provides support, as may
intonation. We have now entered the field of linguistic study that is called pragmatics.
We communicate more than we say explicitly This disparity between what we intend to
communicate and what we actually say is central to pragmatics. It is bridged by what
the speaker implies and what the listener infers on the basis of shared knowledge,.
shared assumptions and the context of the utterance. If a man at a party has no easy
means of getting home and a woman says that she has got a car, she may be offering
him a lift. If so, making the offer entails a number of what may be called implicatures
or presuppositions. She is implying that she is able and willing to drive the man home,
that the car is nearby, that it is not broken down, that she has time to go by way of
his place, and so on. The woman might, of course, be letting others know that she
does not need a lift or even gloating about the fact that she is in a better situation than
the man. But we tend to assume that other people are being helpful, As Paul Grice puts
it„ a co-operative principle tends to apply. He developed four maxims to take account
of how we facilitate communication by being co-operative. A maxim of quality is based
on the tendency for an utterance to be true; the man would assume that the woman
does indeed have a car. A maxim of quantity reflects the tendency for the amount of
information given to be appropriate; the offer of a lift would not be clear if the Ivoman
said that she has a ten-year-old red Volvo, A maxim of relevance reflects the
inclination for the man to assume that what the woman said was relevant to his
situation. And in accordance with a maxim of manner, information is usually presented
in an orderly fashion. By allowing us to take certain things for granted, then, the
assumption of co-operation allows us to say less than would
otherwise be necessary.
Context is a major source of supplementary information. Hearing
Semantics
What did you say? from a frail old man, we are likely to infer that he has not
heard what we said. Hearing it from a young man holding
an iron bar we might infer that we are in a potentially violent situation.
So, too, proximity between speaker and listener minimises what has to be
explicitly said. An utterance like That shop over there is where I bought the pork
yesterday only works if the speaker and listener are together. Without the common
reference points of time and space more information would have to be given: something
like The butcher's shop in Richmond Street is where I bought the pork on Tuesday
April. Even this might leave one wondering which town and which year was being
referred to. Proximity in this sense is generally a matter of shared knowledge. If the
speaker and the listener have common acquaintances the speaker might be able to say I
met Anne in the butcher's; if not, he might need to use an expanded utterance like I
met a friend of mine, Anne Richards, in the butcher's. If a communication is between
people with different cultural backgrounds, then much more might have to be explicitly
stated if the communication is to succeed. We are likely to interpret in two very
different ways notices on the door of a butcher's shop which say Sorry, no rabbits and
Sorry, no dogs. As we do riot eat dogs we assume that the second of these two
notices is telling us that, for reasons of hygiene, we are not allowed to take a dog into
the shop. Somebody from a very different cultural background might, however, assume
that the butcher was apologising for having run out of dog meat. Knowledge might,
then, be shared by virtue of the immediate situation or common experience.
Alternatively, it might have been introduced earlier in the discourse; just as you can
refer to somebody by she if you can point to her, so, too, you can refer to somebody
by she if that person has just been referred to more explicitly earlier in the conversation.
The extent of the knowledge shared by speaker and listener determines the amount of
explicit information required. If the listener does not know Anne Richards we might
need to refer to her dog by saying the dog of my friend Anne Richards. lithe listener
does know her, if he knows that she has a dog and if that dog has
just been mentioned, the speaker can refer to her dog by saying
nothing more than it. On a scale in between we have such forms as Anne
Richards' dog, Anne's dog and her dog.
At the end of the scale where we find it we are dealing with items that are
referentially so vague that they are useless without sub
stantial contextual information. Here we are in the field of deixis, a deictic item
being an item which has very little referential force
without the support of context. As we have seen, the amount of contextual support
required by the items on the above scale varies; as a result, it is arguable what is to be
considered deictic and what is
not. Possessive adjectives like the her in her dog are considered to be cleictic. But
then there arc many Annes in the world. And presumably there is more than one Anne
Richards,
Pronouns like it and she are, then, highly deictic. Having similar reference to
pronouns, possessive adjectives are, as we have just seen, also deictic. So, too, are
demonstrative adjectives like this. The definite article is also considered to be deictic for
we need the context to know which dog the phrase the dog refers to. Also highly
deictic are adverbs or adverbial phrases like there and last week; in a sentence like She
went there last week we do not know where there is
or when last week is without reference points, spatial and temporal respectively.
If we are to investigate the intention and the context of utterances it is useful to

attempt a categorisation of utterances by function, by
speech act.
One basic distinction is that between statements, often called declarative utterances,
questions (interrogative utterances), and requests or commands (imperative or directive
utterances). We may be able to distinguish these by their syntax. A statement typically
has the word order subject-verb-object, as in He shuts the door, a question typically has
the word order verb-subject-object, as in Can you shut the door?, and a command
typically has no explicit subject: Shut the door. Form may, however, be misleading;
social considerations, considerations of politeness may result, for example, in a command
being expressed in the form of a question: Can you shut the door? Thus we need
context as well as form to draw the correct inference from such utterances as Do you
have a spare biro? He apologized could be classed as a declarative utterance. So could
I apologize, But there is a significant difference between the two; whereas the former
describes an action, the latter not only describes an action, declares it to others, but in
fact is the action. By saying we are doing. Such utterances are called performative
utterances. It follows from their nature that performative utterances are, like I
apologize, typically in the first person and the present tense. The use
Semantics
or the potential use of the word hereby is also indicative of a performative

utterance: I hereby sentence you to three years in prison. An utterance is, however,
only a performative utterance if the conditions necessary to make it an act prevail.
I apologize only constitutes an apology if the speaker has done something wrong
and if he acknowledges that he has done so. Nobody would end up behind bars if I
said I hereby sentence you to three years in prison; to constitute an act this
utterance requires such conditions as the listener having been found guilty of a
crime and the speaker having
the authority of a judge. As we saw in the case of the man and the
woman at the party, an offer is only an offer if the speaker is able and
willing to do what is said. Such conditions are called felicity conditions.
Some linguists would claim that most utterances are performative utterances.
By arguing that they can be preceded by performative verbs, with or without
hereby, such linguists claim, for example, that commands and questions are
performative utterances; Shut the door, they argue, is an abbreviated form of r
(hereby) order you to shut the
door, Do you know where the key is? is an abbreviated form of I (hereby) ask you
where the key is.
.9 Discourse Analysis
Closely allied to pragmatics, perhaps part of it, is discourse analysis or conversation

analysis, this being the study of the organisation
and the dynamics of conversation.
We are more at ease when a conversation flows smoothly. We are less at ease
if there are pauses in the conversation or if, conversely, two people are speaking at
the same time (overlap). We dislike people interrupting, beginning to speak at an
inappropriate point. Discourse analysis investigates the mechanisms that facilitate
smooth flow in conversations.
One topic of study, for example, has been turn-taking, how the 'floor' passes from
one person to another with a minimum of disruption. One technique for handing
over the floor is the use of tag questions: it's going to be very difficult, isn't 1t7
Another is the use of a falling intonation contour. If the handover does not go
smoothly, if for example there is overlap, the problem may be resolved by a
.38An Introduction to Linguistics
struggle for domination whereby there is an increase in loudness and a slowing of the
pace of speech.
Another example of the topics investigated within discourse analysis is why some
responses to a proposition reflect more unease than others. Reflecting Grice's co-
operative principle, we are happier going along with a proposition than going against it.
Because going along with a proposition), such as an invitation to go to the cinema, is a
preferred option, the related utterance exhibits no unease: Yeah, fine. Going against the
proposal, on the other hand, exhibits unease, the speaker perhaps hesitating, softening the
refusal, feeling obliged to justify the refusal: UUM, I don't think so. I've
got an essay to finish.
Summary
Words may be divided into content words, those which identify something in the
world around us, and function words, those which serve to specify, link, and so on the
content words. Semantics is
principally concerned with content words. The set of objects, actions,
and so on that such a word denotes is known as the semantic range of that word.
The semantic range of a word can be defined by such techniques as hyponymy

and componential analysis, by reference to synonyms and antonyms. Context contributes
to the defining of the range of a word. In some contexts the meaning of a word cannot
be fully determined without reference to a wider phrase; in such a phrase where there is
a conventional association between the constituent words we are dealing with collocation
or, if the constitutent words give little, if any, indication of the meaning of the phrase,
idiom. It is not always clear whether two words represent two different
meanings of the one lexeme or are two different lexemes; in the case of the former we
are dealing with polysemy, in the case of the latter with homonymy.
We communicate more than we say explicitly. This can be so because we are generally
co-operative, generally want to help the listener to understand. What we need to say is
minimised by the support provided by the context in which the utterance takes place
and by shared knowledge. Such factors fall within the study of
Semantics
pragmatics. Related to pragmatics is discourse analysis, the study of
the organisation and dynamics of conversations.

Exercises
.1 The titles of the books referred to in the bibliography of this

work contain some words which begin with a capital initial and some which do
not. How might one account for when one uses a capital initial and when one does
not in the title of a book?
3.2 Complete the following diagram by (a) devising a category that distinguishes the
word bus from the word car, and (b) giving the appropriate symbol against each
component for the word
motorcycle.
Powered Carries Four-wheeled

people
Bus +++
Car +++
Van + +
Bicycle- +
Motorcycle
.3 Arrange the vehicles in the above exercise, together with some

appropriate superordinates and hyponyms, in a hyponymy diagram.
3.4 How valid do you consider the concept of synonym to be?
3.5 For each of the following pairs of words, state the principal
reason why they may not be considered to be synonyms:
man boy toilet be determined stubborn

pavement sidewalkwalk run
.6 Discuss the problem of distinguishing between homonyms and

polysemic lexernes. Give an example of the practical relevance of this distinction.
.7 What might somebody intend to convey by saying He's got a

gun?
.8 Somebody might say to a soldier Fire! What felicity conditions

are likely to apply before the soldier responds by firing a gun? Account for
the fact that a firefighter would react very differently to this utterance.
3.9 Rewrite the following sentence, using information of your

choice, so that it is less dependent on deixis:
She will try to sell it here tomorrow.

alveolarpalate
ridge vellum
teeth
vocal
cords
Figure 4.1
lips. The sound /g/, like /c1/, is produced by closing and then releasing the flow of
air, but in this case the obstruction takes place further back in the mouth, at the
soft palate or velum. For each sound the vocal cords have vibrated.
When we say the word dog, then, our air flow is manipulated by our lips, by
the tongue being held against different parts of the roof of the mouth, by our
vocal cords, and so on. When we sayfog the use of our teeth is required, for we
start off with our top teeth pressed against our lower lip. Some sounds require that
part of the flow of air is diverted through the nose. Figure 4.1 shows the location
of the various parts of our anatomy that shape our speech sounds. Of these speech
organs the very flexible tongue is of particular
importance. So much so that the word for it is often used as a synonym for
language: English mother tongue, French langue, Russian
I31)IK and so on.
The vocal cords are located in the larynx, in what we perceive as the 'Adam's
apple'. They are like curtains of muscular tissue that can be drawn across the
windpipe. When certain consonant sounds are produced the vocal cords are open
and the flow of air passes unimpeded; such sounds are voiceless. When other
consonants and vowels are produced the vocal cords are closed with the result that
they vibrate as the flow of air forces its way between them; such
Phonetics
sounds are voiced. The sound [f], for example, is voiceless; the voiced equivalent
is lv]. One way of telling the difference is to make
each sound with a finger in each ear; when you do this you hear a buzzing with the
voiced sound that you do not hear with the voiceless sound. You may also be able to
tell the difference by resting your fingers on your larynx as you make each sound.
It should perhaps be noted that what linguists call organs of speech have other
functions even more crucial for the survival of the species such as breathing and eating.
Speech has developed as a secondary use of organs already in place for other
functions.
.2 Consonants
We have seen that the sound id/ in the word dog is produced by vibrating the
vocal cords, by pressing the tip of the tongue against the roof of the mouth behind the
teeth, by damming and releasing the flow of air. We saw that the production of the
sound /g/ differs
in that the tongue touches the roof of the mouth further back, at the velum. If we draw
the vocal cords back to allow the air to flow past for the final sound so that there is
no vibration we produce, instead, the word dock. If instead of damming the flow of air
we allow a
restricted but continuous passage we produce the German word
clod'.
Consonants, then, are determined by whether the vocal cords vibrate, where the air
is impeded and how it is impeded? to use the linguist's terminology, they are
determined by voicing, by place of articulation and by manner of articulation. The
sound [g], for
example, being the product of the vibration of the vocal cords and of
the tongue being held against the roof of the mouth at the velum to allow a
build-up and sudden release of air, is described as a voiced velar plosive. As the
tongue engages the velum, rather than vice versa, the tongue may be referred to as
the active articulator and the velum as the passive articulator.
Figure 4.2 arranges the principal consonant sounds by their place of

articulation and manner of articulation, The further to the right a sound is on the
diagram, the further back in the mouth is its place of articulation. Where two sounds
occur side by side in a box, the first
is voiceless and the second is voiced.
At the top left-hand corner of Figure 4.2 we find fp] and its voiced equivalent
fbl. They are plosives because, as with the two
Place of articulation
cl)
co
Plosive p b td
z Fricative f v 6s z
J
:L,?
TT(
-8' Nasal
±1' Liquid !
Semi-vowel
Figure 4.2
consonants in the word dog, their manner of articulation is the pressurization and
sudden release of the air stream. In this case the air is dammed by the two lips;
thus they are bilabial sounds. The sounds [t] and Id] are called alveolar sounds,
their place of articulation being the upper teeth ridge, otherwise known as the
alveolar ridge. The sounds [k] and fa as we have seen, are articulated further back
at the velum. Plosives are also called stops. Down in the throat, at the larynx, is
produced the glottal stop, the damming
being done by closure of the vocal cords. This occurs, for example,
in place of the / t/ when some Londoners say better. It is akin to a gentle
cough.
A major alternative to plosion is friction, allowing a continuous flow of air but

constricting it so that there is an audible friction as in the consonants in the words
face and vase. As Figure 4.2 shows, such fricative sounds can be articulated at a
variety of places. The initial consonants of face and vase are labiodental, being
produced with the top teeth against the lower lip, while the final consonants are
alveolar. With the tongue either against the upper teeth or between the upper and
lower teeth we produce [0] and 16], these being the sounds represented in English
by <th> in the words thick and this
Phonetics
respectively. If one removes all articulation, if, for example, one starts by saying
[f] and then removes the teeth from the lip, one gets
the 'hi that is often described as a voiceless glottal fricative. But being devoid
of articulation [h] may also be viewed as a voiceless prelude to the following vowel.
Amongst the fricatives we find the first of a number of sounds included in

Figure 4.2 that are not normally encountered in English. The bilabial fricative 1131,
for example, is a feature of the sound
system of Spanish rather than that of English. It employs both lips
like fb1 but, like fv1, it restricts rather than stops the passage of air; it is the
sound represented by the in saber, meaning to know, and
by the <v> in lavar, meaning to wash.
The sounds Ira [n] and [uj, the three consonants in the word meaning, have the
same place of articulation as [b], (di and fgl respectively. What distinguishes them is
that in the case of the latter the velum touches the back wall of the pharynx,
thereby directing the whole airstream through the mouth, the oral cavity, while in the
case of the former the velum falls, allowing the airstream to pass
through the nasal cavity. Thus iml, [ni and [0] are called nasal sounds.
As one goes down the rows in Figure 4.2 the sounds are articulated with less
constriction of the air stream. One can also say that the sounds become more
sonorous. Thus the sounds [l], [r], [w] and as in the words leer, rear, weir and year,
are less distinct from vowels in that the flow of air is less disrupted; these sounds
are called continuants or approxirnants. As the diagram shows, the sounds [11 and
Er] can be distinguished as liquids and [w] and 01 as semi-vowels. The somewhat
vocalic nature of the liquids allows them to stand in a syllable without a vowel. The
word funnel has two syllables but only one vowel; it is pronounced ifAnli. There
may be no contact with the alveolar ridge in which case a vowel sound results; a
Brazilian will refer to his country as ibraziu I, the Dutch equivalents of old and gold
are oud and goud. The Serbo-Croat for Serbian is srpski. Many sources will put [1]
and [r] into two distinct categories. The sound [11 may be called a lateral sound.
The word lateral, derived from the Latin word latus meaning side, describes a sound
where the tongue is pressed against the alveolar ridge but without causing a stop, the
air being allowed to pass the closure on either side. When the closure is further
back, at the palate, the lateral sound is [A], a sound common in some Romance
languages, for
example Italian figlio, meaning son. The character <r> may represent a single tap
of the tongue against the alveolar ridge or a trill whereby the tip of the tongue vibrates
rapidly against the alveolar ridge; the difference between these two sounds is reflected
in the Spanish words pero meaning but and perro meaning dog. Articulated back at the
velum or the uvula is the sound [R], a sound that can be likened to a dry gargle. As
their label suggests, the semi-vowels rw] and [j], also called glides, have properties of
both consonant and vowel; like most consonants they need to be accompanied by a
vowel and unlike vowels they cannot be maintained, but like vowels
they are articulated without audible friction.
The diagram of consonants presented in Figure 4.2 aims to give an introduction to

the principal features of consonants. Inevitably this requires simplification of a complex
situation, The roof of the mouth and the tongue positions each form a continuum and
so the diagram introduces arbitrary divisions. You will encounter additional categories.
The sounds [ITI and [3], as represented by <s> in sure and treasure, are articulated
further back than Is] and [z] and are often called palato-alveolar sounds. These four
fricative sounds have greater acoustic energy and may be distinguished as sibilants. The
initial sounds of the words chair and jump, as the phonetic symbols used in Britain, RI]
and [d3i, suggest, are a combination of plosive and fricative sounds; if they are
considered to be single segments as opposed to combinations of two segments - opinion
varies - they are called affricates. Some sounds articulated particularly far back in the
mouth may be classified as uvular sounds rather than velar sounds. Some sounds have
dual articulation; [w], shown in the diagram as being both bilabial and velar, is a case
in point. The sound represented by a symbol may differ between dialects or languages.
The sound [1] has been presented as an alveolar sound
but many speakers produce a dental [1], their tongue being against
the upper teeth. The sounds [t] and [d] have also been presented as alveolar
sounds, as they are in English, but in many other languages,
such as French, Spanish and Russian, they are dental. It is in this position, with the
tongue against the upper teeth, that most Britons articulate the sound /0/ as in think, but
many Americans do so with
the tip of the tongue between the upper and lower teeth. It is in this latter way
that a Spaniard articulates the sound /0/ in, for example, the word cinco meaning five.
Even within the same variety of speech
Phonetics
a sound may vary depending on its environment. In the word top the first stop is
released before the voicing of the vowel starts, with the result that there is a moment
when voiceless air is expelled; in such cases the It/ is said to be aspirated. In the
word stop, however, the It/ is not aspirated. In the word month the In/ is dental under
the influence of the dental /0/ that follows it. The sound [k] is arti
culated further forward when followed by a vowel that is articulated at the front of the
mouth; the first /k/ of ring Kong is articulated further forward than the second. The
difference between the
position of the tip of the tongue between alveolar [n] and dental In] is easily felt;
it is more difficult to feel the place of articulation further back in the mouth, as in the
case of [k].
.3 Phonemic Notation
In the previous section we have seen sounds represented by familiar characters

such as [d] and [g]. But we have also seen characters that do not form part of the
English alphabet such as [6] and [0]. These characters, familiar or otherwise, form part
of the International Phonetic Alphabet (IPA) published in 1888 in an attempt to
produce a one-to-one correspondence between sound and character. It takes little thought
to realise that the alphabet that we use to write English is not sufficiently precise to
represent speech sounds. Some characters represent more than one sound and some
sounds are represented by more than one character. Thus the character <c> represents
both Isl and [lc] in the word cycle, while the words cycle, sick and psychology all
represent each of these sounds differently. Indeed the character <c> is superfluous in
English and many other languages; the Norwegian word for a bicycle is written sykkel.
Vowels are particularly erratic. In standard English the vowel in the words term, firm
and worm are the same. In the previous section we saw leer, rear, weir and year, four
words in which the same vowel sound is spelt in three different ways. The poor
correlation between sound and spelling in English results in large part from the fact that
the spelling has not kept pace with changes in the sounds. The words write and right
have developed from Old English writan and riht. At that time the <w> of writan and
the <h> of riht were both
sounded, the latter being the palatal fricative that may still be heard in Scotland. The
sounds have coincided but this has not been
reflected in the spelling.
Clearly the variation is all the greater if one compares a number of

languages. Before a front vowel the character <c> represents /k/ in the Celtic
languages; that is why we usually pronounce the word
Celtic as ikeltiki, one exception being when we are referring to the Glasgow
football club. In Spanish <c> before a front vowel represents /0/ or /s/ depending
on the dialect, and in Italian it represents
/j/ or /ti /. The character 1> represents a different sound in each of English,
German, French and Spanish, these sounds being [d3], [jj,
[3] and [xl respectively;
George Bernard Shaw highlighted the lack of precision in English orthography

by spelling the word fish as ghoti; the sound /f/, he argued, could be represented
by <gh>, the sound /i/ by <o> and the sound /f/ by <ti>, his reasoning being based
on analogy with such words as enough, women and rational. We might add that the
spelling of English is so idiosyncratic that the digraph <gh> can represent the sound
[f] at the end of a word but not, as in Shaw's concoction, at the beginning of a
word. This idiosyncracy was something that he felt strongly about, to the extent
that he made provision in his will for the development of a phonemic alphabet for
English. In his play Pygmalion he begins by trying to represent the lower-class
London accent of the flowergirl Liza Doolittle but then apologetically states that
'this desperate attempt to represent her dialect without a phonetic alphabet must be
abandoned as unintelligible outside London' (Shaw, you will note, talks of a phonetic
alphabet - this usage is still common but linguists tend to prefer to use the phrase
phonemic alphabet as the characters represent phonemes).
The symbols are clearly based for the most part on the Roman alphabet or
variants of it such as the <O> that was used in Old English and is still used in
Icelandic. The voiceless equivalent of fa], [01, is one of a few sounds represented
by Greek characters.
International though it claims to be, the IPA is not without its variation. The
palato-alveolar sounds that British linguists generally represent by [51, KJ, [3] and
[d3] are, for example, represented in the United States by [g], Ea [11 and [ji.
A phonemic character may represent a range of sounds. Just as the semantic range
of a lexeme is a compromise between precision and a manageable vocabulary, so,
too, a phonemic character is an arbitrary compromise. We have seen that the
character It] is used to represent a sound articulated at the alveolar ridge and a
sound
cavity, with the result that vowels have a stepless variability. It is the vowels
particularly that give rise to regional and social variations in accent. Mastering the
vowels of French is likely to be more difficult for U.S than mastering the consonants of
that language. There is a tendency for front vowels to be accompanied by lips that are
spread rather than rounded and for back vowels to be accompanied by lips that are
rounded rather than spread. This is the case in English. This applies more to close
vowels; the lower the jaw, the more difficult it is to adopt the extreme lip shapes.
There are,
however, languages with rounded front vowels and, more rarely,
languages with spread back vowels. The pronunciation of French \rowels is made
all the more difficult for us by the fact that French
has a series of rounded front vowels as well as a series of spread front vowels;
if, for example, you hold the sound [i:] and gradually round your lips you get the
rounded equivalent [y], the sound in the
informal form of address tu.
Because of the variability of vowels linguists find it useful to have a set of

reference points to which they can relate the vowels of a language. Such a set of
reference points, proposed by the British phonetician Daniel Jones, is known as the
cardinal vowels. There are eight primary cardinal vowels. One of them, represented by
the symbol [i], is the sound produced when the tongue is as high and as far forward
as possible. Another, represented by [al, is the sound produced when the tongue is as
low as possible at the back. That is why a doctor tells us to say taah' when he wants
to look at our throat. The vowel in the word least approximates to DI, that in the word
last as pronounced in standard British English approximates to fal. The other six vowels
have been plotted between these two extremes, the front vowels being plotted at equal
acoustic intervals, as are the back vowels. Vowels are usually represented by a point on
a trapezium that is a stylised representation of the vowel space in the mouth, the front
of the mouth being understood to be to the left. In fact, the tongue can rise higher at
the front than at the back. The cardinal vowels are represented as shown in Figure 4.3.
As in English, the front vowels are spread and the back vowels rounded. These vowels
are complemented by a set of secondary cardinal vowels in which the front vowels are
rounded and the back
vowels are spread. Thus, in French, the close front vowel li] is
accompanied by a close front vowel [y], and so on.
Phonetics
Figure 4.3
Above I said that the vowel in the word least approximates to ii], not that it is
M. The cardinal vowels are an abstraction, a yardstick to be used by the linguist.
If they represent an actual vowel in some speech variant, that is accidental. If an
actual vowel is close to a cardinal vowel it may be acceptable to use the symbol
for that
cardinal vowel. Thus the vowel in the word dog can be represented by [3] or, in
the case of many Americans, by the more open [o]. Where the actual vowel is not
close to a cardinal vowel or where
greater precision is required another symbol may be used; the vowel in the
word dug as pronounced in standard British English is represented by the symbol
[A].
The vowel in the English word less maybe represented as /e/. The word lace
is pronounced in the same way as less except that there is a glide from /e/ towards
a more close front sound, the tongue rising and the lips spreading. In louse the
tongue rises to a close back sound, the lips rounding, while in lice it rises from a
similar starting point to the close front position with spread lips that was reached
in lace. Such gliding vowel sounds are diphthongs. The diphthongs in the words
lace, louse and lice are /eV, /au/ and fail respectively. The exact nature of a
diphthong is a matter of debate among linguists. Some consider it to be a single
segment, regarding it as a sequence initiated by one stress pulse; others consider it
to be a sequence of two segments. This latter position allows further debate: are the
two segments always vowels or may one be a semi-vowel? Some words may even
have a triphthong, such as the /au/ of hour. Consonants and vowels have been dealt
with in two separate
sections. But, as we saw in section 4.2, the approxirnants, the sounds
[1], [I], [w] and [j], share some of the qualities that we associate with vowels.
It is said that the articulation of a consonant involves the
interruption or restriction of the flow of air, whereas the air flows freely with a
vowel. But such distinctions are a matter of degree, and approximants, having little
disruption of airflow, fall into a middle
ground. So do nasal consonants as, barring a cold, the airflow
through the nasal cavity is unimpeded. The approximants fwi and ijl, the semi-
vowels, are articulated without audible friction; these
are the sounds that result if one closes further the close vowels [ui and [i]. After a
vowel the segment Ell may be close to being a back vowel; this is well illustrated if
you hear a Brazilian naming his native country. The approximants ill and [r], the
liquids, and the nasals Ern] and in] can be maintained like vowels; [rl and fm] may,
for example, be used to imitate motor vehicles and we maintain fm) when we
hum. As we saw, the liquids can stand in a syllable without a vowel; the same applies
to the nasals [m] and In]. In addition, [11], involving no restriction of the airflow, may,
as we have seen, be regarded as a voiceless prelude to a following vowel,
.5 Sounds in Sequence
This chapter has outlined how different speech sounds are produced. Having done
that we must now correct any idea that an utterance consists of a sequence of discrete
sounds. Rather than engage, disengage or change position sharply between segments, our
articulators flow from one to another. We have just seen that diphthongs can be
regarded as one segment that glides or two separate segments, this uncertainty reflecting
the fact that segments merge into each other. The profile of a sound varies under the
influence of its neighbours. We saw in section 4.2 that before a front vowel the sound
[k] is articulated further forward, towards the palate; it is palatalised. The vowel in the
word month, being surrounded by nasal consonants, becomes nasalised as the velum
lowers to allow air into the nasal cavity for these consonants. We saw that the in/ in
month is dental rather than alveolar under the influence of the dental /0/. Some people
pronounce the word exit as /eksiti, some as /egziti; either way the first two consonants
share the same position of the vocal cords. Sharing the same feature in this way, being
homorganic, minimises the movement that the
articulators have to undergo, makes speech easier. Thus there is a tendency for sounds
to adopt a feature of neighbouring sounds, a
Phonetics
process known as assimilation. We saw examples in section 4.3; the plural morpheme of
English is realized as /s/ after a voiceless consonant and as /z/ after a voiced consonant
This is an example of what we might call voice assimilation. In the word unkind the
prefix un- is pronounced /A13/, the nasal sound being velar in anticipation of the
velar /k/. In the word unbelievable it is pronounced
/Am/ to share with the /b/ the bilabial articulation. These are examples of
assimilation of the place of articulation. Such assimilation may have been reflected
in the spelling of a word; if we put somebody into prison we incarcerate or
imprison them. Assimilation to a preceding segment as with the plural morpheme is
called progressive assimilation; assimilation to a following segment, as with the
negating morpheme in unkind, is called regressive assimilation. Assimilation takes
place irrespective of word boundaries; as we saw in section 4.3, in the phrases ten
cakes and ten buns the word ten is pronounced with /0 / and /mil respectively.
Summary
Consonant sounds are defined in terms of where they are articulated, how they
are articulated and whether or not the vocal cords are engaged. Vowels are defined
in terms of the position of the
tongue and the shape of the lips. A diphthong is a glide from one vowel
position to another within a syllable. The distinction between consonant and vowel
is not absolute; consonants vary in the degree to which the air flow is constricted,
in their degree of sonority, and some can stand in a syllable without a vowel.
Sound segments are not articulated in isolation from each other and one may
influence a neighbouring segment to minimise the movement of the articulators in
the course of speech.
The International Phonetic Alphabet allows a more precise representation of
sounds than conventional alphabets do. Phonetic transcription may be broad or
narrow
Exercises
.1 Account for the difference in articulation in each of the

following pairs of words:
coast ghost ghost boast boast most most mist
Phonology
or
Why Should Ephraimites and Spaniards Keep

Away from the Passages of Jordan?
.1 Sound Systems of Languages
In chapter 4 symbols representing speech sounds appeared either between

squared brackets or between slashes. It is the usual practice
to use the former when referring to the speech sound produced by a
particular articulatory combination and the latter when referring to the actual
sound realized in a given context. We might, for example, read that the sound [kl
is a voiceless velar plosive and that the /k/ in the English word kill is articulated
further forward than the /k/ in the English word cull.
This leads us into the distinction between phonetics and phonology. As we saw
in chapter 4, phonetics is concerned with the production of sounds which can serve
as speech sounds in a language. Phonology studies sounds in the context of
languages arid other speech varieties. It is concerned with which sounds a language
uses and how it arranges them. It is concerned with the contribution of sounds to
the task of communication. Viewed from the point of view of phonetics the words
foal and vole differ only in that the initial sound is voiceless in the former and
voiced in the latter. We do not need to know what these words denote or even
that they belong to the vocabulary of English. A phonologist, on the other hand,
needs to know that these are English words for his work is by its nature within the
context of a particular language system. His concern when confronted with a pair
of words that only differs in that one begins with /f/ and the other begins with Iv/
is that the difference is semantically significant, that the choice of initial segment
alters
Phonology
he gives substitutable segments a special status, that of phoneme. A phoneme

is a sound or group of sounds that is functionally distinctive in a language system.
The proof of phoneme status is the substitution test which is nothing more than
seeing, as we have done with the words sue and zoo, whether substituting one
segment for another can produce a different word. A pair of words that are
distinguished by just one segment is called a minimal pair. The sound variants
which are within the one set and so are not mutually contrastive are regarded as
allophones of the same phoneme. The occurrence of an allophone is often
determined by phonetic environment. As we have just seen, in Spanish /z/ is an
allophone of isi, it being the form that takes the place of /s/ before another voiced
consonant. In Spanish /10 / is an allophone of /n/, the form that occurs before
another velar consonant, as in, for example, cinco, meaning five, this being
pronounced /0iijkoi. These are examples of the assimilation that was referred to in
section 4.5 and as such are fairly universal; the English word think has /IV before
the /k/. In English In! and /0/ can, however, also be contrastive, as in the minimal
pair thin and thing, and so are separate phonemes. To us in/ and /9/ are clearly
two distinct sounds. It is much less apparent to us that the two lateral sounds in
the word lateral, are different; our awareness of the difference is obscured by the
following factors: (i) in English they are allophones; (ii) the difference is of no
functional significance to us; (iii) we use the same symbol to represent both. That
they are in fact different is illustrated
by the fact that in Polish these two sounds are contrastive phonemes. When the
occurrence of one sound as opposed to another is determined by their phonetic
environment the sounds concerned are said to be in complementary distribution.
5.3 Phonological Rules
Certain sequences of sounds are acceptable in a word, others are not. We

know that Lewis Carroll's nonsense words like *slithy *toves could be put into
service as English words; this has, in fact, happened to a couple: chortle and
galumph. We know that *stape could be an English word and that *ulape could
not. A phonologist
needs to know why we would accept one and not the other, why one
is in accord with the phonological system of English and the other is not. He
needs to know why, even if we have never seen a word
An Introduction to Linguistic'!4
before, we would agree that the plural morpheme of *toves is realised as /z/ and
why that of *stapes is realised as /s/. We have seen in the English words unkind and
think and in the Spanish word cinco that the symbol <n> may represent the segment /ril,
The English plural morpheme may be realized as /s/, /z/ or iizi. Clearly a phonologist
wants to know the situations in which different allophones occur. We could say that the
plural morpheme is /s/ after such consonants as <f, k, p>, and /z/ after such consonants
as <b, d, g, I, my n>. But it is more efficient and leads to greater understanding if one
identifies the underlying forces. The force in this case is the common one of
assimilation, and the phonologist will very quickly observe that voiceless consonants are
accompanied by the voiceless /s/ and that voiced consonants are accompanied by the
voiced /z/. So efficiency and understanding result from identifying the operative
distinctive feature, in this case voicing. Where more than one rule affects a word one
has to consider rule ordering, for the final form that the word takes may depend on the
order in which the rules affected it. To take the example used by Francis Katamba
(1992, pp. 122-3), the French word an, meaning year, pronounced /6./, must have been
subjected to nasalisation of the vowel by the /n/ before the loss of this final consonant
took place; if the two rules had affected the word in the reverse order there would have
been no consonant to nasalize the vowel. The Latin word focus meaning hearth gave
Italian fuoco and the Latin word plums meaning little gave Italian poco. The
diphthongisation
of the to/ in the former development must have taken place before
the rnonophthongisation of the /au/ in the latter development; if not, the Italian
reflex of paucus would have been *puma For a third example we can consider such
mutated plural forms as teeth, the root vowel having been fronted under the influence of
a front vowel
in a plural suffix that was subsequently lost; clearly the mutation must have taken
place before the suffix that caused it was lost.
.4 The Phonology of English
We have seen that in / and /u/ belong to the same phoneme in Spanish but to
different phonemes in English. We saw a similar thing in respect of /s/ and /z/. English
has, indeed, a phonological
Phonology
system with relatively fine distinctions, a relatively large number of

phonemes: some forty-four. Of these, twenty-four are consonants:
as in pan If! as in fan if / as in shatterMI/ as in mart

/b/ as in ban iv/ as in van /3/ as in seizure in/ as in thin
as in tan /0/ as in thank/tS/ as in chatter //3/ as in thing
as in done /IN as in than /c131 as in junk /1/ as in lung
/k/ as in can/s/ as in sank /w/ as in wort /r/ as in rung
as in gun as in zinc as in young /h/ as in hung
As they are phonemes, we can for each sound find a word that would be changed
to another word by the substitution of that segment; in/ and /JD/ are separate
phonemes because substituting the latter for the former changes the word thin into a
different word, thing. We cannot, however, substitute /Di for In/ in the word not
for not only is there no word pronounced /ijot/ but, more importantly to the
phonologist, there could not be such a word. Nor can we any longer add kJ to
produce /knot/. There are, then, restrictions on how we use our phonemes and the
phonologist has to understand what these restrictions are, in order to identify the
phonological rules affecting the phonetic environments in which phonemes can and
cannot occur. It may be that a phoneme cannot occur in a certain position; in
English /Di cannot occur in initial position. There will be restrictions on which
other phonemes a phoneme may accompany; /kni is not possible in initial position
because initial clusters of consonants must either begin with Is/ or end with a
liquid or a
semi-vowel. Thus the words snot and clot can be pronounced /snot/
and /1<lot/ but the word knot cannot be pronounced /knot/. The king that the
Danes call Knud is called Canute by us„ the insertion of a vowel allowing us to
pronounce both the /lc/ and the In/. An English syllable can have up to three
consonant segments before the
vowel, e.g. Istr/ as in strong, Iskw/ as in square, and up to four after the vowel,
e.g. iksts/ as in texts.
We may identify some twelve vowels and some eight diphthongs in English.
Some vowels are longer than others; the vowel of the word least, for example, is
longer than that of list. Some linguists use this feature to distinguish between the
close vowels, showing length by means of a colon; thus least would be transcribed
as iii:st/ and list as /list/. But these two sounds are also articulated somewhat
differently, the former being more close and more front than the latter, and some
linguists choose to use different symbols, ii/ for the vowel in teas f and /i/ for
that in list. I am using the symbols used by Daniel Jones, which include ii:/ and iii.
Using these symbols, the twelve simple vowels of Standard British English are as
follows:
/ as in least as in learn fu:/ as in Luke

as in list /a/ as in the (unstressed)/u/ as in look
as in less /A/ as in luck /3:/ as in lord
/a.,./ as in lassas in last /3/ as in lost
The vowel represented by 101, known as schwa, is widespread in English, unstressed

vowels being commonly realised as this vowel. The first vowel of the word balloon, for
example, is represented by as opposed to the /3:/ of bali. The words phonetics and
fanatics are only distinguished by the stressed second syllable, the first syllable having
the vowel fa/ in both cases. The two vowels of the word hazard are /m/ and /a/.
Figure 5.1 shows the approximate position of these vowels relative to the eight cardinal
vowels introduced in section 4.4, the cardinal vowels being represented by the dots.
The eight diphthongs are as follows:
fial as in tear (in the eye)Alai as in tour

/0/ as in tray foul as in throw
as in tear (=rip) iliif as in toy
lai/ as in tie /au/ as in town
We have been looking at the sounds of Standard British English, the variant of English
that is said to be spoken with received pronunciation (RP). But there is also Standard
American English, English spoken in Australia and elsewhere, and in each country there
are regional and social variants. Given the variety of accents and the relatively large
number of phonemes in English, variations may overlap different phonemes with the
result that context may be required to identify a word. Without the aid of context a
speaker of Standard British English might misunderstand a Londoner and an Irishman
who tell him that they thought something, believing that the one fought something and
that the other taught something.
Phonology
e
A
a:DL
—410
Figure 51
Americans generally voice It/ after a stressed vowel and before an unstressed
vowel with the result that one might be left wondering whether somebody spent the
morning writing or riding. Indeed letters and ladders may be confused in the United
States for many Americans make little distinction between the first vowel in these words.
For these Americans there will be no audible distinction at all between phonetics and
fanatics. As we know, due to the flexibility of the lips and the tongue, vowels are
particularly variable. We saw in section 2.2 that the American linguist Leonard
Bloomfield, talking of sound symbolism, gave the word clump as an example of the
association between clumsiness and the sequence - ornp, something that seems a little
strange to British readers. Many Americans would pronounce the vowel in lost with a
lower tongue position, producing Ici:/ rather than ID/. The direction of the glide of
diphthongs varies greatly. This is shown by J. D. O'Connor's diagrams for diphthongs.
That for Standard British English /au/ shows a rearward glide in the case of the London
'Cockneys' and a forward glide in the case of the people of Northern Ireland (O'Connor,
1991, pp. 168-9). In Northern Ireland this diphthong often starts at the cardinal vowel [cil
and finishes as a rounded front vowel, a glide similar to the 'Cockney' production of the
diphthong laiI; if a person from
Northern Ireland told you that he found the child's mouse one might be excused for
thinking that he is using the present tense and that he finds more than one mouse.
O'Connor says that in London the surname Poole may be pronounced in such a way as
to sound like Paul (1991, p. 13). As a Poole who grew up in London I can confirm that
I have been thought to have given my name as Paul.
.5 The Phonology of Other Languages
When Spaniards address me by my Christian name a number of problems arise.

The first is the initial consonant cluster istji — even
though.. like Americans, they reduce it to /sti as they are not in the habit of
inserting a semi-vowel before a close back vowel. They
require the support of a vowel; thus Latin stuchum, which gave us the word study,
became estudio in Spanish. A final cluster is unacceptable and so the /t/ is dropped.
And Spanish does not have weak unstressed vowels. The net result is that a Spaniard is
likely to say icsluari rather than rstju:a:t/. Different languages, then, have different
phonological rules and it may require some effort for a speaker of one language to
adapt to the phonological rules of
another language. Remember that, just as Spaniards seek the aid of
a vowel to articulate 1st!, we have inserted a vowel between the /k/ and the hi/ to
produce Canute. Not having the glottal fricative [II], Spaniards often use the velar
equivalent [xl, their jota, when
speaking English.
And having for the most part lost the segment If!, Spaniards might be well
advised to avoid the passages of Jordan in the light of
what we read in the Bible, Judges 12:
5 And the Gileadites took the passages of Jordan before the Ephraimites: and it
was so, that when those Epl-Lraimites which were escaped said, Let me go over; that
the men of Gilead said unto him, Art thou an Ephraimite? If he said, Nay;
6 Then said they unto him, Say now Shibboleth: and he said
Sibboleth: for he could not frame to pronounce it right. Then they took him, and
slew him at the passages of Jordan: and there fell at that time of the Ephrainxites forty
and two thousand.
From this source the word shibboleth has come into English from Hebrew to
refer to a test word or speech peculiarity. Different languages, then, use different
phonemes. French and
German have fewer phonemes than English: some thirty-five. While
the French and the Germans both have initial /S/ they might be in trouble by the
end of the word shibboleth as they do not have the dental fricatives /0/ and IN; hence
the popular perception that
Art Into Linguistics
front. The suffix may be a plural morpheme, a verb morpheme or a morpheme

indicating comparison; thus Swedish has such pairs as
Jot (foot) and Otter (feet), val (choice) and(to choose) and ung
(young) and yngre (younger). Such mutation also accounts for the English pair foot
and feet, but whereas in the Swedish word flitter the vowel remained rounded, in the
English word feet the lips spread.
In a similar way some languages have what is called vowel harmony. The vowels
of a Turkish word are either all front vowels or all back vowels. Thus the plural affix
is -far when the preceding vowel is a back vowel and -ler when it is a front vowel, so
that the
plural of oda, meaning room, is &afar and the plural of ev, meaning house, is mien In
tuziu, meaning with salt, the equivalent of with is
with a back vowel and in sekerli, meaning with sugar, it is -li with a front vowel.
.6 Suprasegmental Features
We now have a good acquaintance with the speech sounds, the segments, of
language. We have seen that by such processes as assimilation and vowel harmony they
affect their neighbours. We now need to widen our perspective further still if we are to
understand the variation in sound patterns between utterances. We need to rise above the
segment, to consider suprasegmental features, A major unit greater than the segment is
the syllable. The concept of the syllable is supported by many phonological rules. As
an example, monosyllabic words in English tend to be of the same length, a vowel
being longer when followed by a voiced consonant than when followed by a voiceless
consonant, the voiced consonants being shorter than the voiceless consonants, and longer
still when no consonant follows; thus in the words seat, seed and see the vowels are
progressively longer. In Spanish the syllable serves as the basis of the rhythm of
speech. When we talk of permissible consonant clusters or the aspiration of voiceless
plosives in initial position, of unvoicing in final position, and so on, we generally mean
syllableinitial and syllable-final rather than word-initial and word-final; the sequence /8dri
in the word withdraw can only be reconciled with our rules for consonant clusters
because we regard /a/ and /dr/ as
belonging to different syllables, a split which in this case, as in many
Phonology
others, coincides with a morphological split, the two syllables coinciding with
semantic units.
Every syllable has a nucleus, its power source. This is usually a vowel but it can
also be a liquid or nasal consonant as in the second syllable of the word people:
/pi:pi/. A syllable may consist of nothing but the nucleus, as in the word owe, but
the nucleus is usually accompanied by at least one consonant. A preceding
consonant or consonant cluster is called an onset and a following consonant or
consonant cluster is called a coda. The nucleus and the coda may be viewed as a
unit, the rhyme. In the previous section we saw the suggestion that there may be a
natural tendency towards open syllables with a single consonant in the onset,
towards a consonant—vowel (CV) structure. In support of this hypothesis we can
cite the fact that pidgins, varieties of speech system at an early stage of
development, tend to have open syllables and to avoid consonant clusters. This
tendency might help to explain why in French an unstressed vowel may be elided
when another vowel follows, as in l'homme, why a final consonant may be silent, as
in !es, /lei, and why such a consonant is heard when a vowel follows in the same
phrase, as in ks hornmes, ilez3rni. In Italian the equivalent
of and, usually e, may be ed before a vowel (paria italiano ed inglese).
We similarly say a pear but an apple and some of us insert r between, for
example, sazv and it.
The force with which a syllable is produced in speech varies. When we say
the word syllable we put more emphasis on the first syllable than on the other two.
As we have seen, the word hazard has the vowels and hi, the latter being the weak
schwa because it
is unstressed. The quality of vowels, then, is affected by stress, by increases in
the pressure of air expelled by the lungs and by
increases in the tension of the vocal cords. We commonly talk of stressed and
unstressed syllables, but stress may be relative, and in words with several syllables
this binary distinction may be too simplistic; in the word opportunity, for example,
there is a primary stress on the third syllable, a secondary stress on the first
syllable, and so on.
The difference that stress makes to the quality of a vowel varies between languages.
As in English, it makes a relatively big difference in Russian; in the word gopora,
meaning road, the stress is on the second syllable with the result that the first
vowel is like the final vowel rather than the second one. In Spanish there is much
less
An Into Linguistics
difference in the quality of stressed and unstressed vowels; the two vowels in
casa, meaning house, are much more similar than are those in the English word
hazard. When learning Spanish at school from a native speaker I was told to open
my vowels more. Incidentally, the teacher unwittingly also impressed on me the
lack of differentiation
between the sounds represented by and <v> in Spanish, for
what struck me most as an adolescent boy was that it sounded as though he
was telling us to open our bowels more. There is, however, diphthongisation of
what had been short /c/ and 1W when they receive the stress. Thus the Latin word
focus became fuego,
meaning fire, in Spanish, while the diphthongisation did not occur in
the reflexes hogar, meaning home, and hoguera„ meaning bonfire, the stress not
falling on the first vowel. As we have seen, a similar development produced ftioco
in Italian..
In some languages the position of the stress in a word is more predictable

than in others. In Russian it is relatively unpredictable; while, as we have just
seen, it is the second syllable in the word gopora, in the word mosiolco, meaning
milk, the stress is on the final syllable. In Polish, on the other hand, the stress
nearly always falls on the penultimate syllable. This is the usual position for stress
in Italian, examples of such parole piane being animale meaning animal, and ancora
iautko:ra 1, meaning again or still, In many words, however, parole sdrucciole
(remember that sdrucciolare means to slide), the stressed syllable is followed by two
unstressed syllables, as in for example povero lipo:verof, meaning poor, and
ancora iraDkorai, meaning anchor. As these examples show, Italian is
one language in which there is a correlation between stress and the length of
the vowel.
As we see from ancora, which can mean again or anchor depending on the
position of the stress, stress can be semantically contrastive. In English we see this
in invalid and entrance which have different meanings depending on whether one
stresses the first or the second syllable. There are many pairs in English where
stress distinguishes
between verb and noun, an example being ilon'trast/ and
1'k3ntrost/ . Talking of contrast, stress is also functional in that it can
indicate emphasis or contrast, as in I wanted the white one, not
the green one.
In some languages the stressed syllable is the basis of the rhythm of the
language. This applies to both Italian and English. In such
Phonology
languages the rhythm requires stressed syllables to occur at fairly regular

intervals. As the number of intervening unstressed syllables may vary, the speed of
delivery has to vary if they are to fill the same period of time; the phrase that
huge dog can be delivered at a more leisurely pace than that enormous alsatian
can. Hence the correlation
between stress and vowel length. In the phrase limes and lemons the
and may well be pronouncedbut in the phrase oranges and
lemons, in which two more unstressed syllables have to be accommodated

between the stressed syllables, the and is more likely to be weakened to /and/ or
/an/. The position of the stress may occasionally change to make things a little
easier; the word unknown said in isolation is stressed on the final syllable but when
it is followed by a word with initial stress, as in unknown soldier, the stress moves
to the first syllable thereby avoiding two stressed syllables in immediate succession.
In French, which is not greatly stressed, or in Spanish, on the other hand, the
rhythm of the language is based on the syllable, whether stressed or not, hence is
syllable-timed rather than stress-timed. It is for this reason that there is no need to
subdue unstressed vowels in Spanish. As we have just seen, when we say I wanted
the white one we put greater stress on the word white if we want to emphasise that
we did not want one of another colour. This alters the pitch variation of the
utterance, produces a different intonation contour, a different 'melody'. If we
emphasised I instead of white we are drawing attention to the person as opposed to
the object. Intonation, then, is functional in that it can alter the message conveyed
by an utterance. You might tell somebody that you met John the postman and old
Charlie. In the written language the lack of a comma suggests that John is the
postman whereas the insertion of a comma suggests that John and the postman are
two different people. In speech, intonation fulfils a similar function. The person you
are speaking to might reply Did you? and from their intonation you can guess
whether they are surprised, whether they disapprove or whether they could not
care less. Intonation also plays a part in the dynamics of discourse, of conversation;
a falling contour may, for example, indicate that the speaker has finished saying
what he wanted to say. Differences in pitch within a word may be the equivalent of
phonemes in that they can serve to distinguish between different meanings of a
word. An example that is often quoted to illustrate
this is the group of Mandarin Chinese words that we transcribe as
ma. When said with a level high pitch it means mother, when said with a
high rising pitch it means hemp,. when said with a fallingrising pitch it means
horse and when said with a falling pitch it means scold. For variety we can also
consider the pair ba with a level
high pitch and ba with a falling-rising pitch, the former meaning eight and the
latter meaning to take, to grasp. It is interesting perhaps, indeed, important if you
are planning to go to a Chinese
market - to note that the pitch distinguishes between buying and selling; to
buy is mai with a failing-rising pitch and to sell is tnai with a falling pitch.
Languages in which pitch is semantically contrastive in this way are called tone
languages.
We conclude by pointing out that stress and pitch are relative. A child has a
higher pitch than a man but both have meaningful intonation as it is the relative
levels that are significant, not absolute levels. Only by pitch ranges being variable
can tone languages also have intonation, as they do.
Summary
Phonology is the study of the sound systems of languages. Phonologists are

particularly concerned with the functional significance of speech sounds. As a result
the phoneme is an important concept, this being a sound that can distinguish
between different words. A variant of a phoneme that cannot be contrastive is
called an allophone.
The sound system of each language is governed by phonological rules that

determine the permissible clusters of consonants, whether a sound affects a
neighbouring sound, and so on. Vowel systems tend to be symmetrical and to
maintain maximtim differentiation.
Segments are grouped in syllables. In words with more than one syllable one
is likely to be stressed more than the others. In some languages the rhythm of
speech tends to be based on all syllables, in others it tends to be based on only the
stressed syllables. Variations in pitch give utterances an intonation. These
suprasegm.enta/ features of stress and pitch can be functionally significant.
Exercises
.1 Identify seven pairs of contrasting phonemes amongst the

following Italian words:
.10 The Danish equivalents of big and bigger are stor and storre.
Account for the iv/ in the comparative form.
5.11 Give three English words where stress distinguishes between

verb and noun.
5.12 For each of the following pairs compare the position of the
stress. Comment.
economy / economic wonder / wonderful
beauty / beautifulacid /acidic
.13 Explain why somebody might choose to stress the following

utterances as indicated by the bold type:
John wanted to do this today.
John wanted to do this today. John wanted to do this today.
Morphology
or
Why Are the Finns People of Few Words?
.1 The Composition of Words
The words manly and virile are both stressed on the first syllable. The noun
manliness retains the stress on the first syllable but in the noun virility, a Latinate
as opposed to Germanic word, the stress moves to the second syllable. The
possessive form of the lexeme wife is pronounced /waifs/ while the plural form is
pronounced /waivzi. Here we have just two examples of the relationship between
phonology on the one hand and the composition of words on the other. Just as
there may be arbitrary boundaries between semantic ranges, so, too, ordering
knowledge may require us to impose divisions across linkages. Having said that, I
shall in this chapter study the composition of words under the usual heading of
morphology. The word black is a minimal unit in so far as it cannot be broken
down into smaller semantic or functional units. The words blackbirds and blackened,
on the other hand, can be broken down into significant elements, three in each
case. In the case of blackbirds two of the
elements, black and bird, stand together to denote a particular kind of bird and the
third, -s, serves to indicate plurality. In the case of
blackened the adjective black is accompanied by the element -en which changes
the adjective into a verb and the element -ed which specifies the past tense or past
participle.
A word must contain an element that can stand by itself, such as black; such an
element is called a root. It may contain more than one root, in which case it is a
compound word (for example blackbird). It may contain one or more elements that
cannot stand by themselves, such as the -en and the -ed in blackened; such elements
that can only
exist when joined to a 'host' are called affixes and the 'host may be called a base.
Affixes may be joined to the beginning of the base, in which case they are called
prefixes, or to the end of the base, in which case they are called suffixes. There are
other alternatives
such as infixes that are significant in some languages. Some affixes like the -en of
blacken create a new lexeme, while others like the -ed of blackened restrict the word
grammatically
The number of such semantic and grarnmatic elements that are typically combined
in a single word varies from language to language. In section 2.1 we saw that the
Finnish equivalent of in my house is talossani: Finnish incorporates in one word the
concepts that we express in three separate words. Finnish is often described as an
agglutinating or agglutinative language, it being a language in which such concepts as
are realised as prepositions and possessive adjectives in English are realised as affixes.
In the case of talossani the element tab- denotes a house, the element -ssa indicates
position within the house and the element -ni indicates that the house belongs to the
speaker. In my house is indicative of an analytic or isolating language, each concept
being expressed by a separate word. On a scale reflecting the average number of
concepts incorporated in a word, then, agglutinating languages lie at one end and analytic
languages lie at the other. In between linguists usually recognise another type, inflecting
languages. Such languages, like French, make greater use of affixes than analytic
languages do; as we have seen, the French equivalent of I shall give is ft don nerai.
Compared to elements in agglutinating languages, inflections may be said to have less of
the semantic substance that we associate with words — to use Sapir's phrase, the affixes
in agglutinating languages tend to have a greater 'psychological independence'. In my
house and I shall give would suggest that English is very much an analytic language.
But in the phrase in my houses the concept of plurality is added without an increase in
the number of
words; it is added instead by inflecting the word house. We might,
then, rather say that English is a fairly analytic language. As an alternative to je
donnercii a Frenchman might use the more analytic utterance je vais dormer. Here too,
then, we are imposing on languages categories that are not clear-cut in the real world.
Morphology
.2 Morphemes
The Finnish word talossani, then, is composed of three significant elements. The
French je donnerai is also composed of three significant
elements: je indicates who is doing the action, donner- indicates what
the action is and -ai indicates that it will be done in the future. Elements like this
that have a semantic or grammatical function are known as morphemes; a morpheme
may be defined as a minimal functional element of a word.
As we have seen, some elements can stand by themselves and some cannot. The
former we call roots, the latter we call affixes. They can also be referred to as free
morphemes and bound morphemes respectively. Of the three morphemes in talossani,
the first is a free morpheme but the other two are bound morphemes. As a marker of
future tense the of je donnerai is a bound morpheme although its
origins are those of the free-standing element ai meaning have. While the Finnish
equivalent of in the house is talossa, the equivalent of in the forest is rnetstissii. Thus the
idea of location inside something is expressed by either -ssa with a vowel closer to
cardinal Eu] or -ssii with a vowel closer to cardinal [al, the choice depending on the
root vowels. In the previous section we saw that the possessive form wifes has a root
that is pronounced /waif/ while the
plural form wives has a root that is pronounced iwaivi. This plural
form ends with the segment /z/ but the plural form aunts ends with the segment
Is/. Thus a morpheme may have different phonetic
realisations. The variation is often determined by phonological environment; the final
segments of the words boat, train and bus determine that the plural morpheme will be
realised phonetically as
Is/, /z/ and iizi respectively, To distinguish the phonetic realisations from the
functional morpheme linguists often call the former a morph. Thus we might say that
the plural morpheme is realised in English by the morphs Is!, /z/ and iizi, As
realisations of a morpheme that are in complementary distribution to each other we can
call /z/ and lizi allomorphs of the plural morpheme, just as we call two or more phones
allophones if they are in complementary distribution.
.3 Derivation and Inflection
In section 6.1 we drew a distinction between the elements -en and -ed
in the word blackened. The element -en created a new lexeme, a label for the
action of making sornethiug black, while the element -ed
restricted functionally the lexeme blacken. The first case is an example of derivation,
the second an example of inflection. Derivation, being concerned with the creation of
new labels, draws morphology towards lexis while inflection, being concerned with
function, draws morphology towards syntax.
The affix -er in teacher is clearly derivational, it creating a new lexeme. The affix
-es in teaches is clearly inflectional: it restricts the
use of the lexeme teach to the present tense, third-person singular. In
some cases, such as teaching is rewarding, the distinction is less obvious. Painting
as in she is painting the girl is clearly inflectional: it
is a syntactically restricted variant of the verb to paint. Painting as in I don't like
that small painting is clearly derivational: it is a lexeme that,
like photograph, denotes an object. Painting as in she has taken up painting is
perhaps more debatable, although the fact that she could
alternatively have taken up photography suggests that it is derivational. The
following are some useful guidelines.
Derived words are syntactically unmarked, whereas inflected words are marked in
some way; inflected nouns may be marked for the plural, inflected verbs may be
marked for person or tense. A derivational affix may produce a related lexeme of a
different word class (as with the adjective weak and the noun weakness); an
inflectional affix does not alter the word class (as with the verb wish and the verb
wished). Derivational suffixes are less predictable in occurence; while weak acquired the
noun form weakness, strong acquired the noun form strength. Derivational affixes are
also less predictable semantically; while fatherhood denotes the state of being a father,
brotherhood denotes an association of men as well as the state of being a brother.
Inflectional affixes are more productive, that is they can be used with new words, this
being facilitated by their greater predictability; as has been illustrated by introducing the
nonsense word 4-wug to children, there is generally agreement about
what the plural form of the word should be even if we have never encountered
the word before.
Inflectional affixes often form paradigms, predictable sets. The Portuguese verb
filar, meaning to speak, for example, has the regular set of affixes associated with the -
Jar conjugation:
productivity is a matter of degree and it varies over time, Similarly, Laurie Bauer
(1988, p. 57) highlights the facts 'that productivity is not all or nothing, but a
matter of more or less' and that 'it is a synchronic notion?.
Over time the full value of morphemes may be lost. Recount and recover can
easily be split into two morphemes in the senses of to count (votes) again and to
put a new cover on, but this is hardly the case with the senses of to tell and to
get better. The word recoil derived from an ancestor that literally meant to fall
back on one's
behind (Latin culus) but now the only free morpheme is the whole
word.
To a greater or lesser extent there may be restrictions on which roots an affix

may be joined to. The affix -ant, for example, can only be joined to a Latinate
root; we have assistants but we do not have *helpants. We have unhelpful but
inconsiderate, the negating prefix ingenerally being used with Latinate roots. The
Germanic suffix -dam accompanies such Germanic roots as king and free. Often, of
course, such pairings pre-date alternative possibilities; the morphs of the
word kingdom were linked in Old English as cynedam before there was extensive
use of Latinate elements in English.
As further examples of the factors that constrain the composition of words we

can refer to the use of the negating prefix un- and the comparative suffix -er. The
former is only used with positive roots; we can be unhappy but not *unsad, unkind
but not *uncrueL The tatter is only used with words of one syllable, such as sad,
or words of two syllables in which the stress is on the first syllable, such as happy;
we cannot be *dejectearer or *deliriouser.
Affixation may result in a phonological change; while divine has the stressed
vowel tail, divinity has the stressed vowel /it. As we saw
with virility at the beginning of the chapter, the stress may change position; as
another example, the noun productivity is stressed on the third syllable, whereas the
adjective productive is stressed on the second. Some loss may occur, as in
negotiate and negotiable.
.5 Problems of Morphological Analysis
The word unhelpful clearly consists of three morphemes which are realised by
three morphs: the semantic root -help-, which derives
the adjective and the negating prefix un-. These are easily identified
Morphology
because each has an obvious function and because one follows another, because
they are concatenated. But there are many cases where the morphological analysis
of a word is less straightforward. We have seen that there is no problem with
recover in the sense of put a new cover on but that there is less obvious
justification for treating as two morphemes recover in the sense of getting better.
Many words have an ancestor that consisted of two or more morphemes but are
now morphologically indivisible, there now being no part of the word that has a
distinct function. The word reject derives from the Latin elements re- and iactare
giving the sense of throwing back, but nevertheless we cannot sensibly divide the
modern English word; as we cannot *ject something, we cannot 'eject something
back or *feet something again. Our analysis must leave us with a root that can
exist by itself.
Several linguists have referred to the problem of analysis presented by the

word cranberry. If one is not content to regard the whole word as a single
morpheme one is left with the problem of accounting for the element cyan-. In fact,
the name of this slender plant might relate to that of the bird that we call a
crane; in Swedish the bird is called trana and the cranberry is called tranbiir. The
words gums and lambs are easily analysed as the composition of a semantic root
and a plural morpheme realised as /z/. But bow do we analyse the words teeth and
sheep? Similarly the word walked is easily analysed as a semantic root plus a past
tense morpheme but the word ran is not How do we analyse the structure of a
word where the morphs are internalised rather than concatenated? In the case of
sheep we can argue that the plural morpheme is realised by a zero morph. But
what about the words teeth and ran which undergo a change to the vowel of the
root? Indeed, what about the word went which phonologically is totally unrelated to
the word go? One morph may realise more than one morpheme. The Italian
word for a chain, catena (concatenation is joining in a row like the
links of a chain), has the plural form catene. Thus the morph -e indicates
plurality. If one also considers that it is morphologically significant that it indicates
feminine gender as well, then it is fulfilling a dual function. If so, this morph
might be referred to as a portmanteau morph.
To address such problems many linguists, such as Stephen Anderson (1995), have
concentrated on the word form rather than segments of the word. Such an
approach is known as word-and
Morphology
Summary
Morphology is the study of the semantic and grammatical structure of words.

A semantic or grammatical element constitutes a morpheme. A morpheme that can
be expressed as a word is a free morpheme, a morpheme that can only be
expressed in conjunction with another is a bo-und morpheme. In agglutinating
languages like Finnish a relatively large number of morphemes is expressed by
affixation, while English makes relatively little use of affixation.
The form in which a morpheme is expressed may be called a morph and any
variation in a form may be called an allomorph. Some morphs may to a greater or
lesser extent have ceased to be productive and the use of morphs may be restricted
in various ways.
A major distinction is between morphemes which create new lexemes
(derivational morphemes) and those which restrict a lexeme grammatically
(inflectional morphemes).
Morphological analysis is complicated by such factors as morphs losing their

significance over time, morphemes not being realised as
identifiable morphs and words developing a variant that is so elided that it needs a
host word.
Exercises
6.1 (a) Identify in the following sentence four bound morphemes.

State the function of each and say whether each is derivational or inflectional:
The teacher's brother considered the project impossible.
(b) Do you consider the word project to contain one morpheme or two?
Account for your opinion.
.2 For each of the following words transcribe phonetically and

account for the allomorphs of the past tense morpheme: waited waved
waded wiped
6.3 (a) Which one of the following Swedish words could be said to
have a zero morph?
(b) What are the functions of the two morphemes that are realised as -en?
bil car barn child
bilen the car barnet the child
bila r cars barn children
bila rnathe carsb am enthe children
.4 The Swedish for a small car is en liten bil. The Swedish for two small cars is tva
snui bilar. Which linguistic phenomenon is represented by the use of sma as
opposed to liten in the plural?
6.5 List as many factors as you can to support the word height being
classified as a derivational form rather than an inflected form.
6.6 Which morphological issues are illustrated by the following

sentence:
I haven't inspected the mice.
Syntax
or
How Does my Wife and her Beauty Like You?
.1 Syntax as Opposed to Morphology
In chapter 6 we considered morphology, the structure of words. Clearly, if we are

to make a sensible utterance we need to know how
to put the words together. The teachers fear the strength of the blackbirds does
not mean the same as The blackbirds fear the strength of the teachers, despite the
fact that we are using the same words in each
case. *Strength fear the the of teachers blackbirds the means nothing. An
utterance requires a certain structure. The study of that structure is syntax, a
word derived from two Greek elements that equated to together and arrangement.
A principal distinction between morphology and syntax, then, is that the

former is concerned with the internal composition of a
word, whereas the latter is concerned with combinations of words.
The distinction hinges on the word being a substantial concept. But in section
2.1 I suggested that this may not be the case and if that
view is accepted it undermines this distinction.
In any event there are problems. A function that is fulfilled by a single word in
one case may be fulfilled elsewhere by a succession of words. As you no doubt
know by now, where we use the three words in the house Finnish-speaking Finns
use only one: talossa. Swedish-speaking Finns use two: i huset. As we saw in section
6.3, the Romance languages of southern Europe do not usually use a subject
pronoun. The Italian equivalent of I give is do. When referring to an action in the
future we speakers of English need an additional word to indicate future tense,
whereas speakers of these Romance languages do not; the Italian equivalent of our
three-word phrase I
will give is the single word tiara. Even within the same language there may
be elements that can be expressed either as words in a
phrase or as morphemes in a word. There is little semantic or
functional difference between more expensive and dearer. Similarly the idea of
something being very expensive can be expressed in Italian by either motto caro ox
carissimo. The Italian equivalent of They
give if to me is Me to dam); the equivalent of They want to give it to me
is Vogliono darmelo, the pronouns being suffixed to the infinitive. Then there
are the clitics that I referred to at the end of the last chapter; we can say They
have given me it or They've given me it. If one regards the word as an
inconsequential distinction it may
be logical to dispense with it and to analyse sentences in terms of
morphemes. One would thereby fuse the study of morphology and the study
of syntax; many linguists do, indeed, consider morphol
ogy to be part of syntax. In any event, both may be grouped together as grammar.
As an aside, we can remind ourselves of the awe in which the common folk
held those with a command of the skill of writing; this is reflected in the fact that
the word grammar is related to the word
glamour. We can similarly refer to the different senses of the word
spell, one denoting the representation of a word by letters, the other denoting
control by means of magic.
.2 Word Classes
*Strength fear the the of feathers blackbirds the is, as we have seen and as we
all knew anyway, not acceptable as an English sentence. The teachers fear the
strength of the blackbirds is acceptable, even though its meaning might strike us as
a little strange. What we are concerned with is grammatical structure. VVhether a
word can occupy a certain
position in a sentence depends on its grammatical category rather
than its meaning. We can replace fear and strength by admire and speed and
the sentence is still grammatical, each word having been
replaced by a word of the same category
The categories concerned, traditionally known as parts of speech, are now generally
referred to as word classes. The words teachers, strength and blackbirds are classed
as nouns. Nouns are words which denote something in the world around us,
something inanimate, something animate like teachers and blackbirds, an attribute
like
Syntax
strength, and so on. Nouns are generally accompanied by a determiner, something which
helps to identify what is being referred to. In the above sentence each noun is
accompanied by the definite article the; other determiners are indefinite articles (a, an),
possessive adjectives (e.g. my), numerals (e.g. three), and so on. The word fear, as used
above, is a verb, a class that has such functions as indicating an action or a state of
being. Of is a preposition, a class that indicates a relationship between other elements
of the utterance.
Now that we can put the words of our sentence The teachers fear the strength of
the blackbirds into classes we can represent the sentence, S, as follows:
S > art., N, V, art., N, prep., art., N.
Other classes are the pronoun, the adjective, the adverb and the conjunction. A
pronoun replaces a noun and anything that may
accompany it; thus the pronoun they could be substituted for the
teachers and the pronoun them could be substituted for the blackbirds. An
adjective provides more information about the thing or person
indicated by the noun; the adjective old may be inserted between the and teachers.
In The very old teachers greatly fear the blackbirds we have
two adverbs: very which is qualifying an adjective and greatly which is qualifying
a verb. If we add to the blackbirds the sequence and the
crows we are using a conjunction, a word which links elements of the
utterance.
Lexemes in two or more word classes may share the same word form; fear, for
example, may, as in our sample sentence, be a verb or it may be a noun. To help us
to identify which class a word belongs to we can enlist the aid of either morphology or
syntax. Morphology shows us that in this case fear is a verb; if the sentence were set
in the past tense it would be fear that acquired the past tense morpheme, realised in this
case by the segment /d I. Here syntax shows us very simply that fear is a verb; a
sentence must have a verb and there is nothing else in our sentence that could fulfil
that function. That the word blackbirds is a noun is suggested by the morphological fact
that the final segment /z/ may represent the plural morpheme and the syntactic fact that
this word can stand alone after the determiner the. As we have seen, we could replace
the verb fear by the verb admire and say The teachers admire the strength of the
blackbirds. We could not, however, substitute the verb sleep because, quite apart from
the fact
that it is semantically nonsensical, the verb sleep cannot be followed by a
grammatical object_ The teachers could be replaced by the pronoun they, as could
the blackbirds if that sequence were being used to
represent the grammatical subject, but if we only had one teacher and one
blackbird they would be represented by a different pronoun
in each case: he or she in the case of the former and it in the case of the latter.
Thus if we are to use word classes as a guide to what is a well-formed uMrance,
finer distinctions are required. When a verb requires an object, as in The teachers
admire the blackbirds, it is classed as a transitive verb; when it does not, as in The
teachers sleep here, it is classed as an intransitive verb. He and she are personal
pronouns: they are used to represent people.
The sequence the very old teachers is acceptable. The sequence *old the very
teachers is not. Using the word classes, we can devise rules to define what is a
grammatically acceptable utterance and what is not.
In English the order art, adv., adj., N is acceptable, the order adj., art., adv., N is
not. The rules which govern the structure of utterances are known as phrase
structure rules. Such rules allow for the generation of grammatical sentences in a
language; they constitute a generative grammar for that language.
Constituent Structure
We can, then, describe the structure of The very old teachers greatly fear the
blackbirds as S> art., adv., adj_, N, adv., V. art., N. This description fails, however,
to reveal the fact that the sentence is not just linear, that it has a more complex,
hierarchical structure. The four words the very old teachers have a coherent
reference, the four words very old teachers greatly do not. The coherence of the
sequence the very old teachers can be illustrated by the fact that it can be
represented by a single word, the pronoun they. It can be illustrated, too, by the
fact that when translated into a language with agreement, the article and the
adjective may exhibit the gender and number of the noun. In the Portuguese
equivalent of the very old teachers, for example, the article, the adjective and the
noun are all masculine plural: as projessores tnuito velhos. Unless, that is, the
teachers were all female in which
case the article, the adjective and the noun would be feminine plural: as professoras
mit° velhas.
The coherence of the sequence the very old teachers is such that it
Syntax
constitutes a phrase. In this case we have a noun phrase, the bead, the key
word, being a noun. The reason for the coherence of this part of our sentence is
that it constitutes a fundamental part of that sentence; the subject. An utterance
typically identifies something and then supplies some new information about that
thing; the 'thing' identified in our case is the veny old teachers, the new information
is
greatly fear the blackbirds. The first element is the subject of the
sentence, the second element is its predicate. Similarly one can use the terms topic
and comment and the terms theme and rheme. , A sequence with a subject and a
predicate constitutes a clause. This may be a whole sentence, as in The very old
teachers greatly fear the blackbirds. Alternatively, a sentence may consist of more than
one clause, as in Many people think that the teachers greatly fear the blackbirds. In this
latter example the second clause, that the teachers
greatly fear the blackbirds, is known as a dependent or subordinate
clause.
A linguist analysing the constituent structure of our sentence will refer to the
subject and predicate as a noun phrase and a verb
phrase, In linguistic notation, S > NP, VP. This represents a deeper
level of analysis than 5> art., adv., adj., N, adv., V, art., N. Analysis at this
deeper level allows us to define more general principles of sentence structure. Thus we
can say that a noun phrase precedes a
verb phrase in a declarative sentence whether that noun phrase is
the very old teachers or simply they.
These two levels of analysis can be related to each other by constructing a tree
diagram as shown in Figure 7.1. If we wanted to
NP VP
NP
/ /\
/A N
art adv. adj. Nadv. V art. N
hTe very old teachers greatly fear the blackbirds
Figure 7,1
add in the park to specify the blackbirds concerned we would be adding a

prepositional phrase to the second noun phrase, thereby producing the structure shown
in Figure 7.2.
NP
P
/1/
\
art, N prep. ait
the blackbirds in thc park
Figure 7.2
The elements on one level are constituents of those on a higher level,
immediate constituents of those on the level immediately
above. Words are immediate constituents of a phrase. In a simple sentence like
The teachers died that phrase may be an immediate constituent of the sentence, a
clause. In more complex sentences
there may be more than one phrase and inure than one clause between word
and sentence.
To understand the grammar of a language we need to know what can serve

as constituents of an element. We need to know that a noun phrase must contain a
noun or a pronoun. We need to know that a noun other than a proper noun like
Fred needs a determiner
which may be the definite article (the), an indefinite article (a or an), a
demonstrative adjective (such as this), and so on. We need to know that a noun
may be qualified by an adjective. We need to know that these all precede the
noun.
Our awareness of what is and what is not a well-formed utterance helps us to

understand almost immediately, and to discard as soon as possible false
interpretations. As soon as we hear the we expect a noun, but not necessarily
immediately. If the next word is very we expect an adjective to follow If the word
that follows that is old we are probably expecting a noun to complete the noun
phrase, but we may already have the complete noun phrase if old is being used as
a noun. If we then encounter fear that might be a noun to complete the
noun phrase or it might be a verb. We need to see if there is still a
verb to come before we can select the correct interpretation. And so the
process of decoding goes on.
Within the clause The very old teachers greatly fear the blackbirds we can insert
another in order, for example, to further determine the teachers; we might say The
very old teachers you met yesterday greatly fear the blackbirds. The structure of this
sentence would be repres
Syntax
NP VP
\
ii S
/
NP NP VP NP
art ad v_ adj. Npron. Vadv.adv. V art,N
The very old teachers you met ye-sterday greatly fear the hlackhinls
Figure .3
ented as shown in Figure 7.3. Incorporating a clause within another clause in this
way is embedding.
A constituent structure representation may clarify an ambiguity in a sentence_ If we

were told that old teachers and priests fear blackbirds, we might wonder whether
all priests or only old priests fear blackbirds. If the adjective old was meant to
qualify priests as well as teachers, the sentence would be represented by the
diagram shown in Figure 7.4, If the adjective old was meant to qualify feathers
NP VPNP VP
,/
/
//
N conj. N V N adj.N conj, N V N
Old teachers and priests fear blackbirds Old teachers and priests feu blackbirds
Figure 7.4Figure 7.5
alone, the diagram would be as shown in Figure 7.5. What one is doing in the
case of the former is like the bracketing that a mathema
tician uses to distinguish 3(y + z) from 3y + z. Indeed linguists may also use
bracketing; this latest sentence might alternatively be represented as follows:
[[[Old teachers] and priestsiffear blackbirds]]
.4 Noam Chomsky
The phrase the teachers is composed of an article and a noun. The
Swedish equivalent, larariza, is simply a noun, the article being suffixed to it. nut
in both cases we are dealing with a noun phrase.
Thus the further one delves below the forms that the realisation of utterances takes in
individual languages, the more one finds commonality between languages_
Incidentally, given my use of the expression delves below, you might like to
consider whether it might be more appropriate to present the constituent structure tree
diagram the other way up. If nothing else, it would then look more like a tree!
Anyway, back to commonality Constituent structure analysis facilitates the quest for
the universals of human language. Understanding and representing the nature of these
universals was a major goal for Noam Chorrtsky, a linguist at the Massachusetts
Institute of Technology who introduced fresh approaches from the second half of the
1950s onwards. He recognised that surface structure is fairly arbitrary. The sentences
John is eager to please and John is easy to please have a superficially similar structure
but they convey a substantially different message; in the former John is doing the
pleasing and in the latter he is on the receiving end. Conversely a similar message may
be conveyed by means of quite different structures. Where we would say Do you like
my wife? an Italian would say Ti piace mia moglie?, literally Does my wife please you?
In Middle English we used a construction like that used in Italian; in The Canterbury
Tales Chaucer wrote How lyketh thee my wyf and hir beautee? , the phrase my wyf and
hit beautee being the subject with the consequence that the verb is in the third person
and the pronoun is in the object form, thee as opposed to thou. For Chornsky it was
essential to get below the surface realisations to the underlying principles, to the deep
structure. The earlier linguist Ferdinand de Saussure, published posthumously in 1916,
had distinguished between the fundamental langue, the language system within which a
person produces his utterances, and the superficial parole, the actual utterances
produced. Chomsky made a similar distinction, using the term competence to denote
a person's underlying awareness of the rules of his language and performance to denote
the actual use of the language, The generation of the surface forms was accounted
for by transformational rules. Converting the active construction My wife pleases
you into the passive construction You are pleased by my wrie, for
example, conforms to the transformational rule
So, too, it is the word order — supported by our experience of the world in
which we live — which tells us that in The child gives the teacher the letter it is the
letter that the child is handing over and that the teacher is the recipient.
Confronted with The child gives the letter the teacher we would be confused, it being
difficult to reconcile what our grammatical understanding suggests with the fact
that we do not usually give teachers to letters.
Word order, then, is very important for the understanding of an utterance.

This is particularly true in the case of a language such as English that is largely
analytic. But there are other devices available for indicating syntactic function.
Japanese marks the function of a word by placing a free-standing particle

after it. Thus, for example, the particle that, using Roman script, is written ga
indicates the subject of a sentence, the particle ni indicates an indirect object and
the particle a indicates a direct object.
Thus the Japanese equivalent of The child gives the teacher the letter is
Kodomo ga sensei ni tegami o ageru, literally Child (subject) teacher (indirect
object.) letter (direct object) gives.
The German equivalent of this sentence is Vas Kind gibt dem Lehrer den Brief.
In this case the form of the definite article helps to indicate
the relationship between the elements of the sentence_ The das of das Kind
indicates either a subject or a direct object, the den of den Brief indicates a direct
object and the dem of dem Lehrer indicates an indirect object. Where there is a
relationship of this kind between form and function we may talk of a case system.
Traditionally the subject form is referred to as the nominative case, the direct
object form as the accusative case and the indirect object form as the dative case.
In our German example only the articles reflect the case. In other languages with a
case system there is usually greater variation in the form of the nouns; in Russian,
for example, the equivalent of teacher is yturre.rib in the nominative case, ytinTarm
in the accusative case and yinitTemo in the dative case.
These languages also have a genitive case which indicates possession; the German
equivalent of the teacher is child is das Kind do Lehrers. A comparison of the
teacher's and des Lehrers shows a common feature, the suffix indicating possession.
English has not always been as analytic as it now is; Old English had a well-
developed case system and we have a vestige of it in the possessive form of the
noun. The Old English ancestor of father was fxder, this having a
Syntax
genitive form fxcleres, We have retained the case system, too, in the personal
pronouns. While The child saw the dogs relies on word order
for its meaning, The dogs saw the child meaning something quite different, we
cannot change the meaning of He saw them by a simple
rearrangement because he and them are marked for the functions of subject
and object respectively.
As an alternative to The child gives the teacher the letter we could say The
child gives the letter to the teacher. As an alternative to the teacher's child we could
say the child of the teacher. In each case we would
be using a preposition to indicate the relationship between the
elements concerned, the preposition to to indicate direction towards an
indirect object and the preposition of to indicate possession. As English became
more analytic, as it relinquished the case system, it
had to rely more heavily on prepositions. Similarly, the Romance
languages came to rely more heavily on prepositions than their ancestor Latin
had, because for the most part they abandoned the case system; the equivalent of
the father's was patris in Latin, the genitive form of pater, this becoming del padre
in Spanish.
Summary
Syntax is the study of the structure of utterances. It is commonly

distinguished from morphology on the basis that syntax deals with combinations of
words whereas morphology deals with the internal structure of words, but this
distinction is not clear-cut. To understand how words may be combined to
construct utterances we need to distinguish between word classes. Using word
classes we can devise phrase structure rules. We must recognize
that a sentence is composed of groupings of words rather than just being a
linear concatenation; by means of tree diagrams constituent
structure analysis illustrates the grouping of words into phrases and clauses.
Noam Chomsky built on such an arrangement to argue that mankind has a
universal grammar and to illustrate how that is
transformed into a language's surface structures.
The means employed by a language to indicate the relationship between the

elements of a sentence include word order, a case
system, prepositions and particles.
An Into Linguistics
Exercises
.1 Do you think that morphology and syntax should be treated as

separate areas of study? Give your views and support them with reasons.
7,2 In the following sentence there is one example of each of the

following word classes: noun, pronoun, verb, adverb, preposition and
possessive adjective.
Her dog always sleeps under it.
Rearrange the list of word classes to reflect the order in which they are
represented in the sentence,
7.3 Write five noun phrases in English. Use a different

combination of word classes in each case and incorporate a clause in one of
your examples.
7.4 (a) Construct a sentence in English that has the following

structure:
S > art., adj., N! V, prep_, art., N.
(b) Represent the structure of your sentence in the form of a tree diagram.
.5 Prepare a tree diagram for the following sentence:

The dog he bought likes young children,
7.6 The dog bit the man in the cat
The phrase in the car could be being used (a) to indicate where
the biting took place or (b) to specify that it was the man in the car that
was bitten. How would the tree diagram for each
differ?
7.7 Explain the difference between the situations represented by

the sentence in (a) and the sentence in (b):
NPVPNPVP
112
NN V ?rep. N NV N prep. N
Copper sham% sink after leak Copper shares sink after leak
(a) (b
Syntax
7.8 The very old fear
Write two sentences that begin with the above, one in which fear is a noun
and one in which it is a verb. Incorporate a
prepositional phrase in each sentence.
.9 *Us visit she on Sundays.
(a) Explain why the above is ungrammatical.
(b) Which personal pronoun in English would have the same form whether it
occupied the position of us or she in the above?
.10 If English still had a case system, which phrase in the following
sentence would you expect to be (a) in the nominative case, (b) in the accusative
case, (c) in the dative case?
My youngest son threw the ball to the white dog.
Regional Variation
or
Div ye Ken Fit he's Spikkin Aboot?
.1 Variations of Variations
'Hello, it's me.' That may be all that we need to hear to recognise somebody
who rings us up if we know them fairly well, for no two people speak exactly alike.
His or her speech profile has been shaped by such factors as where he or she grew
up, by the kind of education that he or she received, by his or her occupation and
by the same range of factors in respect of his or her parents. His or her speech
profile is, of course, affected by whether it is his or her that is appropriate. It is
affected by age. It is affected by many other physical differences that influence the
individual's acoustics, that make up the individual's voice quality. The unique
nature of a person's speech is his idiolect. If, then, we played a recording of a bus
driver in Bolton, Lancashire and a bus driver in Brighton, East Sussex it is likely
that most people would be able to identify which driver was from which town if
presented with the two alternatives. Similarly, most people could distinguish between
a bus driver from Bolton and a judge from Bolton. In principle, then, a person's
idiolect can be plotted on a matrix with a regional axis and a social axis. A
postman from Preston might be plotted as A on the matrix shown in Figure
8.1 while a doctor from Dover might be plotted as B.
In this chapter I shall be dealing with variation in speech between regions.
Social variations are the subject of the next chapter.
If you do not recognise a caller you may at least be able to guess which part
of the country he comes from. The extent of our ability to do so clearly depends
on how well acquainted we are with the speech variety concerned; if the caller is
from Aberdeen in Scotland
Regional Variation
Highest socio-economic group
A
Lowest socio-economic group
North ofSouth of
England England
Figure 8.1
an American might correctly recognise that he is Scottish, while a Scot is

more likely to know that he is from the Aberdeen region and not from, say,
Glasgow. The fact that we can at least to some extent tell where a person grew up
on the basis of his speech suggests that the characteristics of an individual's speech,
while ultimately
unique, are in large part common to those who grew up in the same
region. Where there is a systematic set of features, phonological, lexical, and
so on, that is shared by a subgroup of those who speak a language but not by
others we can refer to the speech of that group
as a dialect. If some features are shared by several but not all the dialects of a
language, we can perhaps distinguish a regional standard (see Figure 8.2).
Any given individual who grew up at C, some point on the spatial continuum AB,
has a unique variety of speech which in principle belongs to a particular dialect,
which in turn belongs to a particular regional standard, which in turn belongs to a
particular language. Thus, for example, a person from Boston, Massachusetts,
speaks the same language as a person from Boston, Lincolnshire. But their
Language
Regional standard
Dialect
A
Figure Si
Regional Variation
, ... ,
-...
,'....
iI
C
S
Figure 8.3
question by walking from Bolton to Brighton and visiting a bus garage every
day. If we found no change at all for the first three days, then on the fourth day
encountered several new sounds, several new words and one or two new syntactic
features, this new set of features persisting for three days, and so on, we might
have little difficulty in identifying several discrete dialects. But it is very unlikely
that we would find such a neat arrangement. Figure 8.3 represents what might be
the spatial distribution of two features which distinguish between the speech of
town A and the speech of town B. In such a situation do the people of town C
speak the same dialect as the people of town A, do they speak the same dialect as
the people of town B or do they have a dialect of their own which shares features
with the dialects of A and B? If these are the only two features that vary in the
region, can we justify any dialect split at all? As one progresses on one's journey
between two points, different features will wax and wane, each doing so at
different places and with greater or lesser abruptness. Thus we do not even have
the precision that the lines on the diagram suggest; they merely represent the point
where one form loses its majority to another, where one form ceases to be spoken
by as much as 51 per cent of the people concerned. We generally experience a
dialect continuum rather than distinct threshholds. One could travel from Lille to
Lisbon, that is from a city in which French is spoken to a city in which
Portuguese is spoken, and find little abrupt change in local speech as one passed
through the areas of such varieties as Occitan, Catalan, Castilian and Galician.
Recognising the elusive nature of dialect boundaries, dialectologists often identify
focal areas and
refer to the areas in which features change their form as transitional
areas. An analogy has been made to the colours of the rainbow;
there will be a point on the spectrum that most people would label orange
and another that most people would label yellow but it is difficult to determine a
point where orange gives way to yellow.
To quote W. N. Francis, 'dialect boundaries are usually elusive to the point of
non-existence' (Francis, , p. 1). Some linguists would
argue that the concept of a dialect is so elusive as to be useless.
If, however, one chooses to persist in the attempt to distinguish dialects, one
has, then, to be prepared for a bewildering array of
boundaries, of isoglosses (lines which delimit the area in which a
particular feature dominates). One can only look for fronts with a relatively
great coincidence of boundaries, fronts where there are what one might consider to
be isogloss bundles. On our walk from Bolton to Brighton we would be likely to
find more changes on the section between Birmingham and Oxford than on the
longer sections
between Bolton and Birmingham and between Oxford and Brighton.
In Birmingham as in Bolton a driver is likely to refer to his vehicle as a
/bus/; in Oxford as in Brighton he would say /bAsi. When talking
of a long bus or the last bus those on our route who say /bus/ are
also likely to say /13n/ andwhile those who say /bAs/
are likely to qualify it with /In/ and ila:sti. If there were only two
dialects in England, we might feel, the boundary between them must pass
between Birmingham and Oxford. And indeed the most substantial dialect boundary
in England does pass to the south of the West Midlands (see Figure 8.4).
In France the Atlas linguistique de la France published by Jules GilHeron at

the beginning of the twentieth century showed a major division between the speech
of the north of France and that of the south, this division running from the region
of Bordeaux north to that of Limoges and east to Lyons (see Figure 85). Like
developments elsewhere, the speech of the south has increasingly approximated to
Standard French but it still retains such distinctive features as the retention of
nasal consonants as opposed to the nasalisation of the preceding vowel; thus bon,
meaning good, is
pronounced /boa/ as opposed to /136/.
In Germany a major isogloss bundle was identified as running from west to east to
the south of Berlin, its significance being that it marked the northern limit of the
consonant shift that distinguished High German, In the region of the Rhine,
however, as Figure 8.6 shows, the isoglosses fan out, the areas in which the
voiceless stops
Regional Vil ria t ion
Bolton
Burlingham
•
Oxford
Bnghicm
Figure 8.4
Figure 8.5
became fricatives being different for each sound; the development

/lc/ > /x/ spread beyond Cologne whereas the development /p/> /f/ stopped
short of Cologne and the development It/ > Is! stopped further south still.
Dialect areas generally surround a centre of cultural and/or political influence. The
boundaries between them may be sharpened by obstacles to communication such as
a range of mountains or a political boundary. It has been suggested that the
boundary that
,
' Cologne '1t
i
...•- -- i
4 1
- Koblenz
Trier.
Figure 8.6
meets the east coast of England at the Wash has settled where it is because there
the pull between the population mass of the north and the population mass of the
south are equalised, a state of equilibrium having been attained. One can add that
inlets like the Wash and obstacles like the Fens minimised the interface between the
two varieties. In the case of the Rhenish fan the isoglosses for vs. If! and for It/
vs. Is/ have been equated with the northern and southern boundaries respectively of
the diocese of Trier. As it flows through the Netherlands the Rhine itself constitutes
a major dialect boundary; reference is often made to speech 'north of the rivers'
and 'south of the rivers', the rivers concerned being the Lek and the Waal,
branches of the Rhine.
As a change spreads the front is often drawn by urban centres. In so far as

the people of the towns are more mobile and exert greater influence, changes tend
to establish themselves in the towns and then spread to the surrounding areas.
Thus the speech of the more dominant region - the south of England, the north of
France and the south of Germany - will to a greater or lesser extent also be found
in the towns of the less dominant region.
.3 Vertical Definition of Dialect
We have seen, then, that it may be difficult to define the dialect that is
spoken by a person in Bolton or a person in Brighton but that we can safely say
that they speak different dialects. At a higher tier we
Regional Variation
can say that they both speak a dialect of Standard British English. One might
arguably propose that one should recognize an intermediate tier and say that a
person from Bolton and a person from Brighton speak a dialect of a northern
England regional standard and a dialect of a southern England regional standard
respectively. In any event the speech of the person from Bolton belongs to a
different regional standard to that which encompasses the speech of
a person from Boston, Massachusetts. An Englishman and an
American, then, speak varieties that belong to different regional standards. But
they do speak the same language; an Englishman and a Dutchman do not.
As we said, we are imposing categories on a state of affairs that is in actual

fact less clearly differentiated and there is scope for debate about any
categorisation. Arguably the distinctions between dialects, regional standards and
languages are as arbitrary as those between hills and mountains, between streams
and rivers. Perhaps it is only
because we have these labels that we seek a differentiation between
concepts. When discussing the relationship between language and thought
Hudson
suggested that the reason why we assume that dialects and languages are
distinct may be because modern English has different names for them; if we had
been brought up speaking English before the Greek word dialect was introduced we
might well not
have made this assumption, and there would have been no need
to question the reality of the distinction. (Hudson, 1996, p. )
Let us nevertheless assume that there is a substantial difference between a

dialect and a language, and consider how one might
determine whether two speech varieties are dialects of the one language or
whether they belong to separate languages. Almost everybody would agree that most
Englishmen and most Americans speak the same language, English, and that
Englishmen and Dutchmen speak different languages. It is generally considered that
Flemish, spoken in Belgium, is a dialect of Dutch but that Afrikaans, a language
of South Africa derived from Dutch, has attained the status of a separate language.
On what basis do we make such
distinctions? One might reasonably reply that most Englishmen can
understand most Americans but they cannot understand the Dutch. This
distinction has a number of flaws but it is about the best that
linguists have come up with; to use the linguists' terminology, the distinction is based
on mutual intelligibility. A language may be considered to be a speech variety that is
spoken by everyone who understands it, while a dialect is a speech variety that is
spoken by only some of those who understand it. To quote W. N. Francis's definition of
a dialect, dialects are 'varieties of a language used by groups smaller than the total
community of speakers of the language' (Francis, 1983, p. 1). If people use varieties
that are not mutually intelligible they would be deemed to speak different languages.
A major deficiency of this yardstick is that intelligibility is a matter of degree, not

all or nothing, An American might on occasion have difficulty understanding an
Englishman. On the other hand an Englishman might on occasion understand something
of what a Dutchman says; a Dutchman who wants you to know that his book is under
the clock might well be understood by an Englishman when he says /man buk iz onclar
d kbki. Swedes and Norwegians generally understand each other without much
difficulty; Swedes and Danes find it more difficult to understand each other. One or
both speakers may accommodate their speech to that of the other
person or to a more standard variety by avoiding certain features that they feel
would be unfamiliar to the other person. Mutual intelligibility is greater when people are
using a standard variety of speech than when they are using a local variety. The
communication
might be assisted by context; a person from Aberdeen who pronounces the word speak
with a short vowel, as ispiki, is less likely
to be understood if he says They cannot speak than he is if he says They cannot
speak English, an extra clue having been given by refer
ence to a language.
The situation is further complicated by dialect continua. People living near the
edge of a language area may understand people in the neighbouring language area better
than they understand people living on the other side of their own language area. People
in the
north of Germany, for example, may well find it easier to
understand a Dutchman than to understand an Austrian despite the fact that the
Germans and the Austrians nominally speak German
while the Dutch do not.
The mutuality may not be in balance; one participant in a conversation may

understand more than the other does. People from a
Regional Variation107
States by a switch. In Britain a motorist puts his luggage in the boot, in the
United States he puts it in the trunk. Inciden_tally, the large number of differences
in the terminology relating to railways (or railroads!) and the motor car is due, it
has been suggested, to the fact that these technologies have been developed since
Britain and the United States became separate countries. What is called a stream in
the south of England is called a beck in that part of northern England once settled
by the Danes, the word beck being of Norse origin. In Scotland it is called a burn,
Where a person from the south of England might say The girl knows the stream a
Scot might say The lass
kens the burn.
There is a general tendency for dialects to become tempered, for them to drift
towards the regional standard. Part of this is the loss of some of their
characteristic vocabulary; lexis, being modular rather than systematic, is relatively
easily substituted. This standardising of varieties is stimulated by social factors such
as mobility and education, these being more significant among some sectors of the
population than among others; we are now ready to deal with social
variation, the subject of the next chapter.
Summary
Each person has a unique speech variety, his idiolect. Yet many millions of
such people speak sufficiently similarly that they are considered to speak the same
language. Between these extremes we may propose more or less local dialects and
regional standards.. It must, however, be appreciated that any attempt to impose
such categories is to a greater or lesser extent an arbitrary representation of the
actual situation. It is difficult to define dialects as different features will change
more or less abruptly and in different areas; there is generally a dialect continuum
rather than a distinct succession of dialects. One can more easily identify a focal
area and a transition area, the former generally being based on a centre of cultural
or political influence. It may also be difficult to determine whether two speech
varieties are dialects of the same language or
belong to different languages. The distinction is usually based on
mutual intelligibility but this can only be a guide.
Exercises
.1 How significant do you consider the concept of a regional

dialect to be?
8.2 Explain the significance of the concept of dialect continuum.
8.3 How useful is the concept of mutual intelligibility to

dialectologists?
8.4 What factors would you consider in arguing whether or not

Scots is a language?
8.5 How relevant is independent statehood to a consideration of the

status of a speech variety?
8.6 Where did you spend the greatest part of your childhood? State
some of the ways in which the speech of that region differed from what you
would consider to be Standard English.
- ,
U AAAAA 1Highest socio-economic group
-11 U AAAA
u u UAAA
Uuii uAALowest socio-economic group
NorthSouth
Figure 9.1
their socio-economic group one could produce a matrix for a feature of the
language such as the pronunciation of the vowel in the word bus. In simplified
form, such a matrix might resemble Figure 9.1. This would show that the higher a
person is on the social scale, the more likely they are to use the southern /A/, this
also being the vowel of the more prestigious standard. It shows that there is less
regional variation at the highest social level; to reflect this, the situation is often
represented by means of a truncated triangle (see Figure 9.2). This latter
representation removes the relationship
between upper-class individuals and the spatial dimension, but it
serves to emphasise the relatively short linguistic distance between individuals
of the upper classes.
Generally, then, people in higher socio-economic groups - and perhaps some who
aspire to be in such groups - have an idiolect that is less conditioned by the region
in which they grew up, that approximates more closely to a standard variety This
is likely to be a variety that they were exposed to by parents and friends. In
Britain it may have been consolidated by a public school education. Their profession
may require the use of a standard form of speech. They
\ Highest socio-economic group
,/
/
/
/Lowest socio-econoirtic group
Figure 9.2
Social Variationill
are likely to be geographically mobile. Thus,, as Figure 9.1 suggests,

it is likely to be easier to identify a bus driver's region of origin than a
judge's. And because the standard language is based on the speech of the south of
England it is likely to be easier to distinguish between a judge and a bus driver in
Bolton than between a judge and a bus driver in Brighton.
We saw in section 8.2 that dialects are not divided by the sharp boundaries
suggested. by isoglosses. A glance at Figure 9.1 makes it
clear that the situation becomes much more complex still when one introduces
the social dimension. The isogloss in Figure 8.4 is less valid for upper-class speech,
for at the higher social levels the southern features will be widely used to the
north of the line. Similarly, High
German is widely spoken to the north of the isoglosses shown in
Figure 8.6. These isoglosses, then, represent a distinction at grassroots level, at
the level of the stable rural population, a state of affairs that is greatly influenced
by social factors.
,2 The Standard Language
The higher one's social standing, then, the less likely one is to use a regional
dialect, and the more likely one is to use a more standardised variety. This is in
part due to the greater mobility of those on the higher social levels; the wider
one's social and geographical horizons, the more one's speech and that of one's
children will lose regional features. Another major factor is the prestige that is
associated with the more standardised varieties; a judge's authority is likely to be
lessened if he speaks a regional dialect, It is this prestige that sets a standard
language apart. Linguistically a standard language is just another dialect; its
origins are usually as humble as those of the other dialects. But socially it has been
elevated, put on a pedestal as the supreme variety. R. A. Hudson (1996, p. 33)
specifies four characteristics of a standard language: it has been selected from
among the varieties of the language, it has been codified, it is suitable for use as
an official, written medium, and it has been accepted by 'the relevant population'.
As it is codified, as it serves as a literary language, as it is perpetuated by the
education system, the standard language tends to be conser
vative, these factors acting as a brake on change. Being codified, it can be used as
a yardstick for assessing a person's 'correctness'. A standard language can also
serve as a symbol of nationhood.
An introduction to Lingui5tics
As the nation states in Europe developed, the centralised governments, assisted
by the invention of printing, spread the use of a particular variety. As one might
expect, the variety that rose to assume the role of the prestigious ideal was often
that of the seat of power. As we have seen, Standard British English is more akin
to the speech of southern England than that of northern England, the seat of power
being in the south. For similar reasons the speech of the region centred on Paris,
Francie-n, became the prestigious variety in France. Often, then, the standard form
of a language is based on the speech of the educated inhabitants of the capital city;
in the case of Danish, for example, 'the most prestigious pronunciation is that of an
educated Copenliagener' (Haugen, 1976, p. 39). In Italy, however, it was Florence
rather than Rome that provided the prestigious variety; Italy has only been a
political entity since the second half of the nineteenth century and the status of the
speech of Florence goes much further back, to the Renaissance. The prestige may,
then, have its roots in cultural rather than political influence. A variety may become
a standard language as the result of being adopted as a religious norm; standard
German and standard Arabic are in large part the result of a particular variety
being selected as the form for the Bible of Martin Luther and the form of the
Koran respectively. Often, of course, the development of a standard will be
influenced
by a greater complexity of factors than the above suggests; the
spread of Luther's East Central German, a variety of High German, was, for
example, made easier by the decline of the Hanseatic
League and with it the influence of Low German.
.3 The Urban Vanguard
At the beginning of the last section 1 referred to mobility and the expectations
of the professions as factors which tend to further the use of the standard language
or varieties close to it. Such factors tend to affect urban areas rather than rural
areas.
This complicates yet further the attempt to represent on a map boundaries between
features, for rather than advancing by pushing forward a front, standardisation
often establishes itself in urban centres and then spreads from them into the
surrounding area, Thus isoglosses may show urban islands surrounded by
unaffected rural areas. Trudgill (1995, p. 149) illustrates this state of affairs by
so, they might move towards the less formal style of speech that has been
more associated with men who, to quote Trudgill, are at a subconscious or perhaps
simply private level very favourably disposed towards non-standard speech forms', a
situation that has been ascribed to a greater concern with group solidarity than
with the desire to rise on the social scale. Men may associate masculinity with the
physical labour of the working class, it may be coarse language that helps you
attain your social goal when that goal is to
be 'one of the lads'.
According to Hudson, this correlation between sex and style of speech must be
regarded as 'one of the most robust findings of sociolinguistics' (Hudson, 1996, p.
194).
.5 Power and Solidarity
Women, then, tend to want to give an impression of high status more than
men do. Women are more concerned than men are with a vertical social dimension,
men setting greater store than women by a horizontal social dimension, by group
identity. In connection with
these two dimensions sociolinguists use respectively the terms power and
solidarity.
Whether it is power or solidarity that is more significant in a particular social

relationship depends not only on who is speaking but also on whom the speaker is
speaking to. Hello, Lizzie is an appropriate way of greeting a sister or friend called
Elizabeth; if you were talking to the Queen of the United Kingdom it would be
more appropriate to say Good morning, Your Highness. Saying Hello, Lizzie to the
Queen would be considered extremely disrespectful. If you said Good morning,
Elizabeth to your sister she might wonder what she had done to upset you. In a
conversation with the Queen it is the rules associated with power that are the most
relevant, there being a substantial difference in social status. In a conversation with
your sister it is the rules of solidarity that apply. The former require a more
formal speech variety than the latter do. If we want to be accepted by those
'above' us or to distance ourselves from those 'below' us we use more formal
speech. Using formal speech with our equals might give them the impression that
we consider ourselves
better than them or, as we have just seen, that they have done some
thing to upset us.
Social Variation
As we saw at the beginning of this chapter, our speech may be marked as

more or less formal by the choice of vocabulary or by whether or not segments
are elided. We may make our speech more formal by such means as avoiding the
imperative and using more
complex formulae; May I see the photograph is more polite than Show
me the photograph.
If we wish to use a person's name when talking to or about them we need to

decide whether it is appropriate to use their given name (e.g. Elizabeth) or their
surname (e.g. Miss Smith). Whatever our social relationship we can, however,
always address that person with you. In most European languages, on the other
hand, there are two equivalents of the singular you, a formal form and an informal
form. In French these are vous and tu respectively and the linguist often refers to
these two varieties as the V form and the T form. The
Spanish equivalents are usted and t, the German ones Sic and du,
the Russian ones BM and Tbl.
Generally the informal form is gaining ground as society becomes more

egalitarian and the criterion of solidarity becomes more
significant at the expense of that of power. In France, for example, it was once
common for children to address their father, a figure of power, with vous, but now
most address their father with. tu, the determining factor being that he is a close
relative.
In some oriental languages such as Japanese and Korean the relationship

between the speaker and the person to whom he is
speaking has to be reflected more extensively in the form of the
utterance. In Japanese verbs have a basic form and a polite form; imasu, an
equivalent of to be, has a deferential form, irasshairnasu, that elevates the person
being spoken to and a humble form, orimasu, that indicates the modesty of the
speaker.
A noun may be prefixed by a- as a mark of respect for the addressee; if a

person is asking somebody what his name is he may
say o-riamae rather than just namae.
Our relationship with somebody may affect not only how we address him but
also what we say. If somebody rings you on the telephone you may feel an
obligation to ask him/her how he/she is if that person is a friend. Here we are
touching on the rules govern
ing a conversation, the requirements that our social values impose on what we say
and how we say it. This is an area of study called discourse analysis.
.6 Registers and Diglossia
If we are talking to a judge we may be in court, perhaps as a lawyer, perhaps as

a witness. If we are talking to a friend we may be playing golf or discussing a personal
problem. Thus, related to the consid
eration of who we are talking to is the consideration of the social context in which we
are talking. We interact with others in many
different situations. I am an employee, a colleague and a teacher, a son, a
husband and a father, a member of an archery club, a neighbour and a dental patient.
While differences reduce as society becomes less formal, my speech will vary somewhat
between roles. A speech variety that is appropriate to a limited social context is known
as a register. Alternating between registers is known as code-switching or style-
switching.
Such alternation means that once again we need to refine the diagrams presented
in chapter 8. We have seen that there are no clear-cut guidelines for grouping the
speech of individuals into dialects and languages. Now we are seeing that even the
speech of the individual, the idiolect, is variable. The postman from Preston and the
doctor from Dover represented in Figure 8,1 will use different registers when talking to
members of their family and when talking to members of the public in the course of
their work. We should, then, now refine that diagram as shown in Figure 9.3 to reflect
the fact that their speech can vary over a range of formality, the <A> being at the
centre of the range of the postman and the 
being at the centre of the range of the doctor.
Trudgill's study of the variation between iii3/ and /in/ in

Norwich showed that it correlated not only with the social class of the speaker but
also with the formality of the situation. Middle-class
people generally use the standard /iu1 in casual speech, whereas working-class people
generally use the non-standard /in/. All groups increase their use of /hi/ in formal
speech, particularly those in the middle class, those who had already preferred it in
casual speech (Trudgill, 199.5, pp. 93-4).
These findings reflect those of the pioneering studies of the American sociolinguist
William Labov who examined the speech of the Lower East Side of New York in the
1960s, looking at the articulation of a number of phonemes that varied between
individuals and, indeed, between the registers of the one individual. In New
Social Variation
Highest socio-economic group
Lowest socio-econornic group

North ofSouth of
England England
Figure 9.3
York the initial phoneme of the word think may be articulated as /t/. Labov found
that the middle class - the highest class that is normally found in the district -
produced /0/ in over 80 per cent of cases when speaking in a formal register, and.
he found that the incidence of t/ increased with people in a lower social position
and at every level increased in less formal registers.
Clearly if a person uses an informal variant of a particular feature there is a

greater chance that they will use informal variants of other features in the same
utterance. If a New Yorker says He's thinking as opposed to He is thinking the
likelihood that he will pronounce the word thinking with /t/ as opposed to /0/ is
increased, There tends, then, to be co-occurrence of features. When such co-
occurrence is so systematic that there are two distinct varieties for formal contexts
and informal contexts respectively, we can say that there is a state of diglossia.
Where there is diglossia the social position of the speaker may not be a significant
factor, the choice of variety being determined principally by the degree of
formality of the context. One may argue that there is diglossia in the south of
France, for example; Dennis Ager (1990, p. 45) tells us that Occitan, the long-
established variety spoken to the south of the isogloss in Figure 8.5, is used in the
home and in informal situations while French is used in formal situations and in
administration, that Occitan is used to speak of basic human needs and during
leisure activities while French is used to speak of intellectual matters, Further
south, on the other side of the Pyrenees, the repressive centralist policies of Franco
in the middle of the twentieth century resulted in Spanish being the lang
Social Variation
King James Version. We can avoid profanity by amending a word or phrase;

thus heck can be used instead of hell, jeepers creepers can be used instead of Jesus
Christ. Such alternatives to unacceptable terms are known as euphemisms. In The
Linguist (no. 5, 1995), the journal
of the Institute of Linguists, the translation specialist Peter Newmark
suggested that a euphemism might be defined as 'an inoffensive or embellishing
word or phrase substituted for one considered offensive, frightening or hurtful, especially
one relating to religion, sex, crime, death, drunkenness, violence, excretion, mental health
and serious diseases'.
The French verb baiser, once meaning to kiss, came to denote a more intimate
activity, with the result that an alternative, ernbrasser, came to be used to denote
kissing. A euphemism may be evasive, discreet, but if it is to be of use it must, like
any other word, be unequivocal* associated with the concept that it denotes. Most people
understand that sleeping with somebody implies something more stimulating than getting
a good night's rest beside somebody else. This being the case, the euphemism can
through time become as tainted as the term that it replaced. Thus a concept may be
denoted by a succession of terms. In a work published in 1974
Geoffrey Leech gave the word nigger as a clenigratory equivalent of
negro (Leech, 1974, p. 52); now, a quarter of a century later, many would be
unhappy with negro and would prefer a term such as black, coloured person, or Afro-
Caribbean.
We have now entered the field of political correctness. As society becomes more
egalitarian, more liberal, we are less sensitive to terms related to sex and other bodily
functions but more sensitive to the concerns of the disadvantaged; people with deficient
eyesight, for example, are now often described by the phrase visually impaired rather
than blind. Crap is on the way in, cripple is on the way out. At the end of section 9.5
I wrote 'to ask him/her how he/she is in the style of those concerned about the
traditional practice of using the masculine term to refer to people of either gender.
Elsewhere I have for the most part followed that traditional practice, partly because I do
not like anticipating the changes in our language, partly because such pairs as he/she are
cumbersome and partly
because I do not like the alternatives suggested; (s)he, for example, cannot be used in
the spoken language and it is alien to the English language to bracket part of a word.
It is perhaps a pity that the
third-person singular pronouns force us to choose one gender or
the other. Finnish uses one pronoun, //du, to refer to both men and women,
and indeed we make no distinction in the plural: they tells us nothing about the
gender of the people referred to. Indeed a few
paragraphs back I resorted to another way out that we use, sub
stituting the plural pronoun they for the singular a person,
.8 Slang
Taboo words are often avoided because they are unacceptable in many social
situations. Other words may not be widely used because they are not even known
to most people. A section of society may wish to reinforce its identity and exclude
other people. The social elite may do so by such means as social etiquette.
Linguistically it may do so by adherence to the standard language. Other sections
of society may strive to achieve the same end by going to the other extreme, by
using a variety that is so different from the standard language that it cannot be
easily understood by the uninitiated. Such a variety is slang. When slang is used to
conceal the activities of a group of people such as criminals it may
be called argot.
One example of slang is the rhyming slang associated with the East End of
London. This refers to an object, action, and so on by using a phrase which rhymes
with the standard term; thus a telephone may be referred to as a dog arid bone.
The exclusivity may be increased by omitting the element that rhymes with the
standard term; thus a Londoner may refer to his friend or mate as his china, the
first element of the phrase china plate. Some examples, on the other hand, have
been more widely adopted to the extent that
people are often unaware that the word has its origins in rhyming
slang; many people will not be aware, for example, that when they say the
phrase use your loaf as an alternative to use your head they are using an
abbreviated form of loaf of bread.
Another example is to be found in French which has a form of slang known.

as verlan. The basis of this form of slang is the inversion of syllables and the word
verlan is itself an example, it being the slang equivalent of revivers meaning the
reverse. Other examples include trome instead of metro and ren pats instead of
parents.
In a similar way certain ethnic groups may revert to a creole. In the s and
1960s immigrants to Britain from the Caribbean brought
Social Variation
with them varieties such as Jamaican creole. Over time this generally approximated
towards standard English. Some, however, particularly the young, who feel
themselves to be marginalised by society, may turn back towards the creole, in so
far as it can be readopted, in order to have an alternative identity.
Some of today's slang words will eventually be elevated to the standard

language. Some will disappear_ Slang words are an aspect of language that is
particularly prone to change. Language change is
the topic of the next chapter.
Summary
A person's speech is influenced not only by where he grew up but also by his
social background. At the higher social levels there are standardising factors which
result in less regional variation. The most uniform and prestigious variety is the
standard language, the variety that is used in official contexts, the variety that
provides the written norm. The standard language has generally been elevated from
amongst other dialects because it was the dialect of a centre of influence. People
who live in cities and women tend to approximate more closely to the standard
language than do people who live in
rural areas and men.
A person's speech is also influenced by whom he is speaking to; if the two

people concerned have a 'power' relationship the speech will be more formal than
if they had a 'solidarity' relatiohship. The significance of this depends on the
formality of the society concerned. Our speech varies between social situations.
Where two distinct varieties exist for formal and informal situations we may say
that there is a state of diglossia.
Some terms are avoided in formal situations because they may cause offence;
such terms are taboo words. Instead, euphemisms or politically correct forms are
used. Some terms are not widely used because they are exclusive to certain social
groups; such terms are slang.
Exercises
9.1 Why is it likely to be easier in the north of England than in the

south of England to determine from their speech whether people are hospital
porters or surgeons?
.2 Why might a person avoid using the standard language?
9.3 There are various suggestions why women tend to approximate

more closely to the standard language than men do. What do you consider to be
relevant factors? Why?
9.4 List five words or phrases which you would say to a friend but
not to your doctor. Against each give an equivalent that you would say to the
doctor.
93 If you wanted somebody to close a window, what might you say

if (a) you wanted your brother to do it, (b) you wanted a lecturer to do it?
or
How Wostow that I Was Nat an Astronaut?
.1 The Diachronic Dimension
Instead of the standard How do you know that ? a native of

Aberdeen in Scotland might say Fit wye div ye ken that .? Art
Aberdonian is more likely to use his local form if he is a bus driver

than if he is a judge. Both are more likely to use the local form with
friends than in a courtroom. As we have seen in the last two chapters,
language varies between regions, between social classes and between social
situations. Thus an item of language - a word, a syntactic structure, and so
on - may occur within a certain area of a
grid that has regional variation and social variation as its axes (see
Figure 10.1).
Such a grid represents a contemporary state of affairs. If we want to

understand how such a state of affairs has come about, if we want
to know whether it is fairly static or whether it is evolving, it is of great
help to look at a third dimension, that of time. It may be that
the situation is very much as it was, say, 500 years ago, in which
case it could be represented by Figure 10.2_ An item may have changed
and expanded its regional and/or social domain (Figure 10.3). It may
not have existed 500 years ago (Figure 10.4). In addition, it may
well have changed its phonological form and its semantic range.
When we study language as it exists at any given point in time we are
looking at it from a synchronic point of view. When we consider how it
has changed over time we are looking at it from a diachronic
point of view. Those who study language from this latter point of view are
working in the field of historical linguistics. Section 10.2
Social variation
Regional variation
Figure 10.1
looks at ways in which language changes. Section 10.3 looks at some of the
forces that give rise to changes.
.2 How Language Changes
Where we say How do you know that. . ? Geoffrey Chaucer might
have said How wostow that . ? From this we can ee two

or three
differences between the English of the fourteenth century
and that of
today. The verb witen, a cognate of German wissen, has
disappeared,
leaving only a few traces in such words as unwittingly.
The syntax of
Figure 10.2Figure 10-3Figure 10.4
questions has changed, the inversion of subject and verb having given way to the
use of do as an auxiliary verb. In section 7.4 we saw that Chaucer wrote 'How
lyketh thee my wyf and hir beautee?' Here again we see a question being formed
by the inversion of subject and verb. In this case the subject is iny wyf, the verb
lyke being used as we use the verb please today. It therefore follows that the verb
is in the third person, this being indicated by the morph -eth. We no longer
distinguish number or case for the second-person pronoun; we use
you throughout. The pronunciation of the long vowels has changed
since Chaucer 's time; he would, for example, have pronounced the vowel in
the word wyf as ii:/, as opposed to the diphthong /aii that most of us use today.
The word beau tee is a French import; 500 years earlier an Anglo-Saxon poet
would probably have used the Germanic fxgernis, the word that has come down to
us as fairness. These are a few examples of a myriad of changes - semantic,
phonological, morphological, syntactic - that have constituted the evolution of the
living organism that is the English language.
The lexemes of a language, the cells of the vocabulary, may expand their
semantic range at the expense of others. The word fowl once had the broad
semantic range that has now passed to the word bird, which once denoted a young
bird in particular; while the King
James Bible tells us that God said that there should be 'fowl that may
fly above the earth' (Genesis, chapter 1, verse 20) the New English Bible tells
us that He said 'and let birds fly above the earth'. The
semantic range of fowl has, then, been subjected to narrowing, that of bird to
extension. The semantic range of a word may undergo a change of status; the
word knight has been elevated, its Old English ancestor cniht having denoted a
youth or servant (cf. German Knecht). A word may fall out of use altogether, as we
have seen with Middle English witen.
The phonological form of words changes. The word bird has evolved from
Middle English brid. This inversion of segments is known as metathesis. It is
particularly common with the liquids [II and Er]; the Latin words periculosus
meaning dangerous, rnarmor meaning marble and formaticum meaning cheese (made
in a mould)
have given peligroso to Spanish, mpamop to Russian and fromage to
French.
Words can change because people lose sight of where they start and finish. The
word adder that denotes a kind of snake was known to Chaucer as nacldre and
Germans still use the word Natter, What has
for in both cases the intervocalic consonant evolved from /t/ to iol; the
Spanish word is a reflex of Latin vita and the Danish word is a cognate of the
Middle English word witem A similar change accounts for the fact that it may be
difficult to know whether an American or somebody from Northern Ireland is saying
writing or riding. Many phonological changes, then, are widespread, occurring commonly
in particular environments. Sometimes a language experiences a wholesale shift in a
large part of its phonological system. This happened to the long vowels of English in
the fifteenth and sixteenth centuries, each vowel becoming more close, the highest
becoming diphthongs as in the words wife and house. We call this shift the Great
Vowel Shift. Long before, German had been subjected to a shift in its system of
consonants, the Second German Consonant Shift. A major part of that shift was a
form of lenition, the voiceless plosives becoming affticates or fricatives depending on
their position in the word; thus what is pipe in English is Pfeife in German, what was
witen in English is wissen in German. With such shifts a whole series of segments
generally changes to perpetuate an evenly spaced series of sounds which efficiently
provides enough distinctive segments. But it may take a century or two for the
symmetry to be restored. The shift may be initiated by a segment at one end of the
series becoming more like its neighbour. They might coalesce or the neighbour might in
turn move to maintain a distinction between the two segments, in which case a so-called
push chain has been set in motion. If, on the other hand, the shift is initiated by a
segment at one end drawing away from its neighbour and the other segments in the
series move to fill the vacated domains we have a so-called drag chain. The vowel
shift in English in the fifteenth and sixteenth centuries is often described as a drag chain
shift (Potter, 1968, pp. 65-6; Barber, 1995, pp_ 191-3), the most close vowels having
been the first to move up. Lexernes may in certain circumstances lose their
independence and be reduced to the status of bound morphs. In the Romance languages
the adverbs like Spanish rapidamente incorporate a bound element that was once the
ablative form of the Latin word mens, meaning mind, the idea being that one is doing
something in a particular frame of mind. As nens was of feminine gender the
adverbial morph is affixed to the feminine form of the adjective. An
equivalent of I shall write to him quickly in Spanish is Le escribire rapidamente.
In the future tense we again see an example of a lexeme
being rnorphologised in the Romance languages for it originated in a construction

in which the infinitive was followed by habere, meaning to have; cf. Spanish escribir
meaning to write and he meaning I have, We may still see traces of the fact that the
adverb and the future tense once comprised distinct lexernes. When two adverbs are used
together Spanish only uses the adverbial_ morph in the second
one; the equivalent of I shall write to him quickly and affectionately it's Le ecriliré
rapicla y carifiosamente. When a pronoun is used with the future tense in Portuguese
the pronoun may be inserted between the infinitive and the future tense morph; thus I
shall write to him may be Escrever-lhe-eL
Conversely, there are many cases in which the Romance languages have developed
free forms_ An alternafive to escreverei is you escrever, you belonging to the verb ir,
meaning to go. I have been writing the book would be Tenho escrito o livro in
Portuguese. With these compound tenses Portuguese, like other Romance languages, has
developed functions for the verbs ir and ter that correspond closely to the way that we
use to go and to have in the future tense and the perfect tense respectively. These
forms are the result of slight semantic adjust
ments; in the future tense to go has lost the implication of physical
movement and in the perfect tense to have has lost the implication of possession,
the origins of this tense being in constructions like I have
the book; it is written.
A major feature of the development of the Romance languages has been the
general loss of the case system of Latin, its function having been assumed by such
mechanisms as prepositions and word order_ Similarly, English was not always as
analytic as it is today. If we hear Christ gives men a sign we know who is giving the
sign to whom because the convention in English that the subject should come
before an indirect object is now more rigid and because the verb is
in the singular. In Old English one also had the guidance of a case system; in the
Old English equivalent, Crist ge-sweotolap mannum, it is clear that it is the men who
are on the receiving end because rnannum is in the dative plural form_
This erosion of the case system facilitated the change in the use of the verb to
like that we saw earlier in this chapter, it becoming less clear who pleases whom. A
major example of syntactic change that we saw was the appearance of the verb do as a
dummy auxiliary in questions and negated statements.
Before leaving this section, we might consider for a moment that an awareness of
the ways in which lexemes change can allow us to amuse ourselves by exploring lexical
labyrinths, by constructing extensive word webs that trace a wide variety of words back
to a common source. Offspring that have survived the Middle English verb witen include
the words unwittingly and witness. The semantic equivalence and an awareness of
the sound shift that changed It! into /s/ between vowels in High German suggests
that our old word witen is a relative of the German word wissen, meaning to know.
From the node wissen we can link to such words as Gewissen which means
conscience and has a very similar composition to this Latinate word: an element
denoting knowledge prefixed by an element indicating comprehensiveness. In
Scandinavia the cognate of witen has been digested in the equivalents of somebody,
the original expression meaning something like I do not know who; this has
produced, for example, nokkur in Icelandic. Our words view and vista are both
derived from the feminine form of a Romance past participle of the equivalent of
the verb to see, the one from French vue, the other from Italian vista. Thus they
are offspring of Latin videre meaning to see. What we might find particularly
surprising is that if we go far enough back we can establish a link between the
web that includes unwittingly and that that includes view. We know because we see.
A word can travel abroad and, shaped by its travels, by the phonology of the
language it stayed with, come back to rejoin its cousins, Old English had weardian,
meaning to protect, this having developed into modern English to ward (off). From a
Germanic cognate of weardian French acquired garder which came into English as
to guard. Swedish has ftillstol meaning folding chair and its welltravelled cousin
feltiiij meaning armchair, the latter being derived from the French lexerne fauteuil
which in turn was derived from a
Germanic lexeme for a folding chair.
.3 Why Language Changes
We have seen that some phonological changes are widespread, occurring in

many languages, The more widespread they are, the greater the justification for
proposing some underlying physiological factor. In many cases such a factor may
be ease of articulation.
Having been introduced in chapter 4 to the mechanics of the production of speech
sounds we are in a position to appreciate that
Since then English has abandoned grammatical gender for nouns and
abandoned most of the case system, We retain the genitive in, for
example, the house's roof, something that we could abandon in favour
of the roof of the house. We retain gender and case with pronouns. We only have
gender in the case of personal pronouns, he and she, and if we can get by without it
in the plural we should be able to do so in the singular. As far as case is concerned,
we retain it in the pronouns,
but again we can argue that we can do without the distinction between I and me if you
can serve both as subject and object. Do we need the indefinite article a? Welsh
manages without an indefinite article. So do Arabic and Russian. The Russian equivalent
of I am not an astronaut is as minimal as He KoemoHaarr, literally I not astronaut.
Steven Pinker (1994, p. 181) gives a nice illustration of redundancy by pointing out that
we can understand what he is writing exvxn xf x rxplxcx xll thx vxwxls wxth xn "x"1.
Arabic script proves the point.
We can see such redundancy causing variations in our contemporary language. I

have used the object form whom in this chapter but it increasingly seems pedantic, at
least when it does not follow a preposition. Between you and me alternates with
between you and 1.
Judging by the misuse (if I may be so prescriptive!) of the apostrophe, it seems that it
may be no more necessary to distinguish in writing between the possessive banana's and
the plural bananas
than it is in speech. Such variation may be indicative of change in progress.
We speak not only of the house's roof but also of the sun's rays, this despite the fact
that the genitive form of sunne was sunnan. Having done away with gender, we have
taken one genitive form and applied it generally. So, too, the nominative plural form of
sunne was sunnan; we now say suns, having generalized the plural morph of such
masculine words as cnihtas, meaning youths. This is an example of analogy, another
development that can be ascribed to the minimisation of effort in that it produces a
system that is easier to learn and remember. Young children go through a stage in
which, having discerned a pattern, they apply it comprehensively, saying things like The
mans goed in the house, before they discover that there are some words to which the
general pattern does not apply. Irregularities tend to persist when they are so common
that there is little chance of us forgetting them, as with the plural form men and the
past tense of the verb to go. But occasionally an analogical form
An Introduclion to Linguistics
establishes itself and remains; in Old English and Middle English

the past participle of the ancestors of the verb to help was hofperi, it having since
been regularised to helped_
When we have reduced the linguistic distinctions to the bare minimum we still
receive guidance from the context of the utterance, from our experience of life. As we
have already observed, if we read.
on a butcher's door Sorry, no dogs we do not conclude that he has run out of mongrel
meat, If we read on the packaging of some product Keep out of the reach of children
we understand that it is the product that is dangerous, not children.
But clearly the loss of distinctiveness may nevertheless reach a point where there is a
danger of confusion. It has been suggested that Cinderella came to lose a glass slipper
because the French word for squirrel fur, vair, had coincided phonologically with the
word for glass, yen-e. This is an example of homonymic clash. Where such a clash is
deemed unacceptable an alternative to one of the words may come into use. English
once used the verb to let both in the sense of to allow and to obstruct; as these senses
are so contradictory there was a real danger of confusion and to let in the sense of to
obstruct gave way remaining only in one or two instances such as the phrase
without let or hindrance.
Language also changes because the communities in which it is spoken change.

Different cultures come into contact, one perhaps dominating another. Society's values
change, New discoveries and inventions have to be named.
The Anglo-Saxons were joined in England by Danes. These Germanic peoples

spoke varieties that had not yet diverged so much that they were mutually
unintelligible. But the differences in inflections caused complications. It has been
suggested that contact with the Danes not only gave the English language new words
such as take and die, the former ousting niman and the latter narrowing the semantic
range of steoifan which became our starve, but also
contributed to the erosion of its case system.
The arrival of the Normans in the eleventh century had a much more significant effect
on English. English was much more different from Norman French than it had been
from the Norse tongue of the Danes. Society was largely split between the Norman
masters and the Saxon subjects, the former continuing to speak Norman French and the
latter continuing to speak English. Both these factors — the
Historical Linguistics133
great linguistic divide and the great social divide - initially limited the extent
to which English was influenced by Norman French. When through time the social
division broke down the English language reasserted itself. But it was now an
English influenced by Norman French, particularly in those fields of activity that
had been the preserve of the Norman masters: administration, the law, chivalry,
hunting, and so on. Thus in the fourteenth century Chaucer wrote jugement as well
as doom, beautee as well as fairnesse, The social divide is reflected in the fact that
animals are generally known by
the Germanic terms, the resultart meats by the French terms: the masters ate what
the peasants reared. In the novel Ivanhoe by Sir Walter Scott we can read the
following:
'Nay, I can tell you more,' said Warnba in the same tone: 'there is old
Alderman Ox continues to hold his Saxon epithet while he is under the charge of
serfs and bondsmen such as thou, but
becomes Beef, a fiery French gallant, when he arrives before the worshipful jaws
that are destined to consume him. Mynherr Calf,
too, becomes Monsieur de Veau in the like manner: he is Saxon when he
requires tendance, and takes a Norman name when he becomes matter of
enjoyment.'
The lexemes derived from French, then, often co-existed with those derived from
Old English, the former perhaps having narrowed the semantic range of the latter.
The semantic range of a lexeme in the one language may be influenced by that of
a lexerne in the other. This happened to the Middle English verb faren which in
Old English had meant to go, to travel but came to mean to behave as well under
the influence of what was to become the French verb faire; in The Canterbury Tales
Chaucer wrote, for example, 'Ye fare as folk that clronken been of ale.' The dialect
in which Chaucer wrote, East Midland, became a prestigious dialect, not because it
was the dialect used by Chaucer so much as because it was the dialect of the
nation's capital city, London.
In the last chapter we saw some of the social factors that are influencing the
development of the language today, such factors as taboo, equality and group
identity. _
In conclusion we can look at a fairly widespread example of how a society's values

may be reflected in its language, In the past people's wealth was closely associated
with the amount of livestock
that they owned and this association is seen in many languages. In Old English
the word feoh denoted both money and cattle; it has
come down to us as the word fee and it is a relative of the German word Vieh
meaning cattle. A more distant relative, the Latin word
pecus, denoted livestock and has given us our word pecuniary. (Here we can see
another word web developing!) In Spanish the word for
cattle, gattado, is also the past participle of the verb ganar meaning to earn. The
Russian word for cattle, CKOT, once also denoted money or property it being cognate
with the German word Schatz meaning treasure.
.4 When Language Changes
When we know what changes took place we can attempt a little linguistic
archaeology to find out when they took place. We may be helped by knowing the date
of a text, but sounds may be obscured by conservative spelling_ The /k/ of Latin
centum has evolved by way of Its/ into the /s/ of French cent without this being
reflected by any change in the character used to represent them. But if we look at
German we see that the word Keller, meaning cellar, was adopted from the Latin word
cellarium and that it was followed by the word Zelle with the affricate segment Its/.
English has adopted the word chief from Old French and its cognate chef from Modern
French_ If we can assume that the pronunciation of the initial segments has not changed
since they came into English we can trace the development of Modern French /1/ by
way of /ti/ from Latin /k/ where it was followed by /a/, the source of these words
being the Latin word caput. The assumption that a sound in a borrowed word has not
changed since its adoption is, of course, hazardous. The word fine was adopted from
French with the vowel ii:/, as was the word machine. Our conservative spelling has
failed to reflect the fact that the vowel in the former has become the diphthong /aii as
a result of the vowel shift in English referred to in section 10.2. Thus if we know when
we started to use these two words we have an indication
of when the vowel shift was initiated by the diphthongisation of the long close vowels.
The resultant form of a word can tell us whether one change happened before or after
another. The French word cent, pronounced /sal, now has a nasalised vowel and no
nasal consonant; as the nasal
Historical Linguistics.5
consonant, In!, caused the nasalization of the vowel, the nasalisation clearly took
place before the loss of the /n/.
10.5 Divergence
At the end of section 102 I linked our word unwittingly with the German word
wissen and I linked our words view and vista. I then suggested that if one delved far
enough one could establish a relationship between all of these words. To find a
lexeme that was a source of both unwittingly and wissen one might have to go back
to a time before the Germanic peoples first came to England in the fifth century.
One would have to go back to a similar point in time to find anything like the
common ancestor of the French and Italian equivalents of to see. In the first case
we are going back to some West Germanic tongue, in the other we are going back
to Latin. To link all of these words together we need to go much further back in
time in search of a source that gave rise to both the Germanic languages and
Latin. The relationship between these languages can be shown in the form of a
family tree (see Figure 10.5). Here, too, reality is less abrupt than our
representation suggests. Synchronically, as we have seen in
chapter 8, there are continua, and diachronically there is a gradual
evolution; French, for example, gradually evolved from Latin rather than
Latin giving birth to French. Thus the more steps we show on these two
dimensions - for example Occitan on the synchronic dimension and Old French on
the diachronic dimension - the truer would be the impression given. Having said
this, the family tree is still very useful as a means of showing the lineage and the
degree of
relatedness of languages; Figure 10.5 shows clearly that Italian has
Proto-Germanic
Latin
West Germanic
//
EnglishGermanFrenchItalian
Figure 10.5
A D.Latin
West East
RomanceRomance
A,D.
A.U.PortSpan _1 CatFrenItalRum
Figure 10.6
evolved from Latin and that it is more closely related to French than to English. An
alternative means of representing the evolution of languages is as shown in Figure 10.6.
Observing that the English word foot is similar to the German word Fufi, that the
English word father is similar to the German word Vater, and concluding that the
similarity could not reasonably be accounted for by coincidence was hardly a major
achievement. So, too, with the French and Italian equivalents pied and piede, pere and
padre. The words in each pair must be derived from a common source or - less likely,
particularly with terms for such basic elements of human life as parts of the body and
kinship - one must be borrowed from the other. What was somewhat more laudable was
the hypothesis that all these words that denote feet and fathers ultimately derive from
common sources. The sound changes behind these correspondences were presented early
in the nineteenth century by the Dane Rasmus Rask and the German Jacob Grimm, he
who wrote fairy tales with his brother. They observed that words with /p/ in Latin
tended to have /f/ in the Germanic languages; thus where Latin had pe5, pater, and
pecus English has foot, father and fee. They observed that, similarly, the voiceless stops
/t/ and /k/ in
Latin correspond to /0/ and /x/ in the Germanic languages. This series of
correspondences became known as Grimm's Law. Having established this, linguists in the
nineteenth century were driven to attempt to delve into the mists of time and reveal the
nature of the common ancestor of Latin and the Germanic languages, the X of Figure
10.5. This common ancestor pre-dated any written evidence. That being so, the linguists
had to rely on the laws of sound change that they were developing, to decide which
sound was most likely to have given rise to those in the subsequent languages. It is
considered, for example, more likely that the Germanic fricatives If, 0, x/ developed out
of the plosives /p, t, k tha.n that the Latin plosives developed out of the fricatives, for
it is more common for consonants to weaken, to undergo lenition, than to strengthen,
undergo fortition. Trask illustrates the process of reconstruction using a selection of
western Romance lexemes (Trask, 1996, pp. 208-16). Having embarked on the quest for
the nature of this source language, it is natural to wonder where its speakers lived. It
has been suggested that linguistic reconstruction - identifying those features of life and
the environment which are denoted by related terms in the subsequent languages -
supported by archaeological evidence, points towards a culture of what is now southern
Russia. Common terms for flora and fauna have been adduced. But Bynon reminds us
that the meaning of terms may change and concludes that the 'location of the Indo-
Europeans in time and space must remain, for the time being at least, an open question'
(Bynon, 1993, p. 280). These speakers of the source language are called Indo-Europeans
because they have left their mark not only on the Germanic languages and the offspring
of Latin but also on most other languages in Europe and on languages as far away as
India. The spread of their language may have been principally due to migrations of
peoples or principally due to cultural transmission as, for example, the culture of
agriculture spread. The source language is referred to as ProtoIndo-European_ A
family tree for all the principal Indo-European languages of Europe is given in the
next chapter, in section 11.1. If Proto-Indo-European produced Proto-Germanic,
Latin and
others, which in turn produced such languages as English and German, French and
Italian, there has clearly been a process of
divergence at work during the course of the several millenia since the days of
Proto-Indo-European, a proces5 similar in some ways to Darwin's theory of
evolution that was launched on the world while
used by medieval doctors, and stool now denotes only the humblest of seats,
the Normans prefering to use the word that has come down to us as chair to
denote what they sat on. Allowing for changes in the real world in the course of a
thousand years, Swedish still uses the words halm, veta, leikare and tol in much the
same way as the Anglo-Saxons used healm, witan, l'ce and to/.
In Middle English the system of inflections was much weaker than it had been
in Old English. Adjectives still added -e in the plural and following a determiner
such as this. We have since eroded even this inflection but Swedish retains
something very similar. Since Middle English we have, as we saw earlier,
introduced the dummy auxiliary do into questions; Swedish still follows the old
practice of inverting the subject and the verb, as we still do with the verb to be
and the modal verbs.
While we say Do you smoke. ? then, we still say Are you rich?, not *Do you
be rich?. There are usually some recesses that changes do not reach. While we no
longer feel that we have to keep the verb in
second position and while the plural form of nouns that were once neuter has long
been assimilated to a common form, we still occasionally say things like Never have
I seen such skinny sheep. While Swedish has abandoned the case system, the phrase
till sjoss, meaning at sea, and others retain the genitive form.
.6 Convergence
We are, then, unable to speak to Swedes because English and Swedish have
generally grown further and further apart over the
course of many centuries.
But Swedes now stick paper together with tejp, their representation of our
word tape. Industrial unrest in Sweden might result
in a strejk. People on the move might listen to music using a freestyle. While we
are exposed to very little Swedish in comparison with Swedes' exposure to English,
we have acquired from Swedish the words ombudsman and orienteering.
So, while the languages have diverged, they have also influenced each other at a
superficial level. This is particularly true in this age of aeroplanes, television and
the Internet. But it is nothing new; the Danes were interacting with the English in
the ninth century. They brought us words like take and die, affected our system of
pronouns, and so on. As the Normans, as their name suggests, had Norse
linguistiche' which serve as guidelines for determining the relative age of

different forms which occur in neighbouring regions. Lateral areas, those with
poorer communications with the centre of influence, will, for example, tend to
have older forms, as we have
just seen with the Romance equivalents of to eat. A larger region will
tend to have older forms, but this norm may be overridden by the one above.
When languages share major features that transcend families it may be useful
to arrange them by such features rather than by family. This is the study of
typology. An example that is often used is the definite article in Albanian,
Bulgarian and Romanian; each of these languages belongs to a different branch of
the Indo-European family and yet each follows the minority practice of suffixing
the
definite article. This commonality has been ascribed to the heritage of Byzantium.
The significance of typology increases where it can be shown that one feature
tends to co-exist with others. Languages with the word order verb-subject-object
tend to put adjectives after the nouns that they qualify; this is seen in languages as
different as Welsh and Mixtec, an indigenous language of southern Mexico.
Languages with the word order subject-object-verb tend to place adjectives
before nouns, this being true of, for example, Turkish and Japanese. The more
typological correlations one can make like this, the greater our progress towards
the quest for linguistic universals.
.7 Pidgins and Creoles
As Figure 10.7 showed us, Middle English was the result of the fusion of Old
English and Norman French as the Saxon and Norman societies merged. Being a
language that developed to facilitate communication between two different cultures,
it is considered by some linguists to be a creole, a creole being a language that
may result when two cultures co-exist.
A creole usually develops out of a more limited pidgin. A pidgin is more limited in
its functions and, consequently, linguistically. It may have arisen, for example, as a
means of communication between traders from different cultures; the various
suggested origins of the term pidgin include corruption by the Chinese of the
English word business or the Portuguese word ocupacao. Or it may have arisen in
the context of slavery, as a means of communication
between slave and master and/or between slave and slave. It often arose in
the context of colonialism and the vocabulary is often largely based on that of the
language of the colonial power concerned: English, Dutch, Portuguese, French, and
so on. A pidgin typically lacks such features as grammatical gender, a copula and
a
passive construction. They generally lack articles and have a poorly
developed system of prepositions. Pidgins have a small vocabulary, with the
result that many things are denoted by circumlocution; to
take an example from Romaine (1988, p. 35), hair is denoted by gras bilong
hed in Tok Pisin in Papua New Guinea. They generally have little inflectional
morphology. As a result of the small vocabulary and the lack of inflections, words
are often used in what we consider to be different word classes.
Where a pidgin is spoken in a stable community such as a slave community it

may become the only language of subsequent generations. Being the only language,
it will need to meet all the linguistic requirements of its speakers, with the
consequence that it will cater not only for such functions as receiving orders but
also for such functions as phatic communion, expressing emotions, and so on. When
this stage of development has been reached the pidgin has
become a creole.
The development of the structure of different pidgins and creoles often follows
a similar path and that path has many similarities to the course of language
acquisition in children; the lack of articles and copulas in pidgins, for example,
reflects the lack of such items in the speech of 24-month-old children. This has led
many linguists
to turn to pidgins and creoles in the hope of finding universal factors
underlying language change and language acquisition.
Summary
The semantic range of a lexerne may change or a lexeme may cease to exist.
Its phonological form may be changed by such processes as assimilation,
palatalisation and lenition_. Lexernes may become
bound morphs. Syntactic changes include the widespread loss of a case system in
western Europe.
Some changes are the result of minimising effort. There is some scope for this
due to the redundancy that exists in language. Some changes are the result of
historical or social developments,
As language has developed it has often diverged, one variety evolving in a

differential way in different regions. It is generally accepted that a Proto-Indo-
European language spoken several thousand years ago increasingly developed
distinctive varieties leading to most of the languages of Europe. Such varieties
continue, however, to influence each other after they have gone their separate ways,
and features can spread between quite different languages.
Exercises
.1 Give four reasons - one lexical, one phonological, one morpho

logical and one syntactic - why an Englishman of the tenth century might find
the following question strange: Do the houses have a good view?'
10.2 The following is from The Canterbuni Tales:
And specially, from every shires ende Of Engelond, to Caunterbury they

wende, The holy blisful martir for to seke,
That hem hath holpen, whan that they were seke.
(a) The first seke is a verb, the second an adjective. What is the modern
equivalent in each case? The adjective had the basic form seek; why is it inflected
here?
(b) Why do we not have the word holpen in Modem English?
10.3 Give an example of semantic narrowing other than any given

in this chapter. Can you suggest why it may have happened?
10.4 Which kind of change has taken place in the case of the
following English words?
apron < Old French naperon burn < Middle English brenne
Account for the changes to the consonants in the following:

Spanish rniiagro < Latin miraculurn
.6 Scots has the words kirk and kist as equivalents of standard

English church and chest. What process has distinguished the initial segment of
the standard English words from that of the Scots words?
10.7 Discuss the merits and limitations of a family tree diagram as

a means of representing the synchronic and diachronic relationship between
languages.
10.8 What arguments might be presented to support a view that

Old English and Modern English are different languages?
The Languages of Western

Europe
or
Hur Vet du att Jag inte Var Astronaut?
.1 The Indo-European Faznily of Languages
Ten chapters have introduced us to the principal features of language. In this

eleventh chapter I shall relate these features to the major languages of western
Europe and thereby build up something
of a profile of each language.
As we have seen, most of the languages of Europe are considered to have a

common source, a hypothetical language spoken several thousand years ago, perhaps
in what is now southern Russia. The process of evolution and differentiation since
then has left us with the languages of today. The nature of this process is such
that each language is more closely related to some languages than to others. French
and Italian are relatively closely related, both having developed out of Latin. To
find an ancestor common to both French and German or to both French and
Russian one has to go much further back in time, This situation can be
represented by a family tree diagram such as that shown in Figure 11.1_ The
Brittonic group of Celtic includes Welsh and Breton, the Goidelic group Irish and
Scottish Gaelic. English, Dutch and German are West Germanic languages;
Icelandic, Norwegian, Swedish and Danish are North Germanic languages. French,
Catalan, Spanish and Portuguese are West Romance languages; Italian and
Romanian are East Romance languages, The Baltic branch includes Lithuanian and
Latviart and the Slavonic branch includes Russian, Ukrainian, Polish, Czech,
Bulgarian and Serbo-Croat.
An Introduction to
Linguistics
Proto-lndo-European
r_
Cell Proto-Germanic Latin Greek alto-S Lavon ic
c Albanian
FL-1 FL-1
iirition ic Goidelic West North West Baltic Slavonic
East
Oen-flank Romance
Figure 11.1
Figure 11_1 represents the principal extant Indo-European

languages of Europe. There are other varieties that might be included such as
Faroese, Frisian and Occitan. We shall be looking at the Germanic languages, the
Romance languages and the Celtic
languages, these being the groups that are relevant to western
Europe. Having done that, we shall briefly make the acquaintance of one of
the languages of Europe that is not a member of this family:
Finnish.
.2 The Germanic Languages
We begin with the Germanic languages, with the group of languages that includes
English_
These languages grew out of the speech of the Germanic peoples in northern
Europe, A change that did much to distinguish the speech of these peoples from
that of other Indo-European peoples was the First Germanic Consonant Shift, the
shift that we saw referred to as Grimm's Law in section 10.5. This shift changed
the voiceless plosives /p, t, k/ into the fricatives /1, 0, x or c/. Thus where Latin
had paler, tres and canis English has father, three and hound. The /k/ that was
retained in canis has also been retained by the Celtic languages. As a result the
Scots Gaelic word for a dog sounds very similar to the Germanic Scots word for a
cow, the one being elo the other coo. The words cannabis and hemp illustrate the
Germanic development of /k/ - they are cognates which denote products from
related plants. As the voiceless plosives developed into fricatives their quality was
adopted by the voiced equivalents. Thus the numbers that the Romans knew as duo
and decern are called two and ten by us_ Our words ignore and agnostic, the one
from Latin and the other from Greek, are cognates of our word know in which
principle but it is still seen in such sentences as Never have I heard

such a thing-.
Through time Proto-Germanic developed regional varieties that we call West

Germanic, North Germanic and East Germanic. North
Germanic, the Germanic of Scandinavia, for example, developed
such features as the loss of initial /j/ (while English has young and German
has (ung, Danish has Ung) and the addition of a determiner
to the end of the noun (while English has the boat and German has das Boot,
Danish has bilden). As we have seen from the family tree of
the Indo-European languages, West Germanic has given us English, Dutch and
German, and North Germanic has given us Icelandic, Norwegian, Swedish and
Danish. East Germanic has died out.
11.2.1 West Germanic
In the fifth century some Germanic peoples crossed to England and settled.
These peoples were for the most part Angles and Saxons. The country was named
after the Angles, Englaland (the Gaels in Scotland call it Sasunn after the Saxons)
and the speech variety that they developed was called Englisc. Since that time, as
we have seen, English has changed a great deal and we refer to their variety as
Anglo-Saxon or Old English. in the late Middle Ages the English spoke Middle
English and now we speak Modern English. (So how, you may wonder, will
linguists of the future refer to any subsequent
stage?)
Clearly there was no sudden break in the development of the language. With
that understood, it is helpful to identify stages in the development. The necessarily
arbitrary dates that we put on these stages are 1100 and 1500. 1100 serves to
mark the transition from Old English to Middle English, a major factor in that
transition being the Norman Conquest. Around 1500, when the Great Vowel Shift,
the shift that raised the long vowels of English, was in progress, we have the
transition to Modern English. Chaucer wrote in Middle English,
Shakespeare in early Modern English.
Further details of the development of the English language have been given in
chapter 10 and we now move on to its dose foreign
relatives_
A particularly close relative, as we might expect given that the

Angles and Saxons came from north-western parts of the continent
of Europe, is Dutch, nederiands. In section 8.3 we saw that a Dutchman might

tell you that his book is under a clock by saying /man buk iz ondr d kloki. That
is an extreme example of similarity; if the similarity were always this great we
would regard Dutch as a dialect of English or vice versa. English has, as we have
seen, changed a lot since Old English; it has lost its case system, it has acquired
Romance vocabulary from the Normans, it has undergone a major vowel shift. But
then to some extent Dutch has followed a similar path. It too has lost the case
system. It too has developed diphthongs from long close vowels; the Dutch
equivalents of five and house, viff and huis, are pronounced with /eV and /Ay/.
Although much less than English, Dutch has acquired lexemes of
Romance origin; thus, for example, the equivalent of holiday or vacation is
vakan tie.
In other respects Dutch is closer to German. Both languages have, for

example, so-called separable verbs, verbs with an element that is separated from
the root in the present and past tenses. Thus the Dutch afbreken, meaning to break
off, to interrupt, becomes brak af in the past tense, as in Hii brak het gesprek a/. He
broke off the conversation. In subordinate clauses the verb tends to come at the end
of the clause. The past participle and the infinitive are preceded by their
complements. Thus the equivalent of He said that my dog had drunk the water is Hij
zei dat mijn hond het water gedronken had. Forming the past participle with the
prefix ge- is another feature that Dutch shares with German. We still had a similar
feature in Middle English; Chaucer wrote Thou hast y-dronke so muchel hony . . A
phonological similarity between Dutch and German is the fact that both devoice
plosives and fricatives at the end of a syllable; hond is pronounced
Monti.
Dutch, then, as one might expect from the geographical location of the region
in which it is spoken, lies between English and German l i nguistically. Dutch nouns
have gender but there is a much weaker
distinction between masculine and feminine gender than there is in German;
the definite article de is used for both. As in German the infinitive ends in <en> in
the written language, but in Dutch one does not usually hear /n/ in this case.
Dutch is in an intermediate position with regard to the retention of /n/ before a
fricative; German retains it inftinf and Mund, English has lost it in its equivalents
five and mouth,
Dutch has lost it in viii but retained it in mond. Dutch makes extensive
use of the plural morph is/.
.50An introduction to Linguistics
At the end of this chapter there are sample texts to illustrate some of the
distinctive features of certain major languages. They are
followed by a translation into English and notes on some of the principal
features. The first of these, text 11.1, represents Dutch.
Two features that distinguish German from both English and Dutch
are its case system and its consonants. The first of these is conservative, the
distinctiveness arising from the fact that German
has retained a system that English and Dutch have abandoned. The second is
progressive in that German experienced a major shift in its consonant system.
German has four cases: the nominative case that indicates the subject of a
sentence, the accusative case that indicates the direct object, the genitive case that
indicates possession and the dative case that indicates the indirect object. Thus in
the equivalent of The baker's child gave the teacher the letter, Das Kind des Bilckers
gab dem Lehrer den Brief the word Kind meaning child is in the nominative case,
Backer meaning baker is in the genitive case, Lehrer meaning teacher is in the dative
case and Brief meaning letter is in the accusative case. Only in the case of Barkers
does the form of the noun indicate the case. With the other nouns in this sentence
only the definite articles indicate the
case and even they are deficient; das Kind can be in the accusative
case as well as the nominative case. So even Germans need to rely somewhat
on word order, assisted by their experience of the world,
to decipher an utterance.
The use of a particular case may be determined by a preceding preposition.

Some, such as durch meaning through, are followed by
the accusative case. Some are followed by the dative case, an example being
nit meaning with. (You may remember that in section
10.5 we saw that mid was followed by the dative case in Old English.)
The shift in the consonant system that gave German much of its phonological
distinctiveness took place in the south around the sixth
century AD. and subsequently pushed northwards. It is sometimes called the
Second German Consonant Shift, the first consonant
shift being that which had given the Proto-Germanic speech distinctive
consonants many centuries before. It is also known as the High German Consonant
Shift, as it helped to distinguish the I Ugh
German variety of the south that came to form the basis of standard German.
This shift affected the voiceless plosives /p, t, k/. Between vowels, for example,
they become the fricatives if„ O. x (or c),/_ In initial position they stopped halfway,
becoming affricates. Thus what is pipe in English and plip in Dutch is Pfeife in
German. Once
this had happened, the vacated voiceless plosives were occupied by their voiced
counterparts, with the result that what is day in English and dag in Dutch is Tag in
German.
In more recent times the long close vowels developed into diphthongs, as in
English and Dutch; the result is seen in such words as Pfeife and Haus.
These various changes to the phonology of High German have affected the north
of Germany only in so far as High German serves as the standard language. The local
vernacular, Low German or plattdeutsch, was not affected by them and so remains
phonologically close to the other Germanic languages_ In Der Schimmelreiter (1888), by
Theodor Storm, who grew up in Schleswig, we can read the following words of
wisdom:
Hest du din Dagwark richtig dan, Da komrnt de Slap von siilvst heran.
Dagwark would be day's work in English, dagwerk in Dutch and Tagewerk in
High German, dan would be done, gedaan and getan, these illustrating the fact that
Low German was not subjected to the change /d/ > It! that formed part of the High
German Consonant Shift. We see, too, that it was not subjected to the change /p I> / fi
, Slap corresponding to English sleep, Dutch slaap and High German Schlaf. That Low
German was not subjected to the later diphthongisation of the close vowel /i:/ is seen
from the possessive adjective
din which shares its form with the Scandinavian counterparts but
not with the High German counterpart, that having become dein. High German is
illustrated by text 11.2.
.2.2 North Germanic
We have seen that in Scandinavia the Germanic tongue developed such

characteristics as a suffixed definite article and the loss of initial /j/. Thus the Swedish
equivalent of year is dr and the Swedish equivalent of the year is aret. Similarly, the
Scandinavian languages often lack an equivalent of our semivowel 1w! in initial
position
before a back vowel; the Swedish equivalent of word is or. Another
An Introduction to Lingilistics
Old Norse
West NorseEast Norse
IcelandicFaroese NOMegian NorwegianSlkedishDanish

(nynorsk)(bokmal)
Figure 11.2
feature that helped to distinguish the languages of Scandinavia was the

development of the reflexive pronoun sik into a verbal suffix that came to indicate
passive or reciprocal action. The Swedish equivalent of The king wrote the book is
Kungen skrev boken; by adding -s to the verb we get one of the two ways of
forming the passive, Boken skrevs at' kungen being an equivalent of The book was
written by the king. We might note just for interest that this development has given
us the verb to bask, its origin being in the Old Norse NT& sik meaning to bathe
oneself. There was fracture or breaking of the vowel le/ that produced, for example,
jeitnn in Swedish where English has even. Such features had begun to give the
Germanic tongue of Scandinavia a distinctive character around the fifth century
A.D. This
Scandinavian variety is known as Old Norse. By the eleventh
century differences had emerged between Norse spoken in the west and Norse
spoken in the east. Thus we can represent the devel
opment of North Germanic in greater detail as shown in Figure 11.2.
Figure 11.2 shows that Icelandic and Faroese are closely related to nynorsk,
the variety of Norwegian that is based on the old dialects of western Norway. But
since that time major differences have arisen
for Norwegian, like the other continental Scandinavian languages, has undergone
many changes, whereas the insular languages have changed relatively little.
Icelandic has, for example, like German, retained four cases for the nouns. It
has kept its vocabulary very pure by coining words for new concepts from its own
elements rather than by introducing foreign words. Thus a kitchen is denoted by
the word eldhiis, literally
fire-house. The Icelandic word for an astronaut is geirnfari, geintur
being space and fara being to travel,
A number of reasons have been proposed for the conservative nature of

Icelandic_ Until modern times the isolated position of
The Lang-uages of Western Europe
Iceland severely limited Icelanders contact with people other than their Danish
masters and a nationalist spirit strengthened their
resolve to resist contamination from Danish. It has been argued that
the tradition of the sagas restrained change. It has been argued, too, that it
would be difficult to adapt foreign words to the case system
with its mutations and inflections.
During the Viking period, in the ninth and tenth centuries, differences began
to emerge between Norse spoken in the west and Norse spoken in the east.
In the east monophthongs developed where West Norse retained diphthongs.

Icelandic retains an Old Norse diphthong in, for example, its word for smoke,
reykur, whereas Danish has reduced it to a monophthong in its equivalent reg.
Icelandic, like nynorsk, has haust, which is the equivalent of our autumn and a
cognate of our
harvest, whereas Danish, like Norwegian bokmal, has host. Swedish has Kik and hijst.
In the west there was more assimilation of nasal consonants before voiceless
stops; thus where Danish has synke (and we have to sink) Icelandic has sokkva.
The Hanseatic League of the late Middle Ages, dominated by Lubeck, had a
greater influence on the speech of the east than on the speech of the west.
Reflecting its fields of activity, the influence of Low German on the vocabulary of
Swedish was particularly marked in the case of vocabulary relating to trade, crafts
and urban life. Thus
while the Icelandic equivalent of to pay, gjaida, perpetuates the Old Norse lexeme,
the Swedish equivalent, betala, came from Low German. Similarly, the Icelandic
equivalent of tow-n or city, borg, is perpetuated from Old Norse, whereas the
Swedish equivalent stad owes this particular meaning to Low German.
By the time of the Hanseatic League differences had begun to emerge within
East Norse, these laying the foundation for the
distinctive characters of Danish and Swedish.
Danish had voiced the voiceless stops following a vowel. We have seen that Danish
has rog where Icelandic has reykur (and Reykjavik which means Bay of Smoke); as
we have also seen, the Swedish word is rok. The equivalent of to buy and a
cognate of cheap is kopa in Swedish and kobe in Danish. To continue the
explanation of the names of Scandinavian capital cities, the Danes call theirs Kobe-
nhavn, this meaning something like Trading Harbour. Further, the alveolar and velar
plosives weakened to fricatives; thus where Swedish has
plosives between vowels or between vowel and liquid consonant remained as

in Latin, whereas in the west there was lenition. Thus Latin amicus meaning friend
became alnico in Italian but amigo in Spanish and arni in French. In the west
vowels following a stressed vowel are less likely to have been retained; as we saw
above, Latin tabula has become tavola, a word with a stressed syllable followed by
two unstressed syllables, in Italian and the shorter table in French. In the west the
plural morphs seem to be derived from those of the accusative case of Latin„
whereas in the east they seem to be derived from those of the nominative case.
Thus, for example, the accusative plural of Latin arnicus, amicos, resembles Spanish
amigos while the nominative plural amici at least looks like the Italian word amici.
in the west a prothetic vowel is used before /s/ plus another consonant; thus Latin
strictus (cf. English strict and strait) evolved into strati] in Italian but etroit in
French, estret in Catalan and estrecho in
Spanish.
We can, then, talk of West Romance and East Romance. In the previous
paragraph I used French, Catalan and Spanish to provide
examples of features of West Romance and Italian to provide examples of features
of East Romance. In fact the dialects of northern
Italy exhibit West Romance features and linguists define this principal
boundary within the Romance world with reference to a
line that runs across the north of Italy, the La Spezia-Rimini line.
11.3.1 West Romance
As differentiation continued we can eventually discern varieties that we can

call early forms of the Romance languages that we know today.
The first to distinguish itself was French which is generally traced back to the
ninth century. More specifically, it is widely considered that the earliest surviving
specimen of French is the Strasbourg Oaths of 842 when Charles the Bald and
Louis the German met at Strasbourg to conclude an alliance against their brother
Lothair. In that text we see, for example, that the intervocalic ipi of the Latin
lexerne sapere, meaning to know, has developed to the /v/ that we
have in Modern French savoir. The word for a thing appears as cosa, it having
lost the diphthong of the Latin causa.
This word is now chose in French; one of the developments that have
distinguished French from both its 'mother Latin and its
The Languages of Western Europe157
'sisters', the other West Romance languages, has been the palatalisation of /k/
before /a/ in Latin by way of /tf / to /1/. Another distinguishing feature has been the
loss of /s/ before a consonant, as in the development of tete, meaning head, from Latin
testa. The word chateau, meaning castle, exhibits both these developments, the Latin
ancestor having been castelium. The segment Is/ ceased to be
pronounced before a consonant around the thirteenth century but
for some time afterwards the word was written as chaste! or chasteau (chasteau
developed by backformation from the plural form chasteaus).
There has been widespread loss of final consonants and widespread loss of final
/a/. This common reluctance of different items to stand in final position has resulted M
severe erosion of the spoken word; the Latin word pon tern, meaning bridge, has, for
example,
become pont in French, this being pronounced /p3/. As this phonetic
transcription indicates, the vowel was nasalised by the following In! before it
disappeared; this is another distinguishing feature of French. Yet another is the
widespread development of Latin /a) to a front mid vowel as in, for example, the
evolution of Latin nzatrem, meaning mother, to French mere.
French is illustrated in text 11.4.
We have been looking at some of the principal features of standard French, a

variety that evolved from the speech of the north of what is now France, more
specifically from the speech of the region around Paris, a variety that is called Francien.
In the south there was a variety that we know as Occitan. The northern and southern
varieties may also be referred to as langue (foil and langue d'oc respectively, oil and
oc having been the equivalent of yes in each variety. The speech of the south was
generally more conservative than that of the north. We have seen that the Latin word
causa has become chose in French, the first segment having been palatalised and the
second having been rrkonophthongised. In Occitan neither of these developments
happened; the word remained causa. in Figure 8.5 we saw a major linguistic division
running across France; that shows the northern boundary of Occitan. One should add
that it is widely recognised that there is a third variety, Franco
provencal, that lies to the north of the eastern end of this boundary.
Just as High German came to dominate Low German, so, too, French increasingly
established itself in the south. As we saw in
section 9,2, the speech of the Paris region gained prestige as the nation state of
France developed, as Paris became an ever more
significant centre of power. As usually happens with the spread of a
standard variety, French established itself in the towns before the rural areas and
in writing before speech.
Today most people in the south speak French, whether or not they also speak
Occitan. As we saw in section 9.6, people who speak Occitan will tend to do so in the
home and in other informal situations; in more formal situations they will be more
likely to speak French. This generalisation obscures the fact, however, that we are not
dealing with a binary situation. Rather, we are dealing once again with a continuum;
the more formal the situation, the closer a
person's speech approximates to standard French. Typically what
people speak is a southern variety of French. In the south, for example, there is
less occurrence of nasalised vowels, the nasal con
sonant being pronounced instead. Unstressed /a/ in final position is more often
pronounced in the south. Given these two features, the
word cornpte, meaning account, pronounced /1C5:t/ in the north, will often be
pronounced /kontal in the south.
To the north of the eastern Pyrenees, in Roussillon, one may hear the word compte
pronounced in a similar way, the only difference being that the nasal consonant is
articulated as a bilabial sound: /komto/. This fine distinction has taken us one more step
along a West Romance dialect continuum; we have now encountered Catalan. Just as
Occitan has been threatened for centuries by the language of the nation state of France,
Catalan, spoken mostly to the south of the Pyrenees, has been threatened by the
language of the state of Spain. Following the death of General Franco in 1975,
however, democracy and devolution were restored in Spain, and as part of that process
the Catalan language was given official status alongside Spanish in the autonomous
region of Catalonia in north-eastern
Spain. It now thrives.
Like Occitan, Catalan is relatively conservative, having changed less in its

evolution from Latin than, say, French and Spanish have.
The Catalan word for a stone, for example, palm, has only voiced the /t/ of the Latin
original petra; the French equivalent pierre has retained no trace of the A/ and,. like the
Spanish word piedra, has diphthongised the stressed vowel.
The Lang-uages of Western Europe
There was a time in the Middle Ages when the speech of the kingdom of Castile was
of no greater significance than Catalan. But, like Francien, Castilian became more
significant as did the power of those who spoke it, with the result that it became the
national language of the state of Spain and is better known to us as Spanish. In section
5.6 we saw that the Latin word focus has given Spanish fuego, meaning fire, hogar,
meaning home, and hoguera, meaning bonfire, and that the short vowel / D / developed
into /we/ in fuego but not in the other two reflexes because only in the case of fuego
was it stressed. These three reflexes of focus also show the widespread loss of initial If/
in Spanish; one environment in which it remains is before the glide /w/.
Another phonological development characteristic of Spanish was the palatalisation

of the initial consonant clusters /f1/, /1(1/ and
/pi/ as /A/. Thus the Latin source of our word flame, flamma, has become llama
in Spanish. This is a homonym of the word llama that forms part of the present tense
of the verb llamar, meaning to call, there having been palatalisation in this case of the
cluster /Id/ of the Latin word clarnare.
The lexis of Spanish has been given a distinctive flavour by the presence of the
Arabs in Iberia for several centuries. Loanwords generally accompany new concepts, and
the areas in which the Arabs contributed significantly to life in Iberia and thereby to
the Spanish language include administration, agriculture and furnishings; thus Spanish has
borrowed from Arabic words such as alcalde meaning mayor, aceituna meaning olive
and alfi)mbra meaning carpet_ As these examples show, the loanwords taken from
Arabic often
begin with /a(1)/, the Arabic definite article having been incorporated into the word.
It was, of course, through Spanish that many indigenous American words came to
Europe; the discovery of the Andes and of the local animals by the Spaniards resulted,
for example, in an addition to the set of homonyms llama.
Text 11.5 is Spanish.
As in Spanish, in Portuguese there has been palatalisation of initial /f1/, /k1/ and
/p1/, the result in this case being If/. In Portuguese
Latin clamare and flamma became chamar and charna. Portuguese has lost intervocalic
/1/ and ini; thus the equivalent of to go out is sair
(cf. Spanish salir) and the equivalent of full is awl° (cf. Spanish Reno).
The loss of /1/ is seen in plural forms; the equivalent of tunnels is taneis. Before
disappearing the /n/ often nasalised the preceding
vowel; thus, for example, what was ?nanus, meaning hand, in Latin became ri-tdo in
Portuguese, the nasalisation being indicated in this case by the tilde, in Portuguese til (),
a mark that may be seen as a vestigial <n>.
One distinctive morphological feature of Portuguese is the personal infinitive. In

the Spanish text, text 11.5, we see the construction al pasar which, like the English
equivalent on passing, gives no indication of the subject, of who or what was passing.
Portuguese, however, can do so; as well as ao passar Portuguese allows, for example,
ao passarmos, this indicating that the subject is first-person plural and so is comparable
to as we passed. Another feature is the positioning of object and reflexive pronouns
after a finite tense in affirmative utterances. As in Catalan and Spanish, pronouns are
suffixed to an infinitive but in Portuguese they can also follow, for example, the present
tense; a Spaniard would tell you that he is
called Jose by saying Me liamo Jost', a Portuguese person by saying
Chanio-me use.
.3.2 Fast Romance
East Romance, represented principally by Italian and Romanian, has in many

respects remained closer to Latin, As we have seen, in the east voiceless intervocalic
plosives have remained voiceless; thus Latin lupus, meaning wolf, has become lupo in
Italian and lup in Romanian as opposed to lobo in Spanish and Portuguese. Vowels that
follow stressed syllables are more likely to have survived in the
east. Romanian retains, albeit in a simplified form, the case system
of Latin.
In the east the plural forms of nouns and adjectives reflect the nominative forms
of Latin; thus the equivalent of wolves is lupi in
Italian and Romanian as opposed to lobos in Spanish and Portuguese.
We have seen that the local speech of the north of Italy belongs to West
Romance rather than East Romance. Italian is, indeed, quite
varied. It is spoken in an area that only became a political unit in the second half
of the nineteenth century, and while a standard
ianguage based principally on the variety of Tuscany has increasingly established
itself/ there is still relatively great variation,
The La ng uages of Western Europe
has the soft form Gymru, the nasal form Nghyrnru and the spirant form Chyrnru,
The corresponding variants of pen, meaning head, are
ben, mhen and phen.
The origins of the mutations are phonological. The soft mutation of initial /k/ and
ip I, for example, reflects the tendency for voiceless plosives followed by a vowel to
become voiced in initial position when the preceding word ends in a vowel. Over time,
however, such mutations became grammaticalised, that is they came to be determined by
function rather than by phonological environment. As an example, the equivalents of the
possessive adjectives his and her in Welsh once had different forms, one of which gave
rise to soft mutation and one of which gave rise to spirant mutation. Now, however, ei
serves for both, with the result that the mutation alone distinguishes the gender of the
owner; while his dog is ei gi, her dog is ei chi. In Gwelodd hi gi, meaning She saw a
dog, the soft mutation of ci results from it standing as an object immediately after a
subject; here, then, mutation helps to distinguish subject and object. Other circumstances
in which soft mutation arises include where a feminine noun is preceded by the definite
article and where a noun is preceded by one of certain prepositions; thus the
equivalents of
garden and pound are gardd and punt while the equivalents of the
garden and for a pound are yr ardd and am bunt.
Breton,. the Celtic language of Brittany in north-west France, is not a survivor of

continental Celtic, of Gaulish. It was taken across the English Channel from Cornwall
around the fifth century as the
Saxons extended their hold on southern England. One indication of the links between
Cornwall and Brittany is Cornouaille, the name of
the region around the Breton town of Quimper.
Migrations also account for the spread of the Goidelic group for Gaelic was taken
from Ireland to Scotland. Irish Gaelic survives,
precariously, on the west coast of Ireland, in areas referred to as the Gaeltacht, and
Scottish Gaelic survives on the islands off the west coast of Scotland,
In Gaelic, as in the Brittonic languages, initial segments vary_ As in Welsh, there
is, for example, mutation in feminine nouns following the definite article; in Irish Gaelic
the equivalents of woman and the woman are bean and an bhean. The verb bris,
meaning to break, becomes
bhris in the past tense. The same thing happens in Scottish Gaelic where, for example,
the equivalent of She broke the chair is Bhris i an
seithear, the initial segment of bris having changed from /b/ to ivi.
.5 Finnish
And finally, Finnish. We end this chapter by looking at some of the features of Finnish,
a representative of the few languages of Europe which do not belong to the Indo-
European family.
Finnish, as we have seen (sections 2.1 and 6.1), is described as an agglutinating

language, items that would be expressed by separate words in English being expressed
by suffixes in Finnish. Thus the equivalent of in Iny house is talossani, -ssa equating to
the preposition in and -ni to the possessive adjective my. The suffix -ssa is one of a
number of suffixes which, like our prepositions, indicate the relationship between items.
As we see, -ssa often corresponds to our preposition in and this construction may be
called the inessive case. Our on often corresponds to the suffix -Ha (the adessive case),
our out of often corresponds to the suffix -sta (the elative case), and so on. The case
suffixes introduced above all have an alternative form with the vowel <a>. Thus the
adessive case has the suffix -Ha, which
has a vowel close to the cardinal vowel [a], and the suffix which has a vowel close to
the cardinal vowel [al. The choice of vowel in
the suffix is determined by the stem vowel of the word to which it is attached.
The suffix -ha accompanies the back vowels represented by <a, 0, u> while -Ha
accompanies the front vowels represented by <a, ö, y> and usually <e, i>; thus on the
chair is tuolilla while on the table is pOydailii. This is an example of vowel harmony,
one of the distinctive features of Finnish.
Other phonological features include a traditional reluctance to admit initial

consonant clusters_ The Finnish word tuo/i that we have
just seen is an equivalent of the Swedish word stoi. The equivalent
of shore is ranta, a cognate of English and Swedish strand. The equivalent of
school is koulu. Finnish has also been reluctant in the past to admit voiced plosives in
initial position; thus the Finnish equivalents of bank, doctor and gas are pankki, tohtori
and kaasu.
Summary
Most of the languages of Europe are considered to be derived from a single

source, Indo-European.
The lndo-Furopean languages of western Europe are grouped as the Germanic languages,
the Romance languages and the Celtic languages. The Germanic languages are
distinguished by such
features as the First Germanic Consonant Shift, vowel gradation and i-mutation.
The Romance languages are those which have evolved from Latin. The Celtic languages,
the remains of a once extensive variety, are distinguished by such features as the word
order verbsubject-object and extensive mutation of initial segments. These groups may be
divided further. The languages of Scandinavia, the North Germanic languages, are
distinguished from the West Germanic languages by such features as their suffixed
definite
articles. The Romance languages are divided into West Romance
and East Romance, distinguishing features including whether or not voiceless
intervocalic plosives have become voiced. The Celtic languages are divided into a
Brittonic group and a Goidelic group.
Finnish is an example of the non-Indo-European languages of Europe.
Texts
Text 11.1 Dutch
Source: The newspaper De Telegraaf (1997).
De man met het witte hondje en zijn vrouw zijn weer actief langs de
doorgaande vakantieroutes in Frankrijk. Bij de politic in Arnhem, rnaar ook bij
andere politiekorpsen in het land zijn aangiftes binnerigekornen van Nederlanders
die opgelicht zijn door het keurig uitziende std. De man spreekt vakantiegangers
aan met het verhaal dat zijn auto opengebroken is of dat hij bestolen is.
The man with the little white dog and his wife are active once
again along the trunk holiday routes in France. To the police in Arnhem but
also to other police forces in the country have come
in reports from Dutch people who have been swindled by the elegant looking
couple_ The man speaks to holidaymakers with
the story that his car has been broken into or that he has been robbed.
De man met het witte hondje (The man with the little white dog): Dutch has two
definite articles,. de and het, the latter being used with nouns
of neuter gender. Dutch makes extensive use of the diminutive
suffixes -tie and. -fe, In. the diminutive form nouns take the article het: de hand but
het honcife.
weer (once again); In many cases intervocalic /d/ is tending to disappear; the equivalent
of again can be either weder or, as here, weer.
zijn aangiftes binnengpkomen (reports have come in): As in several other European
languages, the equivalent of to be is used to form the perfect tense in the case of verbs
expressing motion or change. As in English we see the segment /s/ being used to form
the plural of a noun, as in German we see a past participle formed with <ge>.
De man spreekt vakantiegangers aan (The man speaks to holidaymakers):

Aanspreken is a separable verb.
dat zijn auto opengebroken is of dat hij bestolen is (that his car has been broken
in to or that he has been robbed): In a subsidiary clause the verb tends to come at the
end of the clause. Like the English equivalents to break and to steal, breken and 5telen
are strong verbs, the past
participles showing similar vowel gradation.
Text 11.2 German
Source: Die Verwandiung (1912) by Franz Kafka.
DaR die Veranderung der Stimme nichts anderes war als der Vorbote einer
tuchtigen Verkiihlung, einer Berufskrankheit der Reisenden, daran zweifelte er nicht im
geringsten. Die Decke abzuwerfen war ganz einfach; er brauchte sich nur em n wenig
aufzublasen, und sic fiel von selbst. Aber weiterhin wurde es schwierig, besonders weil
er so ungemeirt breit war. Er hate Arme und Halide gebraucht, urn sich aufzurichten;
stat-t dessen
aber hatte er nur die vielen Beincheri, die ununterbrochen in der verschiedensten
Bewegung waren und die er iiberdies nicht beherrschen konnte.
That the change in his voice was nothing more than the precursor of a bad cold,
an occupational illness of travelling salesmen, he did not doubt in the slightest.
Throwing off the blanket was quite easy; he only had to puff himself up at bit and it
fell off by itself. But after that it was difficult, particularly as he was so extraordinarily
broad. He would have needed arms and hands to
raise himself up; but instead he had only the many little legs that were continually
moving in the most random fashion and that moreover, he was unable to control.
Dafi (That): This exhibits the consonant shift /t/ > /s/, the English and Dutch
equivalents being that and dat. In the future dafi will be written dass, there being a
spelling reform in progress which among other things will replace by <ss> after
short vowels.
die Veranderung der Stimme (The change of the voice): In German nouns
are marked by the use of a capital initial letter. Both these nouns are of feminine
gender. Here the first is in the nominative case and the second is in the genitive case
to indicate possession (of the voice). So,
too, Verktihlung, Berufskrankheit and Reisenden are all in the genitive case.
tachtigen (thorough): This exhibits the shift /d/ > /t/, the English and Dutch
equivalents being doughty and duchtig.
ml geringsten (in the slightest): This is in the dative case; im is a fusion of the
preposition in and dem, a dative form of the definite article.
abzuwerfen (to throw off): Abwerfen is a separable verb, this being indicated by
the embedding of the element zu in the infinitive. Aufblasen and aufrichten are also
separable verbs. The separable particles have the nature of prepositions or adverbs; ab-
indicates movement away (cf. English off) and auf- indicates upward movement (cf.
English up). In inseparable verbs such as verandern and beherrschen the morphs ver-
and be- are less independent, being
bound prefixes. Werfen and blasen are strong verbs, their past tense
and past participle forms being warf and geworfen, blies and geblasen. Werfen
exhibits the consonant shift /p/ > /1/; it is cognate with the English lexeme warp.
fiel (fell): This is the past tense form of fallen, a strong verb with the
same alternation pattern as blasen: fallen — fiel gefallen.
weiterhin wurde es schwierig (beyond that it was difficult): This illustrates the
Germanic principle of word order that the verb should retain second position; as the
first position is occupied by the adverb weiterhin, the subject, es, falls back behind
the verb,
weil er so ungemein breit war (as he was so extraordinarily broad): In a
An Introductkm to Linguistics
subordinate clause the verb is placed at the end. The English translation uses the
adverb extraordinarily to qualify the adjective broad; German rarely distinguishes
between adjectives and adverbs.
hiitte (would have): Here the subjunctive form of the verb is being
used in a conditional construction. The subjunctive form of a verb generally
indicates a hypothetical situation.
Hande (hands): This is the plural form of Hand. The plural form of nouns often
entails mutation.
Beinchen (little legs): -rhen is a diminutive suffix. As with Dutch words with the
diminutive suffix -aye, words with the suffix -then are of neuter gender.
ununterbrochen (continual(10): This is derived from brechen, a strong verb like

to break and breken, its English and Dutch equivalents. Brechen has the same
alternation pattern as welfen: brechen - brach
gebrochen. It exhibits the consonant shift /k!> ic /.
Text 11.3 Swedish
Source: the magazine Vi (1997).
Sedan invigningen 1866 har han statt dar uppe i roken fan alla
Vaxholmsbatarna. P andra sidan vattnet Egger slottet han
byggde, ocksa rokskadat. Tessin Miler en model' av slottet i handen. Man ser
fortfarande att det ar slottet fast marmorn vittrat och losts upp sa att ytlagret är
bara pulver.
Since the inauguration in 1866 he has stood up there in the smoke

from all the Vaxholrn boats. On the other side of the water lies the palace he
built, it too damaged by smoke. Tessin holds a model of the palace in his hand.
You can still see that it is the palace although the marble has weathered and
disintegrated such that the outer layer is just powder.
Sedan invigningen 1866 har han stein (Since the inauguration in 1866 he has
stood): As the sentence begins with the adverbial phrase Sedan
invigningen 1866 the subject ban and the verb har are inverted to keep the
verb as the second concept in the sentence.
invigningen (the inauguration): This is the definite form of the noun
invigning, the suffix -en serving as the definite article. Other words in the
definite form include riiken, vattnei and slottet, -en indicating
common gender and -et neuter gender.
rdken (the smoke): Rok illustrates the rnonophthongisation that took place in
East Norse; cf. Icelandic reykur.
Vaxholmsbatarna (the Varholm boats): Batarna is the definite plural form of

bat.
vattnet (the water): Here we have an example of elision. The unstressed /a/ of
vatten is elided when the definite article -et is suffixed. We see the same thing
happening to ytlager.
slottet (the palace); Slott is an example of the vocabulary acquired from Low
German.
fast marmorn vittrat och losts upp (although the marble has weathered and
been dissolved): Swedish can omit the auxiliary of the perfect tense in a
subsidiary clause.
leists (been dissolved): Like rdic, losa is an example of the

monophthongisation in East Norse. The, final -s indicates that in this case
it is in the middle voice to serve as the passive.
Text 11.4 French
Source: the novel La Porte etroite (1909) by Andre Gide.
Elle me repoussait, m'arrachait &elle doucernen.t — et ce furent la nos

adieux, car ce soir je ne pus plus Hen lui dire, et le lendemain,
au moment de mon depart, elle sienferma dans sa chambre. Je la vis a sa
fenetre me faire signe &adieu en regardant s'eloigner la voiture qui m'emportait.
She pushed me away, gently forced me from her. And that was our
farewell, for that evening I was unable to say anything more
to her and the following day, at the moment of my departure, she shut herself
in her room. I saw her at her window waving goodbye to me and watching as
the vehicle that was carrying me away drove off into the distance.
Elle me repoussait (She pushed me away): Here the verb is in the imperfect tense
which in the Romance languages generally describes
what was happening at a given moment (as with qui m'emportail , that was
carrying me away) or indicates habitual action. In Modern French
a stylistic variation has developed whereby the imperfect tense may
be used for a completed action.
in'arrachait &elle ((arced me from her): Here we have examples of

simplification, the /o / of me and de having been elided before a following vowel. In
section 5.6 we saw the suggestion that such elision may reflect a natural tendency
towards consonant-vowel
(CV) syllables.
doucement (gently): In the Romance languages adverbs are formed by adding a

derivative of the Latin equivalent of mind to the feminine form of the adjective, the
Latin word, limns, having been of feminine gender. This word contains an example
of the nasalisa lion of a vowel, the vowel in the final syllable having been nasalised
by the /n/ before it ceased to be pronounced; the word is pronounced dusma I.
ce furent la nos adieux (that was our farezve11): furent is in the past definite
or past historic tense. This tense is used to refer to a completed action in the past.
It is used in the written language; in the spoken language the perfect tense is used
for this purpose. So, bearing in mind what was said above about the use of the
imperfect tense, there is some variation in the use of tenses for reference to actions
in the past depending on register and style. The phrase nos
adieux illustrates liaison, the situation whereby a final consonant is
pronounced only when a vowel follows, another consequence of the preference
for CV syllable structure.
le lendemain (the following day): Uncertainty about whether segments constitute an

article or part of a noun may lead to an article being incorporated into a noun. In
this way rendemain (cf. Italian indornani) was seen as simply a noun, with the result
that another article was used with it.
chambre (room); Here we see an example of the evolution of Latin

/lc/ before /a/ to /1/ in French. The Latin original was camera.
fene'tre (window): Here we see an example of the toss of /s/ before a

consonant; cf. Italian finestra.
en regardant (watching): En is the only preposition that is not accom

panied by the infinitive; it is accompanied by the present participle.
Text 11..5 Spanish
Source: the novel Nada (1944) by Carmen Laforet.
Salte de la cama traspasada de frio y de sueflo. Tan asustacla, que tenia la

sensacion de no poder moverme aunque, en realidad, no hice otra cosa: en pocos
segundos arra.nque las ropas de la carna y me envolvi en ellas. Tire la almohada, al
pasar, en una silla del
comedor y llegue hasta el recibidor envuelta en una manta, descalza sobre las baldosas
heladas, en el moment() en que Angustias entraba de la calle seguida del chOfer con
sus maletas
y conducienclo a Gloria por un brazo. La abuelita aparecia tambien.
I jumped out of the bed, overcome by cold and sleep. So startled that I felt as
though I was unable to move although, in fact, I did
nothing else; in a few seconds I grabbed the bedclothes and
wrapped myself in them. As I passed I threw the pillow onto a chair in the
dining room and I arrived at the hallway wrapped in a blanket, barefoot on the frozen
tiles, just as Angustias was
coming in from the street followed by the driver with her cases and leading Gloria
by the arm. Granny also appeared.
Salte (1 jumped): No subject pronouns are used in this extract; the

morph -é is sufficient to indicate the first-person singular.
la cama (the bed): Spanish once also had a word cama meaning leg. This
homonymic clash resulted in the word pierna replacing cama as
the word used to denote a leg. In Catalan, where no such dash occurred, the word
calm still denotes a leg.
frio heladas (cold . . frozen): Latin /g/ has often been lost before
a front vowel in Spanish. Frio shows a loss of intervocalic /g/, it being from
Latin frigidus. From the same source Spanish later acquired the word frigicio; such later
borrowings, less subjected to sound changes, are often more learned in nature and are
known as cuitismos in Spanish. fieladas shows a loss of initial igi in an unstressed
syllable, the verb helar being from the Latin gelare. Cf. congeiar, also meaning to
freeze, which has a palatal fricative sound.
Text 11.6 Italian
Source: the magazine Oggi (1997).
Sempre Pioneer ci permise di avere le prime fotografie da distanza ravvicinata di

Giove, ii pianeta pia misterioso e piü gra.nde del sistema solare, e delle sue lune. Da
quando Galileo scopri, circa tre secoli fa, le prime quattro delle sue dodici lune, Glove
e stato accuratarnente studia to dagli astronomi, ma scarsissime e inadeguate sono le
informazioni che slam° riusciti a ottenere sia
pure attraverso i telescopi piit potenti.
Pioneer always allowed us to have the first photographs at close quarters of

Jupiter, the most mysterious and the largest planet in the solar system, and of its
moons. Ever since Galileo discovered, some three centuries ago, the first four of its
twelve moons,
Jupiter has been closely studied by astronomers, but very scanty
and inadequate is the information that we have managed to obtain even by
means of the most powerful telescopes.
ii pianeta pii tnisterioso (the most mysterious planet): Here we see the development /pi/
> /pj I; pifi, meaning more, is a reflex of Latin plus,
oldie sue lune (of its moons): In Italian possessive adjectives are generally accompanied
by a definite article. Here again we see a definite article fused with a preposition.
tre secoli (three centuries): A comparison of secoii and the Spanish equivalent
siglos illustrates three of the principal features that distinguish between East
Romance and West Romance; the Italian
word retains the voiceless intervocalic plosive, it retains the vowel following the
stressed vowel, the first one, and in the Italian word the plural marker reflects the
nominative case of Latin, whereas in the Spanish word it reflects the accusative
case.
scarsissime (very scanty): Scarsissimo is an intensified form of scarso; thus it means

very scanty. In Latin the morph -issim- marked the superlative.
siamo riusciti (we have managed): In Italian most intransitive verbs, verbs
which are not accompanied by a direct object, form the perfect tenses with the
auxiliary essere as opposed to avere. In such cases the
participle agrees with the subject in respect of gender and number.
.4// Introduction to Linguistics
ottenere (to obtain): This illustrates the assimilation of a consonant

immediately before /t/.
sia pure (be it even): Sia is a subjunctive form of the verb essere. The subjunctive
generally indicates a hypothetical situation (Cf. English
even though it be).
Exercises
.1 In the following three sentences (a) identify three examples of

significant difference between Dutch and German and (b) account for the difference
in word order between the English sentence on the one hand and the Dutch and
German equivalents on the other:
English I have seen ten sheep

Dutch Ik heb tien schapen gezien
GermanIch habe zehn Schafe gesehen
.2 (a) Two of the following German words are derived from the Latin equivalent. Which
do you think they are? Give a reason.
Latin EnglishGerman
piper Pepper Pfeffer
piscis fish Fisch
peregrinuspilgrim Pilger
(b) Of the two words derived from Latin, which do you think entered German first?
Give a reason.
.3 From Wittenberg Martin Luther spread his religious teach

ings and with them the local form of High German. (a) Why might it seem
strange that High German was spoken in a town with this name?
(b) What, would you suggest, does this illustrate about the nature of placenames?
The Languages of Western Europe175

.4 Supply the missing Danish word:
EnglishSwedish Danish
pipe pipa pibe
street gala gade
cake kaka
11.5 Account for the differences between the initial segments of
the following words:
EnglishFrenchGerman
tooth dent Zahn
.6 The Latin word planus, meaning fiat, even, evolved into Italian
piano flat, even) and Portuguese chat) (floor). Account for the phonological
differences between these Italian and For words.
11.7 The French word echelle, meaning ladder, scale, has evolved
from the Latin word scalae. Identify in this evolution three features of the
phonological development of French.
.8 (a) What are the principal phonological differences between

Catalan and Portuguese as far as the consonants are concerned? Give examples.
(b) The Latin word for a moon is tuna. What do you think it has become in
Catalan and Portuguese?
.9 (a) It is thought that the word penguin is derived from Breton.

Using the examples of Welsh that you have been given, what feature of a bird
(not necessarily a penguin!) do you think the word alludes to?
(b) The French word bi2ragouin means gibberish, it resulting from the fact that
the French could not understand the Bretons. What do you think was the meaning of
the Breton phrase that gave rise to this word?
.10 Give a pair of Finnish allomorphs.
For each of the following sentences, (a) say which language it is in

and (b) translate the sentence into English.
An Introduction to Lingui5tics
.11 Ao chegarmos o professor deu-nos a mac). 11.12 k il ponte Oil stretto

della citta. 11.13 Han red at baclen sank i denne havn. 11.14 La mere de mon
ami est dans sa chambre. 11.15 MO vrouw is met de auto gekomen.
11.16 Brooir rninn gaf mer Pessa oroabok. 11.17 Tha miathair ag iasgach arms
an loch.
11.18 Der Lehrer hatte das Boot des jungen Kindes gebrochen.
Writing Systems
or
Why Does the Coming Together of Two

Women in China Result in a Quarrel?
.1 Communication across Time and Space
We are near the end of the book and yet we have given little
consideration to the visual representation of language, to writing. Back in chapter
1 we defined language as a system of communication by means of speech sounds
— this reflects the emphasis that linguists place on the spoken language. The
reasons for this emphasis include the fact that we acquire speech naturally while
we have to be taught how to read and write, the fact that we generally speak
more than
we read and write, and the fact that writing develops as an adjunct
to speech; some societies have no writing system and no society has
developed a written language and only subsequently decided to
transmit the utterances orally.
Nevertheless writing has played an important part in the development of our

society Our society requires communication to be made across space and time..
Laws have to reach every part of the country and they have to endure. It would
be difficult to build a power station without specifications and contracts. Almost
as importantly, writing is allowing you to study the subject of language without
having to travel to a particular place at a particular time to listen to me or
some other linguist. Societies have often had to span space and time as best they
could using such means as messengers and law-speakers, but they rarely offered
the same potential that writing offers. In modern times rival means of
communication, of bridging space and time, such as the telephone, the television,
the tape recorder and the video have appeared, but writing still thrives.
This chapter looks at different ways in which humans communicate through
writing.
.2 Morphemic and Phonetic Script
The written representation of an object, of a concept, of an action, and so on may be a

direct representation of the object itself or a representation of the word that represents
the object. The characters or, more technically, the graphs <1> and <5> represent a
concept directly; they are morphemic symbols. The written words <one> and <five>, on
the other hand, are symbols of symbols; they
represent the sounds that represent the concepts, they are phonetic symbols.
In some cases one may see in a morphemic symbol a visual representation of the
object or concept that it represents_ Our character <1> represents the concept of
singularity by a single stroke; the same applies to the equivalent Roman numeral, ,
and the characters used by the Arabs, .0 >, and the Chinese, < >. The
Roman and Chinese equivalents of <2> and <.3> are similarly transparent. Our
characters <2> and <3> become a little more transparent if we turn them clockwise
through 90 degrees and compare them to the Arabic numerals from which they derive,
<Y> and <r>, these being the symbol <%> with the addition of one hook and two
hooks respectively. But it would become increasingly cumbersome if we kept adding
strokes or hooks, and eventually we use opaque characters like <5> which are linked to
a particular concept only as a matter of convention, just as the sequence of sounds
ifaivi is. Thus there is no reason why a character that looks like our <7> should not
represent the concept of six in Arabic, as is the case. Phonetic systems may generally
have a character for each sound segment; the written word <seven> represents the
sequence of sounds isevoini. Such a system is called an alphabetic system_
Alternatively, phonetic systems may have a symbol for each syllable; these are syllabic
systems. In addition to the Chinese symbols for the basic vocabulary. Japanese has other
alphabets which it uses for
its inflections, for foreign names, and so on. These supplementary alphabets are syllabic,
this being quite adequate for a language in which the syllable structure is almost
entirely consonant-vowel
Writing Systems
(CV); it is not necessary to have separate symbols for consonants as

they never occur in clusters.
The Chinese system of writing is primarily morphemic, while the system that we
use is primarily phonetic. Among the more transparent symbols in Chinese are <di >
and , simplified images of a mountain and a child respectively. Our written words
<mountain> and <child> represent most immediately not the objects but the sequence of
sounds that represents the objects. I have said 'primarily morphemic' because there are
phonetic elements in the Chinese script. I have said 'primarily phonetic' because when
we read we generally associate a visual pattern with an object or concept rather than
spelling out the sounds represented by the characters. If we did spell them out the
written words <child>, <one> and <five> might be as likely to have us say chilled, own
and
fever!
Like the pictograms that we see at airports around the world, morphemic symbols,
symbols that represent something without the intermediate element of the spoken word,
have the advantage that a written communication can be understood by people who
speak so differently that they would not understand each other's spoken language.
Without training we would not understand a Frenchman who says quatre, a Pole who
says cztery or a Welshman who says pedwar, but the symbol <4> is internationally
understood. Similarly, Chinese characters are understood throughout China even though
the spoken language varies greatly Given this variety, the writing system contributed
greatly to the cohesion of the Chinese empire. Having adopted the Chinese script, even
the Japanese understand characters that the Chinese use; the words for a mountain are
mutually incomprehensible, the Chinese word, transliterated into Roman script, being
shan and the Japanese word, similarly transliterated, being yama, but both write the
character < Lit >. Morphemic systems do,- however, present problems in the modem
age. Having a different graph for each morpheme clearly means that thousands of graphs
must be used. It is easy enough for us to have a few symbols like <5, 7, +, egz, -=>
as a supplement to our alphabetic system, but imagine us having only such symbols,
having to represent the words dog, eat and thousands of others by such symbols.
A morphemic system introduces problems for typing and word processing. As
international contact increases, the difficulty of rep
resenting foreign names without a phonetic writing system becomes
more important. Foreign names may be represented by using for each syllable the
graph for a word with a similar pronunciation, but
this is cumbersome. To use an example given by Geoffrey Sampson (1985, p.
167), the name Tchaikovsky is represented in Chinese by
five graphs, graphs which if read for their semantic content rather
than their sound would convey 'firewood, suddenly, begin, this, founda tion
Our alphabetic system has advantages of economy and adaptability. Only a modest
number of graphs are needed and they can
be combined to represent any permutations, to represent, for example, our various
consonant clusters.
Generally speaking phonetic systems develop out of morphemic systems. When

humans first wanted to convey visually the idea of an object it was natural that they
should draw an image of that object rather than attempt to represent the sound of
whatever word they used to denote that object. The former required no convention, no
training; the latter did. Through time the image may become stylised and conventions
may develop. The moon might be represented as a
half moon, as a semi-circle, to distinguish it from the circle that
might represent the sun.
Somewhere along the line humans passed from drawing pictures to writing with
morphemic symbols. It is generally considered that the representation is writing once it
can only be understood by those who are party to the culture concerned, who share the
convention. Thus Sampson says that one might suggest that a key distinguishing feature
of writing is that it 'communicates ideas in a conventional manner' (Sampson, 1985, p.
26), He proceeds to say that there is convention in art too, but the fact remains that art
transcends cultural boundaries much more easily than writing does. The American
practice of displaying the word <walk> at pedestrian crossings uses
language, even if people who do not speak English may come to understand the
message by associating it with the shape and
combination of the characters. The European practice of displaying an image of a man
walking is not language.
What that green walking man is is a pictogram or an ideogram. If one considers

pictograms and ideograms to be different one will consider a pictogram to be more of
an image, an ideogram to be more conventional. Of the icons that are displayed in
railway stations, the image of a train that indicates where the trains are to be
Writing Systems
found is more of a pictogram than is the symbol that indicates where one can
obtain information, the latter being derived from a word_ Clearly, concrete objects lend
themselves to representation by pictograms more readily than more abstract concepts do.
One is more likely to use the term pictogram for the Chinese graph for a mountain,
< than for that of the concept of one thing being
located above another, <1 >. But it is not easy to distinguish between
pictograms and ideograms,
Whether pictograms or ideograms, images tend, as we have seen, to become

stylised. A further development may be that the stylised image comes to be
associated not only with a particular object but also with the initial sound of the
word that denotes that object. When this happens the system has begun to develop
from a morphemic system to a phonetic one. In the Middle East a graph that
represented an ox's head came to be associated with the initial sound of the word
for an ox, jaieph. It has come down to us as the letter <A>.
.3 Chinese Script A Morphemic System
As we have seen, the Chinese system of writing is primarily morphemic, the

characters being associated directly with the object or concept that they denote
rather than with the sounds of a spoken word. Images become more stylised and
more angular, but in some cases one can still see a resemblance between the
character and the object denoted. We have already referred to the character that
denotes a mountain, <J>, and the character that denotes a child, ; other
examples are the characters <El > and <k>, which denote the sun and a tree
respectively. So, too, we can see a logic in various characters that denote abstract
concepts; <T).
and <EP> denote the concepts of above, below and middle (the last of these is
found in , the digraph that denotes China, the Chinese referring to their
country as the Central Country). Super
imposing the character for a tree on that for the sun, thereby alluding to the
sun rising behind the trees, gives a character that denotes the concept of east: <*>.
As with the digraph for China, concepts may
be represented by combining or reduplicating characters. The characters for the sun
and the moon together, <EIR>, denote the concept of brightness. The character for
a woman together with that
for a child, <VI >, denotes goodness and love. On a more negative
An Introductiorz to Linguistics
note, the reduplication of the character for a woman, < V tt >, denotes quarrelling.
These may be referred to as symbolic compounds. When referring to the Chinese
writing system I qualified the word morphemic with the word primarily for there
developed a phonetic element. The character representing one concept may also come to
be used to denote another concept that is denoted by a word that sounds similar. Thus
it may be sound rather than concept that is the common thread linking different uses of
a character_ Often the semantic element, the radical, is supplemented by a phonetic
element that indicates the spoken word, this specifying the sense of the semantic
element. Thus, for example, the character for a woman can be accompanied by another
indicating the sounds /ma/ to denote the particular kind of woman indicated by these
sounds, a mother. The same phonetic element is added to the character for a mouth to
denote scolding, that also being pronounced /ma/, albeit with a different tone. As with
our own language, the link
between a phonetic element and pronunciation may become
tenuous as changes in pronunciation fail to be reflected by the written characters.
The Japanese adopted. Chinese characters. We have seen, for example, that the
character for a mountain is used by both. But the Japanese needed to supplement the
system with graphs that would
deal with such features of their language as inflections and grammatical particles_
Thus the morphemic Chinese symbols (kanji) are supplemented by two syllabic scripts
(hiragana and katakana).
En Japanese the Chinese characters generally represent content words, words such
as nouns and verbs, while inflections and particles are represented by hiragana.
Katakana is used mainly to write foreign words; America, for example, is represented by
the katakana graphs for the syllables i a/, /me/, hi/ and /kal: <p,) , ) ,77›.
,4 Roman Script - An Alphabetic System
The writing system that we use came to us from the Middle East by way of the
Greeks and the Romans. It had its roots in a morphemic system, probably Egyptian
hieroglyphs, but the Phoenicians, a Semitic people, associated symbols with the initial
sound of certain words as well as with the objects that those words denoted. As we
have seen, they associated the graph for an ox, < ,>, with the initial
Writing Systems
sound of the spoken word, 'aleph, and, subsequently rotated, this has come to
us as our letter <A>. Similarly, our letter began life as the graph representing
a house, the word denoting a house being beth, the first morph of Bethlehem. These
words were used to denote these alphabetic graphs - hence the word alphabetic. This
development is as though we were to use the symbol <4> to
represent the first sound of our word four, /U. If we did the same with the
symbols <8> and <3> we could write the word faith as
<483›!
As in subsequent Semitic languages, the system developed by the Phoenicians

represented only the consonants; the initial sound of the word 'aleph was a glottal
stop, not a vowel.
The Phoenicians were great traders arid some eight or ten centuries B.C. their
writing sytem was adopted by the Greeks. But the Greeks felt that they had to be
able to represent vowels as well and they did so by using Phoenician symbols that
were not needed for consonants; 'aleph was one such graph. Another change made
by the Greeks concerned the direction in which one wrote; the Phoenicians had
written from right to left but the Greeks eventually wrote from left to right. The
writing in fact became a mirror image of what it had
been, for the letters were also turned round; Semitic < s>,for
example, became Greek <E>. This development took place by way of the
practice of changing direction at the end of the line; this practice was likened to
the passage of an ox ploughing a field and
was consequently called boustrophedon.
The Greek alphabet passed to the Romans, probably by way of the Etruscans.
Certain changes took place in the process; as the Etruscans had no voiced plosive
sounds the Romans had, for example, to restore a symbol for the sound [g], which
they did by amending the character <C>, the symbol for the voiceless equivalent.
Accompanying the introduction of Christianity, the Roman alpha
bet came to Anglo-Saxon both directly from Rome and by way of
Ireland. Thus it was Roman largely influenced by the forms that the Roman
characters had adopted in Ireland. Again certain amend
ments were required to cater for the features of the phonology; thus the
symbol <k> was adopted from the runic alphabet to represent
the sound /0/ which had been unrepresented in the Roman alphabet. The
runic alphabet, called the futhork after its first six characters (<r, h, k, F,
eti ›), had been another development from the
Greek alphabet, it having been adopted by the Germanic peoples
and adapted to facilitate carving. Particularly after the Norman

Conquest, the English alphabet was increasingly influenced by continental
forms.
Another branch of the development took the Greek alphabet to the eastern
Slays. Here again the alphabet accompanied Christianity. In the ninth century the
faith - or, if you prefer, the 483! - and the writing system were taken to the Slays
by two Greek brothers. These missionaries were called Cyril and Methoclius and
the resulting alphabet is known as the Cyrillic alphabet. In this case, too, the
system was altered to reflect differences in phonology; to represent ijui, for
example, the characters and <0> were linked to
produce <10>.
Some Slays received Christianity from Rome rather than Greece and this is
reflected in the alphabet used. The Polish and Russian equivalents of our word
truth are very similar to each other, but the Poles, having received Christianity
from the west, write <prawda> and the Russians, having received Christianity from
the east, write <llpas,Exa>.
As it is a Semitic language - like the language of the Phoenicians - Arabic

uses a script that has also been developed from that of the
Phoenicians. Arabic continues the Semitic tradition of writing from right to
left. Furthermore, it generally only represents the consonants; in Arabic it is the
consonants that indicate the semantic core of words, the vowels merely indicating
derivations or grammatical variations. Whether the Arabic system is alphabetic or
syllabic is debatable given that it does not represent the vowel sounds. The spread
of Islam took the use of the Arabic script to non
Semitic languages such as Persian and Turkish.
.5 Allographs
The form of a graph may vary; instead of <h> we use <H> at the beginning
of a sentence. Just as variant forms of phonemes and morphemes are called
allophones and allomorphs, so, too, such variants of graphs that are determined by
context or position can be called allographs.
The Arabic script is cursive and so graphs are often joined to each other. As
a result there are different forms for each graph depending on its position in a
word; the sound lb/ is represented by < > but
Writing Systems
linked the form is < > in initial position, < > in medial
position and <4.7... > in final position. The word for a house, a cognate of beth,
illustrates the form in initial position (which, remember, is to the right):
And that's another thing - chimpanzees wouldn't be able to write a specification or

an operating manual for a power station!
Summary
As a supplement to their spoken language societies have often developed a system
of writing so that their communication can transcend time and space. If, like Chinese
script, the system primarily represents objects and concepts directly it is morphemic; if,
like our script, it represents'the sounds of the words that represent the objects and
concepts it is phonetic. Phonetic systems may be alphabetic or syllabic. Phonetic systems
generally develop out of morphemic systems, images becoming symbols representing
objects,
some symbols in turn becoming associated with a sound that represents the object
concerned.
Exercises
.1 How and why have the Japanese supplemented the system of

characters that they inherited from the Chinese?
.2 What are the benefits of an alphabetic system over the

alternatives? Explain with examples why these benefits are greater in some
languages than in others,
.3 Outline the history of the letter <A>.
.4 Give examples of how religion has influenced the alphabet that

a society uses.
Gfossary
analogy Analogy is the change to one form under the influence of

another, often change to an irregular form under the influence of a regular form.
As an example, the Middle English past participle holpen became helped.
analytic language This is a language which tends to express each

concept and function by means of a separate word. A good example is Chinese.
English is also fairly analytic; cf. We live in that white house. Also called isolating
language.
Apocope is the loss of the final element of a word. approximant An approximant is a

consonant sound that entails a
minimum of constriction of the airflow, e.g. [14
argot Argot is a type of slang used by a group as a secret means of

communication.
aspiration Aspiration is an expulsion of air as a consonant is articu

lated; the It/ in top is aspirated and as such may be represented in phonemic
transcription as ithi.
assimilation Assimilation is a change to a sound under the

influence of another, as, for example, when the prefix in- became im- in imprison
under the influence of the bilabial /IV. auxiliary verb An auxiliary verb is a verb
which may accompany a lexical verb. It may do so to fulfil a grammatical function, as
do does in negating constructions such as I do not know. It may indicate mood, as in I
should know, in which case it may be called a modal auxiliary.
back vowel A back vowel is a vowel produced with the tongue

drawn back in the mouth, e.g. [u].
backforrnation Backformation is the development of a lexeme by the erroneous

assumption that it formed the basis for another lexerne, e.g. the development of the verb
edit from the noun editor. bilabial A bilabial sound is a sound in which the two lips
are used in its articulation, e.g. [pl.
blend word A blend word is a compound word in which at least

one element has been abbreviated, e.g. motor cavalcade becoming motorcade.
calque A calq-ue is a lexeme adopted in translated form by one

language from another, as the English lexeme honeymoon was adopted by French
as iune de miel. Also called loan translation.
cardinal vowel The cardinal vowels are a set of vowels that serve
as reference points for describing the vowel system of individual languages.
clause A clause is a sequence of words that incorporates a subject

and a predicate. It may be a whole sentence (e.g. He will do it today)
or part of a sentence (e.gthat he will do it today).
clitic A clitic is a variant of a word that is reduced in such a way

that it has no syllable and so is dependent on another word, such as the -is in
He's asleep.
close vowel A close vowel is a vowel produced with the tongue

high in the mouth, e.g. [hi.
code-switching Code-switching is the use of different speech

varieties depending on the social context.
cognate The noun cognate denotes a lexeme that derives from the same source as
another. The adjective cognate similarly describes two or more lexernes that derive from
the same source. Thus, for example, the Spanish word lien() and the Italian word pieno
can be said to be cognates or cognate, both deriving from the Latin word
plenus.
collocation Collocation is the conventional combination of lexernes

as in, for example, auspicious circumstances.
componential analysis Componential analysis is the semantic
analysis of a lexeme in terms of the presence or absence of distinctive features.
compound A compound is a lexeme that is composed of two or

more roots, e.g. boathouse.
concord Concord is the reflection in one word of the grammatical

properties of a word that it accompanies; in These books are mine the words these
and are reflect the fact that books is plural. Also called agreement.
connotation Connotation is an association between a lexeme and

a concept that arises from a person's attitudes, experiences, and so on, as the
lexerne school might connote boredom for some children.
consonant A consonant is ei speech sound produced with total or substantial

obstruction of the airflow, e.g. [IA and [f]. constituent structure Constituent structure is
the hierarchical relation.ship between words, phrases and sentences.
content word A content word is a word that denotes something,
that has a full lexical meaning, e.g. arrow, departure, hopeless. A content word is
opposed to a function word.
conversion The conversion of a word is the process of using it in a
Glossary
explicitly stated in an utterance but which has to be inferred if a communication is to

be successful; Sorry, no dogs on a butcher's
door is only correctly understood if we infer that dogs are rarely eaten in our
society and that hygiene is an important consideration in a food shop.
inflecting language An inflecting language is a language which

makes extensive use of inflection. An example is Spanish in which, for example,
vivimos is an inflected root, as opposed to the analytic English equivalent we live.
inflection Inflection is the grammatical marking of a lexeme by

means of an affix as, for example, the affix -ed in waited indicates past tense.
intonation Intonation is the variation in pitch during the course of

an utterance.
isogloss An isogloss is a line on a map that delimits the area in

which a dialectal feature dominates.
isolating language An isolating language is a language which

tends to express each concept and function by means of a separate word. Also
called analytic language.
lab iodental A labiodental sound is a consonant sound in which the

flow of air is constricted by the upper teeth being held against the lower lip, e.g.
[f].
lateral A lateral sound is a consonant sound in which the airflow

passes to either side of the tongue, e.g. [1].
lenition Lenition is the weakening of a consonant, often under the

influence of neighbouring vowels. This may take the form of voiceless consonants
becoming voiced or plosives becoming fricative; thus, for example, the Old Norse gala,
meaning road, has become gade in Danish, the voiceless plosive it/ having become the
voiced fricative /45/.
lexeme A lexeme is an item of vocabulary which may consist of one or more

than one word, e.g. concede, give in. The term is often used to denote the basic
uninflected form of an item; the lexeme give can, for example, be realised by such
words as gives and gave.
loan translation Loan translation is the adoption of a lexeme from
another language in translated form, e.g. the adoption of honeymoon by French
as lune de mid. Also called calque.
Ioanword A loanword is a lexeme adopted by one language from
another.
manner of articulation The manner of articulation of a consonant
is the way in which the flow of air is constricted in its production. The
manner of articulation of the segment En for example, is
fricative.
metanalysis Metanalysis is change to a word resulting from

erroneous division between words; as an example, an adder evolved from a
naddre.
metathesis Metathesis is the inversion of segments of a word; as an

example, Middle English brid became modem English bird. minimal pair A
minimal pair is a pair of words that are phono
logically the same but for one segment, this serving to illustrate that two
sounds are separate phonemes; thus the words foal and vole are a minimal pair
that shows If! and /v/ to be contrastive in English.
mood The mood of an utterance is its status in terms of expression

of fact (She is doing it today), expression of possibility (She may be doing it
today), and so on.
morph A morph is the form in which a morpheme is realised; thus,

for example, the segment Is! is a morph of the plural morpheme in English.
morpheme A morpheme is a minimal unit that is of grammatical or

semantic significance. A morpheme that can stand by itself is a free
morpheme, a morpheme that cannot stand by itself is a bound morpheme. The
word painters, for example, consists of three morphemes, the free morpheme paint
and the bound morphemes -er and -s.
morphemic script Morphemic script is a writing system in which a character

represents an object or concept itself as opposed to the sounds of the spoken word
that denotes the object or concept. morphology Morphology is the study of the
composition of words. mutation Mutation is change to a sound in certain
morphological or syntactic contexts. Mutation produced the verb feet as a cognate of
the noun foot and it is seen in, for example, the Welsh phrase ei ben, meaning his
head, the basic form of the noun being pen.
narrow transcription A narrow transcription is a more detailed
phonemic transcription whereby, for example, an aspirated It] is represented
as It'].
nasal A nasal consonant is one which is produced by passing air

through the nasal cavity, e.g. mill.
nasalisafion Nasalisafion is the change to a sound resulting from
Glossary
the presence of a nasal consonant as in, for example, the French

word yin.
neo-classical compound A neo-classical compound is a compound

word consisting of Latin or Greek elements, e.g. hepatitis. noun phrase A
noun phrase is a phrase that denotes an object, concept, and so on. It will
generally include a noun and a determiner and may include one or more
adjectives, e.g. that young man.
organ of speech An organ of speech is an organ of the body that

may be used to produce a speech sound, e.g. lungs, lips, the tongue.
palatalisation Palatalisation is the movement of the place of articu

lation of a sound towards the hard palate under the influence of a
neighbouring sound.
performative utterance A performative utterance is an utterance

which is itself an act, e.g. I declare this bridge open.
phatic communion Phatic communion is the use of speech to

maintain social relations, e.g. Pleased to meet you.
phone A phone is a minimal distinct speech sound, e.g. [f]. Also

called a segment.
phoneme A phoneme is a sound that is semantically or
functionally distinctive in a language system.
phonemic alphabet A phonemic alphabet is an alphabet in which

there is a one-to-one correspondence between speech sounds and the
characters that represent them.
Phonetics is the study of speech sounds. phonology Phonology is the study of

the sound systems of
languages.
phrase A phrase is a semantically cohesive sequence of words. An

example is the sequence that old car, this denoting a particular object.
structure rules Phrase structure rules are rules which

determine whether or not a sequence of words is grammatical. pictogram A
pictogram is a stylised visual representation of an
object.
pidgin A pidgin is a means of communication between people with

different first languages.
place of articulation The place of articulation of a sound is the

place where the flow of air is constricted in the production of that sound;
in the case of the sound [p], for example, this is the lips.
plosive A. plosive consonant is one that is produced by completely obstructing

the flow of air, e.g. [p1, [t]. Also called a stop.
polysemy Polysemy is the situation in which a lexerrke has two or
more distinct referents.
popular etymology Popular etymology is the process whereby a

loanword is modified to make it semantically more acceptable. By this process the
French word crevi55e became crayfish in English.
pragmatics Pragmatics is the study of the relationship between
linguistic form and intended communication.
prefix A prefix is a bound morph affixed infrortt of a root, e.g. re- in

rejoin.
push chain A push chain is a sequence of sound changes initiated

by one sound approximating to another; other sounds then change to maintain their
distinctiveness. A push chain is opposed to a drag chain.
redundancy Redundancy is the situation in which an utterance contains more

information than is necessary for successful communication. Given that we need to
indicate the subject, it is unnecessary, as the Scandinavian languages show, to mark the
third-person singular form of the verb as in She drives.
regional standard A regional standard is a more formal speech
variety used by speakers in part of a language area. register A register is a
speech variety, the use of which is
determined by social context.
root A root is a morph that forms the core of a word. It may form a
word by itself, it may be accompanied by another root, by an affix or by
combinations of these, The root of rejoined is -join-.
segment A segment is a minimal distinctive speech sound, e.g. [1].
Also called a phone.
semantic extension Semantic ex tension is the extension of the

semantic range of a lexeme to incorporate a new concept, e.g. the use of mouse
to denote an object used to operate a wordprocessor.
semantic range The semantic range of a lexen-ke is the set of objects

or concepts that that lexeme denotes.
semi-vowel A semi-vowel is a speech sound that resembles a

consonant in that it must be accompanied in a syllable by a vowel but resembles
a vowel in that it does not entail constriction of the airflow, e.g. [w].
Glossary
slang Slang is a variety of speech that may be used to reinforce the

identity of a section of society.
sociolinguistics Sociolinguistics is the study of the relationship

between language and society.
sound symbolism Sound symbolism is a relationship between a

sound and that which is denoted.
speech act A speech act is an action performed by means of an

utterance such as I apologise for my mistake.
standard language A standard language is a codified norm of a

language used in formal contexts.
stop A stop is a consonant that is produced by completely

obstructing the flow of air, e.g. [p], kb Also called a plosive as the flow is then
released in a burst.
stress Stress is the greater prominance given to a syllable by means

of greater articulatory force.
suffix A suffix is a bound morph affixed behind a root, e.g. -er in

painter.
suppletion Suppletion is the use of an etymologically unrelated

word for a grammatically marked form of a lexerne, e.g. went as the past tense
form of to go.
suprasegmental feature A suprasegmental feature is a phonological feature that

applies to a unit larger than a segment.
surface structure Surface structure is a term used in transfor
mational grammar to denote the actual syntactic realisation in a language of an
utterance.
syllabic script Syllabic script is a writing system in which charac

ters represent syllables. An example is the Japanese katakana script in which, for
example, the character < 7/ > represents the syllable / ka /1.
syllable A syllable is a phonological unit consisting of one or more

segments, the basis of the unit being a nucleus, usually a vowel. The word syliabk
consists of three syllables: /sit, /la/ and /bI/.
synchronic A synchronic view of language studies language at a
particular point in time. Synchronic is opposed to diachronic. syncope Syncope is
the loss of an element within a word, e.g. the loss of the segment /1/ in the evolution
from the Latin word doming, meaning lady, to the Italian word donna.
syntax Syntax is the study of the composition of sentences.

taboo word A taboo word is a word that is often avoided because it may cause
offence.
Glossary
which indicate properties of nouns. Word classes have tradi

tionally been known as parts of speech.
word-and-paradigm morphology Word-and-paradigm morphol

ogy is morphology that treats the word as an entity rather than as a succession
of morphemes_
Guide to Exercises
.1 The definition of the term language that I proposed in section 1.2 was
'Language is a form of human communication by means of a system of symbols
principally transmitted by vocal sounds. The element 'human' reflects the fact that it is
generally only human communication that exhibits the productivity and the cultural
transmission that is usually associated with language. 'Communication' reflects a major
function of language (perhaps the major function — see question 1.3). 'System' reflects
the fact that language provides a system or framework for generating utterances.
'Symbols' reflects the fact that the connection between word and thing is usually a
matter of convention. 'Vocal sounds' alludes to the principal means of transmission of on
utterance_ What we understand by the term language may vary and variations on the
above definition may, of course, be
acceptable.
.2 Language helps us to exchange information, to learn, to plan, to co-operate, to establish

and maintain relationships. It helps us to refer to things that are removed from us in
time and space.
1.3 Clearly communication is a major element of the function of human

language. How major an element it is depends on one's definition of
'communication'. There can be little doubt that telling somebody that a train leaves at
ten past four is an example of communication. A performative utterance such as
declaring a bridge open is, on the other hand, primarily
an action rather than communication, even though the utterance, together
with any symbolic action such as cutting a tape, communicates to others that the
action is taking place. More debatable are the likes of emotive utterances; if some
frustration causes you to curse, are you communicating your frustration or relieving your
tension?
2.1 Your definition should acknowledge the fact that a word is a sound or,
usually, a sequence of sounds that is conventionally associated with an object,
concept, and so on. To distinguish a word from a bound morpheme you should allude
somehow to the independent nature of words; this may
be done with reference to, for example, substitution or pausing. You should
acknowledge the problem of distinguishing a word from a phrase with a similar
function, e.g. to appear and to turn up; one might address the problem
Guide to Exercises
by arguing that, while turn and up form a unit in this sense, they can be used
independently of each other in other senses.
2.2 One might refer to the form or the function of a towel. The former
approach might produce a compound like cloths quare or extend the semantic range of
the word cloth. The latter approach might produce something like
handdrier. A combination of the two approaches might produce handcloth (cf.
Swedish handduk) or drycloth. One might borrow a term from a foreign language, e.g.
the French term essuie-mains. To make it more acceptable, this French term might
conceivably be altered phonologically to swimming, particularly as there is a connection
between the object concerned and the activity of swimming; this would be an example of
popular etymology. Alternatively, the elements of the French term might be translated into
English, this being another development that might produce handdrier; this would be an
example of loan translation. The object might be given the name of a person, company or
place associated with it.
.3 The object is a
crowbar.
.1 In book titles content words
are generally written with a
capital initial
letter while function
words are not.
.2 (a) One might distinguish
between buses and cars by
introducing a
category such as
'public' (or, conversely,
'private).
(b) A motorcycle +powered, +carries people,
would be defined
as
-four-wheeled, -
public.
.3 Possible hyponymy diagrams
for these words include the
following:
vehicles vehicles
N
people carrying frelght carrying powered nork-powered
buscar bicycle motorcycle van bus car van
motorcycle bicycle
.4 Your answer should show awareness of the variety of factors that may
prevent words from being completely interchangeable. We have seen that syntax prevents
hide and conceal being considered synonyms. In one context
we might use either of two words but in another context we might only be
able to use one, as we saw with high and tall. Two words might be distinguished by
social register, one being more colloquial than the other, by whether they are associated with
a positive or negative point of view, and SO on.
3.5 The words ?nan and boy are principally distinguished by age, the words walk and run
by speed. The principal distinction between the words toilet and too is one of social register.
Determined and stubborn are largely
distinguished by attitude - a person reluctant to give up is described as determined by

those who sympathise and as stubborn by those who do not. The difference between the
words pavetnent and sidewalk is a matter of geography, the former being used in Britain
and the latter in America_
3.6 Two words with the same form may be different lexernes or may refer
to different elements of the semantic range of the one lexerne. But it is often
difficult to decide which is the case. One guide is the degree of similarity between the
sense of the two words (consideration of foreign equivalents might help). Another guide is
etymology; if the two words come from distinct sources they are more likely to belong to
different lexernes. The decision of whether or not two words belong to the same lexerne
is of
practical significance to lexicographers as it determines whether the words should be
presented under the one headword or as separate headwords. 3.7 In a situation like a
bank robbery the principal intention of this utterance might be to warn. At a hunting
event, on the other hand, it might
be intended to convey that it is not necessary to provide a gun for somebody
.8 For a soldier to respond to the call Fire! he needs to have a gun, the call should come
from a superior officer, and so on.
.9 This sentence might be expanded to something like Mrs Smith will try
to sell the table at the market here m Winchester on Saturday 25 April. Even this
presupposes that the listener knows which Mrs Smith, which table and which year the
speaker is referring to.
.1 The words coast and ghost are distinguished by the fact that the initial segment is
voiceless in the case of the former and voiced in the case of the latter. The words ghost
and boast are distinguished by the place of articulation of the initial segment, /g/ being
velar while ibi is bilabial. Boast and most are distinguished by the manner of articulation
of the initial segment,
/ra/ being nasal. Most and mist are distinguished by the fact that the former has a
rounded back vowel while the latter has a spread front vowel.
.2 1bl is a voiced bilabial plosive, [v) is a voiced labiodental fricative, [kJ is a voiceless
velar plosive.
4.3 The initial segment of the word plosive is a plosive sound, that of the word fricative
is a fricative sound, that of the word nasal is a nasal sound and that of the word liquid
is a liquid sound. (Do we possibly have examples of sound sym.bolism here?)
.4 If one's nose is blocked nasal segments may be articulated like their oral counterparts.
Thus the bilabial [m] may sound like tbl, the alveolar [rt] may sound like [d] and the
velar [IA may sound like [g).
4.5 As [1,1 is a bilabial sound, articulated with the lips, it is more difficult for a
ventriloquist to conceal what he is doing.
Guide to Exercises
4.6 The vowel of the word cheese, 41 , is produced with the lips spread, this resembling
a smile.
4.7 There is a glide from a relatively open back position towards a close front
position, the lips changing from rounded to spread. 4_8 One can say that consonants
need to be accompanied by a vowel, that they cannot be maintained, that they are
articulated with audible friction. The distinction is, however, blurred by the fact that
some consonants are more 'vowel-like' than others; the liquid consonant [lb for example,
can stand in a syllable without a vowel and it can be maintained.
4.9 au a SAIL woz drAok twais last mAnO.
4.10 The words surface and service only differ in that the former has a voiceless /f/
while the latter has a voiced /v/.
.11 Plosive sounds are

represented by straight
strokes, fricative sounds
by curved strokes. Labial
sounds are represented by
strokes that fall to the
right, alveolar sounds by
verticle strokes. Voiced
sounds are represented by
heavier
strokes.
.1 o/a n/r (tfentoiticrto)
(tortalt3rta)
o/a tid (kwantolkwando)
(tortoiOrta)
(konto/kanto) t/k
(tantoikanto)
nil (kontoikolto) tip
(kolto/kolpo)
.2 Voiced plosives become
voiceless in syllable-final
position.
.3 /id/, /d/, /t/, /id/. /id/ follows
another
alveolar
plosive.
Otherwise, Eli follows voiced
consonants and ft/ follows
voiceless
consonants, there
being voicing
assimilation,
.4 kiipte, byggde,
levde, lekte
.5 <s> represents is/ in initial
position, /5/ in final position
and before
consonants and /z/ between vowels. This applies irrespective of word boundaries,
5.6 grawl, tritch
5.7 strike, splash, square, etc. Only Is/ can occur in first position. The second
segment must be a voiceless plosive (/p, t, k/). The third must be a liquid or a semi-
vowel (IL r, w, j/). Not all of these combinations would be
possible; as question 5.6 showed, the combination istwi, for example, is not permissible.
5.8 The sequence inkl/ bridges two syllables.
5_9 Finnish tends to avoid consonant clusters. It has traditionally avoided

beginning words with a voiced plosive sound,
.10 This is an example of i-mutation/ umlaut in the comparative form of the adjective.
.11conflict, confine, abstract, and so on.
.12In adjectives ending in -ie the stress moves to the following syllable,
in adjectives ending in -fu/ it does not.
5.13 The first utterance implies that John was unable to do what he
wanted. The setund implies that he was only able to do something else. The third
implies that he was only able to do it some other day. 6.1 (a) The -er and the -'s of teacher's
are bound morphemes, the former being derivational, as it produces a lexerne that denotes
the person who does
an action, the latter being an inflectional morpheme, as it indicates possession. The -ed
of considered is inflectional, indicating that the action took place in the past. The irn- of
impossible is derivational, producing a new lexeme that denotes the opposite of possible.
(b) In the word project one can identify two elements that had a meaning in Latin,
pro- and -fed indicating movement forwards and throwing respect
ively. But neither element can stand by itself, neither is a free morpheme.
.2 waited, /id/ (id/ would merge with

waded another alveolar
plosive if not separated by
a vowel.)
waved /d / (voiced iv/ is followed by
voiced /c1/)
wiped /t/ (voiceless /p/ is followed by
voiceless /tn.
.3 (a) barns meaning children. (b) The singular
definite article as in hi/en
and the plural definite article as in barne-n.
.4 Sup
pletion
.5 It is grammatically unmarked. Height and high
belong to different word
classes. Its form is unpredictable (cf, tallness). Its meaning is unpredictable

(cf. highness). Any inflection would be more peripheral, e.g. heights.
.6 Haven't illustrates a clitic, a form that might be considered to be a separate word

or an affix, Does inspect consist of one morpheme or two (cf. project)? How does one
describe the realisation of the plural morpheme in the case of the word mice?
.1 The distinction between morphology and syntax may presuppose that

the structure of a word is substantially different to the structure of longer utterances.
Such a distinction is undermined by the fact that semantically
similar utterances may consist of one word or more than one word.
7.2 Possessive adjective, noun, adverb, verb, preposition, pronoun.
7.3 Examples might include a large selection, four very rich men, my young sons, this hat
and the hat that I prefer, the last of these incorporating a clause.
Guide to Exercises
10.5 The /r/ and /1/ have been inverted (metathesis) and the intervocalic
/lc/ has been voiced (lenition).
10.6 The words church and chest have undergone palatalisation; the Scots
equivalents have not.
10.7 A family tree diagram shows the lineage of a language and the degree of
relatedness of different languages. But it suggests abrupt transition rather than gradual
evolution over time and synchronically it suggests distinct entities rather than continua,
the former only being relevant for the standard language. The family tree diagram does
not generally illustrate interaction between branches.
10.8 They are not mutually intelligible; we would not understand King Alfred and
he would not understand us. Direct lineage does not necessarily
mean that two varieties belong to the same language; we do not usually consider Latin
and Italian to be varieties of one language. Indeed, in many respects (e.g. the case
system) Old English is closer to modern Icelandic than to modem English.
11.1 (a) The German words ich and Schafe have the fricative sounds /c/
and If/ where the Dutch equivalents have the plosive sounds /k/ and /p/. The
German word zehn has the affricate its/ in place of the plosive It! of
the Dutch word Hen. These differences are accounted for by the High German
Consonant Shift.
(b) In Dutch and German the past participle of a perfect tense is placed at the end of a
clause.
11.2 (a) The words Pfeffer and Pilger are derived from Latin. Unlike Fisch, they do
not exhibit the First Germanic Consonant Shift.
(b) Pfeffer it was adopted early enough to be subjected to the High German
Consonant Shift, whereas Pilger was not.
.3 (a) If subjected to the High German Consonant Shift the intervocalk

iti in the name Wittenberg, meaning white hill, would have become /s/. (b) This may
illustrate the conservative nature of placenames.
11.4 kap
11.5 The French word dent retains the segment Id/. As part of the First Germanic
Consonant Shift this segment became /1/ in the Germanic languages, as in the English
word tooth. In High German, as part of the High German Consonant Shift, the plosive
iti became the affricate /ts/ in, for example, initial position; hence the Its/ in the German
word Zahn.
.6 In Italian the initial cluster /pi/ has become /pjf, while in

Portuguese it has become /J./. In the Portuguese word there has also been nasalisation
caused by In/ and then the loss of that In/.
11.7 The word ichelle acquired a prothetic vowel before /s/ + consonant.
Bibliography
General
Aitchison, Jean, Linguistics (Hodder &z Stoughton, 1992 [19781). Bloomfield, Leonard,
Language (Henry Holt, 1933). Crystal, David, Linguistics (Pelican Books, 1989 [19711).
Fromkin, V. and Rodman, R, An Introduction to Language (Harcourt Brace,

1993 [1974]).
Hocken, Charles F., A Course in Modern Linguistics (Macmillan, 1958). Jennings,

Gary Personalities of Language (Victor Gollancz, 1967).
Lyons, John (ed.), New Horizons in Linguistics (Penguin Books, 1973 [1970]). Lyons,
John, Language and Linguistics: An Introduction (Cambridge University Press, 1992
119811).
, Bruno, Linguistica (Felice Le Monnier, 1972 11970]). Pinker, Steven, The Language
Instinct (Penguin Books, 1994), Potter, Simeon, Modern Linguistics (Andre Deutsch, 1960
[19571). Potter, Simeon, Language in the Modern World (Penguin Books, 1968 [1960]).
Robins, R. H., General Linguistics (Longman,. 1990 [1964]). Sapir, Edward, Language: An
Introduction to the Study of Speech (Harcourt,
Brace & World, 1921).
Animal Behaviour
Hin.de, Robert A., Ethology (Fontana, 1982).
Manning, Aubrey, An Introduction to Animal Behaviour (Edward Arnold,

1979).
Semantics
Berlin, B. and Kay, F, Basic Color Terms: Their Universality and Evolution
(University of California Press, 1969).
Leech, Geoffrey, Semantics (Penguin Books, 1974).
, Stephen C., Pragmatics (Cambridge University Press, 1983). Mott, Brian L., A Course in
Semantics and Translation for Spanish Learners of
English (Promociones y Publicaciones Universitarias, 1993).
Bibliography
Ogden, C. K. and Richards, L A., The Meaning of Meaning (Ark Paperbacks,

1985 119231).
, E R., Semantics (Cambridge University Press, 1981 [19761). Ullmann, Stephen, Semantics:
An Introduction to the Science of Meaning
(Blackwell, 1962).
Yule, George, Pragmatic (Oxford Universiy Press, 1996).
Phonetics and Phonology
Abercrombie, David, Elements of General Phonetics (Edinburgh University

Press, 1967)
Carr, Philip, Phonology (Macmillan, 1993).
Catford, J. C., A Practical Introduction to Phonetics (Oxford University Press,

1988).
, Francis, An Introduction to Phonology (Longman, 1992 ), Laclefogecl, Peter, A Course in

Phonetics (Harcourt Brace, 1993 [19751). Mott, Brian L., A Course in Phonetics and
Phonolog-y for Spanish Learners of
English (Promociones y Publicaciones Universitarias, 1991). O'Connor, J. D., Phonetics
(Penguin Books, 1991 119731).
Morphology and Syntax
Anderson, Stephen R., A-Morph ous Morphology (Cambridge University

Press, 1995 [19921).
Bauer, L., Introducing Linguistic Morphology (Edinburgh University Press,

1988).
Brown, Keith and Miller, Jim, Syntax: A Linguistic Introduction to Sentence

Structure (Routledge, 1996).
Carstairs-McCarthy, Andrew, Current Morphology (Routledge, 1992). Chorrisky, Noam,

Aspects of the Theory of Syntax (MIT Press, 1965).
Culicover, P., Syntax (Academic Press, 1982).
, Francis Morphology (Macmillan Press, 1993). Matthews, P H, Morphology (Cambridge

University Press, 1991 [19741). Matthews, P. H., Syntax (Cambridge University Press, 1992
[19811). Radford, Andrew, Syntax: A Minimalist Introduction (Cambridge University
Press, 1997).
Regional Variation
Chambers, J. K. and Trudgill, P., Dialectology (Cambridge University Press,

1980).
Francis, W. N., Dialectology: An Introduction (Longman, 1983). Petyt, K. M., The Study
of Dialect (Andre' Deutsch, 1980).
Trudgill, Peter, On Dialect (Blackwell, 1983).
Bibliography
Social Variation
Ager, Dennis, Sociolinguistics and Contemporary French (Cambridge Univer
sity Press, 1990)_
Clyne, Michael, Language and Society in the German-speaking Countries

(Cambridge University Press, 1984)_
Downes, William, Language and Society (Fontana, 1984). Hudson, R. A.,

Sociolinguistics (Cambridge University Press, 1996 [1980]). Sanders, Carol (ed.), French
Today: Language in its Social Context (Cambridge
University Press, 1993).
Truclgill, Peter, Sociolinguistics: An Introduction to Language and Society

(Penguin Books, 1995 119741).
Aitchison, Jean, Language Change: Progress or Decay? (Cambridge University

Press, 1995 119911).
Bynon, Theodora, Historical Linguistics (Cambridge University Press, 1993

[1977]).
Romaine, Suzanne, Pidgin and Creole Languages (Lon.grnan, 1988). Miihlhausler, Peter,
Pidgin and Creole Linguistics (Blackwell, 1986). Trask, R. L., Historical Linguistics
(Arnold, 1996).
, Dennis, Sociolinguistics and Contemporary French (Cambridge

University Press, 1990).
Barber, Charles, The English Language: A Historical Introduction (Cambridge

University Press, 1995 [19931).
Brachin, Pierre, The Dutch Language: A Survey (E. J. Brill, 1985). Clyne, Michael,
Language and Society in the German-speaking Countries
(Cambridge Univ'ersity Press, 1984).
Comrie, Bernard (ed.), The Major Languages of Western Europe (Routledge,

1990).
Donaldson, B. C., Dutch: A Linguistic History of Holland and Belgium

(Martinus Nijhoff, 1983).
Harris, M. and Vincent, N. (ed.), The Romance Languages (Croom Helm,

1988).
Haugen, Einar, The Scandinavian Languages: An Introduction to their History

(Faber & Faber, 1976).
Keller, R. E, The German Language (Faber & Faber, 1978).

Lockwood, W. B., A Panorama of Indo-European Languages (Hutchinson, 1972).
Maiden, Martin, A Linguistic History of Italian (Longman, 1995). Mattoso Camara, J., Jr,
The Portuguese Language (University of Chicago Press,
Index
embedding , 89, 189 Icelandic -4

emotive utterances , 189 ideogram , 181
English idiolect , 190
history of , 92, 124, idiom , 10, 29, 30, 190
language
, 127-35, 138-40, Indo-European , 138, 145, 146
148
phonology -61, 66-8 inflecting , 191
languages
regional variation , 61, 96-107 inflection -7, 191
standard language International
Phonetic Alphabet
euphemisms , 118, 119, 189 , 48
intonation , 70, 191
felicity conditions , 189 isoglosses , 191
Finnish , 63, 163, 164 isolating , 191
languages
First Germanic Italian
Consonant Shift
morphology , 173
focal areas , 190 phonology , 65, 68, 160, 161,
French
history of , 156-8, standard , 160
language language
text , 173
phonology -5, 67, 69
regional variation „ 101 Japanese
standard language , 157, 158 phonology
text , 170 script
function words , 190 syntax , 141
Gaelic , 163 La Spezia- .56
Rimini line
German Labov, W. , 117
phonology -5, 149-51, 167 lateral , 191
regional variation -2 Latin -7
standard language , 113 Leech, G. , 33
syntax , 149, 150, 167 lenition , 127, 191
text -8 lexemes , 191
glottal stops , 190 liquid
Great Vowel Shift loan translation , 191
Greek loanwords , 18, 191
Grice, P. Low German , 151
Grimm, J.
Grimm's Law , 137, 146 manner of , 44„ 191
articulation
meaning
heads , 190 metanalysis , 126, 192
High German metathesis , 192
Consonant Shift
, 150, 151 minimal pairs , 192
Hockett, C. , n morphs ,
F.
homographs , 190 morphemes , 75, 192
homonyms , 31, 190 morphemic -82, 192
script
homonymic clash , 190 mutation , 126, 147, 162,
165, 192
homophones , 190 mutual -5
intelligibility
hyponymy 25, 26, narrow , 192
190 transcription
i-mutation, 147, 1190nasal, 192
Index
nasalisation , 126, 192 Saussure, F. de

Norwegian , 152, 154 schwa
noun phrases , 87, Second German
Consonant Shift
nynorsk , 154 , 150, 151
segments , 194
Occitan , 157, 158 semantic , 194
extension
Ogden, C. K. semantic field
and
Richards, 1.
A.
, 32 semantic range -7,
125,
194
Old Norse , 153 semi-vowels , 194
onomatopoeia Shaw, G. B. , 48, 49, 106
organs of -3, sibilants
speech 193
slang , 121, 195
palatalisation , sociolinguistics -11, 195
126,
193
performative , 36, solidarity
utterances 37,
sound symbolism , 195
phatic ,6, Spanish
communion 193
phones , 193 history of , 159,
language
phonemes , 57, 193 , 172
phonemic -9, lexis , 155, 159
alphabet 193
phrases ,87, 193 morphology , 128
phrase , 193 phonology , 62-6,
structure rules 68,
69,
172
pictograms , 181, 193 text , 172
pidgins , 142, 193 speech acts , 37, 195
place of , 44, standard -13,
articulation 193 languages 195
plathieutsch , 151 stops , 195
plosives ,44, 194 stress -9, 195
political , 120 style-switching
correctness
polysemy ,31, 194 suffixes , 195
popular , 194 suppletion , 195
etymology
Portuguese , 77, 159, suprasegmental -70,
160 features
power
pragmatics -7, 194 surface structure , 195
prefixes , 194 Swedish , 138, 139, 151-4,
168,
productivity , 78
push chains , 194 text , 169
Pygmalion , 48, 106, syllabic script , 195
109
syllables -9, 195
Rask, R. symbols , 22
redundancy , 131, 194 syncope , 195
referents synonyms
regional , 98,
standards 194
registers , 116, 117, taboo , 118, 119, 196
194
Roman script -4 taps
Romanian , 155, 160, tone languages , 70, 196
161
roots , 194 tongue -7, 49-51
Russian transformational , 91,
rules 196
phonology , 64, transitional areas, 196
67,
68
syntax trills
Trudgill, P. -14, 116
Sapir, E. , 4, U, 12 Turkish , 141
Index
typology , 196 vowels -52, 196
umlaut , 147, 196 wave theory , 141, 196
Welsh
velum -6, 196 phonology , 163
verb phrases , 196 syntax , 141, 162
verlan 120 words -12
vocal cords, 196 word classes, 85, 196
voicing , 43 word order -3
vowel gradation, 196 word-and-paradigm morphology
vowel harmony , 196 , 80, 196
A It INDINpitg.14- 1*, -NP Ai

General
Poole, S.An Introduction to Linguistics
Radford, A. M. , et al.Linguistics An Introduction

Robins, .R. Ff. General Linguistics Fourth edition
Yule, G. The Study of Language Second edition
phonetics and Phonolog-yjf
Clayk, J. a al.An Introduction to Phonetics and Phonology
**A -44 et X n
Roach, P. English Phonetics and Phonology; A Practical Course

Second edition
Sntax tIM*.
Cook, V. Chornsky s Universal Grammar: An In
Second edition
Radford, A. Syntax : A Minimalin Introduction
Radford, A. Transformational Grammar : A First Course
St *AIL 41-
Semantics *5*
Lyons, J. Linguistic Semantics: An Introduction
Saeed, J. I. Semantics
j Ivo
Matthews, P. H.Morphology Second edition
Pragmatics WIEIJ*_ _ _
Peccei, j.S.
pragmatics
Verschueren, J. Understanding Pragmatics

"i4 )11
Analysis
Discrytime
Brown, G. et a. Discourse Ana iysis
i4 #4fr
Gee, j P. An Introduction to Discourse Analysis: Theory &
Method
* xi: ;tit
Lexicography -PA*
Hartmann. R. R. K. et al. Dictionary of Lexicography

*A-tirli-dt-f`ht
Stylistics 5t
Thornborrow, J. et al.Stylistics for Students of Language and Literature

: 4t4 ;5t_44-I.X.r1
Wright, L.. ot Stylistics A Practical Coursebook
Typology if
Croft, W.
Sociolinguistics
Fa,sold, R. The Sociolinguistics of Language
Hudson, R . A.Sociolinguistics Second edition
Wardhaugh, R. Introduction to Sociolinguistics Third edition
Psycholinguistics ,Lisfif. if;
Aitchison, J.The Art kulate MammalAn Introduction to
Psycholinguistics Fourth edition

iz-LF1 417 Au
Carroll, D . W.Psychology of Language Third edition
I*
Intercultural Communication
Samovar, L. A. et at.Communication Between Cultures Second edition
5t.
Scollon, R. et al.intercultural Communication
Transiaidugy mi A
Baker, M.In Other Words: A Coursebook on Translation
-t;11014
Functional Linguistics Xi3M4
Halliday, M.A. K.An Introduction to Functional Grammar Second
edition
rn
Thompson, G.introducing Functional Grammar
-Ott 41 n
Historical Linguistics iii_StAg*
Trask, R L.HisOrical Linguistics
rt,
corpus Linguistics A-14** tie*
Biber, D. et al.Corpus Linguistics
M-4-42-*
Kennedy, G.An introduction to Corpus Linguistics
Statistics in Linguistics A
woods, A. et al.Statistics in Language Studies
History of the English L.-anfage 4i11.5t
Freeborn, D.From OW English to Standard English Second edition
Child Larig e JLAILti-

Peccei, J S. Child Language New edition
Cohen, A. I). Strategies in Learning and Using a Second Language
-"9. it
First Laimags_ cry.isitioniN.=Cjr
Goodluch, Language Acquisition: A Linguistic Inerodwrion
11§ Ai Ai ill:
Second LanIpteee Aequisifion
Cook, V. Linguistics and Second Language Acquisition
Cook, V. Second Language Learning and Language Teaching
Second edition
Larsen-Freeman, D. et al. An Introduction to Second Language Acquisition

Research
Research Method liFMMk
McDonough, J. et al.Research Methods for English Language Teachers

- ft 46r M-57
Slade, CC.. Form and Style Research Papers. Reports, Theses
Tenth edition
XRttit..A t*-1f-4-
Academic Writing f
Roberts, W.1-1. et al. About Language A Reader for Writers Fifth edition
AiltAftiltAt
Methodology
Harmer, j_ How to Teach English

I• 44 it38-
Richards, J. et al. Approaches and Methods in Language Teaching
it-ti *if A,*
Trudgill. Pet al.International English Third edition

IFF
Ur, P. A Course in Language Teaching; Practice and Theory
i6-trlYktfir-41;ilit -1L1'1AA
Testing
al a*
Alderson, .C. et (II, Language Test Construction and Evaluation
Ai ir5
Heaton, J.B. Writing English Language Tests New edition
Course Des_jim lila if

Yalden, J.
English CrranlMar ***p
Biber, D. et al- Longman Grammar of Spoken and Written English

OWtikr-7 41" gzit-i-
Dictionary fit *
Bus.stnann, H.
Richards, J. C. et al.

An Introduction To Linguistics Stuart C Poole 1999 205p

Uploaded by

Copyright:

Available Formats

An Introduction To Linguistics Stuart C Poole 1999 205p

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

An Introduction To Linguistics Stuart C Poole 1999 205p

Uploaded by

Copyright:

Available Formats

What topics does the book cover?

What topics does the book cover?

What skills does the author say the reader already possesses?

What skills does the author say the reader already possesses?

iK (1.

(WZ)VR41' 155 -14

i71,A1'1/(*)*(Pooles. c_ )—M40'1* LIEFE

ISBN 7 — 5600 — 1869 — 6

.Cibr---1ff -N. H319, 4

English edition © Stuart C. Poole 1999

This edition © Foreign Language Teaching arid Research Press 2000

L;(11-,Th-gENNLIEff 19 q.• (100089)

650 x 980 i'16

)1** 6 )1* 1 &EPA

1111` : .90 3-E

tInUrP)01 AMMAN:1M MAMMA

Foreign Language Teaching & Research Press is to be congratulated on its

The books are a representative selection of up-to-date writings on the most

Emeritus Professor University of Sydney

kt:.46JA- fN/tlki,_) 11?

T44.4-Mi-fttiksIC ittISIktt4ofdtlki0 PRAtitt

*41/. t 144 TIT t ifT- it- itirt*ki

Alfri .71-Jr vtlt -I4-k±-4-1/T‘

AffIt rktlirf -AttAA11*+4-ao

Mt*Ig41411/1* (Akt11) MAA'qM, 1J-ig

gffIft, iffrovtiEtt, 5A-0111443I4AX{M fA0

PA)Vil4A-1Aif-A*'10,*0.14A“t, SfkrJL ilT.TM4t—t

Sociolinguistics is a branch of linguislics which concentrates on language in

Pq_PL a t AO" 1-*%,

'fM.410 '4P4WINW-kft4VVZ-41144Y-[Wtil ‘)V2N

$0 f-t XE*4.111f litAifitityjnito

camcorder. +1414 --- '3K

Ri6rfq "Nt" P1A—fittita;

*nf- F-13A0 413M11#1.41111 t#thiliiti4/41, A4§.(44,

AuXit(iq honeymoon n -tiFA lune de naiel.

iMykid N.JO, 51- it Niqt" 0

nAE0 - Pfi4"6 *tt,

ifi*T-fFtife ni3W1 6-0 TiffIZWii9rEd-,

WFAAA K610310t1A-fflYIk'Ac. *404VOikiltAftftiVifire-M-aiM, VAIR

(!!U fire 111VG-6 )0

31itill*0 Ai6A-AiPAYMIJERI, kiti*TIRA-A-NteLOW,Z9.10

O ARTi431EM schola, t%**111StftliEricii*W1 T

AT. ecole ll escuelai ifri=13)t,T -

tiMW M *4141r4hrog, °01.1.1r1-r) -

*YL4)5( iiltRTWTs gfWV1-•4.i_

t IM± -A-11=rit fll VI Mit

frEiRA,, -1\-- s,f$J, #

tfliff*MtATiii, xiiifriaiff*oEL i§(}moRAgel

4)AT/coil/x1. atgatIt"ftA4liti ri$W1W,Y1A5 MIELPLIM

iAVZEPV6i4 INtiltili-CiITI5.1410frOtirtiWITT -FAO, tA

MO, AMEVF151.—ftift-Vivial3)145X Tigirtlilioi,t,

fMTVIIif-trAAni tliRtN0 iS-14 131-.4•IYIR CI;

Tr• itONWIZT3K, iCESt- E,

q*.)2jft-, iWti ±ifrfah.,11±tUtifii

±11 ,SC ZIl'-A•17

Itte, TirkEIgilt iff4A4$1[

Preface by Halliday FlU

.1 The Significance of Language

.3 The Functions of Language

i71,A1'1/()(Pooles. c_ )—M40'1* LIEFE

41/. t 144 TIT t ifT- it- itirtki

MtIg41411/1 (Akt11) MAA'qM, 1J-ig

PA)Vil4A-1Aif-A'10,0.14A“t, SfkrJL ilT.TM4t—t

31itill0 Ai6A-AiPAYMIJERI, kitiTIRA-A-NteLOW,Z9.10

tfliffMtATiii, xiiifriaiffoEL i§(}moRAgel