Chapter 7 ——
Speaking, writing, and learning
Differences between speech and writing
Taking and writing, then, are different ways of saying. They are
different medes for cxpressing linguistic sieanings. To go back €0 Our
tri-stratal model of language, from Chapter 2: in language, a network
of meanings (‘seinanic* system) is encoded through a network of
wordings (‘lexico-grammatical’ system), and the wordings are in turn
encased through a network of expressions, The expression, in all
cultures, makes use af the medium of sound (‘phonolog.cal’ sysien);
in some cultwres if also makes us> of the visual medium Corthographic™
or ‘graphological” system)
Can we Say, then, thal talking and’ writing are different ways of
expressing (saying) the same meanings (things)? The answer we are
suggesting must >e: yes, and na, Yes, it the scnss. that the two are
alternat.ve ‘outputs'—alternative realisations of the menning potential
of language; anything that can be said in writing can also be said in
Language" is mare speaking. and vice versn ‘The two are both language; and language
important than either is more important than e:ther, It is a mistake to become too much
spoken language or gbsessed with the medium
writcen Language. Atche same time {and again for the stme reascn), it is important
thar we should recognise that there are specific seiises in which the
anewer must be “no”. It is rather like the principle that what is said
in any one human language can also be suid in any other, AS a general
principle, this is true; and it is a fundamental and significant ruth: all
languages lave the same poventlil for expressing human meanings. But
in practice each language hes cvolved in its cwn culture; and while
every language Is squally well adapted to serve the needs of (i.e. express
the meanings of) the culture within which it itself develcped, it is not
So well equippes for expressiag the meanings of ancther culture,
Aboriginal languages are net, in fact, equipped 10 express the semintics
ot Western societies—nor are European languages suited to the meaning
styles 0: Aborigines. Each would have to adap: itself in order to wert
such different demands,
‘There is an endlogy with speeck and writme We could idenify
three respects in which we woulk have to-qualify the generalisation that
92Spoken language and written language are “saying the samo thingst:
1. Writing does not ineo-porate all the meaning potential of speech:
‘Teleayes out ihe prosodic and paraling sistic contribut ons. There are
also certain reciprocal effects: spoken language does net show
Sentence and paragraph boundaries, or signal te move into direct
quotation. But these are different: itis the signels that are missing,
Not the features themsalves. +
2. Speech and Writing are in practice used in differentcontexts, for
different: ptrpose}—though obviously with a certain amount of
overlap. This is partly a ritual matter, a form of social convertion:
such institutions as the applicatior form (‘please submit your request
in writing. in triplicate’), the office memorandum, and the “spy
for the files’ ere essentially written language rituals—and will renain
50 In ‘the elertronic office of the futtee", cven if uv longer in the
form of rays and cabinets full of paper. But x is mainly for the reason
for which wrting evolved in the first place: namely, that there are
different functions to which each one is appropriate. To that extent,
therefore, they are not saying ihe same things. The meanings that
are expressed in contexts where writing it used are normally not
translated into talk. No doul they could be; but thers would be no
point in jt, We do not go around reading the regulations on the airline
tiekst aloud. (t is a pity, on cie other hand, that we donot go around
reading plays ar retiting poctry aloud—we might get better sounding
pucuy if we did.) Similarly we do net write down “he houserold
maraperrent eptiversations and service encounters of daily life. which
are extremely borng to have to read. Just z¢ in a bilingual culture,
the languages are almost always specialised out—ta different
activitics, different topizs, different personal relationships: the-e is
a point in having two languages bath doing all the same thinge—
30 there would be no pcint in having both speech and writing if the
two simply duplicated she functions af each other.
4. Thirdly there is the point made in the last chapter: that speech ancl
writing impose different grids on experience. There is a sens: in
which they create different realities. Wrising creates a world of things:
talking creates a world of happening, The question arises, then.
wheiher one is in some sense more *basic’, ‘nearer tc reality’ than
the other: and if so, which. This is not an easy question, aid we
shall not solve it here. But itdeserves a spocial—if inevitably very
aketchy—section of the discussivu Ww iuself
Grammatical metaphor
Notice that wher we ‘translate’ between speech ard writing, certain
types of wording frequently tra np. On the writes sidz, we tond to
have rether few different verbs; those that occur, apart from be and
its synonyms (reprecent, comprise, form, add up to, corstituic, ct),
are often verbs such a5 lead to, catise, follow, allow, derive, cover,
an inched, which are the names of cireumsiantiol relations:
applause followed che ‘after the “announcement.
announcement people applauded"
93
Are there differences
in the meanings
expressed by spoken
aie) we.tten
language?Spokea forms are
‘nearer the
‘grammatcal
nietaphor” i a
fenture of written
lag
mitaphorical and
fess aueiapherical
ways o° representing
mocaning in the
‘gram
this development could lead ‘in view af this developmen:
oa different outcome things might turn out
differently’
he derived mach satistaction ‘because of this discovery
from this discovery he was very satisfied’
her specch covered five points ‘she spoke about live points
‘On the spoken side, we find a greater range of verbs, Thess verbs ars
tumed sy the wr.tten version into nouns, many of whick are therefore
the names af processes, Like applause, speech, aterecite. and Seitiyfuc-
tion in the exampies just cited. In a *more spoke" equivalent, aunomtace-
ment, developrent, and discovery might alsu appear a3 Verds:
after they had announced (it). people applanced’
in view of what has developed, things might turn out different
he was very satisfied hecause he had discovered this
There is a sense n which the sooken forms appear to ke nearer
the bone. The writen variants have a sort of metaphorical quality about
them—except that there is no greater degree of metaphor in their choice
of words, which are the same in botk: announce/annonucemsnt,
applauselapplauced, discoverfdiscovery. tnd so on. (Members of sich
pairs ars ‘the same words” in the sense that they are the same LEXIZAL.
vem, differing only in derivation or inflexion ) ‘The metaphor iz in
the grammar, Semetking ‘hat sould typcally be represented, given
the grammatical system of Rnglish, as a verb, has beer represented
instead as a noun.
Why do we say thet such things az “announcing”, ‘appiouding’, ‘dis-
covering’, “turning out’ would typically be represented as verbs? Beccuse
in the semantics of English they are precessos: so¢cbody is doing sume-
thing, or something is happening. There is a similar element of geam-
‘matical metaphor in what the written versions have as telr verbs, If
‘one thing ‘followed’ another, but both ‘thirgs’ were in fact everts, tien
im the vemantics of English the plenumenon 1s registred as" hat
happened, tken this happened’: two processes, related to cach other
in time. The typical way of expressing a time relationship between
Processes in the grarimar is by a hypotactio or parstactie clause
eomplex:
after they had anadunend it, people applauded pe
they anno.need it, then people applauded 12
Bur by the use.of « grammatical metaphor, the time relation is encoded
by a verb, fodlow; a8 if there was ‘ust ane process—a ‘thenring'—instead
of two—one of announcing and ane of applauding
Lotus make it quite clear tha: exp-essitg something metaptoricelly
's intrinsically neicher better nor worse than expressing it congruently.
or ‘literally’. All ase of language embodies a great deal of metnphor,
there is no reason to suppose that larguage has eve: been without it,
since it became Tengnage as distinet from protolanguage. Ove could
soe certain developmen's even in protolanguage as being metaphorical,
0 the extent that Mey involve a tcansfsrence ef meanings. Tlie question
is whether metaphor is involved ia the phenomenon we are considering
hire; and if so, while each will appear as seciaphorical if viewed Irom
the standpoint of the ather, whether there is reason for regarding either
‘one as the more congruent,
94Here, in fact, there are grouncs for interpreting the variation as
metaprocical, ix a way that is quite independent of whether it occurs
in speech or in writing. That is to say, an example such as
‘The council’s proposed seplacement of subsidies Ly a loan is incur-
ring the bitter resontment of the medical arofessiui.
contains a consideratle amount of grammatical metaphor, and we ean
produce various less metaphorical versions of it, such as, for instznee,
Because the councl is propusing “0 lend money ‘nstead at giving
it, (people who work as) dectors are feeling bitterly resentful,
Doctors sitteily vescut il that the council is 2ropesing that money
should be lent ins:ead of being given.
It makes no difference whethe- on a particular oezasion auch a “orm
was written or spokea; the point is that the less metaphorical variants
tend to.oecur in speech and the more metaphorical onet tele te occur
in writing. Hence we can say that written language is assockated with
the uss of grammatical metaphor.
Atthe samettime, there are many instances of grammatical metaphor
that are just as much associated with syeucli, Some of them may iave
beén borrowed from writing in the first place; but not necessarily s0—
and some arc paitivularly characteristle of speech, such as that by Whicit
a process is represented by the combination of a lexically ‘empty’ verb
(ave, do, make, rake, go, give) with a noun expressing the event or
action: have a bath, inake a misiake, give a growl, and sc on. The same
Principle is at work acre as with other nominalising structures: it is
possib.e to expand nouns more casily than verbs, ard to say things like
make the same silly mistace, gave two or three warning grovls, ave
a nice hot bath, whick are wot easy te express with the ve-bs err, growl,
and. bothe.
Nevertheless this kine of c-03s-codirg in its more complex forms
is particularly characteristic of written language, and is ore of the ways
it achisves its high lexical density. One point to note shout ir is that
it is difficult for young children to understand. We must distinguish,
in this respect, ameng three different phenomena: generalication,
abstraction, and metephar, sinee they are. mastered at different times
ir children's larguage development.
1, Genevalisation; from specific to general. Children have no difficulty
‘with this; a5 s00n a3 they get ints the mother tongue, the fact that
they have learnt to name things