Beruflich Dokumente
Kultur Dokumente
,{{
DRT3
1. Negation. KamP
if its discourse
(1)
assertions)"
holiday?'
more-' '
B: "Well,
he is not
.- j*q"
-rrt$*fl
\
by
The effect of denial is to reject somethiag that has been put forth tentatively or confidently
is no map from its
else. The force of denial is that a certain picture "is not corrct", i.e., there
someone
referents to objects such that &ese objects satisfy the depictetl conditions.
fimctiotr
The contcrt of the negative s,rtenct is a ftnstion of the embdded affirmative setrtstce, a
represented as
negated sentence.
(5)
(6)
JanesdoesnotownaPorsche'
on the representation of
tle non-
We slrould see the DRS above as including a discourse referent and two conditions. The familiar Jcnes (x)
and the second consisting of the internal box with its contents, affixed by
'-'.
drstinguishes this DRS from previous onrs is thatit contains a DRS as companent.
The condition will be verified by an embeddingfunctionf ( which associates an individual a to x) only
there is
no yr/ay
if
words f satisfies the condition" if there is no b, b is a Parsche such that f(x) owns b.
The scope ofnegation. The scop ofnegation does not usually include the subjec! as seen by inspecting
(7). There is an asymmetry betweer subjects and objects. Subjects scope above negation, objects mostly
scope inside negation
(7)
a.
s"ry:P#jggilolsH
"?
PH;fl[e; r
b.
U.
He didn't say
d-
e.
0
rro
tJ r:c/:ff
-rR
lw
uJY#17"."
l.L
Syntax
The inventory of PSRs is enriched with a nrte for the auxiliary component, since unlike affirmative
sentstrces,
(8)
PSRS
s+ l\F ^ IP',
VP')
Aux
(8)
Aux^not^\lP
lT
(Motlalxhave+n) (be-irg)
Aux
\--*
.-----IIP
I D", '---
'l
oJ,
own I
$oron"
(e)
x
Jone (x)
[x does not orrnaPorsche ]
(10)
x
Jones (x)
[xowns a Parschel
In expanding a subordinate DRS introdrced by negation rre may draw on discourse referents
processing (12) we may
rntroduced earlier (into a main or higher box), as wi& anaphoric prurouns Thus in
link the pronoun it, \yith the referent
(11)
it
(12)
x,yz
Jones
(x)
Porsche (Y)
x owns y
z=x
- tZ-Ii[AE
(13)
Jones
(x)
Ponche (y)
x or*rl}$y
(14) is not
15, we cannot ao$vert it into (16). This.is in agreement with our imrftion that dismurse
ameptable. The referers y is inacessihle &om the position of
(14)
liks it##
it
(15)
Porsche Cy)
[x owns y]
*(16)
Porsche (y)
xownsy
2. Aecmsibility
a,
In other words only discourse rderents in fhe mrin box or in a box higher than that containing the pronoun
are accessible to pfonorss. Compme (17) qrhich ooiltains Fqper
(17)
he
liks
(18)
Jone {x}
Ulysses (y)
l-ownsy
it
local
Y-
(1e)
Jones (x)
Ulysses (y)
DRS K
Tmth, Accessibility
is true in the wdel M tf there is away of associating members of UMwith the discourse referents
Technical Notions
are
compatibb
tf
they assign
ttr
they are both define4 i.e. for any a which belongs to both Dom(0 and Dom(g), i-e., f(a) = c(a)
nunction g is an extension of f if g is compatible with f and the Domain cf g includes the domain of f.
Since the conditions of negation,
-,
DRS is like.
Definition:
A DRS K confined to V (a vocabulary) alrd R
(:
of a srbset UK (poesibly empty) of R and a set Conr of DRS conditions confined to V an'd R
A DRS condition confined to V and R is an expression of one of the following fonns:
(d)
(e)
(S
x(. where x belongs to R and ( is a unary predicate corresponding to an inhansitive rcrb from V.
x(y. rvhere x, y belong to R and ( is a binary predicate corresponding to a transitit
- i!
from V.
e verb
Conditions may be atomic (simple) or complex ( those which have other DRSs as constituents).
Definition of Verification
Let K be a DRS confined to V and R, let y be a DRS<onditiorl and let f be an ernbeddrng (function) from
R into l\,{, i.e., a function s,hose domain is included in R and whose mnge is included in UM.
r1l\
(i)
(ii)
ifff
(a)
(b)
y is of the forrn
iff
x: y and f
U1"1
Namer"r
(c)
y is of the form q(x) rnd f rnaps x onfo fhe element a of Us srch that a belongs
16
predu(q).
(d)
y is of the forrn
x(
predu(O.
(e)
y is of the fcrm
dy,
and
elements a and b of
Uu such tlrat
<a, b>
belongs to Predu($.
(0
y is of the forrn
that
K'
DomG): Dom(fl
ur
Ur
and g verifieslC
inM
2.3 An iltustration. Considerthe following discourse and let us check whether it isverifred in ldr.
02)
Predla| where
4 e}
&,
<Candide e>}
likeqn.
t <4 dZ <4
e>,
+,
e>)
i0
iir)
(23)
Jones (x)
Ulysses (y)
x likes y
(y)> belong to NameMl and <f(x), f(]IP belong to likes Ml' If we add
thesecondsentence,wewillbedealingwithacomplexcondition-K:
Q4)
Jones(x)
Uysses (Y)
xlike*y
FX
u
referefrs x
This DRS is verified by an embedding f which mape the discourse
andthe discourse refererry onto the individual
u It is straigltforwardto
ad
<Jones'f(x)>and,<Ulyssegf(yPbelongtoNameMI,<(x)'f{y)>belongtolks}vl1andx'zars
map,pd onto the same elemqrt C-:onsider the last mndition now:
{2s}
Dom(f)
The only possibiliry* is that u is mappd orto c, the only indfuidual in the edension of Porsche.
a discourse referent
(fail
neglect,
forget), it will not be blocked by a double negation. While in (1) the pronoun if cannot be interpreted as
dependent on a question and in (2) th6 pfonour cannot deeild ofi an antswer, the definite in (3) roy depnd
on the preceding indefinite and the i, in (4) can be taken to ref,er to an umbrella.
it
(l)
Bill didn't
o)
(3)
(4)
John didn't remember not to briag an umbrella, afthough we had no room for iL
(5)
DRSr
DRS
&y
John (x)
umbrella(y)
bdns(ay)
Double twgfiiofrurplugs
the
as a
*double plugl'-
t0