Discourse Semantics Meets Lexical Field Semantics 
Beate Firzlaff and Daniela S. Kunz 
German National Research Center for Information Technology (GMD) 
KONTEXT Natural Language Systems 
Dolivostr. 15 
D-64293 Darmstadt, Germany 
{firzlaff, kunz}@darmstadt.gmd.de 
Abstract 
The focus of this article is the integra- 
tion of two different perspectives on lexi- 
cal semantics: Discourse Representation 
Theory's (DRT) inferentially motivated 
approach and Semantic Emphasis The- 
ory's (SET) lexical field based view. A 
new joined representation format is de- 
veloped which is exemplified by anal- 
yses of German verbs. The benefits 
thereof are on both sides. DI/T gains 
basic entries for whole lexical fields and, 
furtherlnore, a systematic interface be- 
tween semantic and syntactic argument 
structures. SET profits both from the 
much larger semantic coverage and from 
the fine grained lexical analyses which 
reflect inferential behaviour. 
1 Introduction 
The construction of lexical entries is one of the 
crucial and challenging tasks given in the field of 
computational linguistics. In the ideal case, lexi- 
cal entries fulfill, among others, two requirements. 
First, the representations are suitably fine grained 
such that they capture lexeme-speeific distinc- 
tions. Second, the lexical entries are sufficiently 
general, for reflecting similarities between single 
lexemes. Furthermore, the information they con- 
tain should systematically link various levels of 
description, e.g. syntax and semantics as well as 
referential and inferential potential. The latter is 
of special interest for text analysis as opposed to 
sentence analysis (eft for example (Haenelt, 1994); 
(Haenelt and KSnyves-Tdth, 1991)). 
Corresponding to these requirements, we ex- 
ploit the specific strengths of two distinct seman- 
tic theories. These theories are Discourse Rep- 
resentation Theory (DRT) ((Kamp and Reyle, 
1993); (FraCaS-DS, 1994)) and Semantic Empha- 
sis Theory (SET)((Kunze, 1991); (Kunze, 1993)). 
However, our central goal is an integration of D113' 
and SET. It will be shown that this integration is 
possible and of benefit to both theories as well as 
to the construction of lexical entries. To achieve 
our overall objective, the following four points will 
be exemplified by joined representations of Ger- 
man verbs: 
• DRT profits from SET's systematic deriva- 
tions of thematic roles and of morpho-syntaetic 
features on the basis of predicate-argument- 
structures. These features include both gram- 
matical and prepositional case. 
• DRT gains a purely semantically motivated ori- 
entation towards lexical fields. 
• Dll\[l' covers much more semantic phenomena 
than SET. Therefore, DI{T offers SET the pos- 
sibility to test its results against a semantic 
background that e.g. includes plurals, tenses, 
and attitudes. 
• DRT's fine grained lexical analyses are 
grounded in inferential behaviour. These lex- 
ical distinctions mark possible starting points 
for refining SET's representations. 
The paper is structured as follows: DRT's and 
SET's basic motivations, principles and formal 
means concerning lexical semantics are retraced ~n 
sections 2 and 3. The new joined representation 
format is introduced in section 4 by analysing the 
German verbs leihen (in its variant to lend) and 
verschenken (in its variant to give as a present). 
Moreover, section 4 provides evidence that the 
four main points stated above are backed up by 
the joined analyses. Finally, directions for further 
research are pointed out in section 5. 
2 DRT -- Inferentially Motivated 
Discourse RepT~sentation Theory (DRT) is first 
and foremost a theory about discourse interpre- 
tation, i.e., it is essentially textually oriented 
in natm'e. The meaifing of sequences of sen- 
tences is seen as strongly connected with their 
382 
inferential behaviour. Therefore, work on lexi- 
eal seinantics in tile Dll2F frmneworl¢ ((Kam I) and 
Rofldeutscher, 1994a); (Kamp and Rotldeutscher, 
1994b); (ll.otadeutscher, 1994)) investigate.s the 
role of lexieal information in supporting infer- 
ences. Consequently, lexical distinctions correlate 
with non-equiwflent sets of associated inference.s. 
The following ('.xalnt)les illustrate that the Ger- 
man verb leihen (in its variant to lend) implies in 
(:ontrast to the German vert) versehenken (in its 
variant to give as a present) the lending t)erson's 
belief in a return of the involved object,: 
(la) Calvin lciht IIobbes eine Krawatte. 
(Calvin lends Hob/)es a tie) 
(lb) Calvin glaubt, daft \[lobbes ihm die Krawatte 
zuriick.qeben wird. (Calvil, believes that th)l)l)es 
him the tic will give t)ack) 
(2a) Ualvi'n versehe'nkt ei'n Hueh an Hobbes. 
(Calvin giw'.s as a 1)resenl; a book to Hobbes) 
777 
(2b) Calvin .qlaubt, daft Hobbes ihm alas Buch 
zuriickgeben wird. (Calvin believes that Hot)bes 
him the. I)ook will give 1)aek) 
\[n line with the ret)resentation format (Icy(lolled 
by Kanlt) and l/,ot.~deutseher, the corresl)onding 
lexieal entries are, twofohl stru('.tures: They (',onsis(; 
of a I)resul)l)ositional and an asserlx)ric Diseo'wrs(" 
l~cpresentation Structure (I)R,S). Th(; underlying 
anat)horie notion of presul)position was originally 
t)roposed by (Sand(, 1992). Presupl)ositional in- 
formation is embedded in the discourse eontext 
by a process called justification, whi(:h (:omt)ines 
1)inding (veriJication) with contextual (mri(:hment 
(accommodation) in varying prol)ortions. 
Kamp and I/.o\].~dcutscher model the inl;erface 
l)e~ween syntactic and semanti(: al'glltrlents as a 
list of t)airs. Each pair consists of tim general- 
ized case information and the eorresl)onding the-. 
matic role of the m'gulnent slot under eonsi(lera- 
lion. This mapt)ing offers two starting points for 
an integration of DRT and SET. 
Firstly, the representatioi~ format fin' the gener- 
alized case information is only sketche(l, an algo- 
rithm for case assignment is not given: With each 
verb is associated a .(liven set of so-(ailed theta 
roles Oi. These theta roles are arran.qcd in a fixed 
hierarchy, the theta-hierarehy. \[...\] Those argu- 
ment phrases wh, ich get assigned a theta role also 
get assigned a particular case (Nora, Aec, (etc.). 
\[...\] Case assignment is partially determined by 
kit(', theta hierarchy in that the argument phrase 
which bears tit(" highest theta role (in th<' sense 
of this hierarchy) always gets nominative case. 
((Kamp and l{ot.kleuts(:her, 1994a): Pl). 109f) 
Secondly, the thematic roles are specified indi- 
vidually for each lexical entry, there in no get> 
eralization with respect to lexical fields. As an 
example, the. interface list of verschenken is given 
ill Figure 1, where the eompoimllts of each pair 
m'e displayed vertically. 
e(:: versehellke, n 
< 0l) NOM > < 02, A()C > (< 0a, an + A(X; >)\] 
Agcnt&.Sourcc Theme Goal 
Figure 1. lntcxfaee lint of verschenken. 
The, discourse referent ec and the thematic roles 
of the. interface, are. direct links to the DR, S ret)-- 
resenl;ing the hi(airing of the German verb yew. 
schenken (of. Figure 2). The event comi)lex ec, 
whi(:h stands tor the verb itself, in described as 
a process e, which is caused by an action e* of 
a person p. p tel)resents l;he one wit() gives the 
t)resent u to ;mot, her person q. The giving itself is 
(:hm'aeterized by the concept (HIAN(IE-SIGN. The 
signs changed are those of the disi)osal and owner-- 
ship relations So and sl: p looses the disposal and 
ownershi 1) of u and q gains them. The former cir- 
(:umstances of disposal mid ownershi t) (so and ,st 
abut on ee: ,So ZXZ e.c .sl 73(2 eel m'e t)resut)t)osed , 
the t)ost, statcs (ce bC s2 ec ~ sa) are asserted. 
12(: t) q 8 2 S 3 1l 
e e 
(~* CAtJSEveT,~chcnkc) ~ (! 
p ag(mt(e ' ) 
. __ ( ..... >: \[CHANGE SlGN(p~q,u 
l ) (1 SO sl It soln'c:e(e) 
u = theme(e) 
Is,: ~s~ ~ .... ,g(,,,t ((,c) 
SO X" (!C Sl X" (:(: I) := sotlre(~(ec) 
L : ~ _ A: Z __ q := goal((c) i( p R l,J ,q' 
U : t, henle((!c) 
s,: \[ ~DISP&OWN(1),u)~ 
el: ~(£ S2 eC ~(. 8,2 
1( A,q'H 
Figure 2. DRS of versehenke'n.. 
The (;xarnt)le inf('a'ences (la) to (lb) and (2a) to 
(21)) result froln differences in the lexical DlI.Ss of 
leihen and verschenken. The main t)oint here is 
that tllo. (.;elInan vei'l) leihen intt)lies the lending 
t)crson's belief in a l'eturn of the involved object. 
On the basis of (;his belief it in easy to inR.'r front 
(la) to (lb). ltowever, there is no similm' SUl)I)ort 
for inferring from (2a) to (2b). A detailed lexical 
representation of leihe'n will be given in section 4. 
3 SET - - Lexical Field Based 
Semantic Emphasis Th.eory (SET) has identified 
princit)les that allow to link a prototypical de- 
scription of a situation to a number of proto- 
typical meaning descriptions of con(:rete lexeines 
383 
that; are suitable to refer to that situation. The 
link is based on a set of well-defined and sys- 
tematically occurring mappings (cf. (Firzlaff and 
Kunze, 1995)) rather than on intuitive criteria. 
Given a basic semantic form (BSF) as a cornmon 
starting point, we derive semantic and syntactic 
case frames and construct prototypical meaning 
descriptions of concrete lexemes by refining the 
BSF. Additionally, the rule based interpretation 
of a BSF delivers a prototypical description of the 
corresponding situation. 
The set of lexemes that are suitable to refer 
to the same situation constitutes a lexical field. 
The field as a whole is characterized by a BSF. A 
BSF is a propositional description. It consists of 
a predicate and a nmnber of arguments, each of 
which is either a predicate-argument structure or 
an elementary argunlent. In general, elementary 
arguments are represented by variables that have 
to be filled in by phrases which denote reference 
objects (participants of a situation). 
The number of arguinents, as well as the de- 
cision whether the arguments are elementary or 
propositional, both depend on tim predicate that, 
directly takes these arguments. We derive the 
participants' ttminatic roles (deep cases) in accof 
dance with a set of general rules. Semant;ically, 
each pair of a role and the predicate directly dom- 
inating an elementary argument demands partic- 
ular selectional features for that argument. The 
BSF describing the field of change-of-possession 
(with one object to be transihrred) and the de- 
rived deep cases are given in Figure 3. 
CAUSE r: (agens,act) 
(ACT(r) p:<source,have) 
ET q: (goal,have) 
(BEC(NOT(HAVE(p,u))) u:(from-obj,have) 
BEC(HAVE(q,u)))) u:(to-obj,have) 
Figure 3. BSF and deel) cases. 
From both the syntactic and the semantic point 
of view, the BSF delivers the maximum case frame 
of the lexemes that constitute the lexical field. 
Some of the roles of the maximum case frame can 
be put into the foreground; these are said to have 
cmphasis. Some roles nmst not be verbalized ex- 
plicitly; these are said to be blocked. In the subset 
of roles that are not blocked there are, on the one 
hand, roles referring to obligatory actants and, on 
the other hand, roles referring to optional actants. 
Which roles have emphasis and which do not have 
emphasis, which are the ones that must be verbal- 
ized, and which are the ones that need not be ver- 
balized is determined according to general rules. 
Exploiting the field specific possibilities to make 
some variables denote the same reference object 
(by renaming of variables) results in more specific 
BSFs. These then describe partial lexical fields 
like, e.g. to give or to take. 
By adding infbrmation about, emphasis and 
blocking of roles, a BSF is transformed into a num- 
ber" of prototypieal meaning descriptions. We can 
then derive systematically which are the suitable 
grammatical realizations of each role. However, 
there are two important points concerning the 
determination of which grammatical realizations 
are possible: Firstly, the predicate that takes the 
corresponding elementary argument directly and, 
secondly, the choice of that subset of roles of tim 
maximum case frame that are not blocked. One of 
the three prototypical ineaning descriptiolm that 
constitute the partial field of to 9ire and tile gram- 
rnatical case assiglmmnt of verschenke,n 1 is given 
in Figure 4. (Those parts of l;he description t;hat 
have emphasis are written in bold face. Tile oc- 
curence of a variable preceded by "T" is blocked. 
The grammatical realization of the optional actant 
(an+accusative) is put, in brackets, z) 
CAUSE 
(ACT(p) (agens,act>: nora 
ET 
(BEC 
(NOT(HAVE(Tp,u))) (from-obj,have}: ac(: 
BEC 
(HAVE(q,Tu)))) (goal,have): (an+ace) 
e.g.: Calvin vcrschenkt ein Buch an Itobbes . 
(Calvin gives as a present a book ~o Hobbes ) 
\[pnom uacc qan+acc \] 
Figure 4. Prototypical ineaning description and 
grammatical case assignments. 
However, BSFs do not only provide the ground 
for the derivation of grammatical features. They 
are also suitable to derive prototypieal situation 
descriptions. In order to do so, instantiation rules 
must be applied to a BSF in a recursive way. 
The application of instantiation ruh;s has to be 
regarded as an interpretation of every partial de- 
scription in a BSF. Some of these parts are, then 
represented by variables that have to be filled in 
by objects referring to states or (;vents, and other 
parts deliver relationships between these states or 
events. In addition, some of the instantiation rules 
provide temt)oral and/or spatial constraints that 
are applicable to (tim corresponding parts of) a 
prototypical situation description, e.g., etimc is a 
mapping fl'om the set of events or states to the set 
of temporal entities (etime: g -+ T). 
In general, tile instantiation rules provide struc- 
1Generally, this grammatical case assignment is 
suitable for about 20 verbs of the partial field to give. 
2More precisely, there is a mapping front the set of 
variables into the set of nominal phrases (more gener- 
ally, parts of speech) f: V --+ b r. 
384 
CAUSE 
(ACT(p) 
ET 
(BEC 
(N()T(nAVE(p,.))) 
BEC 
(HAVE(q,u)))) 
Presul)l)ositions: 
(,.: 
,,,: ,.(,f(f(p)): ~;ht,~.,+~(:~ 
(:,.,: ,.,f(f(p)): lo(:~t-h~lv,: 
A ref(f(u)): thenteqtave 
init(e~): nAV~,;(t),n) 
fin(e2~ ): 
,,=: ,-0r(f(q)): ~,,c~-h~v,, 
A ref(f(u)): them(.'-h~lv(, 
init(e~.): -q~aw,:(q,u) 
fin(e~): 
Assert:ions: 
OAUSI.:(e, ,c~) A etime(el)--etinw.(e~):=(~.tinw.(e) 
ACT(p) 
(e2, ,e2~ ) A etinw.(e2,)--etime(e.22)=:ctmm(e2) 
'rl(ANsrrioN (inii;(eua),fin(e~ )) 
- 'IIAVE(pJI) 
TRANSrr~ON (init(e~),fin((',u~)) 
IIAVE(q,U) 
Figure 5. llSF and l)rotoi;ypieal sil;uation (les(:ril)l;ion. 
luring mechanisms in tern~s of t)r('.supposii;ions 
~ttl(1 assert,loire. As an ex;mq)le, (:onsi(lev the 
t)redieate BEC: I1; has one mgmnent whi(:h is ~ 
t)redieate-.a.rgunw.nl,-strucl;ur(~. This sl;ru(:l;ur(~ is 
1;o tie inl;ert)rel;e(1 &s the finnl sl;a.l;e of ;1 tra nsi- 
l;ion. Because of the insta.n|;iation rule of BEC the 
initial state (init(e.))of the l:ra.nsition (c)is l;he 
"()t)t)ositC' of the final state (fin(e)), i.e., BEC(A) 
is inl;erpreted as c: TII.ANSITION (~A,A). Accord- 
ing to (aUilg 3,1l(1 Kiistne.,', 1q90), init(c) (i.e. ,At 
is I;11(; 1)resul)t)osition of c, and " ," does not; ;ffl'e(:t 
A's t)resut)l)osition, e.g. seh~cA;iotm\[ restrict;ions for 
A's elelnenl;&ry arguments. A more (~xtensive ex-. 
mnl)le, of the d(wivation of t)i'ol;otyl)i(:;d sil;ua.tion 
des('.rit)l, ions is given in Figure 5? 
The situai;ion 1)rol;ol;yl)i(:Mly th~s(:ribe(t in Fig- 
llre 5 (:all 1)e reDrred to by al)Otll; 65 (~ertmm verlts, 
i.c.., the elements of the partial fM(t to give includ- 
ing our samph~ verbs vcrschc.nkcn (in its w~ri;mt 
to .qivc as a p'rescnt;) and h:ih, cn (in its vm'iant 
to h'.nd). As far as the degree of speeifieal;ion is 
(:on(:erned the des(:rit)i;ion is at leas(; suil,able as 
(:ommon d(mominator. Since SET's l)rineipal ori- 
ental;toil is l;owaa'(ls the systemal, i(: des('ription of 
le~cical fields rather than of single lexi(:al entries, it 
provides ret)resentations whit:h tend t;o 1)e mtder- 
sttecified with respect to e.g. Dll\[\["s requirements. 
However, due to SITI"s gener;,l eq)pr()ach mty fltr- 
ther spe(:ification of its (h~s(:rif)l;ions lea(Is to an 
enlargement of the ret)resental;ion r~ther t;h;m to tt 
change of the common denominator. The (les(:ril)- 
lions i)rovided by SET are suitable as the basis for 
fine grained representations. Theretbre, one can 
expmM the lexicM enl, ries rather (;turn (:onstrut:t- 
ing l;heln ea(;h and every time flom s(;ral;('h. TO 
exentl)lify l;his, in the next secl;ion, the ret)resen- 
tatioll of lcihe'n (in its w~ria,nt to Ic'nd) is emiched 
by the hmding ltel'SOll'S belief in a return of t, he 
involved object. 
aref(f(x)): V -~ .7 -+ 1). And 7) is the set of 
rel~rence objec|;s. 
4 The Puzzle Fits 
Based on the hypothesis that: SET's proi;ol;ypical 
situation descriptions ca.n be interl)reted in the 
same way as 1)l/Ss we have l)ro(:ee(led to a new 
joined ret)resentation format. Since w~riabh~s in a 
I{SI" have to 1)e filled in l)y r('.ferenc(~ objects and, 
fltrtherm()re, the rtR;llrsive ;q)t)li(:al;ion of insl.m> 
tiation rules provides wu'iM)les of the stone kind 
\[*Of eVOlI\[;S &Ild SIALI;(~S, S\]~r\["S l'efereiIce o})je(',(;s }l.lt(l 
l)l{T's discourse referents are reg~r(ted as etlUiV- 
M(',nl; metals of (!Xl)ression. Tlmreli)re, the joined 
l(~\[)l(iSelll;aA;i()li fOl'llt3\]; ltses I)I/\[F's boxes, llow- 
ever, it is enriched witih among others, n revised 
inLerI, tt:e I;() syntax where the thematic roles m'e 
derived according t;o lISFs. 
The sample rel)resent&tions giwm in Lhis sec- 
tion exploit Karat) a.nd \]{ofideutsc.her's idea of h'.x 
ical axioms (of. (Karat) ~md l{.oB(h~uts(:her, \] 994a); 
(Kaml) and l{oBde.uts(:her, 19(,)4t)); (Ro\[/d(!ul,seher, 
1994)). We (:~i1 ~ (lis(;inguish in a l);4ra, ltlet;ri(; fas\]t- 
ion between I;11(; s('an;mti(: (:Oml)onenl;s of 1;t1(! en- 
Lries that (:hma(:|;(wiz('. (pnrtia\]) lexi(:al \['i(;l(ls and 
the (:on(:et)t si)e(:ific inf()rnt~tion in i;he axioms. 
The first pair of axioms introduc(;d below mir-- 
I'OI'S l;he fact tlud; the configuration a l)brevialxxl 
by eu: (cul ,cue) (ef. Figure 5)is suitabh~ I:o spec 
ify a wlriety of h',xica.l fMds wherein the s('ammti(:s 
of the (~lenw.nts inv()lves a sl)eeinl kind of '(:han/~e'. 
Some exalnples of these fMds are ch, angc-@placc 
(e.g. to travel fro'm o'nc plat(" to another), cha'ngc- 
of-class (e.g. to promol, c somebody to a certo, in 
rank), ",rod changc-@posscssior~,. Ae(:()rdingly, we 
t)roltose a predicate hierarchy, wheret)y the i)re(li- 
(;~tes PI,ACI,;~ \[SA~ \[lAVE ~:~\]e st)ecializations of the 
I)re(ti(:ate STATE. Note, however, that 1;his does 
not affect the ine(:hmdsm of role deriva.|;ion 1)re- 
seilte(l in section 3. 
In the t)rototyI)ical sit;m~tion des(:rit)tion 
((:f. Figure 5), (,'e inchMes eel a.nd c~2. F,a.(:h ()f' 
Lhese (10,nol;es ;1~ TIIANSI'I'ION tl()IIl tttt initiM st, ate 
to a. final state, i.e., from init(s,e:L) and inil,(c.e2) 
(the presupposition) to Ji'n(e.2t) m,d Jin(c.2.~) (the 
385 
assertion). Because of the temporal identity of e21 
and e2~, there are temporal overlaps between the 
initial states as well as between the final states. 
In the axiom defining CHANGE-SIGN's prestate, 
so's consequences sl and s2 correspond to init(e21) 
and init(e22). In the axiom defining CHANCE- 
SIGN's result state, s0's consequences st and s2 cor- 
respond to fin(e~:t) and fin(e22). The axioms have 
in common that they involve the concept CIIANGE- 
SlC, N (cf. c2 in Figure 5). The axioms are given in 
Figure 6 ("O" denotes temporal overlapping). 
r ____ riP0 rl F2 ,SO 
So: | (STATE)) S2: \[~~) 
I (ro,rl,r2) II I t l\] \[ 81 O S2 
I r0 rl r2 s0 Sl s2 
\[STATE0"I,,'=) I S2: t 
(ro,rj.,r2) ij \[ sl O s2 
Figure 6. Axioms for CHANGE-SIGN. 
The concepts defined by means of these axioms 
are, then, used to specify the lexical entry of ver- 
schenken (in its variant to (live as a present). The 
thematic roles and the corresponding grammatical 
realizations result from the derivation presented in 
section 3. PRE(CHANGI,2-SIGN) delivers the first 
part of verschenken's presupposition. The pa- 
rameter STATE is filled in by DISP&OWN which 
is added to the predicate hierarchy sketched al- 
ready as a specialization of the predicate IIAVF. 
Thereby, it is possible to distinguish between the 
pure disposal and the disposal thai; is accompa- 
nied by ownership. 
Furthermore, verschenken's presupposition in- 
cludes the semantic roles delivered by its prototyp- 
ical meaning description. However, the selectional 
restrictions for discourse referents do not differ 
from the restrictions given in the prototypical sit- 
uation description (cf. Figure 5). With respect to 
the semantic interpretation, each of source-have, 
goal-have, and locat-have just means is suitable as 
first ar.qument in a IIAVE-proposition. Generally, 
the predicate directly determines the selectional 
restrictions of its arguments, i.e., the discourse 
referents. Furthermore, for those predicates that 
take more than one argument, it is the order of 
the arguments which additionally determines the 
selectional restrictibns. 4 
In accordance with the prototypical situation 
description given in Figure 5 the DRS for ver- 
seheuken is as follows: 
4Clearly, these are .iust two pieces of information 
for the seleetional restrictions. 
ec: verschenken 
NOV ACC (an + ACC) \] 
< agens, act > < fi'om - obj, have > < goal, have >\] 
p q so u 
PRE(CHANGE-SIGN 
SO: (DISP&OWN)) 
(p,q,u) 
So D(-- ec 
p -- agens-act(ec) 
p = source-have(ec) 
q : goal-have(ec) 
u -- from-obj-have(ee) 
u -- to-obj..have(co) 
ec p q s~ u 
e e* 
OC: e*: 
CHANGE-SIGN(1) q u) c:\[ _____, , )~ 
e ~ CAUSE e 
RES(CHANGE-SIGN 
Sl: (DISP&OWN)) 
(p,q,u) 
ec 23(_ sl 
KASS 
Figure 7. Lexical entry for verschenken. 
Figure 8 clarifies the correspondences between 
DRT's and SET's representation: 
DRT SET 
ec e: CAUSE(el,e2) 
e* el: ACT(p) 
e e~: (e~,,e~) 
so init(e~l): HAVl,;(p,u) A init(e22): ~HAVl,:(q,u) 
sl fin(e21): ~naVE(p,u) A fin(e22): ltAVF,(q,u) 
Figure 8. tl.elevant correspondences. 
The entities constituting ec (the action, the 
transitions, and the causation) are located in a 
common time span. Therefore, the transitions' 
initial states precede ec (so ::)(7_ ee) and the tran- 
sitions' final states follow ec (ec DC st). 
The lexical entry of leihen (in its variant to 
lend) consists of an interface list, whose thematic 
roles are based on SET, and of semantic struc- 
tures, which include and extend versehenken's se- 
mantic components. The inferential behaviom' of 
leihen (exemplified in section 2) motivates a for- 
real description that contains more than the basic 
distinctions provided by the partial lexical field to 
give. Additionally, there is the lending person's 
belief in a return of the involved object, in other 
words, the belief that the CHANGE-SIGN from s o 
to s~ is temporary. Therefore, leihen's represen- 
tat, ions make use of CIIANGE-SIGN's subconcept 
CIIANGE-SIGN-TEMP. This subconcept entails ;t 
transformation of its superconcept's prestate So 
(so DC ec) to its superconcept's poststate s~ (ee 
DC sl) as well as the new poststate s2 (ec ~ s.e), 
i.e. the belief in a return of the involved object. 
ec: leihen 
\[ NOM DAT ACC \] 
< agens, act > < goal, have > < to - obj, have > 
Figure 9a. Interface list of leihen's entry. 
To nmke the description of leihen complete, a 
further lexical axiom which explicitly notes the 
belief in a return of the involved object is ne6ded. 
One of the formal means provided by DRT is the 
possibility to model components of psychological 
386 
attitude states, e.g. beliefs or desires (of. (\[q'aCaS- 
D8, 1994)). ~ p q ,, .~, 
e. e 
p q S 0 tl e * : 
,~o: (D,se)) / ~,: CHANGE SIGN WEMP(p,q,u) 
(1),(I,U) \] e* CAUSE e 
s0 2X~ ec RES(CHANGE SIGN I 
| .... ~go,,,~-~,,I;(,,,:) (.tsP)) I l) 
:= source-have(ec) st: 
u -- from-obj-have(cc) e(', DC sl 
. : to-ob>l.~v,+~) I RES(C"^"GE-S~G"-rEu~ 
ec D(- s2 
\[<ASS 
Figure 9b. Semantic structures of lcihcn's entry. 
This possibility can be used to state the axiom 
which represents the specific semantic contribu- 
tion of C\[\[ANGE-SI(IN-TEMI': it, s poststate ehara(> 
terized by the state s2 of the person r0 being in 
an psychological attitude state one of whose coin- 
ponents (c) is a certain belief. This belief con- 
sists of an inversed (\]IIAN(;I,;-SI(-\]N-eVeIIt C, i.e. a 
return, with its resulting disposal (:onfiguration 
sa. Thereby, the forlner circumstances of disposal 
Sl, that result fl'om the CIIANGE-SI(IN-TEMP-evenL 
itself, are supposed to abut on the return event c. 
c s 2 
su: \[PSYCH-ATT STATE(ro,c)\] 
HEL, 
o s 3 
RES(CHANGE-SIGN 
s3: (D'SP)) I \[ ( l'l 
,,'0,1"2 ) 
e ~X2 Sa 
> 
Figure 10. Axiom for CHANOF-SIGN-TEMP. 
On the one hand, these results mark directions 
tbr the developme, nt of a comprehensive lexical 
theory, that include, s, for example, an elaborated 
concept hierarchy with associated axioms. On the 
other hand, they (:an be used for a detailed recon- 
struction of the inferences inentioned in section 2. 
5 Conclusion and Future Work 
In this article, we have first shown that it is sen- 
sible and promising to comt)ine DRT's and SET's 
perspectives on lexieal semanti(:s. We made use 
of the theory-sI)e(:ifie strengths of the single ap- 
proaches in order to overcome their specific weak- 
nesses and to gain a powerfill means of expres- 
sion for modelling the semantics of lexical entries. 
Second, we have proposed that and described how 
joined representations ('.an be constructed by e,x- 
ploiting tile merits of bo*h theories. 
I%ltm'e work will concentrate on evaluating the 
benefits of this approach fl)r eomtml;ational text 
analysis. Tile joined representation format pro- 
posed he, re is likely to facilitate and improve lex- 
ieal modelling as well as the automatic construc- 
tion of text representations, l%\]rther investiga- 
tions ill otller lexical fields and word classes are 
required in orde.r to aehieve~ a larger lexieal cove.r- 
age. In correspondence with the theory-specific 
strengths, promising subtasks will be reference 
resolution and the construction of conceptual rep- 
re, sentations. 
References 
(looper, Robin et al. 1994. Describing the. Jp- 
pr'oachcs. FraCaS. A 1,5"anmwork for Conltm- 
tational Semantics, Deliveralfle 8. Edinburgll, 
University of l,;dinburgh: t)P. 8-79. 
Firzlaff, Beate and Jiirgen Kunze. 1995. Auto- 
matic Generation of Lcxical Fields. Working 
Papers of GMD, Report 892. St. Augustin, Ger- 
Illall Natiomd Research Cente, r for lnforlnation 
Technology. 
Haenelt, Karin. 1994. Das Textanalyse- 
system KONTEXT. Konzeption und Anwen- 
dungsm6glichkeiten. In Sp'rache und Datcnver- 
arbcitung, 18: pp. 17-31. 
Haenelt, Karin and Michael KSnyves-Tdth. 1991. 
The Textual Development of Non-Stereotypic 
Concet)ts. In ProccediT!gs of the 5th Confl:rcncc 
of the EACL. Berlin: pp. 263-268. 
Jung, Uwe and Herbert K{istner. 1990. Semw,,l,i- 
sche Mechanismen dcr Ne.qation. studia gram- 
matica xxxi. Berlin, Akademie Verlag. 
Kmnp, llans and Uwe Reyle. 1993. From Dis- 
course to Lwic. l)ordrecht, Kluwer Academic 
Publishers. 
Kmnp, tIans and Antje Rogdeutscher. 19943. l\].e- 
marks on Lexieal Structure and DRS Construc- 
tion. In 77u,oretical Linguistics, 20.2-3: 1)t)..(t7- 
164. 
Kmnp, Itans and Antic Rofideutseher. 19q4b. 
DRS-Construction and Lexieally Driven ill- 
f'erence. In Theoretical Linguistics, 20.2-3: 
pp. 165-235. 
Kunze, Jfirgen. 1991. Kasusrelatiorten und Se- 
mantischc Emphasc. studia grammatica xxxii. 
Berlin, Akademie Verlag. 
Kunze, ,liirgen. 11993. Scmemstrukturcn und Ii'eld- 
strukturen, studia grammatiea xxxvi. Berlin, 
Akademie Verlag. 
l~.ofldeutscher, Antje. 1994. Fat Child Meets 
DifF. A Semantic Representation for the Open- 
ing Lines of Kasehnitz' "Das dicke Kind". In 
Th.c.oretical Linguistics, 20.2-3: pp. 237-305. 
Sandt, I{ob A. van der. 1992. Presupposition Pro- 
jection as Anaphora Resolution. In ,lournal of 
Semantics, 9.4: pp. 333-377. 
387 
