The representation of syntactically unexpressed com 
to nouns 
Toni Badia and Roser Sauri 
tbadia@upf, es and roser, sauri@trad, upf. es 
 lements 
Institut Universitari de Lingfifstica Aplicada 
Universitat Pompeu Fabra 
La Rambla 32 -- 08002 Barcelona 
Catalunya - Spain 
Abstract 
We address the representation of nouns hav- 
ing complex argument structures like deverbal 
nominalisations. In particular we address the 
semantic representation of syntactically unex- 
pressed arguments.We put forward a treatment 
of this kind of optional complements in a frame- 
work that combines HPSG syntax and the se- 
mantic approach in GL (Pustejovsky, 1995). 
1 Introduction 
In this paper we address the representation of 
nouns having complex argument structures like 
deverbal nominalisations (and other nouns with 
complements). These classes of nominals very 
often appear in the surface string without their 
complements, even if they are semantically im- 
plied. Such semantic implications are often 
shown in discourse, so that anaphoric references 
cannot be accounted for without taking them 
into account. We put forward a treatment of 
unexpressed complements to nouns in a frame- 
work that combines HPSG syntax and the se- 
mantic approach in GL (Pustejovsky, 1995). 
In the next section we present some data con- 
cerning the semantic implications of syntacti- 
cally unexpressed complements. Our data are 
basically from Catalan, but the discussion is 
clearly applicable to other languages including 
English. In the third section we discuss the de- 
ficiencies of the standard treatments of option- 
ality of complements. Finally, in section four 
we put forward our proposal. We concentrate 
on the representation of deverbal nominalisa- 
tions, but we discuss some other kinds of op- 
tional complements to nouns. 
2 Data 
Most predicative nouns (i.e., those having com- 
plements) may or may not have their comple- 
merits present in the sentence. 1 In other words 
most complements to nouns are optional (in an 
appropriate context): 2 
(1) a. Avui he vist el pare del 
Today have(l-s) seen the father of-the 
Joan 
Joan 
b. Hi ha dos pares que no han 
There are two fathers that not have(3-s) 
vingut a buscar el nen 
come to fetch the child 
(2) a. Comprar6 dos fulls de catrolina 
Will-buy(I-s) two sheets of card 
b. Escriu-ho en un full 
Write-it on a sheet 
(3) a. 
b. 
Aquesta tarda un grup de nens 
This afternoon a group of children 
jugavaa la plata 
played on the square 
E1 grup 1' ha acceptat molt 
The group him/her has accepted very 
b@ 
well 
Of course optionality of complements is not 
limited to nominal complements: it is pervasive 
among verbal complements as well. Even within 
the context of verbal complements the discus- 
sions in the literature show that the distinctions 
between classes of complements are far from ob- 
vious. At the simplest level a twofold distinction 
1Here and throughout this paper we use the term 
"predicative noun" to refer to nouns that can have com- 
plements, thus including deverbal and adjectival nomi- 
nalisatioas and relational nouns. 
2In the following a-examples the complements of 
nouns that are omitted in the b-examples are written 
in bold face. Note that the contexts usually have to be 
different. 
may be proposed between (obligatory) comple- 
ments and (optional) adjuncts. This however 
does not comply with the facts since it groups 
together very different classes of optionality: 
grammatically induced elision (4), optionality 
with semantic implications (5), as a pure ad- 
junct (6)... 3 
(4) a. Avui s' ha comentat la novel-la 
Today ReflPr has commented the novel 
b. Avui els estudiants hail 
Today the students have(3-p) 
comentat la novel-la 
commented the novel 
(5) a. Menja sempre en tornar de 1' 
Eats always when come from the 
escola 
school 
b. Menja sempre pa amb xocolata 
Eats always bread and chocolate 
en tornar de 1' escola 
when come from the school 
(6) a. Avui he estat pintant al 
Today have(l-s) been painting at-the 
parc 
park 
b. Avui he estat pintant 
Today have(I-s) been painting 
These very simple examples show that the 
presence or absence of complements to predica- 
tive heads is not uniform from a semantic point 
of view. There are complements which are se- 
mantically external to the predication, whereas 
there are others that are internal (that is to say, 
that contribute some entity to the relation de- 
noted by the predication). And complements 
may or may not be present according to their se- 
mantic relation to the head. Most external com- 
plements (i.e., adjuncts) are always optional: 
they may be present or absent irrespective of 
their particular relation. 4 But there are many 
internal complements that can be optional and 
some cannot be present except under very spe- 
cific circumstances; these are the complements 
described as default- and shadow-arguments re- 
spectively in Pustejovsky (1995:63f): 
3Some authors have proposed a graded classification 
consisting of 5 to 7 classes (Somers, 1987). 
4However there are a few cases where one might want 
to say that there are obligatory adjuncts, as for example 
in: this suit washes easily. 
(7) a. D-Arg: John built the house out of 
bricks 
b. S-Arg: Mary buttered her toast with an 
expensive butter 
When dealing with a purely syntactic gram- 
mar (or lexical representation) all these as- 
pects boil down to determine the degree of op- 
tionality of the complement and to deal with 
the variation in syntactic behaviour accord- 
ingly. That is to say in many implementations 
of syntax some optionalities are dealt with by 
general rules or principles (this is clearly the 
case of the optionality of the logical subject, 
which is either accounted for by a passivisa- 
tion/impersonalisation rule of the grammar or 
by a lexical operation) and some are accounted 
for by listing them as different entries (or just 
as subentries of the same entry). But except for 
the cases where a grammatical relation clearly 
exists that applies to most lexical items of a 
particular class (as passivisation to transitives), 
the optionality of complements is dealt with by 
listing all the options in the dictionary. Further- 
more such a treatment does not include some 
of the complements that are allowed only un- 
der certain very specific circumstances (e.g. D- 
Arg, S-Arg) since these can only be spelt out 
in semantic terms. 
Not surprisingly the application of these no- 
tions of obligatory and optional complements 
(developed for VPs and Ss) to NPs has not been 
successful: too many arguments in the NP are 
optional and there is too little grammatical con- 
trol (such as the one we find in passivisation) 
for general, syntactically based treatments to 
be successful. Note that even the simple dis- 
tinction between objective and subjective com- 
plements cannot be made operative on syntac- 
tic terms in Catalan and other languages (i.e., 
without taking into account their semantics): 
(8) a. l' estudi de les plantes 
the study of the plants 
b. la soluci6 dels estudiants 
the solution of-the students 
c. 1' avaluaci6 dels inspectors 
the evaluation of-the inspectors 
But in addition there are strong reasons to 
consider that a semantic approach has to be 
taken to predicate-complement representation 
if we consider discourse factors, such as coher- 
ence, anaphora and the recovering of implica- 
tions. Consider the following examples: 
(9) a. La decoraci5 del pont 
The decoration(f-s) of-the bridge(m-s) 
ens ha portat molt de temps, perb ha 
us has taken much of time, but has 
quedat molt ben acabada! 
resulted very vell finished(f-s)! 
b. La traducci6 d' aquest pamflet m' 
The translation of this pamphlet me 
ha costat molt, perb al final 
has costed a-lot, but at-the end 
crec que m' ha quedat molt 
beleive(1-s) that me has resulted very 
natural 
natural 
c. Traduir aquest pamflet m' ha 
To-translate this pamphlet me has 
costat molt, pero al final crec 
costed a-lot, but at-the end beleive(1-s) 
que m' ha quedat molt natural 
that me has resulted very natural 
d. Ha vingut una mare aquest matf. 
Has come a mother this morning. 
Venia a dir que el seu fill no 
Came(3-s) to tell that the her son not 
podr~ venir a l' excursi6 
will-can come to the excursion 
e. Hem amanit l' enciam i l' 
Have(l-p) dressed the salad and it(ac) 
hem hagut de llenqar perque 
have(l-p) hat to throw-away because 
l' oli era ranci 
the oil was rancid 
These examples show that complements not 
explicitly present in the NP or VP can be re- 
ferred to or implied in discourse. Of course the 
subject of acabada (9a) and natural (9b and 9c) 
is only recoverable as the result of the decorat- 
ing and translating act respectively; and the use 
of definite specifiers el seu and I in (9d) and (9e) 
is allowed by the "hidden" complement of mare 
and amanir. 
3 Standard treatment in argument 
structure 
The standard treatment of syntactically ori- 
ented representations introduces some level of 
abstract description of argument structure in 
syntax, so that general grammatical relations 
can be accounted for in a straightforward way. 
Thus, content relations can be stated between 
arguments and their heads that differ from the 
surface syntactic ones. In HPSG, for example, 
control relations are expressed as coindexation 
between syntactic and argument values, in such 
a way that a single valence element realises two 
argument slots; and passive is (lexically) dealt 
with as a change in the correlation between va- 
lence elements and argument ones. 5 
But this level where argument structure is 
usually represented is a direct projection (al- 
though neither simple, nor one-to-one) of the 
surface structure. That is to say there are dif- 
ficulties in incorporating new argument slots 
when no correponding syntactic slot exists. And 
this is so simply because the argument values 
are not thought as full semantic representations 
but simply as the deep representation of syntac- 
tic structure. 6 
Being syntactically oriented, the standard 
treatments of argument structure can only rep- 
resent the types of complements that corre- 
pond to the simple twofold distinction men- 
tioned in the previous section: strongly re- 
stricted obligatory complements and optional 
adjuncts. ~ Nonetheless, there are at least three 
aspects that are not considered in a satisfac- 
tory way: complements to verbs and to nouns 
which are not linked with a strong syntactic and 
semantic relation, but cannot be treated as or- 
dinary adjuncts, i.e., intersectively (as for in- 
stance the kind of complements mentioned in 
footnote 4); complements to nouns generally, 
because they usually are optional (as in (1) 
to (3) above); and some complements to verbs 
which are manifestly optional (as in (5)). This is 
so because in the standard typing of arguments 
5Recall that the syntactic valence is expressed in 
HPSG as a list of descriptions of the complements which 
include their semantics, and that the argument struc- 
ture is obtained by coindexing the semantic part of that 
description with the appropriate argument slot. 
6In HPSG there is some ambiguity here, since the 
argument structure is embedded within the content part 
of the sign, where a lot of terminolgy is used that comes 
from situation semantics. 
7In HPSG for example the representation for the for- 
met is strictly based on the coindexation between syntac- 
tic and argument values, whereas the treatment for the 
latter assumes an event structure for predication and in- 
tersects it with the optional adjunts (see Badia & Colom- 
inas (1996)). 
there is no way to introduce arguments (or other 
kind of slots) that are not induced syntactically 
(or grammatically) in a restrictive way. A full 
theory of lexical semantics is needed to do so in 
a consistent way. For example, in Badia (1997) 
a particular case is considered and solved in an 
ad hoc way: the nominal translation (as in 9b 
above) is said to have an extra argument (to 
those inherited from the verb) so that its re- 
sult reading can have its denotation properly 
derived. This of course is only adequate for a 
small class of the complements we have here in 
mind (see examples in (8) and (9) above) and 
does not derive from a general theory of argu- 
ment structure. 
Such optional complements have only been 
treated in standard treatments by creating dis- 
tinct lexical entries, where the omitted com- 
plement is either just not present in the argu- 
ment structure representation or represented by 
an abstract existential quantification (without a 
correspondence to any syntactically realised el- 
ement). However this treatment neither reflects 
the relation between the different lexical entries 
that exist for a single word, nor is adequate in 
avoiding lexical redundancy. An alternative op- 
tion that could be considered is to treat those 
complements as adjuncts. However the treat- 
ment of adjuncts has not been fully developed 
yet, so that most of the proposals at hand are 
only adequate for intersective adjuncts. 8 
Furthermore, complements to nouns are par- 
ticularly difficult. From a semantic (and argu- 
ment structure) point of view many of them are 
like complements to verbs, in that they con- 
tribute entities to the relation denoted by the 
noun; thus an adequate treatment might intro- 
duce a particular number of arguments for ev- 
ery relation class denoted by nouns (just as it is 
done for verbs) (Badia & Colominas, 1996). But 
from the syntactic point of view the difficulty 
arises because they cannot be simply introduced 
in the valence lists (as verbal complements are). 
And a simple treatment as adjuncts is not very 
satisfactory either since it would increase the 
complexity of the semantic calculations. 
8Non intersective adjuncts cannot be easily treated, 
particularly those that contribute a new entity involved 
in the relation denoted by the predicate (for some 
of them quite complex proposals have been developed 
within HPSG: e.g., Kaspers, 1994). 
4 Proposed treatment 
In the following we attempt to account for some 
optional complements of predicative nouns in 
HPSG, by enlarging the semantic information 
and representing these complements in an in- 
formation level distinct from the valence lists. 
With regards to our semantic representation 
proposal, we maintain the referential informa- 
tion level in HPSG, expressed by means of the 
INDEX attribute. However we follow Puste- 
jovsky (1995) in enriching the argument struc- 
ture information level (expressed in HPSG as 
the RESTRICTION attribute and adapted here as 
ARGUMENT-STRUCTURE (ARGSTR)), SO that D- 
and S-Args can be represented (in addition to 
the strictly obligatory ones). Furthermore we 
also introduce the EVENT-STRUCTURE (EVSTR) 
and QUALIA levels of information estipulated in 
Pustejovsky (1995). The latter describes the 
lexical semantics of the entity and its related ar- 
guments, whereas the former displays the event 
properties of the expressed eventuality. As will 
be seen below, the argument structure level acts 
as interface between the mechanisms that allow 
for the syntactic realisation of the complements 
and the rich lexical semantics -constituted by 
the EVSTR and QUALIA levels. 9 
Let us now consider the syntactic treatment 
of D- and S-Args. Due to their optionality, the 
HPSG standard treatment of obligatory com- 
plements by means of valence lists is not ade- 
quate for them. Nonetheless a recent proposal 
by Sanfilippo (1998) gives a good insight into 
how they can be treated. For independent rea- 
sons he proposes that some complements are 
treated as real adjuncts from a syntactic point 
of view, even if they are thematically bound to 
the relation denoted by the head. We adopt 
this mechanism and represent D- and S-Args as 
thematically bound adjuncts, introduced as set 
members at the nonlocal (NONLOC) information 
level. Thus the basic structure of the sign is as 
follows: l° 
9In fact our implementation follows the lines devel- 
oped by Copestake (1992), Johnston (1996), among oth- 
ers. Particularly we follow the formalisation of Puste- 
jovsky's semantics in feature structures developed in 
Johnston (1996). In his treatment the selection of the ap- 
propriate reading in aspectually-based verbal polysemies 
is expressed by means of a boolean value in the head at- 
tribute of each of the relevant qualia (see figures below). 
l°For reasons of space we do not include the LOC at- 
LOC 
CAT 
SEM 
"HEAD head \] 
VALENCE lists-of-obl-complsJ 
"INDEX index 
EVSTR event-structure 
ARGSTR argument-structure 
FORMAL relation" 
QUALIA L AGENTIVE relation 
set-of-opt-compls NONLOC I INHER I SLASH 
Consider firstly an ordinary process-result 
nominal such as construccid ('building'). It is a 
nominalization from the verb construir ('build'), 
which subcategorises for two obligatory comple- 
ments: the agent and the result of the building 
process. 
As indicated in Pustejovsky (1995) this verb 
has also a third argument expressing the ma- 
terial, a D-Are, which is syntactically optional 
but participates in the logical expression of the 
event (see example (7a) above). In the rep- 
resentation for that verb (figure 1), the first 
two complements are expressed in the valence 
lists, whereas the latter is treated according to 
Sanfilippo's proposal for thematically bound ad- 
juncts; that is, as a member at the nonlocal in- 
formation level. In figure (2) we show the pro- 
cess nominalisation of construir where not only 
the D-Are but also ARG1 and ARG2 are not de- 
clared in the valence lists but included in the 
nonlocal set. Recall that we treat them like D- 
Ares. 
For our implementation we use LKB (Copes- 
take, 1993). We first derive the process reading 
of the nominal from the verbal lexical entry: the 
nominal representation for the process reading 
(in figure 2) is obtained by (non-locally) inher- 
iting from the already existing lexical entry of 
the verb (in figure 1). 11 And secondly the lex- 
ical entry for the result sense of construcci6 is 
created by means of a lexical rule that replaces 
the EVSTR head value and removes the ARG2 
attribute, the value of which is now declared 
as the index value. ARG2 has to be removed 
since, as it expresses the entity resulting from 
tribute in the rest of the structures in the paper. 
nThis is expressable in LKB by using the verbal 
lexical entry as a pseudo-sort (psort) (see Copestake 
(1993:23)). 
?,AT 
SEM 
"HEAD verb 
VALENCE \[SUB,\] 
COMPS 
"INDEX \[~\] 
I 'EVENT 1 EVENT2 EVSTR RESTR \[HEAD 
LD-ARG3 
QUALIA 
FORM 
AGEN 
\[~\] process" 
\[~\] state 
E\] 
\[\] anirnate-ind 
\[\] artifact 
\[\] material 
ARGI 
EVENT 
HEAD -- 
RELN constr-act" 
ARGI \[\] 
ARG2 \[\] 
EVENT ED 
HEAD + 
NONLOC \[ INHER \[ SLASH {PP\[\]} 
Figure 1: construir 
the process of building, it cannot appear as an 
argument of the result noun itself, n 
Thus the treatment of optional complements 
of nominalizations is identical to the treatment 
assigned to D-Ares of verbs. Both kinds of ar- 
gument are logically implied by the semantics 
of the predicate (verb or noun) but are syntac- 
tically optional. They are distinguished from 
Pustejovsky's True-Arguments (T-Arg) -that 
is, the syntactically realized parameters of the 
lexical items (Pustejovsky, 1995:63)- by means 
of the syntactic level at which they are declared: 
in the valence list or in the set of nonlocal ele- 
ments. 
In order to show that other types of nominals 
can also be treated in this way, consider for in- 
nRecall that creation nominals do not allow the result 
nominalisation with the thematic argument: *la con- 
strucci6 de la casa ds molt sblida ('the building of the 
house is very robust'). 
2AT 
SEM 
"HEAD 
VALENCE 
"INDEX 
EVSTR 
ARGSTR 
QUALIA 
noun 
SUBJ 
COMPS 
\[\] 
"EVENT1 
EVENT2 
RESTR 
HEAD 
"D-ARG1 
D-ARG2 
D-ARG3 
FORM 
AGEN 
NONLOC \[ INHER \[ SLASH 
< 
o \[~\] process 
\[\] state 
\[~ animate-ind\] 
\[\] artifact | 
\[~ material J 
RELN ~s 1 ARG1 
EVENT 
HEAD -- 
{PP\[\], NP\[de\]\[\], PP\[\]} 
Figure 2: construccid (process reading) 
stance redescription nominalizations. They dif- 
fer from standard creation predicates in that 
their process reading cannot express syntacti- 
cally the argument denoted by the result nomi- 
nalization. 
Consider the noun decoracig ('decoration'), 
derived from the verb decorar ('decorate'). We 
start by assuming three arguments to that pred- 
icate: a first one corresponding to the agent, a 
second one that expresses the theme (i.e., the 
object being decorated), and finally a D-Arg 
referring to the material. The agent and the 
theme arguments are subcategorized as T-Args 
by the verb, but realized as thematically bound 
adjuncts when appearing in a nominalization. 
As opposed to construir, the direct object of 
decorar does not denote the object resulting 
from the decoration act, but the object being 
decorated. As a consequence both nominalisa- 
tion of decorar (process and result) allow the 
CAT 
SEM 
"HEAD 
VALENCE 
"INDEX 
EVSTR 
ARGSTR 
QUALIA 
NONLOC I INHER l SLASH 
noun 
I SUBJ COMPS 
55\] 
'EVENT1 
EVENT2 
RESTR 
HEAD 
"D-ARGI 
D-ARG2 
D-ARG3 
( 
\[~\] process' 
\[\] state 
\[K\] 
\[~ animate-ind 
\[\] phys-obj 
\[\] material 
ARG I 
FORM EVENT 
\[HEAD -- 
"RELN decor-act 
ARGI \[~\] 
ARG2 \[\] AGEN 
ARG3 \[\] artif 
EVENT 
HEAD + 
RELN part-O i 
CONST ARGi \[~\] 
\[ARG2 \[\] 
PPm, NP\[de\]\[\], PPE 
Figure 3: decoracid (process reading) 
complement that corresponds to the direct ob- 
ject of the verb (recall that this is not possible 
in the case of result readings of creation predi- 
cates). Therefore the presence of the objective 
complement of decoraci6 is not useful for dis- 
ambiguating between the two readings. Thus 
nominalizations like construcci6 and decoraci6 
with the objective complement differ in that the 
first one does not accept the two possible senses 
appearing together, while the second one does. 
Following the treatment of Pustejovsky's dot- 
ted types in Buitelaar (1997), this fact leads us 
to conceive construccid as an open dotted type, 
and decoracid as a closed dotted type. 
The treatment we propose for decoracid is 
6 
CAT 
;EM 
HEAD noun 
SUBJ 
VALENCE 
COMPS 
INDEX \[~3 
EVENT1 
EVENT2 
EVSTR ! RESTR 
HEAD 
D-ARG1 
D-ARG2 ARGSTR 
D-ARG4 
D-ARG5 
FORMAL 
QUALIA 
kGENTIVE 
NONLOC \[ INHER 1 SLASH 
< 
\[~\] process" 
\[~\] state 
\[\] animate-ind 
\[\] 
\[\] language 
language 
ARGi 
EVENT 
HEAD -- 
RELN trans-act 
~RGi \[\] 
ARG2 ~UALIA 
"info.ph-obj 
CONSTITUTIVE 
"info.ph-obj 
CONSTITUTIVE ARG3 QUALIA 
EVENT \[~\] 
HEAD 
{Pp\[\], NPt e\] , PP , PP } 
I RELN ARGi ARG2 
RELN 
ARGi 
ARG2 
~-langl \] 
Figure 4: traduccid 
shown in figure (3). Note that it is adequate 
to deal with facts such as the ones exemplified 
in (9a). In this example the redescription nomi- 
nal (decoraci6) expresses the process reading in 
the first clause, and in the second one it is re- 
ferred to as denoting the object resulting from 
the process. In the representation of the nomi- 
nal decoraci6 in figure (3) this is allowed by the 
third argument of the AGENTIVE qualia, which 
is not bound by any argument in the ARGSTR, 
just because it cannot be syntactically realised 
in any case as an argument. Note that argument 
structure acts as an interface between the rich 
semantic representation (i.e., the QUALIA and 
EVSTR structures) and the surface mechanisms 
that allow for complements of the predicate (va- 
lence lists and non-local set). Only semantic ar- 
guments that may be syntactically realised are 
present in the argument structure (either as T- 
Args or as D- or S-Args). On the other hand the 
7 
rich semantic structure allows to express seman- 
tically implied arguments, and thus provides a 
treatment for semantically motivated discourse 
factors (such as the ones shown in (9) above). 
The proposal outlined here provides an ap- 
propriate treatment for the traducci6 example 
(9b). Instead of the ad hoc solution adopted in 
Badia (1997) the treatment proposed here de- 
rives from a general and systematic approach to 
the semantic structure of predicates and their 
nominalisations. To make things concrete we 
present in figure (4) the full lexical entry of the 
process reading of the nominal traduccid. This 
representation is able to deal with anaphoric ref- 
erence to the unexpressed result semantic argu- 
ment of the process reading (9b). Note that 
a parallel verbal lexical entry also gives an ac- 
count of the anaphoric relation allowed by the 
verb traduir in (9c), and that a similar treat- 
ment is applicable to the verbs menjar and 
amanir of examples (5) and (9e). 
Of course this treatment applies in a general 
way to other redescription predicates, like es- 
tudi ('study'), soluci6 ('solution') and avaluaci6 
('evaluation') in (8). Notice as well that the rich 
semantic information we use allows us to deal 
in a natural way with the distinction between 
subjective and objective complements in these 
data. In (Sa), for instance, estudi does not allow 
plantes ('plants') as the agent of the event since 
tha agent has to be an animate individual: 13 
Other types of nouns with semantically im- 
plied (optional) complements can be similarly 
treated; for example non deverbal nouns that 
express a relation with another entity, like 
nouns denoting partition or sets (full 'sheet' in 
(2) and grup 'group' in (3)) or relational nouns 
(mare 'mother' in (9d)). The arguments of full 
or grup can be treated as D-Args because they 
are not strictly obligatory but their semantics 
is implied. However the semantics of the syn- 
tactically realised complement need not corre- 
spond exactly to the semantic implication. For 
example, the implication of .full realised with- 
out a complement is that it is a piece of paper; 
however the complement may denote any sim- 
*aNote that in the figure for estudi the HEAD values of 
EVSTR and QUALIA levels are underspecified, similarly to 
the value of the INDEX attribute, thus letting the entry 
underspecified between the process and the result inter- 
pretation 
~AT 
SEM 
HEAD noun 
VALENCE \[SUBJ 
L COMPS 
"INDEX index 
'EVENTi 
EVENT2 
EVSTR RESTR 
\[HEAD 
'D-ARGI 
D-ARG2 ARGSTR 
QUALIA 
FORMAL 
AGENTIVE 
< 
\[\] process 
\[~\] state 
el)ent 
\[\] animate-ind\] 
\[\] top j 
RELN 
ARG i 
EVENT 
HEAD 
"RELN 
ARG I 
ARG2 
ARG3 
EVENT 
HEAD 
study-act" 
\[\] 
\[\] 
\[\] 
\[\] 
:h 
NONLOClXNHERISLASH {PPm' NP\[de\]\[\]) 
Figure 5: estudi 
ilar class of material (card, cardboard, tissue 
paper...) (Sauri, 1998). We account for this 
difference by means of a lexical rule that over- 
writes the implied semantics of the complement 
when necessary. 
This treatment however is not adequate for 
relational nouns, like mare ('mother') or pare 
('father'). These nouns do not allow the surface 
realisation of the semantic implication unless it 
is more specific: 
(10) a. *Ha vingut el pare d' un fill 
*Has come the father of a son 
b. Ha vingut el pare d' un nen 
Has come the father of a boy 
canadenc 
Canadian 
c. Ha vingut el pare de la Joana 
Has come the father of the Joana 
This kind of complement can be seen as 
shadow arguments (S-Arg) of Pustejovsky's. In 
our proposal they are treated as D-Args because 
they are optional: their syntactic realisation is 
controlled via the non-local set and the semantic 
implication is expressed in the FORMAL qualia. 
However a restriction has to be formulated upon 
the semantics of the syntactic element within 
the non-local set: namely that it be more spe- 
cific than the semantic implication (see figure 
(6)). 14 
CAT 
SEM 
"HEAD noun 
VALENCE \[SVSJ 
\[COMPS 
"INDEX \[\] 
ARGSTR :D-ARGI 
QUALIA FORMAL 
( 
\[\] son l daughter\] 
ARGI 
ARG2 
NONLOC I INHER I SLASH {NPs\[\]} 
Figure 6: pare 
5 Conclusion 
In this paper we have discussed the represen- 
tation of the semantic structure of nominals 
having complex argument structures. In par- 
ticular we have proposed a treatment for syn- 
tactically unexpressed optional complements of 
nouns. We have shown the relevance of the 
representation of such relations (based on dis- 
course factors) and have developed a treatment 
that takes into account both the syntactic and 
the semantic aspects of the phenomenon. In 
particular we have shown that optional com- 
plements of deverbal nominalizations and other 
classes of nouns (in addition of some optional 
complements to verbs) can be nicely dealt with 
by using Pustejovsky's D- and S-Args being im- 
plemented as thematically bound adjuncts in 
an HPSG framework semantically enriched fop 
lowing the lines in GL. This is possible mainly 
14This restriction is indicated by the function "S" (i.e., 
specific) in the feature structure below. We are currently 
working on an implementation of this restriction. 
thanks to the interface role between syntax and 
lexical semantic information that is performed 
by the argument structure level. Our proposal 
has been implemented by using LKB devices 
which allow a general and systematic treatment. 

References 
Badia, T. (1997) 'Predicative structure of nominals 
in HPSG'. In Syntax and Semantics of Predicates 
(1997 Conference of the TLS). Austin. University 
of Texas. 
Badia, T. & C. Colominas (1996) Predicate- 
Argument Structure. In Final Report of the 
Project MLAP-93-15. CUE. Luxemburg. 
Buitelaar, P. (1987) 'A lexicon for underspecified se- 
mantic tagging'. In Tagging Text with Lezical Se- 
mantics: Why, What and How?. ACL SIG on the 
Lexicon. Washington DC. 
Copestake, A. (1992) The Representation of Lezical 
Semantic Information, PhD thesis, Sussex Uni- 
versity, Cognitive research paper CSRP 280. 
Copestake, A. (1993) The Compleat LKB. Acquilex 
H Deliverable ms. CCL. 
Johnston, M. (1996) 'Semantic underspecification in 
lexical types: capturing polysemy without lexical 
rules'. Acquilez Workshop on Lexical Rules, 1995. 
Cambridge. 
Kaspers, R. (1994) 'Adjuncts in Mittelfeld'. In Ner- 
bonne et al. (eds.) German in HPSG. CSLI. Stan- 
ford. 
Pollard, C. & I. Sag (1994) Head-driven Phrase 
Structure Grammar. CSLI, Standford CA. 
Pustejovsky, J. (1995) The Generative Lexicon. The 
MIT Press. Cambridge MA. 
Sanfilippo, A. (1998) 'Thematically bound ad- 
juncts'. In Balari, S. & L. Dini (eds.) Romance 
in HPSG. CSLI. Stanford CA. 
Sanrf, R. (1998) (ms.) Alternances de sentit con- 
trolades per la comptabiIitat en un ldxic com- 
putacional. IULA, Universitat Pompeu Fabra. 
Barcelona. 
Somers, H.L. (1987) Valence and Case in Compu- 
tational Linguistics Edinburgh Univ. Press. Edin- 
burgh. 
