An Underspecified HPSG Representation 
for Information Structure 
Jonas Kuhn 
Institut fiir maschinelle Sprachverarbeitung 
Azenbergstratte 12 
D-70174 Stuttgart, Germany 
Emaih j onas~ims, uni-stuttgart, de 
Abstract 
hfformation structure can be of great use 
in linguistic applications, especially in those 
involving a speech component. However, fo- 
cus marking by prosody is often ambiguous. 
Existing theories capture this by rules that 
produce alternative focus structures. This 
disjunction is hard to handle computation- 
ally. In this paper, a compact, graphically 
underspecified representation is proposed, 
along with composition principles and a res- 
olution routine based on context informa- 
tion. 
1 Introduction 
This paper proposes an underspecified represen- 
tation for information structure (IS) and HPSG 1 
principles that build up this representation for 
German. The representation is designed as a ba- 
sis for use in applications like text-to-speech and 
speech translation systems. It is obvious that 
for a non-tiring, natural output, especially tile 
prosodic side of IS has to be taken into account. In 
this section, examples from sentence-based trans- 
lation with semantic transfer visualize the role of 
IS in several empirical domains. For one thing, 
ignoring IS may result in confusing translations: 
(2) with default accenting, 2 despite being truth- 
conditionally equivalent, is not a suitable transla- 
tion for (1B) in tile given context. System users 
will probably presume that a serious mistransla- 
tion occured. 
(1) A: Zu unsercm Treffen wird also Frau Otto 
yon der Firma Miiller kommen? 
('So, l~-au Otto of Miiller's will participate 
in our meeting?') 
B: Nein, Frau SCHNEIDER wird Firma 
No, Frau Schneider will company 
Miiller vertreten. 
Miiller represent 
(2) No, Frau Schneider will represent MOLLER'S. 
1See (Pollard/Sag, 1994). 
2SMALL CAPS are used to highlight pitch accents. 
Moreover, IS can be exploited to choose be- 
tween certain translation alternatives on sentence 
level (of. (Eberle, 1996)). The particle noch has 
different translations depending on accentuation: 
(3) G. Maria sucht noch einen BRIEFKASTEN. 
Maria looks@~r still a post-bo~ 
E. Maria is still looking for a post box. 
(4) G. Maria sucht NOCll einen Briefkasten. 
E. Maria is looking for another post box. 
Concluding the set of initial examples, (5) shows 
that focus call play a role in simple ilfferenees 
drawn from context to resolve lexical ambiguities 
(or similarly for ellipsis and anaphora resolution). 
(5) ~%au S<:hulze kiindigte an, die Firma im 
Herbst zu besuchen. 
('lS"au Schulze announced to visit tile com- 
pany in autumn.') 
G. Ich nehme an, dab die Managerin ihren 
I assume that the manager her 
Besuch ~EGldiSST. 
visit/visitors welcomes 
E. I assume that the manager will APPI{ECI- 
ATE her visit. 
(6) Ich kann lh'au Miiller nirgends im Geb/iude 
finden. 
('I can't find Frau Miiller anywhere in the 
building.') 
G. Ich nehme an, (lag die Managerin ihren BE- 
SUCH begriiBt. 
E. I assume that tile manager is giving a wel- 
come to her VISITORS. 
Since tile ambiguous Besuch in (5) is not focus- 
marked, it has to be contextually given. Even 
a simple context model will resolve the reference 
back to tile besuchen event mentioned in tile pre- 
ceding sentence, so the noun visit is chosen as a 
translation. 
The last example showed that certain ambigu- 
ities require inferences from context. However, 
the other examples make clear that focus can be 
effectively applied in sentence-based translation 
exploiting the existing linguistic focus marking 
within the sentence. Costly reasoning can 'thus 
be restricted to doubtful cases. 
670 
A sentence-based account of IS faces one prob- 
lem: in the general case, a certain prosodic mark- 
ing of a sentence does not correspond to a unique 
IS. These cases of focus ambiguity have been dis- 
cussed extensiw~ly in the literature and will be 
briefly reviewed in see. 2. The existing HPSG 
account of IS by Engdahl/Vallduvi (1994) gelmr- 
ales a disjunction of alternative sentence analyses 
for ambiguous sentences, a Since focus ambiguity 
mull;Jiffies with all other kimts of ambiguity, a dis- 
junctive treatment makes the use of IS in compu- 
tational applications unattractiw;. Therefore, an 
underspecified representation of IS is proposed in 
this paper (sec. 3) along with a resolution routine 
on the basis of context information (see. 4). Fi- 
nally, the linguistic principles that compose the 
representation are worked out in HPSG (see. 5). 
The data discussed are Gerlnan, but English is 
very similar. 
2 Focus marking 
The basic data are well-known. 4 A pitch accent 
on a direct object like, Buch in (7), can serve to 
mark a number of constituents as focused: (8). 5 
The focus Dature is usually assumed to proje.ct. 
(7) Karl hat ein BucII gelesen. 
Karl has a book read 
(8) a. Karl hat \[(;in BUCII\]F gelesen. 
b. Karl hat \[ein Buclt gelesen.\]F 
c. \[Karl hat ein Bucx~ gelesen.\]F 
A pitch accent on the subject Karl however can- 
not t)roject focus (9), neither do adjuncts t)roject 
focus (10). 
(9) \[KanL\]~.~ hat ein Buch gelesen 
(1{}) Maria hat \[DRAUSSEN\]F geniest. 
Maria has outside sneezed 
These data can be explained by the h)llowing 
F-Assignment Rules fi-om (Selkirk, 1995):a 
(11) Basic F Rule 
An accented word is F-marked. 
aThe same is true fl)r practically all accounts of 
focus projection that I am aware of- of., e.g., (Selkirk, 
:1995; Schwarzschild, 1995). 
4For German, cf., e.g., (tIShle, 1982). 
5The standard means to determine the focus of an 
utterance is the question test, where the tbcus in the 
answer corresponds to the interrogative constituent in 
the question. However, this test can be misleading 
in cases where the question comes in a wider context 
(ef. (16) below and the discussion in (Knhn, 1996)). 
~Rule (12b) may need some refinement. It is not 
clear whether it is a syntactic condition that con- 
strains indirect F-marking of a head. Jacobs (1994) 
argues that the notion of integration or ir~formational 
autonomy is resimnsible instead, which is baaed on the 
complexity in semantic processing. Unibrtunatcly, the 
formal nature of integration is still ill-understood. 
(12) F Projection 
a. F-marking of the head of a phrase licenses 
the F-marking of tile phrase. 
b. F-marking of an internal argument of a 
head licenses the F-marking of the head. 
According to tiles(; rules, the head itself can 
project focus, which appears to be refuted by (b~ta 
like the following. 
(13) Karl hat ein Buch \[GEIA!:SI'\]N.\]F 
However, (13) can have a wider focus if books 
are contextually given (this effect has been called 
dcaccenting). Note the diiference between (14) 
and (15), both marked according to (11) and (12): 
although their maximal focus domain ix identical, 
tin Buch is F-marked only in (15). 
(14) Karl hat \[ein Buch \[(;ELESEN.\]F\]F 
(15) Karl hat \[\[ein Buc:n\]l ,, \[gelesen.\]FJF 
This is reflected in Schwarzsehild's (1995) in- 
terl)retation of the F-feature. All constituents 
that are not F-marked need to be given, where 
givenness is defined an entaihnent by prior dis- 
course. Entaihnent is extended to expressions of 
other type than propositions by existentially bind- 
ing unfilled arguments. 7 Even non-F-marked con- 
stituents embedded in an F-marked constituent, 
like em Buch in (14) have to pass this givenness 
filter. So, deaccenting is no longer a special case 
for the theory. 
To restrict the optional focus projection from 
(12) fllrther, Schwarzschild assumes an additional 
pragmatic filter Avoid F that selects the tree with 
the least F-marking. s 
Non-F-marked constituents that contain F- 
marked subconstituents need to be giwm as well. 
The context has to entail their F-skeleton the 
7The exact definition is as tbllows ((Schwarzschild, 
1995), pp. 5-6.): 
An expression T (for target) is GIVEN iff it ha.s an 
antecedent, A, such that the existential closure of A 
entails the result of existentially binding F-variables in 
the existentially closed F-skeleton of T \[where the ex- 
istential quantifier binding F-w~riablcs quantifies over 
contextually salient values\]. 
The F-skeleton of an expression is the result of rc- 
placing F-marked elements witl, variables (working 
top to bottom). 
The existential closure of an expression is the result 
of existentially binding all unfilled arguments of the 
expression. 
if c~ is type t, ExClo((~) = c,. 
if c~ is of conjoinable type (a, b), 
l~xCl,,(~) = 3x~,\[ExOo(,~(x°))\]. 
if c~ is type c, ExClo((t) = 3P(~,t)\[P((t)\]. 
8I (:onsider the status of this filter somewhat prob- 
lematic. Why shmdd a tree with less F-marking be 
pragmatically preferred? One could as well argue that 
since a sentence should be ~s informative as possible, 
given constituents should be avoided. The underspec- 
ified account dewfloped here recasts Schwarzschild's 
ideas in a way that makes Avoid F redundant. 
671 
result of replacing embedded F-marked elements 
with variables. 9 This condition allows to explain 
data like (16) - a puzzle for theories based on the 
question test for focus (cf. fn. 5 above). 
(16) A: Paula hat eine rote Rose fotografiert. 
Paula has a red rose photographed 
Was hat sie davor getan? 
What has she before done 
B: Sie hat eine \[WEISSE\]F Rose fotografiert. 
She has a white rose photographed 
A's question leads to expect focus marking of 
the complete VP, but intonational marking plus 
projection rules produce a narrow focus on weifle. 
Schwarzschild (1995) predicts the indicated F- 
marking, since the F-skeleton of the NP eine 
weifle Rose, for instance, (a X rose) is actually 
entailed by the context. 
While non-F-marked constituents have to be 
given, F-marked constituents need not necessar- 
ily be new. This is to account for data like the 
following, where ihn in (17B) is given: 
(17) A: Wen hat Peters Mutter gelobt? 
Whom has Peter's mother praised? 
B: Sic hat \[IHN\]F gelobt. 
She has him praised 
The indicated F-marking follows from the the- 
ory: there has to be some F-marking, since the 
meaning of the complete sentence (Peter's mother 
praised Peter) is not entailed by the context. The 
F-skeleton variant Peter's mother praised X is ac- 
tually entailed by the question (17A), thus the F- 
marking of ihn. 
3 An underspecified representation 
In computational applications, a compact repre- 
sentation is a prerequisite for any successful treat- 
ment of IS. Take the following short sentence with 
two pitch accents. 
(18) Die Direktorin der Firma MULLER BEGRUSST 
The director of company Miiller welcomes 
ihren Besuch. 
her visit/visitors 
Even if functional elements are ignored, the 
rules in (12) produce nine alternative F-markings 
that have to be checked against the context for 
givenness. In order to resolve the lexical ambi- 
guity of Besueh however (cf. the discussion of (5) 
above), some partial information about IS suffices. 
The representation developed here is relatively 
independent of the underlying semantic theory of 
focus. Two semantical partitions for focus (FOC) 
and background (BG) are assumed, each of them 
a set of semantic conditions. Underspecification is 
expressed in a graphical way. The interpretation 1° 
9There is some affinity to the f-semantic value in 
Rooth's (1985) alternative semantics, although the 
specific details are different. 
rathe interpretation is given informally in the fol- 
lowing examples. 
of each syntactic constituent forms a node in tile 
graph, while the directed edges express accessibil- 
ity relations. 
(19) MARIA begrfiBt ihren BESUCH. 
Maria welcomes her visit~visitors 
Maria begriiflt 
ihren Besuch 
v 
,, begriiflt 
', ihren Besuch 
(20) ", begr/iflt 
', /ihren Maria 
', ,'" Bes~// 
.~, 
BG FOC 
The solid line arrows signify obligatory inclu- 
sion in the respective IS partition, the dashed line 
arrows defeasible inclusion. The VP can, e.g., ei- 
ther belong to FOC (via a chain of arrows) or 
to BG. The graph in (20) represents the amount 
of information that is encoded on sentence level 
without reference to context. Additional knowl- 
edge may introduce further solid arrows. If, e.g., 
a begriiflen event is contextually given (like in 
a question Who greets whom?), the arrow from 
begriiflt to BG will become an obligatory arrow. 
This arrow again will overrule the dashed arrow 
from begriiflt to ihren Besuch. Since the latter 
was the only path to access the FOC partition, 
the complete graph will collapse into a fully spec- 
ified representation of IS. Resolution will be dealt 
with in more detail in sec. 4. 
The principles composing the representation are 
worked out formally in sec. 5. Here is an intuitive 
overview: the arrows pointing directly to the FOC 
and BG partition originate from accenting or non- 
accenting of the single words, respectively. The 
additional arrows between the constituent nodes 
are introduced by the grammatical principles of 
F-projection irrespective of the actual prosodic 
marking. This becomes clear when we regard the 
following alternative prosodic marking of (19): 
(21) MARIA BEGRUSST ihren Besuch. 
Maria begr/iBt 
ihren Besuch 
' begrii\]\]t 
'qhren Besuch 
(22) ', , ', , begrfii~t 
\ ', ,, lhren Maria 
",', Besuch \ / 
BG FOG 
Note that here, the dashed arrow from begriiflt 
to ihren Besuch is overruled right away, since 
the accented begriiflt is strictly tied to FOC. The 
672 
phrase ihren Besuch is forced into tile BG parti- 
tion, thus the utterance is correctly predicted to 
be restricted to contexts where Besueh is given. 
As anticipated above in tile discussion of the lex- 
ical ambiguity involved, this conelusioil can be 
drawn even if other points remain underspecified. 
4 Resolution 
An underspecified IS arising from tile prosodic 
marking of a sentence can be resolved by informa- 
tion from the context. As noted above, entaihnent 
by context introduces additional solid line arrows. 
rlb covet" more than just trivial cases, the (:ounter- 
part of Schwarzschild's (1995) F-skeleton has to bc 
kept track of in the underst)ecified representation. 
At first glance, this seems to be incompatible with 
the idea of underspeeification, since the F-skeleton 
that is checked against the context for entailment, 
requires settlement on what the actual F-marking 
is. For instance, to check the givenncss of the VP 
in (23), reads a book about X has to be entailed, 
whereas on the basis of the marking in (24), reads 
Y has to be entailed. 
(23) I%anz liest ein Buch filler \[S('ml"FE.\]F 
l~'anz reads a book about ships 
(24) bYanz liest \[ein Buch filler \[ScHnq"m\]F\]F 
The solution lies in the observation that the lat- 
ter F-skeleton entails the former. So, when a cer- 
tain constituent (e.g., again the VP in the abow~ 
examples) is checked for givenness, it suffices to 
assume F-marking of the maximal potentially F- 
marked subconstituents (I call this the maximality 
assumption). If it turns out that the actual F- 
marking is more restricted, this will be detected 
at a lower level. For example, if (23) is the right 
solution, this will be discovered even if reads a 
book about X is not checked, since in this case, a 
book about X will be contextually given as well. 
To illustrate resolution in tile graph representa- 
tion, take the following example in context: 
(25) a. Amta hat Otto fotografiert. 
Anna has Otto photographed 
l). HANS hat Otto ein Bv(,n gegeben. 
Hans has to-Otto a book given 
Sentence (25t)) produces the following graph: 
(Hans hat Otto ein Buch gegeben, 
(3P, ue C)\[P(u)\]} 
(Ot\[o ein Buch gegeben, 
(3R~ C)(~:~)\[n(x,Otto)\]) 
/ / (ein Buch gegeben, 
(3Q, z e C)(3x, y)\[q(x, y, z)\]} (26) 1 / ,~-. 
I \] .."" (gegeben, 
I i (~zeO)(Sx,y)\[geben(x,y,z)\]) 
(Otto,Otto) / / "'" .-J"" ~' (Ilans,u) 
""-. '/ .'"'-'--'"'- (ein B~ 
BG FOC 
The nodes are now labelled by pairs. The first 
element of a pair is tile semantic wflue of the con- 
stituent corresponding to the node (here again ex- 
pressed inforlnally in natural language); the sec- 
ond eleinent is the F-skeleton based on the max- 
imality assumption - the F-skeleton is obtained 
by (i) replacing the maximal F-marked subcon- 
stituents (or sister constituents, in the case of in- 
direct F-marking of the head, following (12b)) by 
a variable, (it) existentially binding unfilled argu- 
ments, and (iii) existentially quantifying over the 
variables from the F-skeleton (cf. also fit. 7). The 
latter quantification is restricted by the set of con- 
textually salient values C. 
In resolving underspecifieation from context, 
it is checked for each node with access to the 
BG t)artition ~t whether its F-skeleton is en- 
tailed by an antecedent in the context. For 
gegeben and the lowest verb projection ein 
Buch gcgeben, there is no such antecedent in 
the context. However, for the VP Otto cin 
Bueh gegeben the VP from (25a) Otto fo- 
tografiert is a suitable antecedent: after exis- 
tential binding, (~x) fotografieren(x,Otto) entails 
(~R C C)(3x)\[R(x,Otto)\]. Similarly for the con> 
plete sentence: fotogrMieren(Anna,Otto) entails 
(3e, u C C)\[P(u)\]. So, solid line arrows replace 
the two dashed arrows pointing to BG (suggested 
in (26) by tile longer dashes). 
The status of tile lower verb projections in (26) 
is still underspecified. If a wider context was con- 
sidered, antecedents could possibly be found, so it 
makes sense to end 1, t) with such a representation 
after processing the discourse (25). Once we know 
that all salient context has been considered, a rule 
of I'bcus Closure is at)plied. This strengthens the 
arrows pointing towards FO(\] for all unresolved 
constituents, t)redicting eiu Buch .qcgcben as the 
focus for the example. In an at)plication, heuris- 
tics may trigger Focus Closure earlier, to aw)id 
unneecessary inferences. 
The representation proposed here comes close 
to Sehwarzschild's (1995) theory of F-marking in 
coverage, .2 avoiding the computationally ext)en- 
sive disjunction of alternative analyses in favour 
of a single graph reI)resentation that is under- 
specified when based on sentence internal infer- 
n Constraining the givemmss check to nodes with 
access to the BG partition makes sure that narrow 
(eontrastive) focus on given entities (like in (17)) is 
treated correctly. Since all acce.nt on a word intro- 
duces just an arrow towards FOC, narrow focus on a 
word survives the check even in cases where the word 
is given. 
12For examples with several ambiguous accents, the 
modified account collapses some F-markings with min- 
imal diffcrm,ces in inte.rpretation into one. For in- 
stance, two of the twelve F-markings for (i) are in- 
distinguishable: (it) and (iii). A context that wouhl 
enforce the latter in Schwarzschild's theory tlrodu(:es 
(it) in the underspecitied account. 
673 
mation only. l~lrthermore, Schwarzschild's prag- 
matic condition Avoid F that selects the analysis 
with the least F-marking (cf. fn. 8) is no longer 
needed. 
5 HPSG principles 
The representation proposed in sec. 3 and 4 can 
be implemented directly in a sign-based formalism 
like HPSG. In this section, tile central composition 
principles for German are worked out. A binary 
branching structure is assumed. Again, the repre- 
sentation will be fairly open as to which particular 
semantic formalism is chosen. This is reflected in 
tile Semantics Principle I assume, which specifies 
the semantic value of a phrase as the application of 
a two-place function compose to the semantic val- 
ues of both daughters. The function can be spelt 
out in different ways depending on the choice of a 
semantic theory, la Furthermore, a function vari- 
able is assumed that maps a semantic object to a 
new variable of the same type. 
The HPSG type cont (the value of the 
CONT(ENT) feature) has the following four new 
features: O-SEM ('ordinary semantics') and 
I,'-SKEL (F-skeleton) of the type of a semantic ob- 
ject, tile set-valued IS-CSTR (IS constraints) and 
the binary MAX-F (for potential maximal focus). 
The phonological information is enriched by a 
feature PROM (prominence) with values accented 
and unaccented. 
The following principles specify the lS-CSTR set 
for a sign, introducing arrows, or links, between 
(i) Der Direktor der SPARKASSE begr/it3te Frau 
The director of-the savings-bank greeted Frau 
MAIEIt. 
Maier 
(ii) Der Direktor der \[SPARKASSE\]F begrfiftte \[Frau 
MA,E~.\] F 
(iii) \[Der Direktor der \[SPARKASSE\]F\]F begrfiflte 
\[Frau MAIER.\]F 
However, such a context is quite intricate to construct. 
Here is an example: 
(iv) Der Direktor der Firma ABC sat~ im C~f~. 
The director of company ABC sat in-the care 
Der Kellner begriiflte eine Bekannte. 
The waiter greeted a friend 
Context (iv) does not entail der Direktor der X 
begriiflte Y, so (ii) is ruled out according to 
Schwarzschild's system. The resolution process pro- 
posed in this paper is based on the maximality as- 
sumption, and thus checks givenness for the com- 
plete sentence only once, with the complete subject 
F-marked. To rescue the difference between (ii) and 
(iii), it would have to be enforced that resolution of 
the subject NP takes place before the resolution of the 
focus projected from the object. This is a conceivable 
solution - however, declarative perspiguity would be 
sacrificed for a very moderate benefit, considering the 
main point of this paper. 
lain the example below, simple predicate-argument 
semantics is used for illustration. 
semantic objects. (Tim IS partitions/oc and bg 
are here also treated as semantic objects.) The 
h_link relation corresponds to the obligatory (solid 
line) arrows in the graphs, s_link reflects defeasi- 
ble (dashed line) arrows. In the following, the link 
relations are expressed verbally. (27) is the coun- 
terpart to (11); (28) and (29) reflect clauses (12a) 
and (12b). 
(27) Focus Linking Principle 
The O-SEM value of a pitch accented word is 
h_linked to foe. 
(28) Head F-Projection Principle 
The O-SEM value of a phrase is s_linked to the 
O-SEM value of its head daughter. 
(29) Indirect Head F-Marking Principle 
In a head-complement-structure, where none of 
the head daughter's arguments have yet been 
saturated, 14 the O-SEM of the head daughter is 
s_linked to the O-SEM value of the complement 
daughter. 
These three principles establish direct or indi- 
rect links towards foc. For the constituents that 
are not obligatorily focus marked, the underspec- 
ified representation requires additional defeasible 
links to bg: 
(30) Background Linking Principle 
The O-SEM value of every sign that is not ac- 
cented is s_linked to bg. 
The principles just presented compose the rep- 
resentation introduced in sec. 3, with the nodes 
in the graph corresponding to the O-SEM values. 
To provide the input for the resolution routine, 
the representation was enriched in sec. 4 by the 
F-skeleton. This is kept track of in the F-SKEL 
feature (assuming independent existential binding 
of unfilled argmnents and free variables). 
(31) F-Skeleton Instantiation 
The F-SKEL value of a word with \[O-SEM (~\] is 
(i) the function variable(a), if ~ is h_linked to 
foc; 
(ii) a, if the word is marked \[MAX-F __\];15 
(iii) eomposition(a,fl) (where fl is the F-SKEL 
value of the word's complement), 
if the word is marked \[MAX-F J-J, but a is 
not h_linked to foe. 
(32) F-Skeleton Principle 
The F-SKIn, value of a phrase is the function 
compose applied to 
(i) the F-SKEL vahm of a daughter with 
\[MAX-F --\]; or 
(ii) the function variable(a) of a daughter with \[MAX-F J-\] and \[O-SEM O~\]; 
for both daughters. 
14This formulation will be subject to changes once 
there is a clear concept of integration (cf. fn. 6) - (29) 
applies to integrated constitnents. 
lSThe feature MAX-F is actually redundant. It has 
been introduced for convenience, signifying that a 
chain of links to foe exists - a condition that could 
be checked directly in the graph. But it does not hurt 
much to carry the feature along in the principles (27) 
(29). 
674 
"IS-CSTR \[\] U \[~\] U { s3ink(m, b q), sdink(N, \[\])\]\] 
O-S~{M m(= gab'(Bucl{)(Otto')(Hans')) \] 
MAX-F -t- 
• I1 
U{ s_link(\[\] , b9), sdink(\[\] , IN)} 
/ O-SEM \[\](: gab'(Buch')(Otto')) / 
F-SKEL compi~ ,var(IN))(= R2(Otto')) 
MAX-F J- 
_tl 
/\[,s-os'rR(~ =)~ u \[m 
/ I U{s~ink(~, ~g), s~ink(®, \[\])} / Io-SnM IN(= gab'(Uueh')) 
• 4 3 ', / I F-s.~., ~o,,,I,(,,-r(\[\]). ~,a,-(~J))(: Q (z), 
-PROM unacc 
IS-CS'FRIN {h_link(N, foe)} Us-cs'r. \[\] {s_link(IN, bg)} : IS-CSTR\[~ {h_link(IN, foe)} 
O-SEM \[~Hans' /O-SEM •Otto' O-SFM INBuch' 
~-SK~,L var(~)(= ~) IF-SKI, n, \[\] F-sK~L van(m)(= ~) 
MAX-F J- LMAX-F -- MAX-F n t- 
I I 
HANS Otto 
Figure 1: Sainple. 
A sample analysis for (33), a slight simplifi('a- 
tion of (25b), is give.n in fig 1. The graph produced 
by the linking constrnints is the one in (26). 
(an) (Ich weifi, daft) HANS Otto ein BUCH gab. 
I know that Hans to-Otto a book gave 
Let us briefly see how the principles interact 
to produce the phrase ein \]3uch gab (For simplic- 
ity, the NP is treated as if it was a word). Since 
tin Buch bears a pitch accent, the Focus Linking 
Principle (27) applies, introducing an obligatory 
link to FOC (h_link(\[\], foe)). For the unaccented 
gab, the Background Linking Principle applies, 
giving rise to a defensible link (s_link(N, bg)). At 
phrase level, the Indirect Head F-Marking Prin- 
ciple (29) and the Head F-Marking Principle ap- 
ply (introducing s_link(g\],N) for the head, and 
s_link(N,iN) for the phrase, respectively). In ad- 
dition, (27) applies again: s_link(iN, bg). 
As for the F-Skeleton, subclause (i) of (31) ap- 
plies at ein Buch, subclause (iii) at gab, causing 
the function compose to apply to gab's own se- 
mantic value and to its sister's F-SKEL value. The 
phrase is covered by (32), where both daughters 
are marked \[MAX-F +\] and thus fulfil subclause 
(ii). 
6 Conclusion 
This paper shows that a fully expressive under- 
specified representation of IS (:an be effectively 
composed by linguistic principles, circumvening 
the eoinputational problems that the disjunctive 
analyses of existing theories pose. Also, a resolu- 
tion routine was presented. The idea is to leave 
I 
ein BucII 
HPSG analysis 
sJi~k(m,m)} 
O-SEM \[\]gab' l,'-SK~I~ eomp(ra, ~r(iN)) 
(= g~t,'(~)) 
MAX-F J- 
g~l) 
the representation underspecified in applications, 
unless resolution is required for a specific reason. 
References 
Eberle, Kurt (1996): "Disambiguation by information 
structure in DRT," COLING-96 (this conference). 
Engdahl, Elisabet, Enric VallduW (1994): "Infor- 
mation packaging and grammar architecture: A 
constraint-based approach," in: Engdahl, E. (ed.): 
Integrating information structure into constraint- 
based and eategorial approaches (DVANA-2 Report 
R.1.3.B). ILLC, Amsterdam. 39 79. 
HShle, Tilman (1982): "Explikation fiir 'nor- 
male Betonung' und 'nor•ale Wortstellung'," in: 
Abraham, W. (ed.): Satzglieder im Deutsehen. 
Tiibingen: Gunther Narr Verlag. 
Jacobs, Joachim (1994): "Informational Autonomy." 
In: Bosch, P. and R. van der Sandt (eds.): Focus 
and Natural Language Processing. Working Paper 
6, ILL, IBM Heidelberg. 
Kuhn, Jonas (1996): "Context effects on.interpreta- 
tion and intonation". Ms. Institut fib maschinelle 
Sprachverarbeitung, Universit~t Stuttgart. (ftp:// 
ftp. ims. uni-stuttgart, de/pub/papers/kuhn/) 
Pollard, Carl, Ivan Sag (1994): Head-Driven Phrase 
Structure Grammar. University of Chicago Press. 
Rooth, Mats (1985): "Association with Focus," PhD 
thesis. University of Massachusetts, Amherst. 
Selkirk, Elisabeth (1995): "SenteIlce Prosody: In- 
tonation, Stress and Phrasing," In: Gold- 
smith, J. A. (ed.): Itandbook of Phonological The- 
ory, Basil Blackwell. London. 
Schwarzschild, Roger (1995): "On Givenness," Hand- 
out of talk presented at University of Massachusetts, 
Amherst. 
675 
