SPANISH ABSOLUTE CONSTRUCTIONS AND ASPECT
Ma Lluisa He nanz
Uni e si a Au bnoma de Ba celona
This pape is de o ed o he s udy o he aspec ual alue o absolu e cons uc ions in
Spanish. Special a en ion will be paid o he co ela ion holding be ween he
pe ec i e in e p e a ion o hese s uc u es and hei abili y o licensing
an
o e
subjec
NP.
On he basis o such a co ela ion, i will be claimed ha pas
pa iciples oge he wi h p edica es selec ing an e en a gumen p ojec
au
ASP(ec )
head p o iding he adequa e s uc u al con ex o he Case-ma king o hei subjec .
O.
In oduc ion
A
widesp ead assump ion in he s anda d li e a u e dealing wi h he s udy o seconda y
p edica ion is ha all majo syn ac ic ca ego ies may con ain a subjec posi ion unde ce ain
s uc u al condi ions (c . S owell
(1983)).
In o he wo ds, subjec -p edica e ela ions can be
exp essed by means o a a ie y o syn ac ic s uc u es, some o which may
be
non-sen en ial.
The co e o he discussion ega ding asymme ies be ween sen ences and ca ego ies which
exhibi a de ec i e dis ibu ion wi h espec o lexical subjec s has mainly ocused on he
p ope ies o small clauses (SCs). As is well-known, SCs a e p ojec ed om heads which,
al hough hey a e able o assign
a
hema ic ole o a subjec NP, hey ne e heless lack TENSE
and AGR unc ional nodes; hence, hey canno Case-ma k i .
This pape is de o ed o he analysis o a kind o s uc u es, namely,
absolu e cons uc ions
(ACs), whose p ope ies, al hough hey ha e been paid less a en ion han hose o SCs,
illus a e he asymme y ci ed abo e s aigh on a dly. ACs pa e n wi h SCs in ha hey lack a
unc ional s uc u e capable o licensing a NP subjec posi ion. In con as , since hey a e
Ca alan Wo king Pape s in Linguis ics (CWPL), 1991:
75-
128.
Uni e si a Au hnoma de Ba celona
always gene a ed, con a y o SCs, as
non go e ned
(maximal) p ojec ions, he e is no
a ailable
ex e nal
Case-ma ke o hei subjec .
The bulk o his pape deals wi h he s udy o he aspec ual alue o ACs. I will be shown ha
he e is a s ong co ela ion be ween he 'in e na1 empo ali y' o he p edica es heading ACs
and hei abili y o licence a subjec posi ion. Ex ending a ecen p oposa1 abou absolu e pas
pa iciples in Spanish made in de Miguel (1990), we will claim ha p edica es aking a
[+pe ec i e] ea u e, oge he wi h pa iciples, p ojec a ' la ' unc ional s uc u e, ha is, an
ASP(ec ) node, which is esponsible o he Case-ma king o hei subjec . A he sa ne ime, i
will be a gued ha e b mo emen in o COMP and hence subjec in e sion, a he han playing
a cen al ole in Case assignmen o he subjec , may
be
iewed
as
an ins ance o a mo e gene al
phenomenon closely connec ed o empo ally de ec i e con ex s.
The s uc u e o his pape is as ollows. In he i s sec ion, empi ical obse a ions ega ding
he main cha ac e is ics o ACs
as
well as hei di e ences wi h espec o o he ela ed
cons uc ions will be p esen ed. In he second sec ion h ee di e en app oaches dealing wi h
he p oblem o Case assignmen in ACs will be discussed. The hi d sec ion is de o ed o he
s udy o he aspec ual in e p e a ion o ACs. As we will show, despi e he he e ogeneous
cha ac e is ics o p edica es heading ACs, an in e es ing gene aliza ion eme ges conce ning hei
aspec ual in e p e a ion: namely all o hem ha e a pe ec i e eading. Finally, in he ou h
sec ion i will be claimed ha subjec NPs ge Case in ACs by means o a so o 'aspec ual
ag eemen ' ule aking place be ween he ASP head and he
NP
mo ed in o i s speci ie
posi ion.
1.
Absolu e Cons uc ions: A B ie Desc ip i e Accoun
'Absolu e cons uc ions' ha e no ecei ed a p ecise and uni o m analysis wi hin he
GB
amewo k. As is well-known, unde such a heading di e en kinds o s uc u es a e
subsumed. Acco ding o S ump (1985), ACs exhibi he main syn ac ic cha ac e is ics o
ad e bial subo dina e clauses, bu hey a e dis inguished by
a
special peculia i y: hey lack any
so o subo dina ing conjunc ion which o e ly speci ies hei logical connec ion wi h he
clause hey modi y. This peculia i y is shown in he examples in (I), which illus a e he h ee
classes o ACs discussed in S ump's wo k, namel y,
"
ee adjunc cons uc ions" (1 a),
"nomina i e absolu e cons uc ions" (lb), and "augmen ed absolu e cons uc ions" (lc):
(1) a. Walking home, he ound a dolla
b.
His
a he being
a
sailo , John knows all abou boa s.
c. Wi h he child en asleep, Ma y wa ched
TV.
[apud S ump
(
1985)l
Al hough he absence o an o e complemen ize is c ucial in he analysis o ACs as we will
see below, i should be said ha i does no cons i u e he only ele an p ope y o such
cons uc ions. Mo eo e , a close look a he examples ci ed abo e e eals a sha p con as
ega ding he condi ions o licensing hei subjec posi ion. No ice ha he ACs in (I), as well
as hei Spanish coun e pa s in
(2),
all unde wo di e en subse s depending on he na u e o
hei espec i e subjec s: a lexical
NP
in (b)-(c), and an emp y ca ego y in (a):
(2)
a. Jugando
ej
al
b idge, mis iasi se en e ienen
Playing
ei
o- he b idge my aun si se en e ain
'Playing b idge, my aun s keep hemsel es amused'
b. Jugando Juan al b idge, las discusiones se án ine i ables
Playing
J.
o- he b idge, he a gumen s will-be ine i able
'Juan playing b idge, a gumen s will
be
ine i able'
c. Con la en ana abie a, pasa nos mucho no
Wi h he window open we-s and much cold
'Wi h he window open, we a e e y cold'
On he basis o hese da a, i is qui e easonable o suppose ha he examples in (la) and (2a)
pa e n, wi h ega d o licensing condi ions o he embedded subjec , wi h con ol s uc u es,
since i is he ma ix clause ha p o ides he an eceden
NP
o he emp y ca ego y illing he
subjec posi ion in he AC. In con as , such an analysis canno be ex ended o he emaining
examples in (1)-(2), which u n ou o
be
p oblema ic o Case Theo y, as has been o en
obse ed.
A somewha di e en app oach o ACs is gi en in Beukema (1982). Acco ding o his au ho ,
" ee adjunc s" a e no a se o cons uc ions belonging o he b oade class o ACs, as in
S ump's app oach. On he con a y, he claims ha ACs ("absolu e ee adjunc s" in his
e minology) a e a pa icula sample -in ha hey include an o e subjec - o he " ee
adjunc s". The la e cons uc ions, in u n, a e de ined
as
"non ini e ph ases which a e adjoined
o ph ases in such a way ha he ee adjunc modi ies he clause in an ad e bial unc ion" -c .
Beukema (1982:71). Gi en ha he omission o he complemen ize is no concei ed
as
a
ele an p ope y o ee adjunc s, Beukema's classi ica ion does no exclude examples such
as
hose in
(3),
in which an o e subo dina ing pa icle appea s:
(3)
a. I go soaked, (while
)
sailing ac oss he es ua y
b. He w o e his poem while d unk
[apud Beukema (1982)l
In sho , he wo de ini ions ske ched abo e do no o e lap. As o Spanish, no ice ha he
choice be ween he i s c i e ion and he second one yields di e en esul s ega ding he
analysis o he ad e bial in ini i e clauses in
(4).2
Thus, i S ump's de ini ion is adop ed, hese
examples and ACs canno be clus e ed oge he , since he in ini i es -con a y o bo h ge unds
and pa iciples- should be p eceded by a subo dina ing elemen ( om now on we use he
con en ion
(.
.
.)*
ins ead o
*(.
..)
o mean 'ung amma ical wi hou he pa en hesized pa ' when
hese pa en hesis comes a he beginning):
(Después de
)*
habla el p esiden e, odos aplaudie on
(A e o ) o-speak he p esiden , all clapped
'When he p esiden had spoken, e e ybody clapped'
(Al
)*
sali el sol, se pusie on en camino
To- he o- ise he sun, se pu - hey in way
'A sun ise, hey se o '
(Con
)*
se Juan ico, i e mise ablemen e
(Wi h) o-be
J.
ich, li es mise ably
'Al hough Juan is ich, he li es squalidly'
(De
)*
segui las cosas asi, nos a uina emos
(F om) o-con inue he hings so, ou sel es will- uin
'I hings go on like ha , we will go bank up '
(*De
)
can ando la Caballé, el ea o se llena á
(F om) singing he Caballé, he hea e se will- ill
'I
Caballé sings, he hea e will ill up'
(*Po
)
habiendo demos ado el domado su alen ia, 10s espec ado es aplaudie on
(Fo ) ha ing p o ed he a ne his cou age, he audience clapped
en usiasmados
en husias ic
The ame ha ing p o ed his cou age, he audience clapped wi h en husiasm'
("Si i
)
leída la sen encia, el juez se e i 6
(Wi hou ) ad-pp he sen ence, he judge se e i ed
'(Wi hou ) ha ing ead he sen ence, he judge e i ed'
(*Po
)
e ec uado el examen, 10s alumnos espi a on
(Fo ) e ec ed he exam, he s uden s b ea hed
'Ha ing inished he exam, he s uden s b ea hed'
In his wo k we will ake o g an ed -in line wi h Beukema's app oach- ha he mos
ele an p ope y o ACs consis s in including an o e subjec
NP.
Howe e , we will ollow

S ump's main assump ion ha complemen ize s ne e eme ge in his lund o cons uc ions.
Pu ing aside o he momen he ques ion abou he de ini i e o ma ha he de ini ion o ACs
mus adop , le us conside a salien peculia i y o such cons uc ions, namely he
he e ogeneous na u e o hei p edica es. No ice ha , besides pa iciples and ge unds, adjec i es
can also head an AC, as in (7)-(8):
(7)
a. Limpias las a m as...
Clean he weapons
'Once he weapons clean.
..I
b.
Llenos 10s equisi os legales..
.
Full he equi emen s legal
'Once he legal equi emen s ul illed
...I
[apud Bel10 (1874)l
(8) a. Ha os de es udia 10s colegiales, la maes a decidi6 lle á selos al pa que
Fed-up o o-s udy he pupils, he eache decided o- ake- hem-away o- he pa k
The pupils being ed-up wi h s udying, he eache decided o ake hem away o
he pa k'
b.
Fija la mi ada en sus hijos, Ped o omaba el sol
Fix he gaze on his child en, P. was- alung he sun
'His eyes ixed on his chlld en, Ped o was baslung in he sun'
c. (Una ez) llena la sala, e a casi imposible espi a
(Once) ull he lounge, i -was nea ly impossible o b ea he
'Once he lounge was ull, i was nea ly impossible o b ea he'
d. Tenso el ges o, el ga o nos obse aba desde la silla3
Tense he pos u e, he ca us obse ed om he chai
'In a ense a i ude, he ca s a ed a us om he chai '
In iew o such da a, i could be a gued ha his is so because he adjec i es
limpias
'clean',
llenos
' ull', e c., being he 's ong' o ns4 co esponding o he egula pa iciples
limpiadas
80
.
'cleaned' and
llenados
' illed', espec i ely, beha e in a sense as ue pa iciples; hence hey
a e allowed o head an AC. No e, howe e , ha his conjec u e is no bo ne ou by he
examples in
(9)
and (10), which show ha bo h ad e bs and PPs can also play he ole o a
p edica e in ACs:
(9)
a. Asi las cosas, el Gobie no dec e 6 el es ado de excepción
So he hings, he Go e nmen dec eed he s a e o eme gency
'Things being like ha , he Go e nmen dec eed he s a e o eme gency'
b. Lejos 10s buques, la población isleña pudo eg esa a sus casas
Fa -away he ships, he popula ion insula could e u n o hei houses
'Once he ships we e away, he insula popula ion could e u n back home'
[apud Gu ié ez
(
1985)]
(10) a. En malas condiciones las conse as, las in oxicaciones se sucedie on du an e a ios
In bad condi ion he canned- ood, he in oxica ions
se
ollowed du ing se e al
días
days
The canned ood being in bad condi ion, poisoning wen on o se e a1 days'
b. Bajo mínimos el ae opue o, es casi imposible a eniza
Below minimums he ai po , i -is almos impossible o land
The ai po being below minimum sa e y-condi ions, i is almos impossible o
land'
c.
Fue a del alcance de 10s ada es 10s a iones, las incu siones aé eas esul a on
Ou o - he scope o he ada s he planes, he ai aids esul ed
de as ado as
de as a ing
'Planes being beyond he scope o he ada s, he ai aids we e de as a ing'
On he basis o he analysis jus ou lined, we will p oceed o summa ize he main o mal
p ope ies o ACs:
(i) ACs a e no lexically selec ed by he p edlca e o he clause hey modi y.
(ii) They adop an ad e bial in e p e a ion which does no co ela e wi h any o e
complemen ize .
(iii) They ha e a lexical
NP
subjec , co e e en o no wi h a
NP
in he main clause.
(i ) They a e headed by an he e ogeneous class o p edica es, anging om non ini e e bal
onns o adjec i es, ad e bs, and PPs.
As has been unde lined by many g amma ians in hei accoun o ACs, a u he p ope y may
be
added o he in en o y illus a ed in (i)-(i ), namely in ona ion: ACs, indeed, a e cu o om
he emaining pa o he sen ence wi h a melodic in iec ion. Ne e heless, such a phenomenon
is by no means exclusi e o hese cons uc ions. Obse e ha bo h adjec i es and pa iciples
can also be sepa a ed om he con ex by a pause al hough hey a e no ACs, as shown in
(
1 1):
(1 1) a.
La
meda de p ensa, con ocada a las es, empezó con una ho a de e aso
The p ess con e ence, summoned a h ee, s a ed wi h one hou o e a d
'The p ess con e ence, which had been summoned a
3
pm, s a ed one hou la e'
b. Temblo osa y aci u na, la mendiga pedía limosna
T embling and aci u n, he begga asked- o alms
c.
La
endedo a, amable
y
jo ial, egalaba ca amelos a 10s niños
The saleswoman, kind and chee ul, ga e swee s o he child en
Despi e he ac ha he examples in
(1
1) and ACs pa e n alike om he melodic poin o iew,
a as a ay o di e ences can be d awn be ween bo h so s o cons uc ions, which we
summa ize below:
1) The so-called 'inciden al's adjec i es (and pa iciples) in (ll), con a y o ACs, a e
connec ed o
an
NP
in he ma ix clause, as he ag eemen holding be ween emblo osa
' embling- em
'I
aci u na ' aci u n- em
'
and la mendiga ' he begga - em 'in (1 lb), e c.
shows.
2)
While ACs may co ela e wi h a pa allel clause in ol ing aQ inciden al adjec i e, he opposi e
si ua ion is no always allowed, as he con as be ween (12a) and (12b) illus a es:
(12) a. Con ocada la uda de p ensa a las es, 10s pe iodis as u ie on
Summoned he p ess con e ence a h ee, he epo e s had
que al e a su agenda de abajo
o al e hei agenda
The p ess con e ence ha ing been summoned a h ee o'clock, he epo e s had
o al e hei agenda'
b. *Amable y jo ial la endedo a, egalaba ca amelos a 10s niños
Kind and chee ul he saleswoman, she-ga e swee s o he child en
3) Inciden al adjec i es ha e
a
neu al seman ic in e p e a ion closely ela ed o ha o non-
es ic i e ela i es. Thus, he examples in (13), whe e an ad e bial subo dina e clause appea s
as a subs i u e o he inciden al adjec i es in (llb) and (llc), canno be unde s ood as
pa aph ases o he la e examples:
(13) a.
#??La
mendiga, po es a emblo osa y aci u na, pedía limosna
The begga , o o-be embling and aci u n, asked- o alms
The begga , because she was embling and aci u n, asked o alms'
b.
#??La
endedo a, una ez amable y jo ial, egalaba ca amelos a 10s niños
The saleswoman, once (she was) kind and chee ul, ga e swee s o he child en
Seemingly, al hough he inciden al p edica e in (lla) is a pa iciple, i has a depic i e
in e p e a ion a he han an ad e bial one, con a y o wha happens in (123). This is illus a ed
in (14a) and (l4b), espec i ely:
Indeed,
as
hese examples show, adjec i es ((27a)),
as
well
as
ad e bs ((27b)) and PPs ((27c))
can play he ole o a p edica e in ACs: h ee ca ego ies which canno
be
collapsed, ob iously,
in o a single 'A-' i em.
90
explo ed in hzzi's wo k, canno
be
cap u ed by he p inciple o nula ed in (21), nei he in
Spanish no in I alian. In o he wo ds, since no auxilia y p ecedes he NPs in i alics in (26), he
way hey ecei e Case emains unclea unde Rizzi's analysis:
(26) a. Resol iendo
el juez
absol e al acusado
...
Deciding he judge o absol e he de endan
...
b. Dispues o
Juan
a colabo a ..
.
Ready Juan o collabo a e..
.
Finally, a hi d p oblem should be men ioned. We ha e al eady no ed ha p edica es eme ging
in ACs exhibi e y he e ogeneous syn ac ic cha ac e is ics. As a consequence, i u ns ou ha
in many cases he elemen p eceding he subjec
NP
is no an auxilia y e b. This en ails a
se ious limi a ion conce ning he scope o he ule s a ed in (21):
(27) a.
Fu iosa
Man'a con su ma ido..
.
Fu ious Ma ia wi h he husband
...
'Man'a being u ious a he husband
...'
b.
Asi
las cosas.
..
So he hings
...
Things being like ha
...I
c.
En es ado de ale a
la población.
..
In s a e o ale he popula ion
...
The popula ion being in a s a e o ale ..
.'

3.
The hi d app oach we will ake in o conside a ion was pu on a d by de Miguel(1990) in
he analysis o Spanish absolu e (pas ) pa iciples. Following he assump ion ad oca ed in
Pollock (1989) ha bo h empo al and ag eemen ea u es commonly associa ed o he ca ego y
INFL
should a he
be
seen
as
wo independen unc ional heads, T(ense) and AGR(eemen ),
de Miguel assumes ha a u he unc ional p ojec ion
-
headed by aspec
-
can
be
p esen in
he clause. Mo e p ecisely, she a gues ha absolu e pa iciple clauses a e aspec ual ph ases
p ojec ed om a head ASP(ec ) aking he ea u e
[+
pe ec i e]. This is due o he pe ec i e
aspec ual alue o absolu e pa iciples as well
as
hei lack o empo al ea u es.
Depa ing om hese assump ions, de Miguel p oposes o de i e absolu e pa iciple clauses
h ough a mechanism ha - o pu i in an in o mal way- collapses bo h he s a egy o
Aux-
o-Comp and Raposo's (1987) app oach o Po uguese in lec ed in ini i es. He analysis can
be
summa ized as ollows. Conside (28), whe e he ele an pa o he ull clause s uc u e is
in i alics:
(28) Con ocada
la
eunidn, 10s polí icos in e mmpie on sus acaciones
Summonned he mee ing, he poli icians suspended hei holiday
The D-S uc u e ep esen a ion o he AC in (28) looks, in i s essen ials, like (29):
I
COMP'
ASPP
I
ASP'
ASP VP
con oca la eunión
As shown in
(29),
he su ace subjec
la eunidn
is gene a ed
as
he objec o he ansi i e e b
con oca .
A S-S uc u e,
V
aises o he ASP head o inco po a e he aspec ual su ix
-do
;
once his mo emen has aken place, he e b becomes a
nominal
ag eemen elemen ( ha is, i
ca ies gende and numbe ea u es bu lacks hose o pe son). F om ASP posi ion, he
complex
N[V-do
]
aises in u n o COMP. On he o he hand, since he
NP
la eunidn
canno
ecei e Acusa i e Case unde objec posi ion, i should mo e o he i s a ailable speci ie
node, namely he speci ie o ASP. Hence, he esul ing s uc u e will
be
(30):
COMP'
A
COMP ASPP
con ocada
i
k
la eunión ASP'
ASP
A
VP
0
/
V ASP
I
V'
[apud de Miguel(1990), p. 17 11
Gi en he analysis ske ched abo e, an answe o he ollowing ques ion has o be p o ided:
how is Nomina i e Case assigned o subjec NP in absolu e pa iciple cons uc ions? No ice
ha , as he s uc u es ou lined in (29)-(30) lack bo h Tense and AGR nodes, he e is no clea
candida e sui able o his pa icula pu pose. Acco ding o de Miguel's p oposal, he e b, a e
being aised o COMP, go e ns ASPP and hence i s head.
As
a consequence, he ea u es
associa ed o ASP, by i ue o being go e ned, become -on a
pa
wi h nominal ag eemen -
's ong' enough o go e n and assign Case:9 indeed, he ASP head, once go e ned, will be
coindexed wi h he subjec NP loca ed in he speci ie posi ion o ASPP by means o he
amilia ule o Speci ie -Head Ag eemen ; ha is, a mechanism simila o he one which has
been ad oca ed o in canonical Nomina i e Case assignmen ia AGR.
De Miguel's analysis is in e es ing as i ies o cap u e a phenomenon ha is c ucial unde ou
concep ion o ACs, namely he ela ionship holding in hem be ween Case ma king and he
aspec ual alue o he pas pa iciple, syn ac ically ins an iq ed as he unc ional p ojec ion
ASPP. As a mino ema k abou his linguis 's hypo hesis, i has o be said ha he
assump ions conce ning he main ac o igge ing he mo emen o
V
o COMP
-
he necessi y
o he ea u es in ASP o become, h ough go e nmen , s ong enough o license a subjec
NP- migh be e o mula ed in di e en ems,
as
we will see below.
Be o e de eloping he co e o ou analysis, i should be no ed, howe e , ha he p oposa1 jus
ske ched gi es ise o some p oblems when ex ended o ACs non headed by a pas pa iciple.
An impo an issue in his ega d conce ns (absolu e) ge unds.10 A i s sigh , hese
cons uc ions look e y simila o absolu e pa iciples, inso a
as
hey p eclude any so o
subo dina ing conjunc ion in COMP posi ion. Ne e heless, hey di e om pa iciples in e y
signi ican ways, which we summa ize immedia ely.
To
s a wi h, no ice ha ge unds, con a y o pas pa iciples, lack he aspec ual ea u e
[+pe ec i e]. The e o e, while he o me e e o du a i e si ua ions, he la e imply ha a
gi en ac ion is comple ed. Examples in (3 la)-(3 1b) and hei espec i e pa aph ases in (3%)-
(32b) illus a e he dis inc ion jus made:
(3
1)
a. Jugando Juan
al
b idge, la pa ida se ala ga d
Playing Juan o- he b idge, he game will-las -longe
'Wi h John playing b idge, he game will las o longe '
b. Dispu ado el p ime encuen o, la elimina o ia quedó decidida
Con ended he i s ma ch, he p elimina y ound emained decided
'The i s ma ch ha ing been played, he p elimina y ound was decided'
(32) a.
Si
/
dado
que
Juan juega
al
b idge, la pa ida se ala ga á
I
1
gi en ha J. plays b idge, he game will-ge -longe
(32) b. Una ez
I
después de que se hubo dispu ado el p ime encuen o, la elimina o ia
Once
I
a e ha had been played he i s ma ch, he p elimina y ound
quedó decidida
eamained decided
'Once 1 a e he i s ma ch had been played, he p elimina y ound was
decided o '
Secondly, i has o
be
no ed ha , as a consequence o hei lachng gende , numbe and pe son
mo phemes, ge unds eme ging in ACs canno ag ee wi h a (subjec )
NP
(33) Jugando
él
1
ella
al
b idge
...
Playing he
1
she b idge
...
This amoun s o saying ha li le mo phological e idence can
be
adduced in o de o suppo a
mechanism o licensing he subjec
NP
pa allel o he one ha has been a gued o wi h espec
o absolu e pa iciples (i.e., he amilia ule o Speci ie -Head Ag eemen ).Howe e , his
di icul y migh be dispensed wi h by assuming -wi h Belle i
(1990)-
ha a so o abs ac
ag eemen akes place in his case, due o he de ec i e na u e o he ge undi al mo phology.
Recall, inciden ally, ha he same si ua ion a ises in in ini i e clauses wi h an o e lexical
subjec (c . exs. in (4)).
Finally, a hi d di e ence holding be ween ge unds and pa iciples in ACs in ol es nega ion.
Obse e in h~s ega d ha only he o me
can
be p eceded by a nega i e pa icle:
(34) a. No diciendo Ma ia la e dad, se engaña a si misma
No elling Ma ia he u h, se decei es o he sel
'In no elling he u h, Ma ia decei es he sel

(34) b.
Ni
siquie a
can ando la Caballé, se llena d el Liceo
No e en singing he Caballé, will- ill-up he Liceo
'E en i Caballé sings he e, he Liceo hea e will no ill up'
(35)
a. *No eunido el Cong eso, el p esiden e enunció a su in e ención
No assembled he Pa liamen , he p esiden ga e-up o his in e en ion
'The Pa liamen no ha ing been assembled, he p esiden ga e up his allocu ion'
b. *Ni con ocada la meda de p ensa con an elación, 10s pe iodis as llega on a de
No -e en summoned he p ess con e ence in ad ance, he epo e s a i ed la e
'In spi e o he ac ha he p ess con e ence had been summoned in ad ance,
he epo e s ani ed la e'
I we assume, in line wi h Pollock (1989), ha nega i e sen ences can be analyzed as
con aining a nega i e ph ase (NEGP) si ua ed be ween he unc ional nodes AGRP and TP in
hei unde lying ep esen a ion, a s aigh o wa d explana ion o he con as s in (34)-(35)
na u ally sugges s i sel . As has been poin ed ou by Belle i (1990), nega ion should no
be
allowed o appea in absolu e pa iciple cons uc ions because NEGP ills a posi ion ou side he
node ASPP domina ing such cons uc ions. In con as , he well o med sen ences in (34) seem
o indica e ha ge unds su ace om deep ep esen a ions p ojec ing up o he node ha NEG
akes as i s own complemen , ha is, TP. The ele an pa o he s uc u e is syn hesized by he
ee
in
(36):
...
NEGP
NEG TP
T ASPP..
Since a de ailed analysis o ge unds is beyond he scope o his pape , we will no pu sue his
issue he e. Le us pu hem aside and come back o ACs headed by a non e bal p edica e
illus a ed in (7)-(10). As han been no iced abo e, hese cons uc ions sha e mos o hei
gene al p ope ies wi h e bal ACs. Mo eo e , hey ha e he la ou o absolu e pas pa iciples
in ha hey ake a pe ec i e alue. This being so, hey a e clea candida es o be explained
ollowing de Miguel's p oposals. Howe e , some di icul ies a ise in a emp ing o p o ide a
uni ied accoun o bo h kinds o cons uc ions. A c ucial di e ence holding be ween hem is
ha ACs headed by a non e bal p edica e lack mo phological aspec ual ma ks such as
-do
o
-
ndo.
Acco ding o de Miguel, who deals wi h he p oblem only ma ginally, clauses including
adjec i es like
con en o
in (37), ins ead o ACs, should be me ely conside ed 'adjec i e
cons uc ions':
(37) Con en o
61,
con en os odos
Happy he, happy e e ybody
'I he is happy, e e ybody is happy'
Ac ually his linguis , who ollows in his espec a sugges ion made in Belle i (1981) -c .
also Belle i (1990)-, does no clea ly s a e wha "adjec i e cons uc ions" a e mean o
be,
no
he way hei subjec ecei es case, specially when i shows up
as
a non p onominal Np.ll On
3.
The Aspec ual Value o Absolu e Cons uc ions
As is well known, wha is adi ionally called
aspec
e e s o a b oad a ay o ac s, including
" ense, mo phological aspec ual ma ke s, ad e bial exp essions, lexical meaning o p edica es,
and he syn ac ic o seman ic na u e o hei a gumen s" -c . Tenny (1987: 16). Gi en ha
aspec is a non uni a y phenomenon, i should no be cons ained o e bal mo phology. In his
wo k we will ake o g an ed -in line wi h Tenny (1987), Bosque (1990) and de Miguel
(1990), among o he s- ha he e is no sha p dis inc ion be ween
aspec
'sensu s ic o' and
ak ionsa .
As Bosque (1990: 177) has poin ed ou , bo h no ions a e no as independen om
each o he as would appea in iew o hei espec i e de ini ions.
Tu ning back o ACs, le us ocus, o he ime being, on he cases in which
an
adjec i e heads
hese cons uc ions. I could
be
a gued ha he abili y o adjec i es o play he ole o a p edica e
98
he o he hand, gi en he examples in (27b) and (27c), bo h 'ad e b cons uc ions' and 'PP
cons uc ions' should be added o he syn ac ic pa adigm which encompasses "adjec i e
cons uc ions".
Summa izing, his sec ion has been de o ed o he discussion o one c ucial p oblem in ol ing
ACs, namely he way hei lexical subjec is licensed. A e a sho look a di e en p oposals
which deal wi h his ques ion, i is possible o conclude ha , as a as Spanish is conce ned, de
Miguel's app oach p o ides e y in e es ing insigh s in his ega d. Ne e heless, as has been
o mula ed, i does no co e ACs headed by p edica es o he han pas pa iciples.
In he ollowing sec ion we will add ess he issue o he aspec ual alue o ACs in mo e de ail,
and we will y o show ha ACs can be gi en an uni ied accoun ega dless o whe he hey
include a pas pa iciple o a non e bal p edica e. Acco ding o ou hypo hesis, bo h kinds o
cons uc ions, despi e hei ex e na1 di e ences, can ecei e an analysis in ol ing a unc ional
p ojec ion ASPP on he basis, c ucially, o hei common
pe ec i e
aspec ual alue.
ollows ei he om he ea u e
[+
VI
hey ake, o om hei close link wi h he pa iciples.
Howe e , such an app oach is misleading, as can
be
shown in (38):
(38) a.
"F anceses
10s pe umes, las en as se dispa a on
F ench he pe umes, he sales boomed
The pe umes being F ench, he sales boomed'
b.
"Ti nido
Juan, no ab ió la pue a
Shy Juan, he did no open he doo
'Being shy, Juan did no open he doo '
c.
"Zn e esan e
el pa ido, odo el mundo es aba pendien e del ele iso
In e es ing he ma ch, e e ybody was hanging on he TV-se
'The ma ch being in e es ing, e e ybody was hanging on he TV-se '
No e ha he ill o medness o he examples in (38), which a i s sigh seem e y close o hose
in
(7)
and
(8) abo e, is by no means due o seman ic ac o s. This
can
be
p o ed on he basis o
he ollowing examples, whe e he adjec i es in i alics in (38) su ace associa ed wi h
a
e bal
o m ( ensed o un ensed), he esul ing sen ence being g amma ical:
(39) a. Dado que 10s pe umes e an anceses
...
Gi en ha he pe umes we e F ench
...
b.
Siendo Juan ímido
...
Being Juan shy
...
c. Como el pa ido e a in e esan e
...
As he ma ch was in e es ing
...
A ca e ul look a he p edica i e adjec i es eme ging in (39) and in (7)-(8) - epea ed he e
as
(40)-
clea ly sugges s ha only he la e a e closely ela ed o pa iciples, as has been
obse ed by some g amma ians -c . Bel10 (1874:s 1176); in o he wo ds, hey belong o he
se o so-called
"
pe ec i e adjec i es" (c . Bosque
(
1990)):
The analysis ske ched he e c ucially elies on he ac ha , in empo aily de ec i e con ex s like
ACs, a subjec -p edica e ela ion
can
only
be
licensed i he i em playing he ole o he p edica e
akes a pe ec i e alue. As we will see below (c . sec ion 4), his equisi e may
be
accoun ed
o in a p incipled way on he assump ion ha some kind o 'aspec ual ag eemen ' holding
be ween subjec and p edica e is a wo k in he cons uc ions unde discussion. Fo he ime
being, le us conside a las piece o e idence in suppo o he signi icance o aspec ual ac o s
in he syn ac ic beha iou o ACs. No ice he con as illus a ed in (49), whe e he ele an
clauses a e in i alics:
(49) a. Con el niño en e no, no se puede abaja
Wi h he child ill, no se can wo k
'Wi h one's child ill, one canno wo k'
b. Con el niño ubio, no se puede abaja
Wi h he child blond, no se can wo k
The examples in (49) sha ply di e in ha he o me is in e p e ed as an AC wi h a condi ional
eading (i.e., 'I he child is
ili...'),
whe eas he la e can only
be
pa aph ased as an ad e bial
complemen (i.e., 'In he company o he blond child
...I).
This con as can easily be de i ed
om he di e en aspec uai alue o bo h adjec i es en e mo 'ill' (a s age-le el p edica e
selec ing an <e>), and ubio 'blond' (an indi idual-le el p edica e). Thus, since he adjec i e
en e mo has a pe ec i e alue, i can p ojec an ASP node; his possibili y, howe e , is
p ecluded in (49b) due o he
[-
pe ec i e] ea u e selec ed by ubio. In iew o such da a, he
ques ion may now be aised whe he he p esence o a Case assigne , namely he p eposi ion
con
'wi h' in he examples illus a ed abo e, cons i u es i sel a su icien condi ion in o de o
license he subjec
NP,
o some u he equi emen is a wo k. No e ha i he i s al e na i e
is chosen, he e is no na u al way o uling ou (49b) in he ele an sense discussed he e, ha
is, wi h
an
absolu e eading.

The analysis ou lined abo e has ocused on empi ical e idence conce ning adjec i es. No ice,
howe e , ha an in e es ing p edic ion conce ning he aspec ual s a us o PPs and ad e bs
heading ACs ollows om ou hypo hesis. Despi e he he e ogeneous na u e o hese elemen s,
all o hem sha e he common p ope y o subca ego izing an e en a gumen . I his
equi emen is no ul illed, he
ASP
node canno be p ojec ed, and he e o e he esul ing
sen ence will
be
ung amma ical. A ca e ull look a he
da a
shows ha his p edic ion holds ue.
Conside again
(10)
and (l8b) - enumbe ed as
(s)-,
as well
as
(51):
(50) a. En malas condiciones las conse as...
In bad condi ion he canned ood
b. Bajo mínimos el ae opue o..
.
Below minimums he ai po
c. Fue a del alcance de 10s ada es 10s a iones
...
Beyond he scope o he ada s he planes
d. En baja o ma las nadado as alemanas
...
In low shape he Ge man swimme s
(51) a. (Una ez) en Ó bi a el sa éli e..
.
(Once) in o bi he sa elli e
b. A cinco me os de la cima 10s alpinis as
...
Fi e me e s om he summi he alpinis s
c. En pe ec a o mación 10s soldados
...
In pe ec o ma ion he soldie s
No unexpec edly, p edica i e PPs eme ging in he ACs in (50)-(51) ha e simila cha ac e is ics
o hose exhibi ed by i ems aking an e en a gumen : hey cooccu wi h
es a ,
allow aspec ual
ad e bial modi ie s, and e e o ansi o y p ope ies. This is shown in
(52):
(52) a. Las conse as
es aban
/*e m en malas condiciones
The canned ood was in bad condi ion
b. El ae opue o
es a'
/*es bajo mínimos
The
ai po
is below minimums
c.
To almen e
en baja o ma las nadado as alemanas..
Comple ely o shape he ge man swimme s
...
d.
Ya
a cinco me os de la cima 10s alpinis as..
Al eady i e me e s om he summi he alpinis s
e c.
The si ua ion illus a ed in (50)-(51) has no pa allel in (53). No ice in his espec ha he
inabili y o he PPs
de buena amilia
' om a well-o amily',
de pla a
'o sil e ', e c. o head
an
AC co ela es signi ican ly wi h he ill o medness o he examples in
(54):
(53) a. *De a mas oma la ia Julia
...
O
bad empe Aun Julia
b. *Debuena amiliaJuan...
F om a well-o amily Juan
c. *Depla aelcolla ...
O sil e he necklace
e c.
(54)
a.
La
ia Julia
e a
/*es aba de
mas
oma
Aun Julia was bad kmpe ed
b. *Siendo
Juan
comple amen e
de buena amilia..
Being Juan comple ely o well-o amily
c.
*
Una
ez que
el colla ue de pla a..
.
Once
he necklace was o sil e
A u he piece o e idence suppo ing he hypo hesis ske ched he e is p o ided by NPs,
which, in con as wi h o he syn ac ic ca ego ies, a ely ac as p edica es in ACs. Pu ing aside
he ac ha NPs a e a gumen s, his is so because o hei non empo al na u e. Consequen ly,
such a p ope y makes i impossible o hem o include an e en <e> in hei hema ic g id. I
should be no ed, ne e heless, ha unde ela i ely excep ional condi ions NPs can also play
he ole o a p edica e in ACs. The examples in
(55)-(56)
illus a e he scope o he
phenomenon:
(55)
a. "P esiden e Feli
pe
González, la si uación del país cambió
P esiden Felipe González, he si ua ion o he coun y changed
'F.
G.
being he p esiden , he si ua ion o he coun y changed'
b. *Reina Ma ía, decidió ac ua como al
Queen Mana, she decided o ac as such
'Being he queen, Ma ia decided o ac as such'
(56)
a.
Una ez
p esiden e Felipe GonzAlez..
.
Once p esiden Feli pe González..
.
'Once
F.
G. has become p esiden
...'
b. Reinapo
un dia
Ma ía
...
Queen o a day Mana..
.
'Being queen o a day, Ma ía
...'
The con as s adduced in hese examples can
be
in e p e ed
as
ollows: due o exp essions such
as
una ez
'once', po
un dia
' o a day',
a la sazón
'on ha occasion', e c. he complexes
NP-
NP
in
(56)
a e endowed, o a ce ain ex en , wi h a ' empo al' a chl ec u e capable o licensing
a p edica i e link be ween bo h NPs. In o he wo ds, he exp essions jus quo ed ul ill a
unc ion somewha simila o ha o he in lec ed mo phemes adjoined o he copula in he
sen ences in
(57):
(57)
a. Felipe González es/e a/ ue/se á p esiden e
'Felipe González islwaslwill be a p esiden '
b. Ma ia esle al uelse á eina
'Ma ia is/was/will
be
a queen'
Summa izing, he discussion de eloped so a leads us o conclude ha a sys ema ic connec ion
can
be
pu on a d be ween he a i s sigh he e ogeneous na u e o p edica es heading ACs
and
a
pe ec i e aspec ual alue: such a alue
can
su ace ei he mo phologically
-
by means o
a non ini e e bal o m- o lexically, ha is, h ough he, so o say,
ak ionsa
o p edica es
selec ing an e en a gumen .
4.
ASPP
and
Case-Ma king
We ha e a gued in he p eceding sec ion ha he licensing o ACs is cons ained by hei
aspec ual alue. As o he issue conce ning he way he e en a gumen <e> selec ed by he
p edica e is syn ac ically p ojec ed, we ha e assumed, ollowing ecen app oaches o mula ed
in Bosque (1990), de Miguel (1990) and Suñe (1990), among o he s, ha an a gumen a1
s uc u e including an e en su aces as
an
ASP(ec ua1) node. In o he wo ds, he e en <e>
p ojec s un i1 a unc ional node ASPP, exhibi ing an in e nal s uc u e simila o he one which
has been pos ula ed o he emaining syn ac ic ca ego ies wi hin he
X'
heo y. This amoun s o
saying ha i is he e en a gumen ha licenses he p ojec ion ASPP. Wi h his p oposal in
mind, le us mo e he co e o he discussion in his sec ion o he mechanisms which make i
possible o subjec
NP
o ACs o be go e ned and case-ma ked, as well as o he way such
mechanisms in e ac wi h he ASPP node.
In
sec ion
2,
when discussing Rizzi's and de Miguel's analysis, we ha e no ed ha bo h
p oposals -pu ing aside he impo an di e ences holding be ween hem- sha e he
assump ion ha a close connec ion has o be es ablished in ACs be ween aising o he p edica e
(o he auxilia y e b) o COMP on he one hand, and go e men and case assignmen o he
lexical subjec NP, on he o he .15 Al hough ou analysis is, o a la ge ex en , compa ible wi h
such an assump ion, we will p opose, o conclude, a somewha di e en app oach.
Speci ically, we would like o sugges , in a p elimina y way, ha he well-known phenomenon
o subjec in e sion in ACs can be concei ed as no necessa ily connec ed o go emmen and
Case assignmen o he subjec . In o de o show in mo e de ail how his hypo hesis wo ks,
wo equi emen s should be ul illed. Fi s ly, an al ema i e explana ion o aising o he
p edica e in o COMP mus be p o ided; secondly, i is necessa y o elucida e which elemen
wi hin he p ojec ion headed by ASP go e ns and Case ma ks he in e ed subjec .
Be o e add essing bo h ques ions, some gene al conside a ions conce ning he syn ac ic
beha iou o ACs a e in o de he e. To s a wi h, no e ha p edica e aising o COMP in ACs
is closely linked o a peculia i y o hese cons uc ions which has been s a ed abo e (c . sec ion
1). Recall in his ega d ha ACs lack any so o subo dina ing conjunc ion which o e ly
speci ies he na u e o he ad e bial connec ion holding be ween hem and he ma ix clause hey
modi y. Such a 'gap' ob iously co ela es wi h he ac ha a he S-S le el he s uc u al space
belonging o COMP is no longe a ailable, as i has been illed by he aised p edica e. As has
been no iced by Belle i (1990), his assump ion p o ides a na u al explana ion o he
o he wise unclea phenomenon o ung amma icali y o bo h absolu e ge unds in
(5)
and
absolu e pas pa iciples in
(6)
- enumbe ed as
(58)
and (59), espec i ely. A he same ime, i
d aws a sha p di e ence be ween hese cons uc ions and he ad e bial in ini i es in
(4)
[=
(60)],
which, in con as , allow a subo dina ing pa icle p eceding hem:
(58)
a.
(*De
)
can ando la Caballé
...
(F om) singing he Caballé
b. (*Po
)
habiendo demos ado el domado su alen ia..
.
(Fo ) ha ing p o ed he ame his cou age
(59) a.
(*Sin
)
leída la sen encia, el juez se e i 6
(Wi hou ) ead-p p he sen ence, he judge se e i ed

(59)
b. (*Po
)
e ec uado el examen, 10s alumnos espi a on16
(Fo ) e ec eú he exa n, he s uden s b ea hed
(60)
a. (Después de
)*
habla el p esiden e, odos aplaudie on
(A e ) o-speak he p esiden , all clapped
b. (Al
)*
sali el sol, se pusie on en camino
(To- he) o- ise he sun, hey se o
c. (Con
)*
se Juan ico, i e mise ablemen e
(Wi h) o-be
J.
ich, li es mise ably
'Al hough Juan is ich, he li es squalidly'
d. (De
)*
segui las cosas asi, nos a uina emos
(F om) o-con inue he hings so, ou sel es will- uin
'I hings go on like ha , we will go bank up '
On he basis o such ac s, an in e es ing possibili y could
be
explo ed
,
ha is, he absence o
an o e subo dina ing conjunc ion in ACs is due o he ac ha he aised p edica e beha es o
a ce ain ex en as a complemen ize . Appealing e idence in suppo o his supposi ion comes
om some his o ical da a conce ning he de elopmen o ACs. As has been obse ed in Bel10
(1874:s
1185 ) among o he s, he ACs in (61), which we e ully g amma ical in Old Spanish,
a e a he sou ce o he examples in (62), whe e ag eemen be ween
NP
and sal o o excep o is
no longe possible:
(61) a.
*Sal es
10s
niños, odo el mundo es aba cansado
Sa e-pl he child en, e e ybody was i ed
'Excep o he child en, e e ybody was i ed'
b. *Excep a Man'a, odos se queúa on en casa
Excep - em Ma ia, all s ayeú a home
'Excep o Ma ia, e e ybody s ayed home'
(62) a. Sal o 10s niiios
...
b. Excep o Mm'a
...
Mo eo e , he close connec ion holding be ween aised p edica es in
A
Cs and complemen ize s
ecei es addi ional con ima ion om he ac ha many ad e bial complemen ize s in Spa ush,
as well as in o he languages, de i e om absolu e pas pa iciples: supues o (que) 'in case
( ha )'(< pas pa iciple o supone ' o suppose'), dado (que) 'gi en ( ha )'
(<
pas pa iciple o
da ' o gi e'), pues o (que) 'since'
(<
pas pa iciple o pone ' o pu '), e c.
Secondly, he e a e in e es ing pieces o e idence in a ou o he idea ha p edica e aising o
COMP in ACs can
be
concei ed as a pa icula ins ance o a gene al phenomenon whose scope
goes beyond go e nmen and Case ma king o he subjec . Conside he ollowing exa nples:
(63) a. Si (Ma ia
)
u iese (Ma ía
)
dine o, se comp a ia la casa
I (Ma ia) had (Ma ía) money, se would-buy he house
'I Ma ía had he money, she would buy he sel he house'
b. *Ma ia u iese dine o, se comp a ía la casa
Ma ía had money, se would-buy he house
=a
c. Tu iese Ma ia dine o, (y) se comp a ia la casa
Had Ma y money, (and) se would-buy he house
=a
In hese cases, as has been obse ed in Rizzi (1982), (1988) o simila examples in I alian and
English, i he complemen ize si 'i ' is omi ed, he esul ing s uc u e (i.e., (63b)) is
ung amma ical, unless subjec - e b in e sion has applied, like in (63c). Gi en he accep abili y
o bo h al e na i es illus a ed in (63a) conce ning he subjec posi ion,17 he ques ion may now
be aised why in e sion is he only a ailable s a egy o p ese ing he condi ional eading o
he sen ence when conjunc ion si does no su ace.
A
possible (in o mal) answe is ha in
(63c) he e b, once mo ed in o COMP, 'abso bs', in a sense, he ole played by si in (63a).
No ice, on he o he hand, ha V- aising and hence subjec in e sion canno be a ibu ed o
go e nmen and Case ma lung equi emen s, since in he sen ence unde discussion (i.e.,(63c))
he e b, c ucially, is in lec ed.
Finally, ele an da a conce ning he analysis o ACs come om F ench and English, whe e no
subjec in e sion phenomenon akes place in such cons uc ions, as is well-known:
(64)
a. Ceci di
...
This said
...
b.
His
a he being a sailo ..
.
[=
(
1
b)]
On he assump ion ha subjec in e sion applies because he e is no in e na1 Case assigne
capable o licensing he subjec o ACs, i appea s o
be
a pa adoxe ha F ench o English,
ha ing a much weake INFL han Nul1 Subjec languages, lack a s a egy such
as
V- aising in o
COMP in o de o p o ide he adequa e con ex o subjec go e nmen in hese cons uc ions.
In o he wo ds, e en i he amilia app oaches accoun ing o Case ma king o lexical subjec s
in ACs u n ou o be ue, he e is no na u al way o ex end hem in o de o co e cases like
hose in (64). In con as , ou p oposal, al hough no p o iding a s aigh on a d solu ion o
he p oblem unde discussion, is open o explaining seemingly dispa a e p ope ies exhibi ed by
ACs in di e en languages.
Tu ning back o he mechanisms which igge mo emen o he p edica e in o COMP, i seems
plausible o assume ha such mechanisms, gi en ha he COMP node is an impo an axis o
pa ame ic a ia ion, a e ela ed o non i ial aspec s ega ding di e ences be ween he
beha iou
o
bo h Nul1 Subjec languages and English (o F ench). As a as English is
conce ned, clea e idence can
be
gi en in o de o show ha he na u e o COMP exhibi s sha p
con as s in his language wi h he co esponding node in Spanish:
(65)
a.
When
(while
)
igh ing in F ance, he was aken p isone
b. Ape lea ing In e s a e
45,
I
no iced a sign
...
(65)
c. She looked pleadingly a he pa en s
as
hough
en ea ing o gi eness
[apud S ump
(
1985)l
(66)
While
d unk
I
a he beach
I
s anded in enemy eni o y, he was aken p isone
[apud S ump
(
1985)l
As is illus a ed in
(65)-(66),
he e is no incompa ibili y in English, a leas in s uc u es close o
ACs, be ween an o e complemen ize and empo ally de ec i e domains. In con as , such an
incompa ibili y exis s in Spanish:
(67)
a. *Es a soluci6n,
si
necesa ia
...
Ths solu ion, i necessa y
...
b. *Juan,
mien as
en baja o ma..
.
Juan, while in low shape
...
c.
*Ped a,
cuando
luchando
...
'Ped o, when igh ing
...'
e c.
In iew o hese ac s, i may be a gued ha he COMP posi ion is less empo ally 'ine ' in
Spanish han in English; hence, i is capable o a ac ing ei he a non ini e e bal o m o a
p edica e selec ing an e en a gumen .18 The co e o his sugges ion is ela ed o a p oposa1
made
in
S owell (1982:563), acco ding o which " he COMP posi ion is whe e ense ope a o s
may appea , a some le el o g amma ical ep esen a ion". Tense ope a o s ix, in S owell's
wo ds, " he unde s ood ime a ne o he complemen clause ela i e o he ense o he ma ix".
As well-known, he choice o he complemen ize co ela es wi h he empo al amewo k o he
complemen clause. The e o e, some mechanism is needed in o de o cap u e he ma ching
be ween complemen ize and he ense ope a o ype. Recall in his espec ha ACs a e
de ec i e in a double sense,
as
hey lack bo h a complemen ize and a TENSE node. In iew o
such ac s, i could
be
conjec u ed ha he only s a egy in o de o make ' isible' he in e na1
empo ali y o he complemen clause is V- aising in o COMP, hus allowing ASP o ake scope
See Bosque (1990) o a mo e de ailed analysis o hese adjec i es.
5
This enn is used in Gu ié ez (1986: 143).
6
In his pape we will no ake in o conside a ion ACs in oduced by con 'wi h'. C . Suñe
(1988) o a de ailed analysis o hese cons uc ions in Spanish.
7
Acco ding o Reuland (1983: 126 ), his condi ion accoun s o he con as be ween (ia) and
(i
b)
:
(i) a. John kep wallung slowly, he ain d enching he oad
b. *John kep walking slowly, while he ain d enching he oad
8
As he examples in (24b) and (25b) show, habe ' o ha e' and es a ' o be' do no beha e in
an homogeneous way ega ding he Aux- o Co np ule. This may be due o he ac ha he
o me auxilia y, con a y o a e e in I alian, has comple ely los i s o igina y meaning, while
he la e s ill keeps i s loca i e alue. Hence, i could be a gued ha i is he lexical weakness o
hbe ha p ecludes i om go e ning and Case-ma king he subjec NP in ACs.
Ca alan p o ides a u he piece o e idence gi ing suppo o his conjec u e. No ice ha in his
language, which has a much mo e es ic i e use o non copula i e es a , he examples pa allel
o (25b) a e clea ly ma ginal:
(i) a. ??Es an en Joan disposa a ajuda -nos
...
es a -ge he Joan eady o helpus
Joan being eady o help us..
.
b. Es an disposa en Joan a ajuda -nos.
..
es a -ge eady he Joan o help-us

9
Raposo (1987:92), o mula es he mechanism o Case assignmen in Po uguese in lec ed
in ini i es in he ollowing way:
"In absence o
[+
Tense], In l (o Ag in In l) is capable o assigning nomina i e Case
o a lexical subjec only i i is i sel speci ied o Case".
10
We will no explo e he e whe he bo h in ini i e clauses wi h an o e subjec -see he
examples in
(4)-
and ACS
can
be gi en a uni ied accoun . In ac , he o me s uc u es di e
om he la e in e y signi ican ways which we will no a emp o in es iga e in his wo k.
Conce ning his ques ion, c . Fe nández Lagunilla (1987) and Rigau (1991), among o he s.
11
Acco ding o de Miguel(1990: 180-181),
él
can
be
analysed in (37)
as
a con as i e p onoun
which, being loca ed in a non a gumen a1 posi ion, does no need Case. No ice ha , on his
assu np ion, he e is no na u al way o explain how he
NP
Juan ge s Case in (i):
(i) Con en o Juan, con en os odos
Happy Juan, happy all
'I Juan is happy, e e ybody is happy'
12
We label his class o adjec i es as 'pe ec i e e bal adjec i es' (PVAs)
in
o de o
dis inguish hem om 'pe ec i e adjec i es' (PAS), ha is, adjec i es which, al hough no
de i ed om pa iciples, ake a pe ec i e alue. See he discussion below.
13
Needless o say, some adjec i es, being neu al o hei inhe en aspec ual alence, allow
bo h se and es a in Spanish. Thus, hey a e si ua ed be ween he wo opposi e poles o he
scale. See He nanz (1988).
14
I should
be
said ha some o he adjec i es ma keú wi h he ea u e [-pe ec i e] can appea
wi h es a in he app op ia e con ex . Since hese di e ences a e no ele an o ou analysis,
we will no pu sue he issue he e (c . He nanz (1988)).
15
See
Belle i (1990) o a somewha e ised e sion o Rizzi's (1982) analysis.
16
On he basis o he examples in (i), i should be a gued ha absolu e pas pa iciples allow
o e complemen ize s:
(i) a. Una ez leída la sen encia, el juez se e i 6
Once ead-p p he sen ence, he judge e i ed
b. Después de e ec uado el examen, 10s alumnos espi a on
A e ca ied-ou he exa n, he s uden s b ea hed
No ice, howe e , ha -as he ill o medness o he examples in (59) shows- such
a
possibili y is exclusi ely limi ed o i ems ha ing a empo al o aspec ual alue. In iew o such
da a, i may be assumed ha he pa icles in i alics in (i), a he han illing he s uc u al space
co esponding o COMP, a e a so o empo al o aspec ual in ensi ie s loca ed in
an
adjunc
posi ion.
17
Bo h al e na i es a e allowed since Spanish is a Nul1 Subjec Language.
18
Conside , in his ega d, he ollowing examples adduced in Bosque
(1990:204):
(i) a. An es de e minada la película..
Be o e o ended he ilm
'Be o e he ilm endslended.
..'
(i) b. Has a muy a anzado el e ano
...
Un il e y ad anced he summe
'Un il well in o summe ..
.'
c. Después de mue o
...
A e dead
'A e (s)he diesldied
...'
(ii) a. *An es de limpia la sala
...
Be o e clean he lounge
'Be o e he lounge islwas clean'
b. *Has a muy en enno su abuelo
...
Un i1 e y i11 his g and a he
'Un il his a he islwas qui e i11
...'
c. *Después de descalzo
...
A e ba e oo
'A e ha ing aken o hislhe shoes
...I
No ice ha he con as s be ween (i) and (ii) co ela e in a signi ican way wi h he di e en
aspec ual s a us o pa iciples ( e minada, a anzado, e c.) and pe ec i e e bal adjec i es
(limpia, descalzo, e c.), espec i ely (c . sec ion
3).
19
PPs beha ing as p edica es in ACs -c . he examples in
(50)-(51)-
aise an ob ious
p oblem o ou analysis, gi en ha he whole p ojec ion a he han P mo es in o COMP. On
he basis o some signi ican ac s whose analysis we will no pu sue he e, we could assume
some kind o inco po a ion p ocess holding be ween
P
and i s complemen NP. No ice, in his
ega d, ha no a icle is allowed o appea wi hin he
NP:
(i) a. *En las malas condiciones las conse as
...
In he bad condi ion he canned ood
(i)
b. *Bajo 10s mínimos el ae opue o..
.
Below he minimums he ai po
20
Acco ding o To ego (1983), "que -dele ionU
in
Spanish, con a y o I alian, is no es ic ed
o subjunc i e complemen s. Howe e , as his linguis no es, "gi en he s ylis ically ma ked
cha ac e o
his
cons uc ion, accep abili
y
judgmen s a e necessa i1
y
sub le".
I
should
be
said,
on he o he hand, ha he examples she gi es in which subjunc i e does no appea a e
cons uc ed, signi ican ly, ei he in u u e o in condi ional.
I
will no pu sue his issue he e.
21
Conce ning he omission o I alian complemen ize che ' ha ', see Rizzi (1982). On he ac
ha he possibili y o subjec in e sion in ol es he 'selec edlnon selec ed' dis inc ion, c . Rizzi
&
Robe s (1989).
22
C . Fa kas
(
1990:
544).
23
An in e es ing con as
in
Spanish p o iding empi ical suppo o Fa kas's app oach is
adduced in Bosque
(
1985):
(i) a. Ped o y Mm'a coinciden el uno con e2 o o
Ped o and Ma ía ag ee he one-masc wi h he o he -masc
Ped o and Ma ía ag ee wi h each o he
b. *Ped o
y
Ma ía coinciden el uno con la o a
Ped o and Ma ía ag ee he one-masc wi h he o he - em
Belle i,
A.
(1981) 'F asi ido e assolu e', Ri is a di G amma ica Gene a i a, 6,3-32.
Belle i, A. (1990) Aspec s
o
Ve b Syn ax, Ms. Uni e si é de Genk e.
Bello, A. (1874) G amú ica de la lengua cas ellana,
[Ed.
con no as de R.J. Cue o], Sopena,
A gen ina, Buenos Ai es, 1970.
Beukema, F.H. (1982) 'On he In e na1 S uc u e o F ee Adjun s', in Daalde , S.
&
M.
Ge i sen (eds.) Linguis ics in he Ne he lands
1982,
No h Holland, Ams e dam,
7
1-
82.
Bosque, I. (1985) 'Sob e las o aciones ecíp ocas en español', Re is a de la sociedad española
de lingüís ica, 15,59-96.
Bosque, I. (1990) 'Sob e el aspec o en 10s adje i os y en 10s pa icipios', in
I.
Bosque (ed.)
Tiempo y aspec o en español, Cá ed a, Mad id, 177-214.
Da idson, D. (1967) 'The Logical Fonn o Ac ion Sen ences', in D. Da idson (1980), Essays
I
I
on Ac ions and E en s, Cla endon P ess, Ox o d.
Fa kas, D. (1990) Two cases o Unde speci ica ion in Mo phology', Linguis ic Inqui y, 21,
539-550.
Fe nández Lagunilla, M. (1987) 'Los in ini i os con suje os léxicos en español', in
V.
Demon e
&
M. Fe nAndez lagunilla (eds.) Sin axis de las lenguas o nánicas,
Ed.
El
A que o, Mad id, 125- 147.
Gu ié ez, S. (1985) 'Cons ucciones a ibu i as absolu as', in Lecciones del
I
y
I1
Cu so de
Lingüís ica Funcional (1983 y 1984), Uni e sidad de O iedo, O iedo, 35-6 1.
Gu ié ez, S. (1986) Va iaciones sob e la a ibución, Uni e sidad de León, Mn.
He nanz, M.L. (1988) 'En o no a la sin axis y la semán ica de 10s complemen os p edica i os
en español', in Es udi Gene al, 8,7-29.
Higginbo ham, J. (1982) 'The Logic o Pe cep ual Repo s: an Ex ensional Al e na i e o
Si ua ion Seman ics', Cen e o Cogni i e Science, Occasional Pape , 21, MIT,
Ca nb idge,
MA.
K a ze , A. (1988) 'S age-Le el and Indi idual-Le el P edica es'.
Ms.,
Uni e si y o
Massachuse s a Amhe s .
Miguel,
E.
de (1990) El aspec o e bal en una g amá ica gene a i a del español, Unpublished
Doc . Diss., Uni e sidad Au ónoma de Mad id.
Napoli,
D.
J.
(
1989) P edica ion Theo y, C.U.P., Camb idge.

Pollock,
J.Y.
(1989) 'Ve b Mo emen , Uni e sal G amma and he S uc u e o IP', Linguis ic
Inqui y, 20,365-424.
Raposo,
E.
(1987) 'Case Theo y and In l- o-Comp: The In lec ed In ini i e in Eu opean
Po uguese', Linguis ic Inqui y, 18, 85- 109.
Reuland,
E.
J.
(1983) 'Go e ning -ing
',
Linguis ic Inqui y, 14, 101- 136.
Rigau,
G.
(
1991) 'Les cons uccions empo als d'in ini iu: una p opos a d'andisi', Ms.,
U.A.B.
Rizzi,
L.
(1982) 'Lexical Subjec s in In ini i es: Go e nmen , Case and Binding', in
L.
Rizzi
Issues in I alian Syn ax, Fo is, Do d ech , 77-1 16.
Rizzi,
L.
(1988) 'Ace ca de la uni o midad es uc u al de las ca ego ías sin ác icas', Re is a
A gen ina de Lingüís ica, 4,3-23.
Rizzi, L.
&
I. Robe s (1989) 'Complex In e sion in F ench', P obus, 1, 1-30.
S owell, T.
(
1982) The Tense o In ini i es', Linguis ic Inqui y, 13,561 -70.
S owell, T. (1983) 'Subjec s ac oss Ca ego ies', The Linguis ic Re iew, 2, 285-3 12.
S ump, G.T. (1985) The Seman ic Va iabili y o Absolu e Cons uc ions, Reidel, Do d ech .
Suñe , A. (1988) 'Suje os con p eposición', in Es udi Gene al, 8,81-112.
Suñe , A. (1990)
La
p edicación secunda ia en español, Unpublished
PhD
Diss., Uni e si a
Au bnoma de Ba celona.
Tenny, C.L. (1987) G amma icalizing Aspec and A ec edness, Doc o al Diss., MIT,
Ca nb idge,
MA.
To ego,
E.
(1983) 'Mo e E ec s o Successi e Cyclic Mo emen ', Linguis ic Inqui y, 14,3,
56
1-565.
Depa amen de Filologia Espanyola
Facul a
de
Ue es
EdiJci B
Uni e si a Au dnoma de Ba celona
E-O81
93
Bella e a
128
--