ES 25 (2003-04) - pp. 103-122
SEMANTIC CONSTRAINTS ON
THE DISTRIBUTION OF
ENGLISH HIGHER ADVERBS1
JOSÉ MARÍA GARCÍA NÚÑEZ
Uni e sidad de Cádiz
The p esen s udy ocuses on he di e en dis ibu ion o a g oup o cu en English
ad e bs and ad e b occu ences, iz highe ad e bs. Rega ding dis ibu ion, hese ad e bs
can be di ided in o monope iphe al and bipe iphe al highe ad e bs. Del ing in o he
seman ics o hese wo classes o ad e bs leads o a dis inguishing seman ic ea u e:
e en i i y. The wo k pu s o wa d a numbe o da a and a gumen s a ou ing he
hypo hesis ha only bipe iphe al highe ad e bs a e e en i e. Finally, he hypo hesis is
explo ed ha ocaliza ion, an e en - ela ed phenomenon acco ding o cu en s udies, lies
a he sou ce o he di e en dis ibu ion o monope iphe al and bipe iphe al highe
ad e bs.
1. ENGLISH HIGHER ADVERBS
Rega ding hei s uc u al seman ic p ope ies, ad e bs and ad e b
occu ences can be di ided in o hose ha can and hose ha canno ake scope o e
o he sen ence cons i uen s. The ad e b always, o example, can ake a quan i ied
NP in i s scope. Con a iwise, i is impossible o an ad e b like well o scope
ou side a quan i ie . The con as is obse able in (1)-(2), espec i ely.
(1) Ma y always g ee s e e yone.
(2) Lewis ea s e e yone well.
One in e p e a ion o (1) has i ha , on e e y ele an occasion, Ma y g ee s
he se o indi iduals deno ed by he quan i ie e e yone. On he o he hand,
wha e e he deno a ion o well in (2), i will be ue as many imes as indi iduals
a e picked ou by he quan i ie . On he p emise ha scope ela ions ha e a
con igu a ional basis (Chomsky 1977; May 1977; 1985), I will call always a highe
ad e b and well a lowe ad e b.
Highe ad e bs also in e ac wi h he scope o o he ca ego ies. They can ake
scope o e o he ad e bs, a possibili y no open o lowe ad e bs. (3)-(4) (wi h he
b acke s p o iding a ough analysis o he ele an con igu a ions) illus a e his
poin .
(3) Ma y always [g ee s e e yone wice]
(4) * Lewis [g ee s e e yone wice] well
1 An ea lie e sion o his pape was p esen ed a he 2002 LAGB Sp ing Mee ing. I am
g a e ul o pa icipan s in his con e ence o in e es ing commen s and c i icisms.
JOSÉ MARÍA GARCÍA NÚÑEZ
104
In mos o he cases, i is no he ad e b p ope , bu one o i s s uc u al
occu ences ha is ei he highe o lowe . The di e en occu ences o luckily in
(5)-(8) p o ide ele an examples.
(5) An hony luckily sa ed bo h goals.
(6) An hony sa ed bo h goals luckily.
(7) An hony luckily [sa ed a goal wice]
(8) * An hony [sa ed a goal wice] luckily.
Fo he sake o b e i y, I will hence o h use he e m highe ad e b o e e o
bo h ad e bs and ad e b occu ences.
Highe ad e bs cons i u e he objec o s udy o his wo k. Mo e p ecisely, I
will deal wi h he linguis ic con as s be ween wo ma kedly di e en classes o
highe ad e bs. In he ollowing sec ion I will ou line and exempli y hese
di e ences.
2. THE PROBLEM NEEDING ANALYSIS
The e is a e y impo an dis ibu ional di e ence among he ad e bs ha
come ou highe by he es pu o wa d abo e. Unlike lowe ad e bs, whose
dis ibu ion seems o be highly idiosync a ic,2 highe ad e bs exhibi a e y egula
dis ibu ional asymme y. Some o hem can occu bo h in he le and in he igh
pe iphe y, while o he s a e con ined o he le pe iphe y. Le us see wha his
means.
The highe ad e bs in (1)-(8) abo e a e dis ibu ionally “well beha ed”.
Nei he always no luckily can occu o he igh o he main e b (i.e. he igh
pe iphe y) wi hou ende ing he sen ence ung amma ical o aking on a lowe
eading, espec i ely.
(9) * Ma y g ee s e e yone always.
(10) An hony sa ed bo h goals luckily.
Unlike hese ad e bs, highe ad e bs like eluc an ly o wice can occu on
bo h pe iphe ies.
(11) Ralph ( eluc an ly) g ee ed e e yone wice ( eluc an ly).
(12) Ralph ( wice) g ee ed e e yone eluc an ly ( wice).
I will adop he e minological con en ion o calling he ad e bs in (9)-(10)
monope iphe al highe ad e bs (MHAs hence o h), and he ones in (11)-(12)
bipe iphe al highe ad e bs (BHAs hence o h).
2 A lowe ad e b like pe ec ly seems awkwa d in i s le -pe iphe al occu ence in (i).
(i) She pe ec ly lea n wo languages
Howe e , he VP in (ii) seems o conside ably imp o e i s le -pe iphe ali y.
(ii) She pe ec ly unde s ood he ques ion.
The dis ibu ional di e ences ha will be deal wi h in ela ion o highe ad e bs a e mo e
SEMANTIC CONSTRAINTS IN THE DISTRIBUTION OF ENGLISH HIGHER ADVERBS
105
(13) shows a possible dis ibu ional classi ica ion o highe ad e bs. The
no ional classes as well as he gene al di isional lines a e bo owed om
E ns (2000).3
(13)
MONOPERIPHERAL HIGHER ADVERBS BIPERIPHERAL HIGHER ADVERBS
Discou se-o ien ed ( ankly, hones ly)
E alua i e (luckily, eg e ably, unbelie ably)
Epis emic (p obably, ob iously, clea ly)
E iden ial ( epo edly, p esumably)
Agen -o ien ed (wisely, udely, gene ously)
Exocompa a i e (simila ly, likewise)
Domain (poli ically, his o ically, s ylis ically)
Quan i ie (always, usually)
Aspec ual (al eady, ye )
Mos du a ion (b ie ly, momen a ily)
Mos loca ion- ime (immedia ely, now, hen)
Men al a i ude ( eluc an ly, in en ionally)
F equency (o en, pe iodically, wice)
Some ques ions a ise gi en he abo e cons ella ion o ac s:
(a) why a e MHAs dis ibu ionally di e en om BHAs?
(b) a e dis ibu ional di e ences due o he exis ence o wo ad e b classes
(MHAs and BHAs) main aining deepe and mo e a - eaching
g amma ical di e ences?
(c) i so, wha a e hese di e ences and wha is hei na u e?
Ques ions (b) and (c) seem he igh place o s a ou que y, o i seems clea
ha a p ope answe o hese ques ions, i he e u ns ou o be one, should su ice o
p o ide an answe o ques ion (a).
3. MHAS AND BHAS
The li e a u e on ad e bs ha will be e iewed in he nex sec ion is almos
unanimous on he heo e ical need o se up some hing like he pa i ion in (13).
Howe e , gi en ha , as will be shown below, some o he heo e ical a gumen s pu
o wa d o suppo he pa i ion in (13) a e clea ly objec ionable, i seems
con enien o eso o heo y-neu al c i e ia o es he solidi y o he MHA/ BHA
3 Unlike E ns (2000), I make oom among MHAs o quan i ie ad e bs such as always and
usually, which clea ly canno occu in he igh pe iphe y. I also depa om E ns ’s classi ica ion
in a ing domain ad e bs among le pe iphe al ad e bs. E ns ’s decision is based on his
judgemen (E ns 1985) ha sen ences like (i) and (ii) a e synonymous.
(i) In ellec ually, hey won him o e .
(ii) They won him o e in ellec ually.
I do no belie e ha he ad e b in (ii) is iden ical o he one in (i). Fo example, unlike he
o me , he la e ails o ake scope o e nega ion.
(iii) In ellec ually, hey didn’ win him o e .
They didn’ win him o e in ellec ually.
JOSÉ MARÍA GARCÍA NÚÑEZ
106
di ision. In he es o his sec ion I will u n o e y simple ac s abou he
classi ica ion o ad e bs in o de o es he alidi y o he di ision in (13).
Ad e b no ional classes like he ones depic ed in (13) (e.g. epis emic, agen -
o ien ed, e iden ial, e c) a e se led on he basis o cooccu ence es ic ions. In
gene al e ms, ad e bs belonging in he same no ional ca ego y do no cooccu .
This is allegedly he eason why he sen ences in (14)-(16) a e ill- o med.
(13) * F ankly, hones ly, she kicked he ball i s
(14) * She ob iously will p obably kick he ball i s
(15) * She in en ionally had eluc an ly kicked he ball i s
I we del e u he in o he cooccu ence c i e ion, we see ha ad e b classes
a e no only occu en ially compa ible o incompa ible. When hey a e cooccu en ,
hey ha e o espec ce ain o de ing es ic ions. Some ad e bs ha e o occupy
mo e p ominen s uc u al posi ions han o he s in o de o hei cooccu ence o be
g amma ical. Assuming o easons ha will become clea e in sec ion 6 below ha
in ad e b clus e s he ad e b o he le is s uc u ally mo e p ominen han he one
o he igh , i ollows om he di e en accep abili y o he (a)/ (b) op ions in (17)-
(19) ha he ad e b classes in ol ed ha e di e en s uc u al p ominence.
(16) a. Ob iously, she o una ely kicked he ball i s
b. * Fo una ely, she ob iously kicked he ball i s
(17) a. She in elligen ly had always kicked he ball i s
b. * She always had in elligen ly kicked he ball i s
(18) a. She always eluc an ly kicked he ball i s .
b. * She eluc an ly always kicked he ball i s .
These s uc u al cons ain s on cooccu ence ha e led linguis s o analyse
ad e b classes as hie a chically a anged.
The hie a chical a angemen o he ad e b classes in (13) shows an
in e es ing di e ence be ween MHAs and BHAs. The di e en classes o ad e bs
alling wi hin he o me g oup main ain a s ic hie a chial ela ionship be ween
one ano he (see (17)-(18) abo e) and o BHAs (see (19)), which a e hie a chically
less p ominen han hem. On he o he hand, he hie a chical o de ing o he ad e b
classes lis ed unde he label BHAs is much less clea . The o de ings in he (a)-(b)
op ions o (20)-(21) do no seem o ha e consequences o he han he di e en
scope o he ad e bs in ol ed.
(19) a. She soon kicked he ball eluc an ly
b. She eluc an ly kicked he ball soon
(20) a. She had al eady kicked he ball i s
b. She had o en al eady kicked he ball
The upsho is ha BHAs a e occu en ially di e en om MHAs. Mo e
speci ically, he cooccu ence da a sugges s ha BHAs cons i u e a mo e
homogeneous ad e b g oup han MHAs. As poin ed ou abo e, his allows us o
seek o an explana ion o he di e en dis ibu ion o highe ad e bs a he
bounda y be ween BHAs and he es o highe ad e bs. Gi en he pa allelism
SEMANTIC CONSTRAINTS IN THE DISTRIBUTION OF ENGLISH HIGHER ADVERBS
107
be ween he cooccu ence and he dis ibu ional beha iou o highe ad e bs, we
can p o isionally assume ha dis ibu ion is bu he g amma ical mani es a ion o
deepe di e ences be ween he ad e b classes unde s udy. I also seems clea ha
only by spelling ou hese unde lying di e ences can we aim a ully unde s anding
why highe ad e bs a e dis ibu ionally di e en . This is he p og amme ha I will
ollow in he es o his wo k. The BHA/ MHA opposi ion (o some hing close o
i ) has been he ocus o a en ion o pa o he cu en li e a u e on ad e bs. Be o e
pu suing a new app oach o he opic, I will assess he explana o y capaci y o
cu en heo e ical accoun s o he BHA/ MHA opposi ion.
4. PREVIOUS ACCOUNTS AND THEIR PROBLEMS
4.1. Syn ac ically cen ed app oaches
I is common place in cu en gene a i e syn ax ha he le -pe iphe al
occu ences o ad e bs a e de i a ionally mo e basic han he igh -pe iphe al ones.
Righ -pe iphe al occu ences o highe ad e bs a e scope-ambiguous in a way
ha can be o e ly exp essed in he le pe iphe y. (22), o example, can ecei e he
in e p e a ion o bo h (23) and (24).
(21) Liz didn´ knock on he doo wice eluc an ly.
(22) Liz eluc an ly didn’ knock on he doo wice.
(23) Liz didn’ eluc an ly knock on he doo wice.
The upsho is ha hese highe ad e bs ha e di e en s uc u al occu ences,
he s uc u al a ia ion being o e ly ealised in he le -pe iphe y bu only co e ly
exp essed in he igh -pe iphe y.
These ac s ha e led syn ac icians o de end ha highe ad e bs a e base-
gene a ed in hei le -pe iphe al posi ions and su ace o he igh as a esul o
mo emen .4 The e a e wo majo and compe ing app oaches o he base-gene a ion
o ad e bs, he so-called igh -adjunc ion (And ews 1983; E ns 1984, 2002), and
le -adjunc ion (Kayne 1993; Alexiadou 1997; Cinque 1999) accoun s. Basically,
igh -adjunc ion and le -adjunc ion heo is s di e in he way hey ha e syn ac ic
cons i uen s su ace o he igh . The o me allow o a ce ain amoun o symme y
in he o e all asymme ic design o X’ s uc u e and de end ha ca ego ies
adjoining o syn ac ic p ojec ions a e allowed o symme ically b anch o he igh .
The la e a e concep ually opposed o igh -adjunc ion. On hei iew, linea iza ion
demands ha cons i uen s be s ic ly o de ed in e ms o c-command, which allows
only o s uc u es ha a e adically asymme ic. In o de o accoun o he igh -
4 Fo an al e na i e iew wi h BHAs base-gene a ed as e b complemen s, see La son (1990)
and Alexiadou (1997).

JOSÉ MARÍA GARCÍA NÚÑEZ
108
pe iphe y ac s e iewed, hey posi a gene alised cons i uen mo emen ac oss
adjunc s o highe p ojec ions.
Righ -adjunc ion ad oca es end o a ou an analysis o ad e b dis ibu ion in
e ms o cons i uency. Jackendo ’s (1972) classical analysis makes he
I(n lec ional) P(h ase)/ V(e b) P(h ase) di ide esponsible o he di e en
dis ibu ional beha iou o ad e bs. Speci ically, Jackendo a gues ha VP ad e bs
a e allowed o occu on bo h pe iphe ies whe eas IP ad e bs a e dis ibu ionally
es ic ed o he le pe iphe y. Empi ically, he analysis ails o cap u e he di ision
in (13), o some MHAs (e.g. agen -o ien ed ad e bs) ha e been shown o be base-
gene a ed in VP (F ey 2000). Theo e ically, he accoun does no p o ide a
mo i a ed answe o ques ion (a) abo e. Bo h he le -pe iphe ali y/ IPhood and he
bi-pe ipheali y/ VPhood associa ions ollow om s ipula ion.
Rega ding le -adjunc ion accoun s, he di e en dis ibu ion o he ad e bs in
(13) is held o ollow om hei di e en s uc u al p ominence. The highe in
s uc u e, he less likely an ad e b will be o le i s c-command domain be aised o
p ojec ions high abo e. As Cinque (1999) poin s ou , his is desc ip i ely
unobjec ionable in iew o he pa allelism be ween ad e b hie a chy and
dis ibu ion. Howe e , he e is, o he bes o my knowledge, no accoun o which is
he p ecise unc ional ca ego y p o iding he s uc u al di ide behind he pa i ion
in (13). Mo e impo an ly, he e is no indica ion as o why his ca ego y should lead
ad e bs abo e i o block cons i uen mo emen . In ailing o p o ide such an
explana ion le -adjunc ion accoun s also lea e ques ion (a) abo e unanswe ed.
4.2. Modi ica ion app oaches
The e is a g oup o s udies (Williams 1994, 23-24; Pe e son 1997; Shae
2000) ha , wi h mo e o less o une, ha e ied o ollow Heny’s (1973) ea ly
indica ion ha he syn ax and seman ics o highe ad e bs should be ackled in
e ms o he es ic i e/ non es ic i e opposi ion. Among hem, Wiliams (1994) is
he only one ha p o ides an explana ion o he dis ibu ional puzzle posed by
highe ad e bs. He obse es ha , as (26) shows, non es ic i e modi ie s gene ally
canno occu in he igh pe iphe y, unless sepa a ed om he es o he sen ence by
a pause.
(24) The G eeks as philosophical as Bob...
(25) The G eeks, as philosophical as Bob...
He a gues ha (27)-(28) depic s he ad e bial coun e pa o he pa e n in
(25)-(26), o , as (29) and (30) espec i ely show, he ad e b in (27) is a es ic i e
modi ie while he one in (28) is a non es ic i e one, i.e. he la e can all wi hin
he scope o nega ion while he o me canno .
(26) John le eluc an ly.
(27) John le , p obably.
SEMANTIC CONSTRAINTS IN THE DISTRIBUTION OF ENGLISH HIGHER ADVERBS
109
(28) He didn´ eluc an ly lea e.
(29) * He didn’ p obably lea e.
The idea is ha ad e bs a e inhe en ly ei he es ic i e o non es ic i e. I
hence ollows ha some ad e bs can be used ei he es ic i ely o non es ic i ely.
(30) He didn’ lea e udely.
(31) * He didn’ udely lea e quickly.
Williams’ p oposal does no i ially link up o he ac ha non es ic i e
ad e bs a e con ined o he le -pe iphe y. Howe e , gi en he ac ha he
s uc u al p inciple ex ensionally ca ies o e o he adjec i al domain, such an
explana ion may ye be hough o be wo h pu suing. The e is, ne e heless, s ong
empi ical e idence ha Williams’ es ic i e/ non es ic i e classi ica ion o
ad e bs does no eally ex ensionally co ela e wi h he dis ibu ional classi ica ion
in (13). (33)-(34) clea ly show ha e iden ial ad e bs a e adically monope iphe al.
(32) He epo edly le .
(33) * He le epo edly.
Howe e , Williams’ nega ion es dubs hese ad e bs es ic i e.
(34) He didn’ epo edly lea e.
The da a sugges s ha he dis ibu ion o ad e bs does no co ela e wi h hei
es ic i e impo , hence unde mining he basis o a possible explana ion o he
dis ibu ion o ad e bs along he lines sugges ed by Williams.
The abo e e ision shows ha he cu en ly a ailable accoun s o ad e b
syn ax and seman ics p o e incapable o cap u ing he dis ibu ional pa e n o
highe ad e bs and/o o p o ide a mo i a ed accoun o i .
In wha ollows I will pay a en ion o he s uc u al seman ic p ope ies o he
ad e bs a issue in he hope ha his will p o ide a mo e con incing answe o
ques ions (a) and (c) abo e.
5. MHAS AND BHAS: SEMANTIC DIFFERENCES
The hypo hesis ha I will explo e in he es o his wo k is seman ic in na u e
and is oo ed in he adi ional iew (Thomason and S alnake 1973; Belle 1977)
ha he ad e b hie a chy is based on di e ences in he seman ic ype selec ed by
ad e bs.
The ac ha BHAs a e hie a chically subo dina e o MHAs bu no o one
ano he sugges s ha he g amma ical homogenei y o he o me class migh be
due o a coincidence in he seman ic ype equi ed by hese ad e bs.
I is in e es ing o obse e ha he ad e b classes included in he g oup o
MHAs ha e equen ly been ela ed o di e en ypes. Thus, o example, agen -
o ien ed ad e bs can be ela ed o he exp ession o ac s, o speake -o ien ed
ad e bs o he illocu iona y con en o he u e ance. This is he eason why he
o me ad e bs a e in elici ous in non ac i e con ex s (see he di e en
JOSÉ MARÍA GARCÍA NÚÑEZ
110
accep abili y o (36) and (37)) o why he la e a e awkwa d in illocu iona y
opaque con ex s (compa e (38) o (39)).
(35) ? Ma y hinks ha John oolishly knocked on he doo .
(36) Ma y knows ha John oolishly knocked on he doo .
(37) * Ma y said ha John, ankly, knocked on he doo .
(38) F ankly, Ma y said ha John knocked on he doo .
BHAs do no demand such highe ypes, i.e. hey can eplace he ad e bs in
(36)-(39) wi hou gi ing ise o unaccep abili y. The upsho is ha whiche e he
ype selec ed by BHAs, i mus be lowe han hese wo. A possible candida e is he
p oposi ion exp essed by he u e ance. Unlike agen -o ien ed and speake -o ien ed
ad e bs, BHAs a e pa o he p oposi ional con en o u e ances. This can be
shown on he basis o In an idou-T ouki’s (1993) es o p oposi ionali y.
In an idou-T ouki (1993) sugges s ha an in e es ing es o de e mine which
ad e bs con ibu e o he p oposi ional con en o he u e ance is o embed hem in
a condi ional clause and o check whe he o no hey scope wi hin he condi ional
ope a o . I hey do, hey a e pa o he p oposi ion exp essed. Acco ding o his
es , he BHAs in (40) and (41) a e ully p oposi ional, i.e. he condi ions unde
which Ch is will be sco ned include his in en ionally o o en kissing Sandy
passiona ely, espec i ely.
(39) I Ch is in en ionally kisses Sandy passiona ely, he will be sco ned.
(40) I Ch is o en kisses Sandy passiona ely, he will be sco ned.
Con a iwise, agen -o ien ed and speake -o ien ed ad e bs a e no
p oposi ional acco ding o he condi ional es , i.e. hey a e no pa o he
condi ions ha will lead o Ch is’s being sco ned in (42) and (43).
(41) I Ch is oolishly kisses Sandy passiona ely, he will be sco ned.
(42) I Ch is, ankly, kisses Sandy passiona ely, he will be sco ned
Howe e , he p oposi ion canno be he minimal ype equi ed by BHAs, o
some MHAs a e also p oposi ional acco ding o he condi ional es . As poin ed ou
by In an idou-T ouki (1993), e iden ial ad e bs a e ully p oposi ional, i.e. hey
scope wi hin he condi ional ope a o (e.g. in (44), Ch is will be sco ned i he is
epo ed o kiss Sandy passiona ely).
(43) I Ch is epo edly kisses Sandy passiona ely, he will be sco ned.
I some MHAs a e pa o he p oposi ion, p oposi ionali y canno be he ac o
dis inguishing BHAs om MHAs. The e o e he ype modi ied by BHAs mus be
pu sued below he p oposi ional le el.
Acco ding o he cu en li e a u e on e en seman ics (Da idson 1980;
Pa sons 1990) he seman ic ype close o and lowe han he p oposi ion is he
e en . The poin o be made in he es o his pape is ha he e en 5 is he seman ic
ype equi ed by BHAs. Fo una ely, he empi ical e idence on he linguis ic
5 My use o he e m e en will be as b oad as in he Neo-Da idsonian adi ion (Pa sons
1990), i.e. i will co e ins ances o genuine e en s as well as s a es.
SEMANTIC CONSTRAINTS IN THE DISTRIBUTION OF ENGLISH HIGHER ADVERBS
111
ealisa ion o he seman ic ype e en is la ge enough o allow us o es he alidi y
o his claim. In he sec ions ha ollow, I will check he beha iou o MHAs and
BHAs in ela ion o some g amma ical phenomena gene ally assumed o be ela ed
o he exp ession o e en s.
5.1. Pe cep ual epo s
The exac syn ac ic impo o e en s is he ocus o much cu en empi ical
esea ch. An e en i e deno a ion has been aced in nominal (Pa sons 1990, 132-
135; G imshaw 1990, 47-63), unc ional (T a is 1988; K a ze 1994) and clausal
p ojec ions (Higginbo ham 1983; Vlach 1983). Especially con incing a e he
a gumen s in a ou o he e en deno a ion o he so-called naked in ini i e
complemen s o pe cep ion e bs.
As Kea ns (2000: 193-195) no es, he complemen s o he e b see in (45) and
(46) a e no o he same seman ic ype. Speci ically, he la e deno es a p oposi ion,
while he o me clea ly does no : a p oposi ion canno be physically pe cei ed.
(44) Mau ice saw Lydia alk o Ralph wice.
(45) Mau ice saw ha Lydia alked o Ralph wice.
Pos ula ing e en s as he seman ic deno a ion o naked in ini i e complemen s
o pe cep ion e bs is no only in ui i ely appealing bu also empi ically adequa e. I
has, o example, been no iced ha he complemen clauses in (45) and (46) a e no
equally anspa en o e e ence. Re e en ially iden ical de ini e desc ip ions b ing
abou di e en u h condi ions in he complemen clause o (46), bu no in ha o
(45). This can be seen by compa ing (45) and (46) wi h (47) and (46), espec i ely.
In a con ex whe e Ralph is he Dean, (45) and (47) a e synonymous, while (46)
nei he en ails no is en ailed by (48).
(46) Mau ice saw Lydia alk o he Dean.
(47) Mau ice saw ha Lydia alked o he Dean.
Higginbo ham (1983), Vlach (1983) and Pa sons (1990: 15-17) a gue ha he
di e en e e en ial anspa ency o he complemen clauses o see in (45) and (46)
is due o he ac ha he o me deno es an e en while he la e deno es a
p oposi ion.
Tu ning o highe ad e bs, i is in e es ing o obse e ha in ini i al
complemen s o pe cep ion e bs disc imina e be ween he wo classes in (13). As
(49) and (50) show, only BHAs a e accep able in hese con ex s.
(48)
Mau ice saw Lydia immedia ely/ in en ionally/ b ie ly alk o he Dean
wice.
(49)
* Mau ice saw Lydia ob iously/ epo edly/ in elligen ly alk o he Dean
wice.
I hence seems ha , unlike MHAs, BHAs a e ully compa ible wi h, and
indeed demand, he exp ession o an e en . The possibili y a ises o ela ing BHAs
JOSÉ MARÍA GARCÍA NÚÑEZ
118
6.2. Spanish –men e ad e bs
As al eady poin ed ou , igh displacemen o ad e bs is an op ional syn ac ic
ope a ion in English. This allows he ad e b o be supe icially placed in wo
di e en posi ions, one o hem ma ked o ocus, he o he a opicalised posi ion. A
di e en cons ella ion o ac s is obse ed in he case o Spanish -men e ad e bs.
These ad e bs a e like p eposi ional ph ases in equi ing igh -displacemen .10 A
he same ime, i ollows om he abo e pic u e o ac s ha monope iphe al highe
–men e ad e bs will block igh -displacemen o seman ic easons ( iz hey canno
be ocused). The combined ou come o his mo phosyn ac ic and seman ic
cons ain on linea isa ion is ha Spanish MHAs will be con ined o explici ly
opicalised posi ions in he le pe iphe y. This is bo n ou by he accep abili y
pa e n in (96)-(101).
(95) * Ca men descuidadamen e ol idó las lla es a la is a de odos
Ca men ca elessly o go he keys wi hin e e yone’s sigh
(96) * Ca men elizmen e había deposi ado las lla es en un luga segu o.
Ca men happily had le he keys in a sa e place
(97) * Ca men ol idó las lla es a la is a de odos descuidadamen e
Ca men o go he keys wi hin e e yone’s sigh ca elessly
(98) * Ca men había deposi ado las lla es en un luga segu o elizmen e
Ca men had le he keys in a sa e place happily
(99) Descuidadamen e, Ca men ol idó las lla es a la is a de odos
Ca elessly, Ca men o go he keys wi hin e e yone’s sigh
(100) Felizmen e, Ca men había deposi ado las lla es en un luga segu o
Happily, Ca men had le he keys in a sa e place
On he basis o he abo e analysis, we expec Spanish highe –men e ad e bs
o pa e n wi h English highe –ly ad e bs in con ex s whe e hei di e en
mo phosyn ax does no b ing abou a di e ence in dis ibu ion. One such con ex is
p o ided by passi e cons uc ions. As is well known, Spanish –men e ad e bs a e
allowed o eely occu in he le pe iphe y o passi e sen ences.
(101) Las p egun as habían sido (cuidadosamen e) con es adas (cuidado-
samen e).
The ques ions had been (ca e ully) answe ed (ca e ully).
Unde he ci cums ances, he heo y pu o wa d abo e p edic s ha he
seman ic cons ain s posi ed abo e will be in o ce, yielding he dis ibu ional
pa e n iden ical o he one ound in English, i.e. allowing highe eadings o –men e
MHAs when hese ad e bs occu o he le o he lexical e b and blocking hem
when hey occu in he igh pe iphe y. The p edic ion is bo n ou by ac s. Only in
10 This may be due o –men e being like a p eposi ion in mo phosyn ac ically igge ing igh -
displacemen . Fo an an isymme ic compa ison o –ly and –men e along hese lines c . Déchaine
and T emblay (1996) and e e ences he ein.

SEMANTIC CONSTRAINTS IN THE DISTRIBUTION OF ENGLISH HIGHER ADVERBS
119
(103) and (104) is i possible o impose a highe (non manne ) eading on
descuidadamen e and elizmen e, espec i ely.
(102) Las lla es habían sido descuidadamen e ol idadas a la is a de odos
The keys had been ca elessly o go en wi hin e e yone’s sigh .
(103) Las lla es habían sido elizmen e deposi adas en un luga segu o
The keys had been happily le in a sa e place
(104) ?? Las lla es habían sido ol idadas a la is a de odos descuidadamen e.
The keys had been o go en wi hin e e yone’s sigh ca elessly.
(105) ?? Las lla es habían sido deposi adas en un luga segu o elizmen e.
The keys had been le in a sa e place happily.
7. CONCLUDING REMARKS
This s udy has a gued o he g amma ical plausibili y o he class o BHAs. I
ha e pu o wa d e idence conce ning pe cep ual epo s, ocusabili y and
quan i ie ad e bs ha wha dis inguishes BHAs om he es o highe ad e bs is
he ac ha hey p edica e o he e en a iable. I has also been shown ha , once
he connec ions be ween e en i i y and ocusabili y a e p ope ly spel ou , he
special dis ibu ion o BHAs ceases o be a mys e y. The bipe iphe al occu ence o
BHAs ollows s aigh o wa dly om hei e en i i y and om he economy o
igh -displacemen . Likewise, i has been shown ha he ac ha he es o highe
ad e bs a e no ela ed o he seman ic ype e en is he eason o hei no being
igh -displaceable.
Looked a om a heo e ical pe spec i e, he p oposed seman ic heo y has he
me i o p o iding a mo i a ed accoun o ac s ha emain obscu e unde cu en
syn ac ically o modi ica ion o ien ed app oaches. Ob iously, he possibili y exis s
ha he seman ic ype e en syn ac ically co ela es wi h some o o he unc ional
p ojec ion. This issue is he ocus o much cu en esea ch (K a ze 1994,
Pus ejo sky and Tenny 2000). Howe e , e en hough his u ned ou o be he case,
a s ic ly syn ac ic accoun o he dis ibu ion o highe ad e bs would s ill be
expec ed o o e he kind o i ial connec ion be ween ad e b ypes and
linea isa ion which he analyses p esen ed abo e p o ides.
In a b oade heo e ical pe spec i e, he wo k bea s ou some ideas and
p oposals in cu en syn ac ic and seman ic li e a u e. On he one hand, he analysis
clea ly suppo s he widesp ead posi ion ha igh -pe iphe al occu ences o
ad e bs a e de i ed om le -pe iphe al ones and ha , in he absence o
mo phosyn ac ic igge s o igh -displacemen , his mo emen ope a ion is
seman ically a he han syn ac ically cons ained. On he o he hand, he empi ical
esul s o his wo k suppo Da idson’s claim ha e en s a e a undamen al
cons i uen o he seman ics o na u al language. Mo e speci ically, he da a and
a gumen s handled abo e suppo He bu ge ’s (2000) posi ion ha he seman ic
JOSÉ MARÍA GARCÍA NÚÑEZ
120
impo o ocus is o a ange he in e nal s uc u e o he e en quan i ica ion
exp essed by he sen ence.
REFERENCES
Alexiadou, A emis. Ad e b Placemen : A Case S udy in An isymme ic Syn ax. Ams e dam: John
Benjamins, 1997.
And ews, Alexande . “A No e on he Cons i uen S uc u e o Modi ie s.” Linguis ic Inqui y 14 (1983):
695-697.
A las, Jay. “Topic-commen , P esupposi ion, Logical Fo m, and Focus S ess Implica u es: he Case o
Focal Pa icles Only and Also.” Jou nal o Seman ics 8 (1991): 127-147.
Belle , I ena. “On he Seman ic and Dis ibu ional P ope ies o Sen en ial Ad e bs.” Linguis ic
Inqui y 7 (1977): 337-351.
Chomsky, Noam. “Deep S uc u e, Su ace S uc u e and Seman ic In e p e a ion.” Seman ics: an
In e disciplina y Reade in Philosophy, Linguis ics and Psychology. Eds. Danny S einbe g and
Leon Jakobo i s. Camb idge: Camb idge Uni e si y P ess, 1971. 183-216.
Chomsky, Noam. Essays on Fo m and In e p e a ion. New Yo k: No h-Holland, 1977.
Cinque, Guglielmo. “A Null Theo y o Ph ase and Compound S ess.” Linguis ic Inqui y 24 (1993):
239-297.
Cinque, Guglielmo. Ad e bs and Func ional Heads. New Yo k: OUP, 1999.
Da idson, Donald. Essays on Ac ions and E en s. Ox o d: Cla endom, 1980.
Déchaine Rose-Ma ie, and Mi eille T emblay “Ad e bial PPs and P eposi ional Ad e bs in F ench and
English.” Uni e si y o Calga y Wo king Pape s in Linguis ics 7 (1996).
Emonds, Joseph. A Uni ied Theo y o Syn ac ic Ca ego ies. Do d ech : Fo is, 1985.
E ns , Thomas. Towa ds and In eg a ed Theo y o Ad e b Posi ion in English. Blooming on: Indiana
Uni e si y Linguis ic Club, 1984.
E ns , Thomas. “On he Odd Syn ax o Domain Ad e bs.” Language Resea ch 21/ 2 (1985): 167-177.
E ns , Thomas. “M-command and P ecedence.” Linguis ic Inqui y 25 (1994): 327-334.
E ns , Thomas. “Seman ic Fea u es and he Dis ibu ion o Ad e bs.” ZAS Pape s in Linguis ics
17 (2000): 79-99.
E ns , Thomas. The Syn ax o Adjunc s. Camb idge: Camb idge Uni e si y P ess, 2002
F ey, We ne . “Syn ac ic Requi emen s on Adjunc s.” ZAS Pape s in Linguis ics 17 (2000): 107-135.
G imshaw, Jane. A gumen S uc u e. Camb idge, Mass.: MIT P ess, 1990
Heim, I ene. The Seman ics o De ini e and Inde ini e Noun Ph ases. PhD hesis, Uni e si y o
Massachuse s a Amhe s , 1982
Heny, F ank. “Sen ence and P edica e Modi ie s.” Syn ax and Seman ics Vol. 2. Ed. John Kimbal, New
Yo k and London: Semina P ess, 1973. 217-245.
He bu ge , Elena. Wha Coun s: Focus and Quan i ica ion. Camb idge, Mass.: MIT P ess, 2000.
Higginbo ham, James. “The Logic o Pe cep ual Repo s: an Ex ensional Al e na i e o Si ua ion
Seman ics.” Jou nal o Philosophy 80 (1983): 100-127.
In an idou-T ouki, Ellie. “Sen en ial Ad e bs and Rele ance.” Lingua 90 (1993): 69-90.
Jackendo , Ray. Seman ic In e p e a ion in Gene a i e G amma . Camb idge, Mass.: MIT P ess,
1972.
Kayne, Richa d. The An isymme y o Syn ax. Camb idge, Mass.: MIT P ess, 1993.
SEMANTIC CONSTRAINTS IN THE DISTRIBUTION OF ENGLISH HIGHER ADVERBS
121
Ka unen, Lau ie, and S anley Pe e s “Con en ional Implica u e.” Syn ax and Seman ics 11. Eds.
Choon-Kyu Oh, and Da id Dinneen, New Yo k: Academic P ess, 1979. 1-56.
Kea ns, Ka e. Seman ics. New Yo k: S . Ma in’s P ess, 2000.
K a ze , Angelika. “On Ex e nal A gumen s.” Uni e si y o Massachuse s Occasional Pape s in
Linguis ics 17 (1994): 103-130.
La son, Richa d. “Double Objec s Re isi ed: Reply o Jackendo .” Linguis ic Inqui y 21 (1990): 589-
632
Lewis, Da id. “Ad e bs o Quan i ica ion.” Fo mal Seman ics o Na u al Language. Ed. Edwa d
Keenan, Camb idge: Camb idge Uni e si y P ess, 1975. 3-15.
May, Robe . The G amma o Quan i ica ion. MIT, PhD hesis, 1977.
May, Robe . Logical Fo m: i s S uc u e and De i a ion. Camb idge, Mass.: MIT P ess, 1985.
Pa sons, Te ence. E en s in he Seman ics o English. Camb idge, Mass: MIT P ess, 1990.
Pe e son, Paul. Fac s, P oposi ions, E en s. Do d ech : Kluwe , 1997.
Pus ejo sky, James, and Ca ol Tenny, eds. E en s as G amma ical Objec s. Ox o d and New Yo k:
OUP, 2000.
Qui k, Randolph, Jan S a ik, Geo ey Leech, and Sidney G eenbaum. A Comp ehensi e G amma
o he English Language. London: Longman, 1985.
Reinha , Tanya. “In e ace S a egies.” OTS, U ech Uni e si y, 1995.
Rizzi, Luigi. “The Fine s uc u e o he Le Pe iphe y.” Elemen s o G amma . Ed. Liliane Haegeman
Do d ech : Kluwe , 1997. 281-337.
Roo h, Ma s. “Focus.” The Handbook o Con empo a y Seman ic Theo y. Ed. Shalom Lappin, Ox o d:
Blackwell, 1994. 271-298.
Shae , Benjamin. “Syn ac ic Posi ion and he Readings o Manne Ad e bs.” ZAS Pape s in
Linguis ics 17 (2000): 265-287.
Thomason Richmond, and Robe S alnake . “A Seman ic Analysis o Ad e bs.” Linguis ic Inqui y 4
(1973): 195-220.
T a is, Lisa. “The Syn ax o Ad e bs.” McGill Wo king Pape s in Linguis ics. Special Issue on
Compa a i e Ge manic Syn ax (1988): 280-310.
Vlach, F ank. “On Si ua ion Seman ics o Pe cep ion.” Syn hese 54 (1983): 129-152.
Williams, Edwin. Thema ic S uc u e in Syn ax. Camb idge, Mass.: MIT P ess, 1994.
Wyne , Adam. Boolean E en La ices and Thema ic Roles in he Syn ax and Seman ics o Ad e bial
Modi ica ion. Ph.D. hesis, Co nell Uni e si y, 1994.