Abs ac
As Kayne (1994) has shown, he heo y o an isymme y o syn ax also p o ides an explana ion
o a s uc u al p ope y o mo phological complexes, he Righ hand Head Rule. In his pape we
show ha an an isymme y app oach o he Righ hand Head Rule e en ually is o be p e e ed
on empi ical g ounds, because i desc ibes and explains he p ope ies o a se o hi he o puzz-
ling mo phological p ocesses —known as discon inuous a ixa ion, ci cum ixa ion o pa asyn-
hesis. In conside ing hese and a numbe o mo e s anda d mo phological s uc u es, we a gue ha
one di e ence bea ing on he p ope balance be ween mo phology and syn ax should be e-ins-
alled ( e- wi h espec o Kayne), a di e ence be ween he an isymme y o he syn ax o mo -
phology and he an isymme y o he syn ax o syn ax p ope .
Key wo ds: an isymme y, Righ hand Head Rule, ci cum ixa ion, pa asyn hesis, p e ixa ion,
ca ego y-changing p e ixa ion, discon inui ies in mo phology.
Resum. L’an isime ia i el cos a esque e en mo ologia
Com Kayne (1994) mos a, la eo ia de l’an isime ia en la sin axi ambé ens dóna una explicació
d’una p opie a es uc u al de complexos mo ològics, la Regla del Nucli a la D e a. En aques
a icle mos em que un ac amen an isimè ic de la Regla del Nucli a la D e a es p e e eix e en-
ualmen en dominis empí ics, pe què desc iu i explica les p opie a s d’una sè ie de p ocessos
ins a a mo ològics —conegu s com a ixació discon ínua, ci cum ixació o pa asín esi. Conside an
aques es i al es es uc u es mo ològiques més es ànda ds, p oposem que una di e ència que é
a eu e amb l’equilib i p opi en e mo ologia i sin axi s’hau ia de ep end e ( e- espec e a
Kayne), una di e ència en e la an isime ia de la sin axi de la mo ologia i la an isime ia de la sin-
axi de la p òpia sin axi.
Pa aules clau: an isime ia, Regla de Nucli a la D e a, ci cum ixació, pa asín esi, p e ixació,
p e ixació amb can i de ca ego ia, discon inuï a s en mo ologia.
*The esea ch epo ed he e is pa o he esea ch p og am de ined and inanced by he U ech
ins i u e o Linguis ics-OTS. I wish o single ou Pe e Ackema, Hans den Bes en, Reineke Bok, Hans
B oekhuis and Jan Don o inspi ing discussion: wi hou hem, sec ion 4 would no ha e exis ed.
Ca WPL 7, 1999 71-87
An isymme y And The Le hand In Mo phology*
F ank D ijkoningen
U ech Ins i u e o Linguis ics-OTS. Depa men o Fo eign Languages
K omme Nieuweg ach 29. 3512 HD U ech . The Ne he lands
[email p o ec ed]
Recei ed: Decembe 13 h 1998
Accep ed: Ma ch 17 h 1999
Ca WPL 7 071-087 13/6/00 12:26 Página 71
1. In oduc ion
This pape is inspi ed by Kayne’s wo k on he an isymme y o syn ax (Kayne
1994), and mo e in pa icula he sec ion in which he claims ha his syn ac ic
heo y de i es he o de ing ac s ha a e o dina ily accoun ed o by he mo pho-
logical Righ hand Head Rule. In his pape we look in o a numbe o s anda d
mo phological s uc u es which a e gene ally discussed by mo phologis s wo king
wi h he Righ hand Head Rule. The basic ques ion is whe he he e a e empi ical
ac s ha migh a ou o dis a ou he an isymme y app oach. We ha e pu
heemphasis on he empi ical side because —a e all— Kayne shows ha he
Righ hand Head Rule (hence o h RHR) can be subsumed unde he an isymme ic
p inciples, bu he does no ully elabo a e he mo phological issues conce ning
igh hand / le hand membe s o mo phological s uc u es.
In he i s sec ion we will discuss wo classical o mula ions o he RHR,
and we show ha an isymme y canno deal di ec ly wi h mos o he adi ional
le hand membe s o a mo phologically complex wo d. The conclusion o his
sec ion is ha he syn ac ic an isymme y app oach is —un ill u he no ice—
empi ically in e io o he mo phological app oach. In he second sec ion we will
discuss one pa icula g oup o mo phological a ixes, discon inuous a ixes.
They canno be explained by he RHR, while we show ha he discon inui y can
be explained by he an isymme y app oach. The conclusion o his sec ion is
ha he syn ac ic an isymme y app oach is empi ically supe io o he mo pho-
logical RHR app oach. In elabo a ing he heo y o hese pa icula discon i-
nuous a ixes, we encoun e ed a numbe o p oblems, bo h o a concep ual and a
echnical na u e. Sec ion 4 discusses hese in a b ie way; e en ually he solu-
ions p oposed in his sec ion bea on all p ope ies o mo phology. In sec ion 5,
inally, we show ha he p oposals made o independen easons in sec ion 4
also allow us o explain he p oblema ic cases o he i s sec ion in a ela i ely
s aigh o wa d way. The g and o al o all his will be ha he an isymme y
app oach is empi ically supe io o he mo phological RHR, bu on he o he
hand also ha he mo phological RHR is s ongly con i med as a p ope y ele-
an a a su ace-le el —as a p ope y o he s ing passed o he phonological
componen .
72 Ca WPL 7, 1999 F ank D ijkoningen
Table o Con en s
Ca WPL 7 071-087 13/6/00 12:26 Página 72
1. In oduc ion
2. Two classical o mula ions o he
Righ hand Head Rule
3. Ci cum ixa ion
4. P oblems and solu ions
5. P e ixa ion
6. Conclusion
Re e ences
2. Two classical o mula ions o he Righ hand Head Rule
2.1. Williams’ RHR
The i s o maliza ion o he (ancien ) idea ha he su ixal pa o mo phologi-
cally complex wo ds is he elemen ha de e mines he ca ego y o he wo d can be
ound in Williams (1981), c . (1). Two examples a e gi en in (2).
(1) In mo phology, we de ine he head o a mo phologically complex wo d o be
he igh hand membe o ha wo d.
The ule in (1) is pa o he o malism, in he sense ha i de e mines he label
o he op node, in a «bo om-up» manne . In mo e ecen e minology we would say
ha , du ing he me ge o wo mo phological cons i uen s, he igh hand cons i-
uen is he cons i uen ha is de e mining he label o he cons i uen c ea ed by
me ge.
On he empi ical le el, Williams (1981) acknowledges he exis ence o «sys-
ema ic excep ions», like ennoble and enla ge — he mo e adi ional class o «ca e-
go y-changing p e ixa ion». I is clea ha hese o m a a he undamen al p oblem
o he RHR: hey also exis in he Romance languages, and in Ge manic langua-
ges hey a e abundan .
On he empi ical le el, i is also well-known ha English and Ge manic com-
pounding is di ec ly accoun ed o . Compounding in Romance, howe e , p esen s
ano he class o a he undamen al p oblems. Examples a e imb e-pos e (li .
S amp-pos , pos -s amp) o essuie-glace (li . Whipe-windshield, windshield-whi-
pe ).
Finally, he ule is limi ed o p e ixa ion and su ixa ion, bu does no deal wi h
he adi ional cases o «in ixa ion» o «ci cum ixa ion». We will e u n o ci -
cum ixa ion below —i is also known as «pa asyn hesis».
2.2. Selki k’s RHR
The c i icisms di ec ed agains Williams (1981) led Selki k (1982) o he o mu-
la ion o he Righhand Head Rule gi en in (3).
An isymme y And The Le hand In Mo phology Ca WPL 7, 1999 73
(2) a. A b. A
V Asu --p e A
ead able im p ope
(3) In a wo d-in e nal con igu a ion:
Xn
P XmQ
whe e X s ands o a syn ac ic ea u e complex and whe e Q con ains no ca ego y
wi h he ea u e X, Xm is he head o Xn.
Ca WPL 7 071-087 13/6/00 12:26 Página 73
The examples gi en abo e in (2) a e ea ed in he ollowing way. I he op-
ca ego y is A in (2), hen he head o his cons i uen is he igh mos elemen wi h
he ca ego y A, hence he su ix (in 2a), o he wo d (in 2b). This ule is o mally
di e en om he one p oposed by Williams. I wo ks in andem wi h an X-ba
ule sys em, which wo ks in a « op-down» manne . E en ually, i is his op-down
p ope y ha enables Selki k o accoun o ca ego y-changing p e ixa ion and a
numbe o he Romance compounds, as illus a ed in (4).
Wi h espec o (4a), i he op node is known (V), hen i is necessa y o
X-ba heo y ha he e be a e bal head, and i is ue ha his e bal head ( he
p e ix) is he igh mos elemen ha has a ca ego y ma ching he ca ego y o
he op node. In essence, ca ego y-changing p e ixa ion is sol ed, bu he o malism
looks e y ich (i le s in e ene bo h X-ba heo y and RHR). Wi h espec o
(4b), his example is sol ed i one akes only in o conside a ion he ca ego y.
Howe e , i does no explain he mo e in e p e a i e p ope ies o he compound
( he windshield emains he objec a an in e p e a ional le el). In addi ion,
Selki k’s app oach does no sol e all Romance compounds: i he ca ego ies o
he wo membe s a e iden ical, as in he case o imb e-pos e, he p edic ion is
w ong (pos e would be he head acco ding o (3)). Hence, Romance compoun-
ding is no eally sol ed.
2.1.3. Di Sciullo & Williams’ Rela i ized RHR
A e hese wo classical o mula ions o he RHR, Di Sciullo & Williams (1987)
de ined ano he one, a Rela i ized RHR. In hei analysis, he mo phological o de
in (4b) e lec s he syn ac ic wo d o de ; bu hey do no add ess examples o he ype
in (4a). We will no u he discuss his ule he e. The eason o he absence o
discussion he e is ha he no ion o « ela i e» head in mo phology is incompa ible
wi h syn ac ic s uc u es in he i s place (syn ax in he GB sense, he si ua ion
could be di e en in a HPSG amewo k). In o he wo ds, « ela i eness o heads»
p esen s an issue ha goes beyond p ope ies o linea o de .
2.4. Kayne’s app oach o he RHR
Kayne (1994) claims ha he an isymme y app oach ha he de eloped o ully
independen syn ac ic s uc u es subsumes he RHR. As an illus a ion he gi es
he s uc u e in (5).
74 Ca WPL 7, 1999 F ank D ijkoningen
(4) a. V b. N
Vp e A V N
en noble essuie glace
whipe( ) windshield
Ca WPL 7 071-087 13/6/00 12:26 Página 74
The s ing can only be linea ized p ope ly (in he o mal in e p e a ion) i he
esul o mo ing J o K esul s in he label K o he mo phological complex. I is
his Linea Co espondence Axiom ul ima ely ha assu es (5). On a less abs ac
le el, he LCA assu es (6a). In p ac ice, wo ds like he one in (6b) de i e om a
deepe sou ce wi h wo independen heads.
(6) a. [H1P H1 [H2P H2]] => [H1P H2i+H1 [H2P i]]
b. mode n+ize <= [ ize [ mode n ]]
The claim is a e y in e es ing one, in he sense ha he qui e gene al obse -
a ion ha all syn ac ic mo emen is upwa d-le wa d, is di ec ly ele an o he
igh hand headedness o s uc u es making use o Head-mo emen . I canno be
acciden al ha le wa d-upwa d mo emen o Heads gene a es o de s ha mo -
phologis s end o explain by he RHR.
Le us i s econside he cases ha Williams (1981) claimed o be «sys ema-
ic excep ions» —ca ego y-changing p e ixa ion. Basically, he e a e wo possibi-
li ies, each illus a ed in (7).
(7a) gi es he co ec ca ego y o he p e ixed e b: i is a e b de i ed om an
adjec i e. Howe e , Kayne’s o malisa ions wo k in a way in which only le -
adjunc ion is allowed. Righ -adjunc ion is ba ed, and he e o e he co ec o de
en+noble canno be de i ed. Con e sely, i one eso s o he pe mi ed le -adjunc-
ion, (7a) su aces as he ung amma ical noble+en. In o he wo ds, his class o
e bs canno be de i ed by using (7a). (7b) on he o he hand can gi e he igh
o de o elemen s h ough le -adjunc ion (en+noble); bu (7b) is no he co ec
ep esen a ion because he ca ego y o he whole is no co ec . All in all, ca e-
go y-changing p e ixa ion canno be deal wi h in his ashion.
In addi ion, ques ions may be aised wi h espec o all p e ixes. One o he
a ac i e p ope ies o Williams’ ule was ha i explained why i is impossible o
de e mine he ca ego y o a p e ix — he p e ix is always o he le o he head.
Conside hen (8).
An isymme y And The Le hand In Mo phology Ca WPL 7, 1999 75
(5) Q
K/*R S
J K T
o e u n
(7) a. VP b. AP
V AP A VP
en noble
A V
noble en
Ca WPL 7 071-087 13/6/00 12:26 Página 75

(8a) illus a es he same p oblem as (7a). Mo ing p ope o he p e ix in he
allowed way —le -adjunc ion— gi es he w ong linea o de : * p ope +im. Righ -
adjunc ion is no allowed, so he o de ing p oblem emains. In addi ion, a decision
has o be aken as o he ca ego y o he p e ix; pe haps Neg o his case (bu his
is qui e oublesome: imp ope does no seem o unc ion in syn ax as a NegP; i
simply seems o unc ion as an adjec i e). In ac , (8b) a es be e : le -adjunc ion
de i es he co ec o de , and he ca ego y o he complex wo d is co ec . Nex o
a decision o ake o he ca ego y o he p e ix, one would be o ced o conclude
ha p e ixes a e in ac he o iginal head o he complemen — ha a NegP is nei he
an adjunc , no a speci ie , bu a complemen . This has a coun e -in ui i e la ou
we will no elabo a e he e (e.g. he scope o nega ion). And he coun e -in ui i e
la ou speci ically holds o his class o p e ixes, no o all cases (i can be a ia-
ble solu ion o a numbe o cases (see las sec ion o his pape ).
As a conclusion o his sec ion, Kayne’s an isymme y explana ion o he RHR
is a he e y bes empi ically equi alen o Williams’ o mula ion o his ule. I is
subjec o he same ype o coun e -examples, and looks less alid i he o dina y
p e ixes a e no ea ed mo e ex ensi ely. In sec ion 5, we will e u n o he p e ixes.
Fi s , in sec ion 3, we will show ha he an isymme y app oach is capable o
ea ing a class o examples which has been beyond he scope o he RHR o begin
wi h.
3. Ci cum ixa ion
In a ph ase s uc u e ule sys em, ci cum ixa ion (also called «pa asyn hesis») has
been desc ibed as a mo phological ins an ia ion o he gene al schema o (9).
(9) A -> B A C
One o he clea es examples is Pas Pa iciple Fo ma ion in Du ch and Ge man.
Ini ially, his mo phological ule simply looks like (10). One akes a e b, adds a p e-
ix and simul aneously also a su ix.
(10) Pas Pa iciple Fo ma ion: V -> ge + V +
ex. pak -> ge + pak +
I is obliga o y o add bo h a ixes, as illus a ed in (11). This equi emen can
be o mula ed in a gene alized way, as in (12).
(11) ge+huil+d *ge+huil *huil+d (‘c ied’)
ge+we k+ *ge+we k *we k+ (‘wo ked’)
(12) B A C * B A *A C
76 Ca WPL 7, 1999 F ank D ijkoningen
(8) a. ?P b. AP
? AP A ?P
im A p ope ?
p ope im
Ca WPL 7 071-087 13/6/00 12:26 Página 76
The subjec has been commen ed as ollows, by Spence (1991).
(13) Some mo phologis s ega d such due s as a special kind o discon inous
a ix, a ci cum ix. Many linguis s would a gue ha all cases o alleged ci -
cum ixa ion can be educed o su ixa ion and concomi an p e ixa ion.
Wi h espec o hese commen s, which end o deny he ele ance and he me e
exis ence o ci cum ixa ion, i is ele an o no e wo hings. Al hough he in o -
mal desc ip ion «concomi an » is close o he in ui ion o he na i e speake s, i is
no o malized. Tha is, al hough one migh ag ee on he ac ha su ixa ion and
p e ixa ion a e concomi an in ci cum ixa ion, mo phological heo y ne e has
gi en a speci ic o malisa ion o he ela ion be ween he wo a ixes ha appa-
en ly exis s. In ha sense, «many linguis s» should y o o malize he e y no ion
o «concomi an ». We will a gue in wha ollows ha «concomi an » can be in e -
p e ed as a Spec-head ela ion.
In hese commen s i is also a he clea ha he p oblem posed by ci cum i-
xa ion s ems la gely om he discon inui y o he p ocess. I is known ha mo -
phological s uc u es canno deal wi h discon inui ies. Bu —c ucially— syn ac ic
heo y can deal wi h discon inui ies; in ac , one migh say ha la ge pa s o mo e-
men heo y a e mo i a ed by (linea ly) discon inuous ela ions in s uc u es. A
his poin , one o he c ucial p ope ies o Kayne’s an isymme y app oach is impo -
an : i is based on mo emen . Hence, de i ing he RHR om mo emen heo y
opens he oad o he explana ion o mo phological discon inui ies. This is in ac
wha we a gue o he e.
Pas pa iciple o ma ion in Du ch and Ge man has been he opic o discussion
in syn ax, in phonology and in mo phology. Al hough no all p oblems ha e been
sol ed, he syn ac ic p ope ies o pas pa iciples a e clea in isola ion; he same
holds o he phonological p ope ies o pas pa iciples. I is he mo phology (and
he posi ion o he mo phological componen ) ha is subjec o discussion.
In syn ax, pas pa iciples a e conside ed o consis o one o m, bu wi h se e-
al in e p e a ions. In his espec , conside (14).
(14) a. Hij hee he boek gelezen (pe ec i e pa iciple)
He has ead he book
b. He boek wo d gelezen (passi e pa iciple)
The book is being ead
c. He boek is aangekomen (pe ec i e pa iciple e ga i e)
The book has a i ed
d. He doo iede een geci ee de we k (pe ec i e pa iciple passi e/adjec i al)
The [by e e yone ci ed] wo k
Cu en syn ac ic heo y pos ula es a unc ional head PART (o P.M.) —inde-
penden ly o issues o checking s. building. In mos analyses, an addi ional AGR
ela ionship is pos ula ed —independen ly o he issue o whe he his AGR is an
An isymme y And The Le hand In Mo phology Ca WPL 7, 1999 77
Ca WPL 7 071-087 13/6/00 12:26 Página 77
independen head o no . In gene al, he in e p e a i e di e ences a e o mula ed in
e ms o abso p ion o case / he a ea u es —no abso p ion in pe ec i es, he
Bu zio-gene alisa ion o he o he s. The e is also discussion abou he ca ego ial
s a us, V o A, which we do no engage upon he e ( o sake o conc e eness, assu-
me A o PART, licensing an E en - a iable in he sense o G imshaw (1990) o
E en -nominaliza ions, see also D ijkoningen (1997)).
In phonology, on he o he hand, PART is nei he one head no one elemen .
The e a e simply wo a ixes. As a as (mo pho-)phonology p ope is in ol ed,
he e is no e idence wha soe e o he link be ween he a ixes. All ules unc ion
p ope ly by aking he a ixes in isola ion. In his espec , conside (15).
(15) a. De gewe k e (*gewe kde) dagen. (‘ he wo ked days’)
He e elde (* e el e) e haal. (‘ he old s o y’)
b. Hij hee he hou bewe k (*ge+be+we k+ )
He has he wood be-wo k-ed (GE+BE+wo k+D)
= He has wo ked on he wood
Fo (15a) one needs a phonological ule ha deals wi h he choice be ween /d/
and / /. This ule is dependen on he oice o he inal consonan o he e bal oo ,
bu igno es he p e ix. In ac , he ule gene alizes wi h he oice-al e na ion ound
wi h he su ixes o pas ense (wi hou p e ix in he i s place). In (15b) he e is
a phonological ule ha may dele e he p e ix pa o he pas pa iciple; his ule
igno es he su ix, and only akes in o conside a ion he o he p e ix in he imme-
dia e en i onmen .
To sum up, syn ax unc ions p ope ly while ha ing one unc ional head
(al hough he e a e wo a ixes), while phonology also unc ions p ope ly while
ha ing wo independen a ixes (al hough hey ela e o one head). In o he wo ds,
pas pa iciple o ma ion is a ypical mo phological p oblem —o an in e ace
p oblem.
Sho ly abo e we no ed wo impo an p oblema ic aspec s o ci cum ixa ion
— he «concomi an » na u e o he a ixa ion p ocess and he «discon inui y» o
he a ixes. Ano he p oblema ic aspec can be added o hese. In gene al, i has
been impossible o d aw a s uc u e o he s anda d ype. Tha is, one migh p o-
pose [[ge+V]+ ], bu his one canno be co ec because [ge+V] is no an indepen-
den cons i uen . Al e na i ely one migh p opose [ge+[V+ ]], bu his one also
canno be co ec o he same eason, [V+ ] is no an independen cons i uen
ei he . Should one deny he ele ance o he c i e ion — he cons i uen hood o he
pa — hen he p oblem comes back in aking a decision as o which one o he wo
should be co ec . The cause is simple: Bo h subpa s in isola ion a e ung amma-
ical —as shown in (11). As a consequence o his, one migh p opose e na y b an-
ching. Bu e na y b anching is no longe s anda d, i.e. heo e ically disallowed.
Hence, he e is no enable s uc u e a ailable o pas pa iciples.
Now, i he e is no s uc u e a ailable, pas pa iciples migh illus a e a ypical
in e ace si ua ion. This ack has been de ended by Don (1993). He gi es he ollo-
wing pic u e.
78 Ca WPL 7, 1999 F ank D ijkoningen
Ca WPL 7 071-087 13/6/00 12:26 Página 78
As a as syn ax goes, he e is no p oblem, he e is one unc ional head. As a
as phonology goes, he e is no p oblem ei he , he e a e wo independen a ixes. The
p oblem is he linking be ween he wo componen s. Don’s p oposal illus a es
c ossing associa ion lines. Fo a heo y o his ype, i is wo hwhile o no e ha c os-
sing associa ion lines a e no a es ed elsewhe e and p obably only occu wi h
ci cum ixa ion. Again, he solu ion does no look p incipled.
We conclude by saying ha al hough phonology and syn ax unc ion p ope ly
in hei own independen way, he e is no p incipled uely mo phological analysis
o ci cum ixa ion.
As we ha e indica ed abo e, de i ing and explaining mo phological o de ing
by linea iza ion axioms includes he possibili y o mo emen —hence, ha mo e-
men migh explain discon inui ies in mo phology. As we ha e indica ed also abo e,
using syn ac ic ela ionships like Spec-Head Ag eemen in mo phology migh
explain he «concomi an » na u e o he ci cum ixa ion p ocess. This gi es us he
ollowing p oposal —which is subjec o addi ional commen s in sec ion 4.
As discon inui y is explained by mo emen , he e a e necessa ily wo ep e-
sen a ions. Mo emen alls wi hin he bounda ies se by he heo y (Head-Mo emen
Cons ain , and o an isymme y: le -adjunc ion). Fo syn ax, he one unc ional
p ojec ion ha is needed is p esen ; i has he complemen i no mally akes in syn-
ax. Fo phonology, he e a e wo a ixal pieces, as desi ed. The ela ionship be ween
he wo pieces is encoded by Spec-Head Ag eemen , by using he Speci ie posi-
ion o PART. This means ha we o malize he in ui ion «concomi an » as a syn-
ac ic ela ion be ween wo mo phological pieces, no a phonological one.
As a esul , he wo mos impo an p oblems dissol e. In addi ion, he addi-
ional mo phological p oblem also dissol es. The sugges ion o e na y b anching
need no longe be en e ained; he s uc u e is bina y. Wha led o he sugges ion o
e na y b anching ( he p oblem o he cons i uen hood o he pa s) also has a clea
answe : a su ace-le el he s uc u e is [ge+[V+ ]]; [[ge+V]+ ] is excluded.
An isymme y And The Le hand In Mo phology Ca WPL 7, 1999 79
(16) Mo pho-syn ax V-STEM PART
Mo pho-phonology P e V-STEM Su
(17) P oposal ( i s e sion)
PART” →PART”
SpecjPART’ SpecjPART’
p e ix p e ix
PARTjVP PARTjVP
su ix V V PART V
huil huilisu ix i
Ca WPL 7 071-087 13/6/00 12:26 Página 79
complemen . Mo e in pa icula , an upda e o he analysis by Hoeks a e al. (1987)
gi es he possibili y in (27a), o he example in (27b).
(27) a. [ Xp e + Y [ x]]
b. da hij [AGROP de wegi[bejwandel ] [SC i j]]
ha he he oads BE +walks
‘ ha he walks he oad’
Second, he di e ence in compounding be ween Romance and English /
Ge manic ha emained p oblema ic also in p e ious accoun s, can be linked up
o a di e ence be ween o e and co e mo emen , unde bo h Kayne’s and my
app oach. Unde ou in e p e a ion o mo phological s uc u es (as discussed in
sec ion 3), he o de di e ence in (28a) can be analyzed by (28b) and (28c).
(28) a. sc ewd i e - ou ne- is
b. [N0max sc ewj d i eie [V0max i[N0max j]]
c. [N0max ——— ou nie [V0max i[N0max is]]
Tha is, compounding o his ype illus a es N0min mo emen o he Speci ie
o a highe N0max, o e in F ench, co e in English. The basic di e ence be ween
Kayne’s analysis o hese and ou s is ha we do ag ee wi h e.g. Di Sciullo &
Williams (1987) in conside ing mo phologically complex wo ds di e en om
objec s cons uc ed by syn ac ic ules a some le el. Wi h espec o (28), da a o he
ype in (29) p o ide he dis inc ion be ween Kayne’s in e p e a ion o he LCA
wi h espec o he RHR and ou s, gi en in sec ion 3.
(29) a. I bough a * [lo o sc ews - d i e ]
b. J’ai ache é un * [ ou ne-beaucoup de is]
c. I need i . (i = sc ewd i e ) (*i = sc ew)
d. J’en ai besoin. (en = ou ne- is) (*en = is)
In sum, sugges ions as o he exis ence o «maximal» heads s. «minimal»
heads (e.g. Chomsky 1995) p o e use ul o mo phology —and allow us o make
use o he LCA inside mo phological s uc u es while main aining a heo y in which
mo phology and syn ax emain dis inc modules.
Re e ences
Ackema, P. (1995). Syn ax below ze o. U ech : Ph.D.Diss.
Bok-Bennema, R. (1996). «Du ch Ve baliza ion and he Theo y o Mo phology».
Yea book CLCG 1995, G oningen.
Chomsky, N. (1995). The minimalis p og am. Camb idge Mass: MIT P ess.
86 Ca WPL 7, 1999 F ank D ijkoningen
Ca WPL 7 071-087 13/6/00 12:26 Página 86

Di Sciullo & Williams (1987). On he de in ion o Wo d. Camb idge Mass: MIT P ess.
Don, J. (1993). Con e sion. U ech : Ph.D. Diss.
D ijkoningen, F. (1997). «Mo phological S eng h: NP posi ions in F ench». In
Bee mann, D.; Leblanc, D.; an Riemsdijk, H. (eds.). Righ wa d Mo emen .
Ams e dam: Benjamins, p. 81-114.
G imshaw, J. (1990). A gumen S uc u e. Camb idge Mass: MIT P ess.
Hoeks a, T., e al. (1987). «Complexe e ba». GLOT 10.
Kayne, R. (1994). The an isymme y o Syn ax. Camb idge Mass: MIT P ess.
Neeleman, A.; Schippe , J. (1992). «Ve bal p e ixa ion in Du ch: Thema ic e idence
o con e sion». Yea book o Mo phology 1991. Do d ech : Fo is.
Selki k, E. (1982). The syn ax o Wo ds. Camb idge Mass: MIT P ess.
Spence , A. (1991). Mo phological Theo y. Ox o d: Blackwell.
Williams, E. (1981). «On he no ions “lexically ela ed” and “head o a wo d”». Linguis ic
Inqui y.
An isymme y And The Le hand In Mo phology Ca WPL 7, 1999 87
Ca WPL 7 071-087 13/6/00 12:26 Página 87