We will now review further arguments in favor of the hypothesis that what we call NegP originates as an internally layered XP encoding more than one feature These arguments have to do with some exceptions to the Jesperson cycle which can be explained only by assuming that the express ID: 671007
Download Presentation The PPT/PDF document "What’s in a “NegP”?" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
What’s in a “NegP”?Slide2
We will now review further arguments in favor of the hypothesis that what we call “NegP” originates as an internally layered XP encoding more than
one feature
.
These arguments have to do with some exceptions to the Jesperson cycle, which can be explained only by assuming that the expression of negation is syntactically and semantically complex.Slide3
The negative cycle has be originally defined by Jespersen in the following way:
Neg
V
Neg
V
neg
V
Neg
The basic property of the cycle is that the sentential negative marker is renewed by means of a doubling stageSlide4
Several
authors
have
proposed
a
more
refined
view
of
the
JC
which
has
up
to
six
different
stages
depending
on
which
the
negative
marker
is
obligatory
.
At the same time reasons have been proposed for its
activation
beyond the original proposal made by Jespersen that the original negator is phonologically weakened. Slide5
For
instance
van der
Auwera
(2011)
assumes
the
following
steps
(
Neg
=
obligatory
,
neg
= optional):
A)
Neg
V (
Spanish
)
B)
Neg
V
neg
(Northern
Italian
)
C)
Neg
V
Neg
(
standard
French)
D)
neg
V
Neg
(
spoken
French)
E) V
Neg
(
Piedmontese
, Quebec French)Slide6
Other languages
The phenomenon is found also in other language groups, not only in Indoeuropean languages. For instance the Bantu languages have been reported to have a similar cycle in a pretty similar way to Romance (i.e. preverbal negation, minimizer doubling, and loss of the original negator). Slide7
The
literature
on
the
factors
triggering
the
JC
is
vaste
, in
principle
it
can
be
claimed
that
the
cycle
is
activated
by
A)
phonological
factors
(
weakening
of
the
negative
marker
non>
no
> ne,
see
French)
B)
Semantic
factors
:
the
original
negator
requires
a
context
of
„
emphasis
“
whatever
that
means
.
C)
Syntactic
factors
:
change
in
word
order Slide8
To
provide
a (partial)
answer
to
the
complex
question
concerning
the
factors
enhancing
or
restraining
the
JC
and
their
interplay
,
let
us
compare
languages
that
have
undergone
the
cycle
with
similar
languages
that
have
not.
Within
the
IE
domain
we
see
Germanic
,
that
has
undergone
the
cycle
rather
quickly
,
Romance
,
where
we
partially
see
the
cycle
, but
much
more
slowly
and
Slavic
,
where
the
cycle
has
not
been
activated
at
all. Slide9
The Jespersen cycle
The JC has been completed in some spoken French varieties, North Western Italian dialects, is in progress in standard French and North Eastern Italian dialects,
but has
never progressed
beyond stage B in van der Auwera’s proposal in
Southern Italian.
What is the specificity of Southern Italian negation that stops the development of the cycle?Slide10
The advantage of using closely related languages is that you control for the basic properties of the language like basic word order, verb raising properties, type of morphology, etc.
The generalizations obtained are free from interference from other factors and more reliable, since you have a much wider set of languages to investigate. Slide11
In a general view of diachronic change similar to the one formalized in van Gelderen (2011), where every “cycle” is the effect of an economy principle which reduces complex elements to heads, which become so small that another XP has to be used in order to express the relevant features, the JC is simply one instance of the Head Preference Principle. Slide12
The problem is that there are languages that are perfectly stable, like German, others where the cycle applies very rapidly, or very slowly.
Comparing the way Low German has completed the cycle according to Breitbarth (2014) with the way French has developed clearly shows the point. Slide13
Even if we take into account sociolinguistic or language policy considerations, we cannot explain
cases where similar dialects ha ve had
a similar history
and
have
been subject to
the same standardization
pressure
coming from
the standard
like for instance, Lombard, which
has undergone the cycle completely
and Emilian, which
has not. Slide14
What
are the factors that enhance or slow down the JC?
Can
the JC
be interpreted as evidence that negation is a complex
XP?Slide15
I ARGUMENT Slide16
We investigate the morphosyntactic properties of the sentential negative marker in SI in order to establish the factors that block or allow for the activation of the JC.
It will turn out that one of the reasons is the fact that the speaker has empirical evidence that the negative marker is morphologically complex in SI but not in NI. Slide17
The empirical field
We focus on the
interaction between the preverbal negative marker and object clitics in old and modern Italian
varieties:
Il
prossimo
tuo
non
ucciderai
e
no
l
fedirai
the
neighbor
your
not
will.kill
and
not. him hurt
e
no
li
farai
(…)
alcuno
rincrescimento
and
not
to.him
you.will.do
any
damage
‘
You will not kill you
neighbor
, nor hurt him, nor do him any harm
’ (
Vizi
e
Virtudi
17)Slide18
Old French and Old Italian
Old French displays a preverbal negative marker and an optional
postverbal
negative marker (
gote
,
mie
,
pas
, etc.)
Si
fait
oscur
,
ne
veient
gote
,
refl
= makes dark not
they.see
not
(…)
ne
veient
clarté
ne
soleil
not
they.see
light not sun
‘
It's so dark, they do not see, (…) they do not see any light nor the sun
’
(
Enéas
195
)
Mais
a
bataille
n
’oset
il
pas
venir
but
to
battle
not=dares
=he not come
‘
But he does not dare to come to battle
’ (
Guillaume
81)Slide19
A comparison
Old Italian also displays a preverbal and an optional
postverbal
negative marker
certo
no
llo
darebbe
tosto
così
sure
not it=
he.would.give
soon so
he
would not answer so soon
’ (
B.
Latini
,
La
Rettorica
)
“Certo
,
sire” disse
elli “io
non
ve
lo celerò
mica
.”
sure sir said
he I not
to.you
=it
I.will.hide
not
‘He said “Of course, sir, I will not hide it from you” ’ (
Tristano
Riccardiano
75
)
Why has Italian remained stable and French has not?Slide20
The hypothesis
Italian has not changed because there was/is evidence that the negative marker is morphologically complex, i.e. it contains two morphemes, while French (and some Northern Italian dialects) did not provide the speaker with any evidence for this.Slide21
Since a complex morphology is an indication of a complex set of semantic features and of several syntactic projections, this can be seen as positive evidence for the big NegP hypothesis.
This means that the JC can be derived by the need to maintain morphological complexity since (the licensing of) negation is semantically and syntactically complex. Slide22
The bi-morphemic status of the negative marker is clearly not the only factor blocking or favoring the Jespersen cycle, since other means can be indicated that negation is complex, but it is at least one of the possible types of evidence which block the activation of the cycle.Slide23
Four systems
A) Old Florentine and Old Sicilian, where the alternation between
non
and
no
is a purely syntactic phenomenon.
B) Modern Western Sicilian deletion, where the alternation between
non
and
n
is a phonological rule of deletion conditioned by syntactic proximity.
C) Modern Eastern Sicilian, where the assimilation rule
n+l
nn
is phonological but still conditioned by syntactic proximity.
D) Modern Florentine, where the assimilation rule has become a purely phonological assimilation phenomenon found in all contexts where
n
+l
ll
.Slide24
Modern Florentine has lost the clues for morphological complexity of the negative marker and is thus potentially subject to the JC unless other independent evidence for complexity is present. Slide25
The Old Italian alternation
The Old Italian alternation between
non
and
no
is conditioned by the presence of a following clitic:
No
ti
vo
’
qui
mostrare
e
aprire
not
to.you
I.want
here
show
and
explain
‘
I do not want to show and explain to you (that
)’ (
Fiore di
Rettorica
65
)
Il
prossimo
tuo
non
ucciderai
e
no
l
fedirai
the
neighbor
your not
you.will.kill
and
not.him
will.hurt
e
no
li
farai
(…)
alcuno
rincrescimento
and not
to.him
you.will.do any damage
‘You will not kill you
neighbor
, nor hurt him, nor do him any harm’ (
Vizi
e
Virtudi
17)Slide26
Old Sicilian
Old Sicilian displays the same alternation between a clitic and the negative marker
:
...
ki
Deu
non
possa
cuntentari
ad
unu
so
bon
sirvituri
,
that
God
not can satisfy
at one
his
good
servant
et
ki
no
li
possa
donari
la
visioni
sua
perpetua
and
that
not him
can
give
the vision
his
eternal
et
perpetua
sequranza
di
no
li
mancar
mai
and eternal
certainty of
not
to.him
disappoint never
‘...
that God cannot satisfy one of his good servants, and give him his
eternal
vision
and the eternal certainty that he will not be disappointed
’
(
Sposizione
del
Vangelo
della
Passione
secondo
Matteo
1.7)Slide27
A syntactic phenomenon I
As no other element can delete the coda of the negative marker, I propose that the phenomenon is not phonologic but syntactic in nature. For instance a verb starting with a liquid does not delete the nasal coda of
non
:
…cui
liga
la
navicella
a la
rocca
non
who
binds the boat to
the rock not
liga
per
firmari
la
rocca
…
binds
to
snub
the
rock
‘…
who binds the boat to the rock does not do that to snub the rock
(
Sposizione
del
Vangelo
della
Passione
secondo
Matteo
1)Slide28
Negation/clitic interaction
Other cases in which the negative marker and clitics interact are known in the literature. For instance,
Friulian
deletes some subject clitics in the presence of a negative marker:
a.
tu
as
mangiat
you have eaten
‘You ate’
b. No as
mangiat
not have eatenSlide29
Analysis
We propose that speakers of Old Italian had evidence that the negative marker was bi-morphemic,
because the lower morpheme alternates only with a clitic:
[[
no
]
[
n
]]
A
purely
syntactic
analysis
might
be
stated
along
the
following
lines
:
both
the
negative
marker
and
the
clitic
have
an existential
feature
,
which
can
be
checked
by
either
of
the
two
elements
.
[NegP
no
[
ExistP
n
[ Clit
]]] →
[
NegP
no
[
ExistP
n
Clit
[
Clit
]]]
Slide30
A syntactic phenomenon II
The phenomenon cannot be analyzed as purely phonological deletion because:
A) it does not occur any time a nasal is followed by a liquid or by a stop, but only in this context.
sensitivity to structural proximity.
B) it is only the object clitic that can delete the coda of the negative marker.
sensitivity to the category of the following item.Slide31
Eastern Modern Sicilian
Eastern Modern Sicilian also presents an alternation between two forms of the negative marker
:
Avissi
statu
cchiù
attentu
,
non
fussi
a
stu
puntu
.
had
been more careful not
would.be
at this
point
‘
If you had been more careful, you would not be in this situation
.’
Penzu
chi
n
o
pottu
dumani
.
think that
not.it
I.bring
tomorrow
‘
I think that I will not bring it tomorrow
.’Slide32
Phonology or syntax?
The alternation is between /non/ and /n/.
The clitic has the form of a simple vowel and provides the vocalic nucleus for the negative marker.
(14) [non
]
V
→ [
n
o
n
]
V
→ [n]
V
Hence, in this case the two negative morphemes are to be analyzed as follows:
(15) [n [
o
n
]]Slide33
Phonology...
Deletion can also occur when
the following element is the verb (if it starts with a vowel
):
(16) N’
assicutunu
picciriddi
, ma
cani.
not follow
kids
but
dogs
‘
They do not follow kids, but dogs
.’
(17) Non
assicutunu
picciriddi
, ma
cani
.
not
follow
kids but dogs
‘
They do not follow kids, but dogs.’
deletion
is no longer sensitive to the category of the following
element.Slide34
...conditioned by Syntax
Deletion
is still sensitive to syntactic
proximity and not simply to linear adjacency:
Don
Antonio → *
D’Antonio
‘Don Antonio
’
Deletion
could still be sensitive to the category of the negative
marker, although it is not sensitive to the category of the following word.Slide35
Other similar phenomena
We propose that Eastern Sicilian has a purely phonological rule of deletion conditioned by structural proximity, as in well known cases like:
(19) la
arancia → l
a
arancia → l’arancia
[
laˈ
raɲtʃa]
‘
the
orange
’
But not:
(20) sono
andat
a
a casa → *sono
andat
a casa
I.am
gone
to home
‘
I went home’Slide36
There exist other cases in which structural proximity is known to count: for instance
raddoppiamento
fonosintattico
(syntactic doubling),
liason
.
We hypothesize that the Eastern Sicilian speaker still has evidence that the negative marker is bi-morphemic.Slide37
Western Sicilian
In the varieties belonging to Western
Sicilian,
the form of the negative marker lacks the initial nasal in the onset and is lexicalized as
un
.
(21)
Si
fussi
statu
chiù
attentu
,
un
fussi
accussì
.
if
were
been more
careful,
not would.be so
‘If he had been more careful, he would have not been in this situation
.’
We still find interaction with object clitics, but of a different type.Slide38
Third person object clitics have two forms:
(22)
U
va
cciercu
.
it go
search
‘
I’m going to search for it.’
(23)
Piensu
ca
pi
curreggi
llu
t u
puortu
rumani
.
think that
to
correct=it you=it bring
tomorrow
‘
I think that I will bring it to you tomorrow in order to correct it.’
In
proclisis
the clitic is a vowel, in
enclisis
a C+V form.Slide39
Assimilation
The liquid of the clitic assimilates to the nasal coda of the negative marker:
(24)
Pensu
ca
rumani
unn
u
puortu
. (Palermo)
think that
tomorrow
not=it
bring
‘
I think that I will not bring it tomorrow
.’Slide40
If another
clitic
starting with an obstruent intervenes, then the third person object
clitic
surfaces in its vocalic form providing the nucleus for the other
clitic
, which becomes the onset of the new syllable:
(25) Piensu ca rumani
un
t
u
puortu
(Palermo)
think that tomorrow not you=it bring
‘I think that I will not bring it to you tomorrow.’Slide41
If the following verb starts with a vowel, no assimilation is observed between the clitic and the negative marker, the clitic exploits the nucleus of the following verb to surface as a liquid:
(26)
Un
l
’accattasti
ancuora
?
not
it bought
yet
‘
Haven’t you bought it yet?’Slide42
Assimilation
is only possible when the clitic whose underlying form has a liquid and the negative marker
are:
a
) adjacent and
b
) the liquid does not already surfaces as the
onset
of the nucleus provided by the
following
verb.Slide43
The
phenomenon is
phonological in nature, but restricted
to the clitic field, where the negative marker and the clitic are located on adjacent heads: the assimilation rule /n/+/l/→/n:/ does not apply in contexts that do not involve clitics and the negative marker:
(27)
Don Luigi → /
Dɔl:uidʒi
/ */
Dɔn:uidʒi
/
It does not provide any evidence that the negative marker is bi-morphemic.Slide44
Relevance for the Jespersen cycle
The syntactic rule of alternation of Old Italian varieties always provides evidence for the bi-morphemic status of the negative
marker.
The purely phonological rule of Modern Florentine never provides evidence for the
bi-morphemic
status of the negative
marker.
The
phonological
rule conditioned by syntax can but need not provide evidence for this (in Eastern Sicilian it does, in Western Sicilian it does not).Slide45
Proposal
If a rule is sensitive to the category of the words involved, it is not a phonological rule, but a syntactic one. Interactions between clitics and negation have to be explained on the basis of the formal features they possess and can check in the clitic field. (in our case an existential feature)
This rule provides the speaker with evidence that the negative marker is bi-morphemic and contributes to block the Jespersen cycle.Slide46
If
the
JC
is
the
effect
of
a
general
uniformity
principle
,
which
tends
towards
a
similar
semantic
,
syntactic
and
morphological
complexity
,
it
is
most
probably
blocked
or
favored
by
several
factors
. Slide47
One
of
these
factors
is
morphological
complexity
:
if
there
is
still
evidence
that
the
negative
marker
is
morphologically
complex
,
it
is
not
necessary
to
„
reinforce
“
negation
activating
the
cycle
.
In a sense, negative doubling makes up for the lack of morphological complexity in the original negative marker. Slide48
II ARGUMENT Slide49
The Southern side of the puzzle
In the Southern
I
talian dialects (SIDs) no standard
postverbal
negative marker has been developed.
SIDs can display a series of non standard strategies to mark sentence
implicatures
but these elements, but nowadays they generally do not involve
postverbal
elements. Slide50
The Southern Italian puzzle
Southern Italian dialects have developed new preverbal negative markers without undergoing any doubling stage, i.e. violating
the Jespersen
cycle.
The exception can be shown to be only apparent if we assume that negation is syntactically complex construct.Slide51
Rionero
mankə
In the dialect of
Rionero
in Vulture the usual form
non
has been replaced by ‘
mankə
’, related to the verb ‘lack’ and the adverb ‘less’, in this sense its development follows one of the known types of evolution of negative markers:
Vivə
spessə
se
mankə
vu
caré
malatə
drink
often if not want.2sg fall.inf
illSlide52
mankə versus
non
Mankə
behaves like
non
in
Occurring after the subject and before object clitics:
Mankə
tə
rə
dikə
pecché
mankə
rə
saccə
.
not
you=it=tell.1sg because not it=know.1sg
b) Being incompatible with true imperative forms
Mankə
u
piglià
!
not it=take.
infSlide53
mankə versus
non
c) Requiring
negative concord with
postverbal
n-words
Mankə
je
venutə
nesciunə
.
not
is come nobody
d) Not being the pro-sentence negation
Hai
vistə
a
Pierə
?
No / *
Mankə
.
have.2sg
seen to P. noSlide54
The
neighboring
dialects
still
display
the
standard
negative
marker
non
:
Nov
volənə
sce’nə
.
Venosa
not want.3pl go=thereSlide55
There exist residues of a negative marker of the standard type:
Sə
nə
mə
nə
vù
rà
,
kə
sə
pòzza
strafucà
.
if not me=of-it=want.2sg give.inf that you may chock
‘If you do not want to give it to me, may you chock.’
However, there is no evidence that there has been a doubling stage. Slide56
How
can
this
exception
be
explained
?
[
Focus/Operator
NO/
manco
[
ScalarP
non
[
MinQ
mica
[
ExistentialP
(
ni
)-
ente
]]]]
The
change
has
occurred
internally
to
the
complex
negation
,
nevertheless
,
since
the
element
moves
as
a
whole
,
the
syntactic
properties
of
the
whole
have
not
changed
.Slide57
Manco
might still be
syntactically complex
, including one of the lower
internal projections of the big XP.
It is not not clear whether it is morphologically complex or not. Slide58
We
conclude
that
the
big
NegP
hypothesis
is
superior
to
the
idea
that
NegP
can
occur
in a
lot
of
different
positions
,
because
it
can
explain
at
least
some
exceptions
to
the
JC.