Transcript Session 4

Negation and diachrony
• Another problem that has not been solved is
the one of the fact that very different
elements can turn into negation by means of
different evolutionary paths.
• Furthermore, only some n-words can turn
into the standard negative marker, not all of
them.
• The big NegP hypothesis is more precise in
this sense, because it leads us to expect that
only those elements which lexicalize one of
the features contained in the big NegP can be
reinterpreted as the standard negative marker.
• The big NegP hypothesis can be shown to
explain the evolutionary path of different
standard sentential negators.
• I will adopt a new perspective on
grammaticalization proposed by Kayne in
recent work: he assumes that
grammaticalization is not simply a process
that changes the category of the original
elements, but can be explained by assuming
that it is the silent elements that go with the
grammaticalized element that change, not the
element itself.
• When an element like Franco-provencal ‘ren’,
which is etymologically nothing else than an
existential light N meaning THING, comes to
represent negation, we can interpret this
change by assuming the following:
• since the big NegP contains an existential
feature/projection, it can be represented by
its existential portion as a whole, although its
visible portion is much smaller.
• This means that an element like ren is
structurally ambiguous, since it can be
interpreted as an existential, or as the only
lexicalized portion of the big NegP which
contains the existential.
• This might help to settle a long debated
question whether n-words are really negative
or not.
• The big NegP hypothesis is compatible with
the idea that n-words can be ambiguous
between a polarity and a negative reading
(depending on their morphological make-up)
but their internal structure in the two cases
must be different.
• The basic idea is still the same: the whole
NegP can be spelled out by a single formative
or by more than one, depending on the
language, but not by elements that do not
encode one of these formatives.
• Here I will compare two types of evolutionary
paths of negative markers to give the jest of
the argumentation.
• I will show that the positions of the different
standard negators correspond to the original
position of the element when it was not a
standard negator, i.e. the four NegPs Zanuttini
has discovered in Italian dialects are not
NegPs, but the original quantifier positions of
the elements from which the negative marker
has taken its form.
• This means that we do not have to postulate
more than one NegP position in the syntax,
but that negative formatives remain in their
original position, the position of the
completed semantic process of negating a
clause could still require a null element in the
CP.
Minimizer-negation and Q-negation.
• It is a well known fact in typological studies that
different “new negators” have a different
evolutionary path when turning into standard
negation.
• Since two of them are minimally different, one of
them coming from a n-word, i.e. the work
meaning ‘nothing’, while another from a
minimizer, both originally occurring inside the vP
as (part of ) arguments of the verb, we will
briefly observe their two distinct evolutionary
paths.
Examples of minimizer negation
• ‘step’: Piedmontese/Valdotain pa;
‘crumble’: Emilian brisa; Milanese
miga/minga; Veneto mina/mia; Italianmica;
‘bite’: Romansh bucca; Livigno Lombard ca
‘point’: Florentine punto;
‘thread’: Salentino filu;
‘flower’: Old Florentine fiore;
‘drop’: Old Venetian gozo.
• M-negations behave like a natural class of
elements, since they display common
syntactic properties. Usually, they appear
higher than adverbs encoding Aspect and
Tense Anterior, this position is occupied by
both standard negative markers and
reinforcers of negation. (but see lesson II for
some exceptions).
•
a.
A l’ha pa gia ciamà.
(Piedmontese, from
Zanuttini 1997)
SCL SCL has NEG already called
‘He has not already called.’
b.
I n’an briza beli ciamà. (Emilian, from Colombini
2007, § 5.6.1)
SCL NEG have NEG already called
‘They have not already called.’
c.
Non hanno mica già chiamato. (Italian, from Cinque
1999)
NEG have NEG already called
‘They have not already called.’
d.
Elts an buca magliau trasora.
(Romansh, from
Manzini-Savoia 2005)
SCL have NEG eaten already
‘They have not already eaten.’
• According to Jespersen, this type of negative marker
develops from emphatic contexts with a special class of
verbs: i.e. pas should come from an expression with a
verb of movement of the following sort
Non ha bevuto neanche una goccia di alcool.
NEG has drunk not-even a drop of alcohol
‘He did not drink any alcohol at all.’
Non ha mosso un passo.
NEG has moved a step
‘He did not budge’
• In Old French I was not able to find any of
these contexts.
• The same observation is made by Larrivée
(2011:11), who notices that in Old French:
“Little evidence supports the hypothetical
pathway from measure to polarity to
negative.”
• In particular Larrivée argues that there is no
empirical evidence of the use of measure
phrase of pas.
• The polarity status of pas is also called into
question, since it is only found in
interrogatives and vernaculars do not provide
any evidence for a polarity use of pas.
• However, Old Italian provides clear cases in
which minimizers used indeed as a measure
phrase.
• Minimizer negation starts out indeed as a
measure element indicating a small amount of
something, and as such can also occur in
positive contexts:
Mica di
•
•
•
•
•
•
a.
On sté de scisceri e miga de vin d’intrà.
(Lancino Curti 6-14)
one staio of chickpeas and MIGA of
wine of income
‘One staio (20 l) of chickpeas and a little
of wine as income…’
b.
Là no se sente miga de male.
(Barsegapé 2430)
there NEG REFL feels MIGA of pain
‘There one does not feel any pain.’
• In these cases mica is clearly still a noun,
although a functional one, since it requires a
complement nominal expression in the
genitive (i.e. introduced by the preposition di)
• Another minimizer brings additional
arguments that this was really the diachronic
path followed by this type of negative marker.
• Garzonio (2010): In Old Florentine punto can
be used as a measure phrase in various
polarity contexts and as such occurs with a PP
complement.
Punto di
El tempio d’ Ercule […] a chi ha punto delle
the temple of Hercules to whom has PUNTO of.
storie romane letto, è manifesto. (Commento
a Ovidio 471)
the histories roman read is manifest
‘The temple of Hercules is manifest to whom
has read a little of roman histories.’
Punto di
E quando hai punto di tempo quieto pensa a’
and when have PUNTO of time quiet think
benifici di Dio. (G. delle Celle – Lettere 1392.10)
about benefits of god
‘When you have a little of quiet time, think
about the benefits of God.’
Punto di
Il re d’Inghilterra fu a gran pericolo con sua oste […]
che 8
the king of England was at great peril with his army
that 8 days
dì stettono, che non ebbono se non poco pane né
punto di vino.
stayed that NEG had but NEG little bread and-NEG
PUNTO of wine (G. Villani Nuova Cronica 13.66)
‘The king of England was in great peril with his
army, since for 8 days they had not but a little of
bread and had no wine.’
Punto is a light noun
• The fact that punto can select a complement
with the usual preposition di ‘of’ means that it
is still a noun, with the usual properties of
nouns, i.e. at this point of its evolution, the
element is still sitting in the object position.
• In some dialects traces of the original complex DP
structure can still be found. For instance, in the
Alpine Lombard variety of Quarna Sotto, when
the m-negation mia is used, the object can
appear in the genitive (partitive) case, even if it
expresses a singular non-quantifiable entity:
Nə caman mia d əu te frial. (Quarna Sotto, from
Manzini-Savoia 2005)
NEG-SCL call NEG of the your brother
‘They do not call your brother.’
• However, already in Old Florentine minimizers
can be used as adverbials as well as
arguments.
• When they are used as adverbials, they
occupy the same position the standard
negative marker occupies today in dialects
that only have a minimizer as its standard
negative marker:
Elleno non poteano punto ancora essere
trovate. (Difenditore della Pace)
they NEG could PUNTO yet to.be found
‘They could not be found yet.’
Here punto occurs in front of the adverb
ancora, as minimizer negation does.
• Hence, the position minimizers that have
developed into negative markers tend to occupy
is not an effect of grammaticalization of these
elements into negative markers, as put forth by
Roberts and Roussou (2003), it is an effect of
their original properties when they were not yet
standard negative markers.
• Grammaticalization of negation does not involve
any movement of the negative marker up the
clausal spine.
Modern Florentine
La radio la un funziona punto. (Florentine)
the radio SCL NEG works PUNTO
“The radio does not work at all.”
un l’ho mandato in punti
posti. (Pisano)
NEG him-have sent in PUNTO.M.PL places
“I have sent him nowhere.”
 Punto works like a reinforcer of negation but
can also be generated inside the VP in an
argumental position.
Punto agrees with the N
• An indication that minimizers come indeed
from inside an argumental position in the VP
comes from modern Florentine:
Un vedo punti
libri.
(Florentine)
NEG see PUNTO.M.PL. books
“I see no books.”
• In the modern variety, punto has become a
measure adjective, not a noun.
Un
l’ho punta
sentita questa storia.
NEG her have PUNTO.F.SG heard this story
‘I have not heard at all about this story.’
• The fact that there is agreement between
punta and questa storia but the past participle
intervenes shows that even when it comes
from the VP, the element punto can raise to
the adverbial space.
• This shows that the adverbial position of
punto is available even to the argumental
punto, not only to the adverbial one.
French
• The same hypothesis, namely that pas comes
from the object position has been put forth by
Rowlett (2003): he proposes that pas
originates in a position similar to beaucoup
and then is raised by the same mechanism of
split quantification that we have to assume for
Qs in French.
• This would account for the fact that pas also
requires a genitive object.
• Hirschbühler and Labelle (1998) argue against
this hypothesis and bring several arguments in
order to show that pas and beaucoup do not
behave the same.
• Notice however, that it is not necessary to
assume that pas originates in the VP
sinchronically, this can be just the diachronic
path of the negative marker, which has never
changed its position.
• The diachronic path would thus not be one of
movement up the structural tree, but simply a
first versus second merge difference: pas is
directly merged in the low IP and not in the VP
and then moved.
• We conclude that the position of the modern
standard negative minimizer negation is NOT a
negative position per se, but a quantificational
one which is already available when the
element has not turned into a (standard)
negative marker yet.
The development of Q-negation
I (n) mandj (pa) nia nkoe
I not eat not not today/ I am not going to eat
I (n) mandj (pa) nia soni nkoe. (S. Leonardo
Ladin)
I NEG eat NEG NEG potatoes today/ ia do not
eat potatoes today
Q-negation derives from the n-word ´nothing´
First stage
• Bayer notices that the usage of Southern
German nichts in constructions like „ich habe
nichts geschlafen“ can be derived by assuming
that the n-word is located in the empty object
position and then raised.
• Breithbart (2014) notices that in Old Low
German the spread of nicht as a negative
marker is related to the same class of
intransitive verbs singled out by Bayer.
• The etymological origin of nicht and nia/nen is
the same: there is n-morpheme plus an
existential light noun. In German it is related to
the word for ‚body‘, in Italian the element ente
has been either analyzed as ‚people‘ or as ‚thing‘.
In both cases it is an existential light noun.
• Although Bayer (2009) notices that nichts in
Southern German is only possible with verbs that
have an empty direct object position, this is not
always the case in Italian varieties.
Second stage
a.
Nol lavora gnente.
(Venetian)
NEG-SCL works nothing
b. Nol dorme gnente.
NEG-SCL sleeps nothing
c.
*Nol leze gnente i libri.
NEG-SCL reads nothing the books
d.
Nol leze gnente libri.
NEG-SCL reads nothing books
a.
No la crese gnente. (Venetian)
NEG SCL grows nothing
b. Nol me piaze gnente.
NEG-SCL me likes nothing
c.
*Nol riva gnente.
NEG-SCL arrives nothing
 Q-negation is only compatible with nontelic predicates.
Old Italian
Sì che non era nostra intenzione essere che ce
ne sia neente renduto
so that not was our intention be.inf that us
of.it is nothing given.back
‘So that we did not want that anything of it
would be given us back’ (Giachino 17)
The n-word already occurs in front of the past
participle
• In French some bare Quantifiers move higher
than their object position in the VP:
Il a tout beaucoup apprecie´.
'He has everything much appreciated.
Il n´a rien bien compris
He not has nothing well understood
 The position of the negative marker is still the
same position of the n-word inside a field for
quantifiers like tout/tous.
• Q-negation starts ist history as the non
standard negative marker from a different
class of constructions with respect to
minimizers.
• However, also in this case, we notice that the
position of niente found in the modern
dialects is the same Q-position found in Old
Italian.
Focus and Negation
• Another type of negative marker found in
macro and micro-variation work is the one
related to Focus, and stemming from different
elements that can encode Focus in various
languages. Here we analyze examples: the
case of Sicilian neca and the case of Northern
Italian no.
• Sicilian has developed a negative marker
coming from a cleft clause, as it is clear from
its form: neca where the n-corresponds to the
negative morpheme, the vowel is the third
form of the present of the verb ‘be’ and ca is
the complementizer. Since Sicilian is a pro
drop language, there is no subject pronoun
present.
Neca ti vogliu imbrogliari! (Corleone)
“NEG you want th.cheat”
[ASIt 159]
Neca ci cridi!
(Corleone)
“Not it believes”
[ASIt 197]
• That neca is a negative marker and not a
whole clause is shown
• A. by the fact that the form of the verb cannot
be inflected for any other person or tense
• The form of the negative marker is reduced
from nun/un to n• The fact that no element can intervene
between the verb and the complementizer
They did not want to go:
*a. N era ca ci vonsi jiri.
(Mussomeli)
not was that there want to.go
b. Neca ci vonsi jiri.
(Mussomeli)
NEG there want to.go
c. Unn’era ca ci vonsi jiri.
(Mussomeli)
not was that there want go
• That neca is located in Focus is shown by the
fact that it is incompatible with a preverbal
Focus:
a. *I piatta neca purtavu.
(Mussomeli)
the dishes not brought
‘It is the dishes that I did not bring.’
b. *cu Mariu neca parlavu.
with Mario not spoke
‘It is with Mario that I did not speak.’
• As expected by the layering of Topic and Focus
in Italian, neca is compatible with a left
dislocated XP and must follow it.
a.
I piatta, neca i purtavu.
the dishes not them brought
‘I did not bring the dishes.’
b. *Neca i piatta i purtavu.
• Once again we conclude that the position of
the negative marker is not a NegP, but Focus,
which is one of the basic components of the
big NegP.
Postverbal no
• Another negative marker related to Focus is
the one found in Lombard, Veneto and
Trentino varieties which corresponds to the
sentential negative marker and is similar to
Brazilian Portuguese nao.
• No ghe vado NO!
Veneto
• Not there go NO
• ‘I won’t go there’
• The same element is also used in sentence
initial position
• NO che non ghe vado
Veneto
• NO that not there go
• ‘Iwon’t go there’
[CPFocus NO [FinP [Fin° che …[IP no ghe vado]]]
The two constructions can be analyzed as
stemming from a single derivation where the
negative marker is always in the Focus
position and the following clause can raise or
remain in situ.
[SpecGroundP [IP no ghe so ndà] [Ground°
[CPFocus NO] [FinP [IP no ghe so ndà]]]
• When the clause raises to GroundP it goes
through FinP and deletes the complementizer:
• [SpecGroundP [IP no ghe so ndà] [Ground°
[CPFocus NO] [FinP [IP no ghe so ndà]]] [Fin°
[IP no ghe so ndà]]]
• Independent evidence that the two
constructions are related comes from cases
where the clause is spelled out twice
•
•
•
•
No ghe vado NO che no ghe vado
Not there goe NO that not there go
I am really not going!
This shows that the position of NO is always
Focus and that the sentence final or sentence
initial position is derived by movement.
• The relation between Focus and negation has
been noticed by many scholars (see lesson II the
part on relativized minimality, lesson I the part on
negative markers in typological variation)
• The position of neca and NO is the one usually
related to Focus in Italian (see Cruschina 2012),
i.e. also these negative markers do not occupy
NegP, but an independent position that could
correspond to one of the semantic operations
involved in creating negation in natural
languages.
• We conclude that the position where negative
markers occur are not NegP positions, but the
position where the original element could be
moved.
• The same can be shown to be the case for the
sentence final negative marker of Afrikaans,
which according to Biberauer (2007)derives
from a negative tag located after the whole
clause.
• Hence, we do not need to postulate any
special movement when the element is
„grammaticalized“ into the negative marker,
and we do not need to postulate several
distinct NegP projections in the clause.
• „negative markers“ are just what they look,
i.e. minimizers, n-words, Focus etc.
• This analysis differs from the one proposed by
Zanuttini (1997) in one aspect: she assumes that
there is indeed a real NegP/PolP in the sentence,
and that this is located higher than TP but lower
than CP.
• I propose that each of the possible negative
markers we have investigated encodes one of
several features of negation, each of them
spelling out one (or more) semantic component
of the complex necessary to achieve negation.
• In a sense, this is a cartographic refinement of
Zeijlstra’s proposal that some negative
markers correspond to the negative operator
(as in non-strict negative concord languages)
while others do not (like in strict negative
concord languages).
• Some negative markers pied pipe the whole
complex structure, while others are extracted
out of it and move indipendently.