elements on the borders a colored approach to vowel reduction in lunigiana dialects gen and num...
TRANSCRIPT
Elements on the borders
a colored approach to vowel reduction in Lunigiana dialects
gen and num distribution in Lunigiana nominal expressions
Edoardo Cavirani
Italian Dialect Meeting 2015 and CIDSM X
22-24 June 2015, University of Leiden
Overview
02/27
EXPLANANDUM
Variation in fem and pl distribution in the nominal expressions of Lunigiana dialects: Carrara, Colonnata and Ortonovo
MAIN HYPOTHESES• The burden of the variation is taken by the phonological
module
no Vocabulary difference (vs Taraldsen 2009)no morphological operation (vs Manzini & Savoia 2005)
• The linearization of fem and pl is taken care of by the phonological module
no ‘special’ syntactic derivation (vs Taraldsen 2009)no morphological operation (vs Distributed Morphology)
The geolinguistic domain: Lunigiana
03/27
Data: NP
GEN and NUM DISTRIBUTION in NPs
Carrarese lup-epl.f vs lup-af
Colonnatese lup-ipl-af vs lup-af
Ortonovese lup-apl.f = lup-af
04/27
Data: NP
SYNTACTIC STRUCTURE
Word final vowels are morpho-syntactically complex: they can represent the phonological exponent of more than one
morpho-syntactic terminal (DM & nS)
05/27
Analysis: previous accounts
Distributed Morphology•Vocabulary insertion targets terminal nodes•Subset Principle•Morphological operations: Fusion and Impoverishment
Carr. morphological operations Carr. Vocabulary ItemsFusion: {pl{f}} => {pl,f} /ø/ {m} (lup)
{pl{m}} => {pl,m} (?) /a/ {f} (lupa)/i/ {pl}/e/ {f.pl}
06/27
/e/ {pl.f}
lup-e ‘wolves f.pl’
/i/ {pl.m}
lup-i ‘wolves m.pl’
Analysis: previous accounts
07/27
/i/ {pl}/a/ {f}
lup-i-a ‘wolves f.pl’
/i/ {pl}/ø/ {m}
lup-i ‘wolves m.pl’
Analysis: previous accounts
08/27
Analysis: previous accounts
09/27
/a/ {f}
lup-a ‘wolve(s) f(pl)’
/i/ {pl}
lup-i ‘wolves m.pl’
Analysis: previous accounts
10/27
/e/ {pl{f}}
lup-e ‘wolves f.pl’
/i/ {pl{m}}
lup-i ‘wolves m.pl’
Analysis: previous accounts
11/27
/a/ {f}/i/ {pl{m}}
lup-i-a ‘wolves f.pl’
/i/ {pl{m}}
lup-i ‘wolves m.pl’
Analysis: previous accounts
12/27
Analysis: previous accounts
13/27
/a/ {pl{f}} /a/ {f}
lup-a ‘wolves f.pl’
/i/ {pl{m}}
lup-i ‘wolves m.pl’
Analysis: previous accounts
SUMMARY
Distributed Morphology
Carrarese /e/ {pl,f} {pl{f}} => {pl,f}Colonnatese /a/ {f}; /i/ {pl}Ortonovese /a/ {f} {pl} => ø (if gen: f)
•problems with Colonnatese Ellipsis•problems with the sensitivity to gen of Impoverishment in Ortonovese•variation both in the Vocabulary and in the Morphology
nanoSyntax
Carrarese /e/ {pl{f}} Colonnatese /a/ {f}; /i/ {pl{m}}Ortonovese /a/ {pl{f}}
•no problem with Colonnatese Ellipsis•no problem with Ortonovese Impoverishment •variation only in the Vocabulary; no need for Morphology
14/27
Analysis
From PROTO-ROMANCE to LUNIGIANA DIALECTS
“[…] unstressed vowel deletion […] is (almost) normal in the Higher Magra Valley […], while the more you descend the valley, the more the
deletion alternates with the reduction” (Luciani 1999)
Word-internal unstressed vowels have been reduced to schwa or deleted in the whole geolinguistic domain (Cavirani 2015)
SILVĀTICU(M) ‘wild sg.m’ Carr./Col./Ort. [səlˈvat(ə)k] Pontr. [sarˈvadg]
LIBERU(M) ‘free sg.m’ Carr./Col./Ort. [ˈlib(ə)r]LIBRU(M) ‘book sg.m’ Pontr. [ˈlibar]
15/27
Analysis
From PROTO-ROMANCE to LUNIGIANA DIALECTS
“[…] unstressed vowel deletion […] is (almost) normal in the Higher Magra Valley […], while the more you descend the valley, the more the
deletion alternates with the reduction” (Luciani 1999)
Word-final vowels display a better resistance to reduction (Cavirani 2015)
16/27
LŬP- ‘wolf’masc fem
sg pl sg pl
Carrarese lup-ø lup-i lup-a lup-e
Colonnatese lup-ø lup-i lup-a lup-j-a
Ortonovese lup-ø lup-i lup-a lup-a
Pontremolese luv-ø luv-i luv-a luv-ø
Analysis
PHONOLOGICAL STRUCTURE
unstressed nuclei graduallya lose their melodic contentb, namely their ability to license phonological structuresc (Cavirani 2015)
•the diachronic change is mimicked by the diatopic variation
•defined in terms of Elements (Backley 2011)
[a] [i] [e] [ɛ]
|A| |I| |I| |A| |A| |I| |I| |A| |A| |I|
•structural complexity hierarchy (|A| > |I|/|U|; Pochtrager 2015)
|X| |X| > > >
|X| |Y| |X| |X| | |
17/27
Analysis
PHONOLOGICAL STRUCTURE
Word final vowels are phonologically complex objects …
[a] [i] [e] [ɛ]
|A| |I| |I| |A| |A| |I| |I| |A| |A| |I|
aka
|A| |I| |IA| |IA|
… translating complex morpho-syntactic structures, viz subtrees (nS)
|A| {f}a |I| {pl{m}}b
a. Rather than {pl{f}} as seems to be required for Ortonovese by nS Superset Principleb. Because of Colonnatese’s ellipsis
18/27
Analysis
PHONOLOGICAL COMPUTATION
Colored Containment Theory (Van Oostendorp 2007)
•the morphological affiliation of phonological objects must be visible on the surface•“Coloring […] grants the bare minimum of accessibility phonology may have to morphological structure” (Trommer 2015: 83)
Constraint set (Cavirani 2015)
*V|Str| : unstressed vowels cannot license complex structures
(*V|XY| >> *V|X| >> *V|X| >> *V| |)
Express-|X|α : elements belonging to the underlying representation of an affix (α) must be (phonetically) interpreted
*MixColor : elements belonging to the underlying representation of an affix (α) cannot be interpreted by the root (ρ)
19/27
Analysis
PHONOLOGICAL COMPUTATION: lup, lupi, lupa
•√s spell out {N{m}} (Taraldsen 2009)
•√s display a CVCV template (Lowenstamm 2008; Lampitelli 2014)
• √-final (empty) V slot hosts the floating elements |I|α and |A|α translating {pl{m}} and {f}, respectively
20/27
Analysis
PHONOLOGICAL COMPUTATION: lupe, lupja, lupa•Carrarese: faithful spell-out of both {pl{m}} and {f}
*MixColor, Express-|X|α >> *V|XY| (… >> Express-|X|ρ)
•Colonnatese: spell-out of both {pl{m}} and {f} + |I| lands on the preceding C (palatalizing it)
Express-|X|α, *V|XY| >> *MixColor (… >> Express-|X|ρ)
•Ortonovese: only {f} is spelled out
*MixColor, *V|XY| >> Express-|X|α (… >> Express-|X|ρ)
21/27
Analysis
CONCLUSION• no Vocabulary difference (vs Taraldsen 2009): the three dialects
display the same Vocabulary Items
Carrarese = Colonnatese = Ortonovese
/i/ {pl{m}} /a/ {f}
• no morphological operation (vs Manzini & Savoia 2005)
• the burden of the variation is taken by the phonological module
Carrarese: *MixColor, Express-|X|α >> *V|XY|
Colonnatese: Express-|X|α, *V|XY| >> *MixColor
Ortonovese: *MixColor, *V|XY| >> Express-|X|α
22/27
Why |I|{pl{m}}-|A|{f} linearization (contra Mirror Principle)?
1.√-rising without pied-piping (Taraldsen 2009; no head-adjoining)
“successive movement of the N to the specifier of the pl head through the specifier of f […] That is, I assume that the heads pl and f appear in their underlying order. One reason for assuming this is that it seems natural to assume that the gender marker, being directly selected by the N, should be below Number (pl)” (Taraldsen 2009: 114)
NumP
NPi Num’
√pl GenP-i
ti Gen’
f ti
-a
Affixes linearization
23/27
Affixes linearization
Why |I|{pl{m}}-|A|{f} linearization (contra Mirror Principle)?
2.Merge under adjacency (MUA; Harley 2010; post-syntactic)
“MUA applies to adjacent terminal nodes, adjoining one to the other even across phrase boundaries, enabling the appearance of affixation of a structurally superior element to a structurally inferior one” (Harley 2010: 174)
TP TP
DP T’ DP T’she she
T VP VP-ed
V DP V DPkick it it
V Tkick -ed
24/27
Why |I|{pl{m}}-|A|{f} linearization (contra Mirror Principle)?
2.Head movement + MUA
Once Gen/n is merged to √, √ moves and adjoins to (its c-commanding) Gen/n head (left-adjoining head movement; Baker 1986’s Incorporation). Then, Gen/nP moves to spec-NumP and MUA right-adjoins Gen/n to Num
NumP NumP
Gen/nPi Num’ Gen/nPi Num’
Gen/n’ Gen/n’Num ti Num
√ Gen/n -i √ -a Num Gen/n
-i -a
Affixes linearization
25/27
Why |I|{pl{m}}-|A|{f} linearization (contra Mirror Principle)?
3.Phonology (my proposal)
a) Gen/nP moves to spec-NumP (with pied-piping)b) Vocabulary insertionc) Phonological computation
NumP Carr. Col. Ort.
|I|{pl{m}} |I|{pl{m}} |I|{pl{m}}
Gen/nPi Num’ |A|{f} |A|{f} |A|{f}
Gen/n’Num ti l u p l u p l u p
√ Gen/n -i -a C V C V C V C V C V C V
Affixes linearization
26/27
Why |I|{pl{m}}-|A|{f} linearization (contra Mirror Principle)?
3.Phonology (my proposal)
a) Gen/nP moves to spec-NumP (with pied-piping)b) Vocabulary insertionc) Phonological computation
NumP Carr. Col. Ort.
|I|{pl{m}} |I|{pl{m}} |I|{pl{m}}
Gen/nPi Num’ |A|{f} |A|{f} |A|{f}
Gen/n’Num ti l u p l u p l u p
√ Gen/n -i -a C V C V C V C V C V C V
lupe
*MixColor
Express-|X|α
*V|XY|
Affixes linearization
26/27
Why |I|{pl{m}}-|A|{f} linearization (contra Mirror Principle)?
3.Phonology (my proposal)
a) Gen/nP moves to spec-NumP (with pied-piping)b) Vocabulary insertionc) Phonological computation
NumP Carr. Col. Ort.
|I|{pl{m}} |I|{pl{m}} |I|{pl{m}}
Gen/nPi Num’ |A|{f} |A|{f} |A|{f}
Gen/n’Num ti l u p l u p l u p
√ Gen/n -i -a C V C V C V C V C V C V
lupe lupja
*MixColor Express-|X|α
Express-|X|α *V|XY|
*V|XY| *MixColor
Affixes linearization
26/27
Why |I|{pl{m}}-|A|{f} linearization (contra Mirror Principle)?
3.Phonology (my proposal)
a) Gen/nP moves to spec-NumP (with pied-piping)b) Vocabulary insertionc) Phonological computation
NumP Carr. Col. Ort.
|I|{pl{m}} |I|{pl{m}} |I|{pl{m}}
Gen/nPi Num’ |A|{f} |A|{f} |A|{f}
Gen/n’Num ti l u p l u p l u p
√ Gen/n -i -a C V C V C V C V C V C V
lupe lupja lupa
*MixColor Express-|X|α *MixColor
Express-|X|α *V|XY| *V|XY|
*V|XY| *MixColor Express-|X|α
Affixes linearization
26/27
ReferencesBackley P. (2011). An introduction to element theory. Edinburgh University Press.Cavirani E. (2015). Modeling Phonologization: vowel reduction and epenthesis in Lunigiana dialects. Utrecht: LOT
publishing. Dissertation, Leiden University.Barbiers S. (2013). “Where is syntactic variation?”. In P. Auer, J. C. Reina, & G. Kaufmann (eds.), Language Variation - European Perspectives IV. (Vol. 14). Amsterdam/Philadelphia: John Benjamins Publishing Company,
pp. 1-26.Harley H. (2010). Affixation and the mirror principle. Interfaces in Linguistics, New Research Perspectives, Oxford
Studies in Theoretical Linguistics 31, pp. 166-186.Lampitelli N. (2014). “The Romance plural isogloss and linguistic change: A comparative study of Romance nouns”.
In Lingua 140, pp. 158-179. Loporcaro M. 1994. “Sull’analisi del plurale femminile la doni8 a ‘le donne’ nei dialetti della Lunigiana”. L’Italia Dialettale 57, pp. 35-42.Lowenstamm J. (2008). “On n, nP and √”. In J. Hartmann, V. Hegedus & H. van Riemsdjik (eds.), The Sounds of
Silence: Empty Elements in Syntax and Phonology. Amsterdam: Elsevier, pp. 105-144Luciani L. (1999). Il dialetto Carrarese. Suoni, forme, costrutti, parole. Carrara: Aldus.Manzini M.R. & L.M. Savoia (2005). I dialetti italiani e romanci. Morfosintassi generativa. Alessandria: Edizioni
dell’Orso. Van Oostendorp M. (2007). Derived environment effects and consistency of exponence. In S. Blaho, P. Bye & M.
Kramer (eds.), Freedom of Analysis? Berlin/New York: Mouton de Gruyter, pp. 123-148. Samuels B. (2010) “Phonological derivation by phase: evidence from Basque”. Proceedings of PLC 33 (PWPL 16.1),
pp. 166-175. Taraldsen T. (2010). “Lexicalizing number and gender in Lunigiana”. Nordlyd 36, pp. 113–127. Trommer J. (2015). “Moraic Affixes and Morphological Colors in Dinka”. Linguistic Inquiry 46, pp. 77-112.
27/27
Xxx 3
Why |I|-|A| linearization (contra Mirror Principle)?
•|I| is not the PL’s exponent of a NUM functional projection, but of CLmax.
“… all nouns, in all languages, are mass, and are in need of being portioned out, in some sense, before the can interact with the ‘count’ system. This portioning-out function, accomplished in laguage like Chinese through the projection of classifiers, is accomplished in languages like English by the plural inflection, as well as by the indefinite article. Put differently, plural inflection is classifier inflection, thus accounting for the complementary distribution between classifier inflection and plural inflection, now reduced to the fact tht they are simply distinct instantiations of the classifier system” (Borer 2005: 93)
Borer’s Extended Projection
{Ex[N]}: {D, Q, #, CL}, order universally fixed
“every ExP segment is optional, but its presence/absence has interpretational Consequences” (Borer 2013)
28/27
yyy
Why it’s |I| that is delinked in DP?
•Derivation by phases (defining the relevant domains; Marantz 2001; Samuels 2010)•n(= gen) +√ are spelled-out first, and cannot be further modified (PIC: Samuels 2010). As a consequence, ellipsis (viz delinking of elements) can affect only |I|, which is inserted in/belongs to the next cycle/phase). Check the other domains (see in Bridget conclusion)
Derivationally, little x’s determine the edge of a cyclic domain (a “phase” in Chomsky’s recent terminology). Thus the combination of root and little x is shipped off to LF and PF for
phonological and semantic interpretation, and the meaning of the root in the context of little x is negotiated, using “Encyclopedic” knowledge. Heads attaching outside a little x take as
complements a structure in which the root meaning (and pronunciation) has already been negotiated. (Marantz 2001: 6)
head x
=> LF; PFx √
29/27