Tobias Scheer
CNRS 6039, Université de Nice
[email protected]
this handout and some of the references quoted at
www.unice.fr/dsl/tobias.htm
OCP-6
Workshop on privativity
21 January 2009
Edinburg
PRIVATIVITY IN GRAMMAR AND IN THE COGNITIVE
SYSTEM
(1)
goal
to point out that privativity is also a key issue beyond melodic representation:
a. in the procedural communication between morpho-syntax and phonology
b. in the representational communication between morpho-syntax and phonology
c. in the exchange of information among different cognitive systems
Conclusion: massive evidence for privativity in grammar and other cognitive systems
The workshop should rather have asked whether the non-privative project is still worth
pursuing.
==> everybody is privative today, even regarding melody.
1. Privativity in melodic representations
(2)
underspecification
a. privative representions in non-privative systems (that are based on binary features)
is called underspecification.
b. since Archangeli (1988), Pulleyblank (1988), Avery & Rice (1989) and others, all
"binary" theories discuss, and to some extent implement privative representations.
c. (generally) without however making explicit that underspecification is the import
of Anderson & Jones's (1974) privative idea into regular SPE-descendent Feature
Geometry.
See Anderson & Durand (1988) on this issue.
d. the result often is a ternary system: a melodic object can be
- marked +
- marked –
- absent
-2(3)
today
a. either representations do not matter at all.
1. This is the OT-mainstream (meaning also the mainstream of OT) where
rudimentary SPE-features such as [± labial], [± dorsal] etc. can do the job
without any geometric structure: constraints such as *labial, *dorsal etc. work
with interchangeable labels. All work is done by computation: this is the
representations-are-decoration-if-anything tropism of OT (mainstream OT, to
be precise).
de Lacy (2007a) in general, and de Lacy (2007b) in particular, is a good witness
of this move.
2. In the introduction of her book on segmental phonology, Lombardi (2001:3) for
example provides a very honest and lucid statement of the fact that
representations are useless and interchangeable in OT as we know it:
"[t]he tenets of OT, regarding constraint violability and ranking, make no
particular claims about phonological representations. We could, for example, do
OT with any kind of feature theory: SPE feature bundles or feature geometric
representations, privative or binary features, and so on."
b.
or the question is not if, but how much privativity is suitable
Hall (2007) provides an overview. A given feature may be more or less well suited
for a privative status: while [spread glottis] and [constricted glottis] for example are
good candidates for privativity (negative values are not manipulated by
phonological processes), [voice] is rather not (since [-voice] appears to be
sometimes active in rules/constraints).
2. Phonology is underfed
(4)
How morpho-syntax talks to phonology
[Scheer forth]
There are two channels
a. representational
juncture phonemes, SPE-style #, the Prosodic Hierarchy
b. procedural
cyclic derivation, i.e. inside-out processing (which is a generative invention)
(5)
underfeeding of phonology is a hard observational fact
a. only a small subset of the available morpho-syntactic information is actually
relevant for phonology: most of it has no phonological effect at all.
b. this basic observation was first formulated by Chomsky et al. (1956), who conclude
that phonology is underfed by translational activity: if only a small subset of
morpho-syntactic information is used, the translational device has fed phonology
only with this subset – phonologically irrelevant information has never been
translated and is thus absent from phonology.
c. despite this situation, (generative) theories have always chosen to ship everything
to the phonology, which then does cherry-picking. Both regarding representational
and procedural communication.
-3-
3. Privativity in the representational communication with phonology
(6)
on the non-privative side I
structuralist juncture phonemes
Moulton (1947) uses "+" in order to anchor the beginning and the end of each word in
the phonemic transcription, irrespectively of whether it will have an effect or not
(7)
on the non-privative side II
SPE
a. The rigid mapping algorithm of SPE also sends all edges of major categories and
higher projections thereof to the phonology without discriminating between those
that will and those that will not be used.
b. That is, the "syntactic" # boundary restores full morpho-syntactic information in
phonology: the syntactic distance of two neighbours in the linear string is a direct
function of the number of intervening #s.
c. The input to phonological computation is thus a string made of lexical items and
clusters of hash marks. After the application of phonology, remaining hash marks
are erased by rule at the end of the derivation.
(8)
on the non-privative side III
Lexical Phonology
Lexical Phonology is also on the non-privative side, at least those versions that use
Mohanan-type brackets (Mohanan 1986), which mark the beginning and the end of
every morpheme in phonology.
(9)
on the non-privative side IV
Prosodic Phonology
a. Prosodic Phonology also promotes non-privative translation, even if it is true that
not all morpho-syntactic information is projected onto phonology (the Translator's
Office makes readjustment decisions, cf. non-isomorphism).
b. Non-privativity in fact is a by-product of the Strict Layer Hypothesis: all strings are
exhaustively parsed at all prosodic levels independently of whether or not there is
an associated phonological effect. That is, the full six-layered prosodic constituency
is always constructed no matter whether there is evidence for particular divisions or
not.
c. all strings are exhaustively parsed at all prosodic levels independently of whether
there is any associated phonological effect or not (Hayes 1989 [1984]:220,1990:86,
Nespor & Vogel 1986:11). That is, the full prosodic arborescence is always
constructed, and all expressions in all languages have the same number of levels of
embedding
-4(10) on the non-privative side V
models of representational continuity between morphology and phonology
Oostendorp's (2006) theory of morpho-syntactic colours
Orgun's (1996 et passim) declarative Sign-based phonology
a. This is an automatic effect of their non-modular architecture, which allows
phonology to directly see and refer to all elements of morphological (and syntactic)
structure.
b. Bermúdez-Otero (forth:21ff) discusses this approach in greater detail; he calls
Orgun's view on the matter isomorphism (of morphological and phonological
structure).
(11) on the privative side I
Chomsky et al. (1956)
a. The only representatives of privative translation among the classical theories are
Chomsky et al. (1956). Noam Chomsky and Morris Halle reverted back to nonprivativity twelve years later in SPE, a move that determined the attitude of the
field up to the present day.
b. It is therefore reasonable to think of their privative stance as a mere accident that
was due to the rhetorical demand of the time: Chomsky et al.'s (1956) strategy was
to introduce generative boundaries into the hostile structuralist environment by
appealing to the well-reputed structuralist notion of economy: juncture exists in
order to simplify phonemic transcription.
(12) on the privative side II
Government Phonology (CVCV): the initial CV
a. the only objects that qualify for the representation of morpho-syntactic information
in phonology are syllabic constituents (Scheer forth) because
1. they are not diacritic (#, φ etc. are), i.e. they are part of the basic vocabulary
that phonological computation processes for "domestic" purposes, i.e. in
absence of any morpho-syntactic conditioning.
2. they are not melody
the area below the skeleton, i.e. melody, is incommunicado with morphosyntax. This is a hard observational fact, cf. below.
This is expressed in Zwicky & Pullum's (1986) generalization regarding
Phonology-Free Syntax.
b. the phonological identity of the beginning of the word is an empty CV
(Lowenstamm 1999)
1. Lowenstamm's initial proposal is non-privative: the initial CV is always
present, but sometimes "invisible". It regulates the distinction between #TRonly languages (only #TR attested) and anything-goes languages (#TR and #RT
occur).
2. extension to two other phenomena:
- first vowels of the word that (do not) alternate with zero
- strength/weakness of word-initial consonants
in a privative interpretation of the initial CV, which is either inserted into the
phonological string or not: Scheer (2000), Pagliano (2003), Seigneur-Froli
(2003, 2006), Ségéral & Scheer (2008), Scheer (forth).
-5c.
typological predictions made by the parameterisation of the initial CV
in a language where the
in a language where the
initial CV is present
initial CV is absent
1. word-initial consonants are strong
word-initial consonants are non-strong
2. initial clusters are restricted to #TR there are no restrictions: #TR, #RT, #TT
and #RR clusters may occur
3. first vowels of words may not
first vowels of words may alternate with
alternate with zero
zero
4. Privativity in the procedural communication with phonology
(13) what is privativity in procedural communication?
the question whether all nodes of the morpho-syntactic tree are spelled out, or only a
subset thereof.
(14) on the non-privative side I
SPE
a. interpretation runs through the bracketed string (that is inherited from S-structure)
from inside out (Chomsky & Halle 1968:15ff)
b. every morpheme break defines a cycle
except for sequences of morphemes that belong to the same major category: these
cohabitate in the same cycle:
[[[theatr]N ic + al]A i + ty]N
(Chomsky & Halle 1968:88f)
c. párent, parént-al vs. párent-hood
representational analysis:
- class 2 -hood comes with a #: /[[parent] #hood]/, against /[[parent] +al]/
- the stress-shift rule is blocked by #
(15) on the non-privative side II
Lexical Phonology
a. interactionism: interpretation occurs upon each concatenation or, alternatively, at
the end of each level, which is remisiscent of SPE: morphemes that are somehow
related (by their major category or their affix class) cohabitate in the same cycle.
b. in any event, all levels undergo phonological (and semantic) interpretation.
c. derivation of párent - parént-al vs. párent-hood in Lexical Phonology
parent
parént-al
párent-hood
lexicon
parent
parent
parent
level 1
concatenation
—
parent-al
—
stress assignment párent
parént-al
párent
level 2
concatenation
—
—
párent-hood
rule application —
—
—
-6(16) on the non-privative side III
Distributed Morphology
a. ground rule: interpretation occurs at every xP
Marantz (2001, 2007)
Applied to phonology by Marvin (2002)
b. example
condensation vs. transportation: contrast based on direct vs.
indirect merge of -ate
a. to transpórt
b. transp[ə]rt-át-ion
c. cond[ɛ]ns-át-ion
nP
nP
vP
v
|
ø
PF
n
|
-ion
√
|
transport
PF
vP
v
|
-ate
n
|
-ion
PF
vP
PF
v
|
-ate
PF
vP
√
|
transport
v
|
ø
PF
√
|
condense
(17) on the privative side I
Halle & Vergnaud (1987)
a. Selective spell-out
Halle & Vergnaud dispense with the classical definition of what an interpretational
unit is: they propose to grant cyclic status only to a subset of morpho-syntactic
divisions. That is, some nodes trigger interpretation, others do not.
b. triggering or not triggering interpretation is a lexical property of affixes that
projects to dominating nodes and is read off by the spell-out mechanism.
[==> piece-driven phase, as opposed to node-driven phase in current syntactic
Phase Theory]
c. Halle & Vergnaud (1987): analysis of affix class-based stress
spell-out occurs
- at the root node
- at nodes that dominate class 1 affixes
but NOT at nodes that dominate class 2 affixes
parént-al
párent-hood
β
phon
al1
α
n
β
phon
parent
spell-out
[[parent] al1]
hood2
α
n
phon
parent
spell-out
[parent] hood2
-7(18) on the privative side II
Government Phonology (Kaye 1992, 1995)
a. Kaye applies Selective Spell-out
b. but
1. with distinct implementational choices
2. introduces no look-back, i.e. what today is called Phase Impenetrability
c. differences between Halle & Vergnaud (1987a) and Kaye (1995)
1.
2.
3.
4.
5.
6.
Halle &
Vergnaud
the root is an interpretational unit
yes
the word is an interpretational unit
no
morpho-syntactic terminals may
no
be interpretational units
interpretation-triggering affixes
their own
trigger the spell-out of
node
English affix-classes: type that
class 1
triggers interpretation
underapplication is achieved by
cycles
7. no look-back device
- type of device used
- derived environment effects
built in
- no look-back is relevant for
underapplication
SCC-K
Kaye
no
yes
yes
their sister
class 2
cycles and no lookback
yes
modificationinhibiting
no
no
yes
(19) on the privative side III
current syntactic Phase Theory
a. Halle & Vergnaud's (1987) selective spell-out is exactly what modern (syntactic)
phase theory is about: in more familiar terminology, nodes may or may not be
phase heads, hence their material may or may not be an interpretational unit.
b. As far as I can see, the phonological heritage is left unmentioned in the syntactic
literature since derivation by phase was introduced by Epstein et al. (1998),
Uriagereka (1999) and Chomsky (2000,2001 et passim).
c. [==> difference: syntactic Phase Theory practises node-driven phase, as opposed to
phonology, which is based on piece-driven phase. But den Dikken's (2007) Phase
Extension is a step into the direction of piece-driven phase, see Scheer (2008)]
-8-
5. Communication among modules of the cognitive system is privative
(20) translation is fractional, and the choice of translated pieces is arbitrary
a. A pervasive property of intermodular communication appears to be the fact that
translation is never complete. Rather, only a subset of the structure of the sending
module is made available to the receiving module through translation. Also, it
appears that the pieces which are chosen for transmission cannot be predicted.
b. Ray Jackendoff's work regularly draws attention to the underfeeding of the
receiving module.
"Correspondence rules perform complex negotiations between two partly incompatible spaces of
distinctions, in which only certain parts of each are 'visible' to the other." Jackendoff (1997:221)
"The overall architecture of grammar consists of a collection of generative components G1, …, Gn
that create/ license structures S1, …, Sn, plus a set of interfaces Ijk that constrain the relation between
structures of type Sj and structures of type Sk. […] Typically, an interface Ijk does not 'see' all of
either Sj or Sk; it attends only to certain aspects of them." Jackendoff (2002:123)
c.
The amount of structure that is visible for interface processors in a given module
may be small or large, and this is unpredictable: the translational channel between
two modules may have a more or less narrow "information bottleneck"
(Jackendoff's 2002:229 term).
(21) evidence for the fractional character of translation from other cognitive activity
a. cognitive modules may draw on information that comes from a range of other
modules (many-to-one); conversely, the output of a given module may be used as
the input to a range of other modules (one-to-many). Jackendoff (2002:223f)
reviews a number of relevant cases.
b. multiple usage of (different parts of the) structure that is created in a given module
Audition for example is an information-provider for a number of very different
modules: sound is processed
- by all-purpose audition (e.g. the perception of sound that is produced by animals)
- voice recognition (the identification of humans according to their voice)
- auditory affect perception (emotion detector)
- and of course the perception of linguistically relevant phonetic material.
Nobody knows how the uniform auditory signal is chopped into the pieces that are
relevant for these four (and other) clients (among other things, there may of course
be overlap). But the fact is that the linguistic system receives all relevant
information that is needed for linguistic computation.
c. the same module receives input from different sources
In perception for example, phonology is fed at least by acoustic-phonetic and visual
information. The latter is documented by the so-called McGurk effect (McGurk &
MacDonald 1976, Ingleby & Azra 2003): when exposed to auditory and visual
information that simultaneously provide conflicting information, subjects
consistently perceive something different from what reaches their ears: either the
visual input overrides the auditory stimulus ([ba] is perceived when the subject is
exposed to auditory [da] and visual "[ba]"), or the perceived sound is a compromise
([da] is perceived from auditory [ga] and visual "[ba]"). It may be assumed that this
kind of "lip reading" enhances perception in noise-impaired environments.
-9d.
complementary distribution of morpho-syntactic and visual input:
- morpho-syntax can only manipulate phonology above the skeleton: no melody
- vision can ONLY manipulate melody
Interestingly, the McGurk input into the phonological module appears to be the
complementary set of what morpho-syntax can provide: melodic primes. Recall that
morpho-syntax and melody are incommunicado. The reverse seems also to be true:
McGurk (i.e. visual) input does not bear on phonological properties above the
skeleton, which is the (exclusive) area of morpho-syntactic influence.
(22) intermodular communication through lexical access
module
A
acousticphonetic
vision
McGurk
phonological lexicon
a p
a p
a p
a p
v p
v p
v p
module
B
lexicon of
module D
A D
A D
A D
B D
B D
p D
p D
module
D
phonology
module
C
p E
p E
p E
C E
C E
lexicon of
module E
module
E
(23) the fractional character of translation is well supported by linguistic evidence
a. melody and morpho-syntax are fully incommunicado:
- no melodic property ever bears on morpho-syntax
- morpho-syntax never manipulates melodic units
b. morpheme-internal phonological properties are not accessible for morpho-syntactic
manipulation
- 10 -
6. Conclusion: five arguments in favour of privativity
(24) argument 1
If the output of translation is non-diacritic, translation must be privative
a. diacritic morpho-syntactic information is always accessed by a specific proviso in
the formulation of processes (constraints or rules) of the kind "only within the
Prosodic Word", or "if a hashmark precedes". A hashmark or a Prosodic Word
alone are perfectly inert: they have no effect at all. By contrast, if non-diacritic, i.e.
phonologically contentful objects are the output of translation, they have an
immediate effect on phonological computation without needing to be referred to by
processes. This is what I call the Direct Effect.
b. In turn, this means that every single object that is inserted into phonological
representations as the output of translation will have an effect: inserting "sleeping"
morpho-syntactic agents that only have an effect when some rule or constraint calls
on them is impossible. Therefore, if carriers of morpho-syntactic information are to
be non-diacritic, their distribution is necessarily privative.
(25) argument 2
phonology is underfed, the minimalist perspective is privative
a. the bare observation that only a small subset of morpho-syntactic information is
phonologically relevant at least puts the burden of proof on those who wish to
argue for non-privative translation. Why should the grammatical system bother
putting the translational mechanism to work in order to translate things that will
never be used? And why should useless structure (boundaries, prosodic
constituency) be built on the phonological side? The minimalist philosophy is
certainly privative.
b. That is, GB syntax in the 80s was full of useless structure: phrase structure rules
were relentlessly building functional categories regardless of whether they would
be relevant or not. The minimalist perspective is privative: phrase structure is only
built when it serves a purpose in the derivation. Minimalist privativity is the result
of a concern for extra-linguistic resources such as active memory, which nonprivative structure wastes.
(26) argument 3
phonology is underfed, the minimalist perspective is privative
Along the same lines but on the procedural side, Bermúdez-Otero (forth:21ff) discusses
the unwarranted empirical consequences of untempered proliferation of cycles. He
concludes that selective spell-out is required for that reason.
(27) argument 4
current syntactic Phase Theory
if this perspective is on the right track, and if there is only one spell-out mechanism, the
fact that derivation by phase relies on selective spell-out excludes systems where all
nodes are interpreted.
(28) argument 5
transmission of information between cognitive systems is fractional.
- 11 -
References
Anderson, John & Jacques Durand 1988. Underspecification and Dependency Phonology.
Certamen Phonologicum, edited by Pier Marco Bertinetto & Michele Loporcaro, 3-36.
Torino: Rosenberg & Sellier.
Anderson, John & Charles Jones 1974. Three theses concerning phonological representations.
Journal of Linguistics 10, 1-26.
Archangeli, Diana 1988. Aspects of underspecification theory. Phonology 5, 183-208.
Avery, Peter & Keren Rice 1989. Segment structure and coronal underspecification.
Phonology 6, 179-200.
Bermúdez-Otero, Ricardo forth. Stratal Optimality Theory. Oxford: Oxford University Press.
Chomsky, Noam 2000. Minimalist inquiries: the framework. Step by Step. Essays on
Minimalist Syntax in Honor of Howard Lasnik, edited by Roger Martin, David
Michaels & Juan Uriagereka, 89-155. Cambridge, Mass.: MIT Press.
Chomsky, Noam 2001. Derivation by Phase. Ken Hale: A Life in Language, edited by
Michael Kenstowicz, 1-52. Cambridge, Mass.: MIT Press.
Chomsky, Noam & Morris Halle 1968. The Sound Pattern of English. Cambridge, Mass.:
MIT Press.
Chomsky, Noam, Morris Halle & Fred Lukoff 1956. On Accent and Juncture in English. For
Roman Jakobson. Essays on the occasion of his sixtieth birthday, edited by Morris
Halle, Horace Lunt, Hugh McLean & Cornelis van Schooneveld, 65-80. The Hague:
Mouton.
de Lacy, Paul (ed.) 2007a. The Cambridge Handbook of Phonology. Cambridge: CUP.
de Lacy, Paul 2007b. Themes in Phonology. The Cambridge Handbook of Phonology, edited
by Paul de Lacy, 5-30. Cambridge: CUP.
den Dikken, Marcel 2007. Review of Frascarelli: Phases of interpretation. Journal of
Linguistics 43, 440-449.
Epstein, Samuel, E.M. Groat, R. Kawashima & H. Kitahara 1998. A Derivational Approach
to Syntactic Relations. Oxford: OUP.
Hall, Tracy 2007. Segmental Features. The Cambridge Handbook of Phonology, edited by
Paul de Lacy, 311-334. Cambridge: CUP.
Halle, Morris & Jean-Roger Vergnaud 1987. An Essay on Stress. Cambridge, Mass.: MIT
Press.
Hayes, Bruce 1989 [1984]. The Prosodic Hierarchy in Meter. Manuscript circulated since
1984, published 1989 in Rhythm and Meter, edited by Paul Kiparsky & G. Youmans,
201-260. Orlando, Florida: Academic Press.
Ingleby, Michael & Ali Azra 2003. Phonological Primes and McGurk Fusion. Proceedings of
the 15th International Congress of Phonetic Sciences, 2609-2612.
Jackendoff, Ray 1997. The Architecture of the Language Faculty. Cambridge, Massachusetts:
MIT Press.
Jackendoff, Ray 2002. Foundations of Language. Brain, Meaning, Grammar, Evolution.
Oxford: Oxford University Press.
Kaye, Jonathan 1992. On the interaction of theories of Lexical Phonology and theories of
phonological phenomena. Phonologica 1988, edited by Uli Dressler, Hans
Luschützky, Oskar Pfeiffer & John Rennison, 141-155. Cambridge: Cambridge
University Press.
Kaye, Jonathan 1995. Derivations and Interfaces. Frontiers of Phonology, edited by Jacques
Durand & Francis Katamba, 289-332. London & New York: Longman. Also in SOAS
Working Papers in Linguistics and Phonetics 3, 1993, 90-126.
- 12 Lombardi, Linda 2001. Introduction. Segmental phonology in Optimality Theory. Constraints
and Representations, edited by Linda Lombardi, 1-9. Cambridge: Cambridge
University Press.
Lowenstamm, Jean 1999. The beginning of the word. Phonologica 1996, edited by John
Rennison & Klaus Kühnhammer, 153-166. La Hague: Holland Academic Graphics.
Marantz, Alec 2001. Words. Handout from the 20th West Coast Conference on Formal
Linguistics.
Marantz, Alec 2007. Phases and words. Phases in the theory of grammar, edited by S.-H.
Choe, 191-222. Seoul: Dong In.
Marvin, Tatjana 2002. Topics in the Stress and Syntax of Words. Ph.D dissertation, MIT.
McGurk, H. & J. MacDonald 1976. Hearing Lips and Seeing Voices. Nature 264, 746-748.
Mohanan, Karuvannur 1986. The Theory of Lexical Phonology. Dordrecht: Reidel.
Moulton, William 1947. Juncture in Modern Standard German. Language 23, 212-226.
Nespor, Marina & Irene Vogel 1986. Prosodic Phonology. Dordrecht: Foris.
Oostendorp, Marc van 2006. A theory of morphosyntactic colours. Ms., Meertens Instituut.
Orgun, Cemil Orhan 1996. Sign-based morphology and phonology with special attention to
Optimality Theory. Ph.D dissertation, University of California at Berkeley.
Pagliano, Claudine 2003. L'épenthèse consonantique en français. Ce que la syntaxe, la
sémantique et la morphologie peuvent faire à la phonologie. Ph.D dissertation,
Université de Nice.
Pulleyblank, Douglas 1988. Underspecification, the feature hierarchy and Tiv vowels.
Phonology 5, 299-326.
Scheer, Tobias 2000. De la Localité, de la Morphologie et de la Phonologie en Phonologie.
Habilitation thesis, Université de Nice.
Scheer, Tobias 2008. Spell out your Sister! Proceedings of the 27th West Coast Conference
on Formal Linguistics, edited by Natasha Abner & Jason Bishop, 379-387.
Somerville: Cascadilla.
Scheer, Tobias forth. A Lateral Theory of Phonology. Vol.2. How morpho-syntax talks to
phonology. A survey of extra-phonological information in phonology since
Trubetzkoy's Grenzsignale. Berlin: Mouton de Gruyter.
Ségéral, Philippe & Tobias Scheer 2008. The Coda Mirror, stress and positional parameters.
Lenition and Fortition, edited by Joaquim Brandão de Carvalho, Tobias Scheer &
Philippe Ségéral, 483-518. Berlin: Mouton de Gruyter.
Seigneur-Froli, Delphine 2003. Diachronic consonant lenition & exotic word-initial clusters
in Greek: a unified account. Studies in Greek Linguistics. Proceedings of the 23rd
annual meeting of the department of linguistics, edited by M. Stavrou-Sifaki & A.
Fliatouras, 345-357. Thessaloniki: University of Thessaloniki.
Seigneur-Froli, Delphine 2006. Le Statut phonologique du début de mot grec. Lénitions
consonantiques et libertés phonotactiques initiales dans la diachronie de la langue
commune et dans le dialecte de Lesbos. Ph.D dissertation, University of Nice.
Uriagereka, J. 1999. Multiple spell-out. Working Minimalism, edited by Samuel Epstein &
Norbert Hornstein, 251-282. Cambridge, Mass.: MIT Press.
Zwicky, Arnold & Geoffrey Pullum 1986. The Principle of Phonology-free Syntax:
introductory remarks. Ohio State University Working Papers in Linguistics 32, 63-91.