|Iterativity in Grammar
Doreen Georgi (University of Potsdam)
Nancy Kula (University of Essex)
Gereon Müller (Leipzig University)
Andrew Nevins (UCL, London)
Milan Rezac (CNRS-IKER)
Rachel Walker (University of Southern California)
Monday, December 2 2019
Tuesday, December 3 2019
Maria Kouneli: The Typology of Determiner Spreading
Determiner Spreading (DS) is the phenomenon in which multiple determiners appear in the DP in the context of (usually adjectival) modification. The terms polydefiniteness and definiteness agreement are also used in the literature, reflecting analytical choices for a given language. An example is given below from Greek.
|(1) ||to ||forema ||to || kokkino ||to ||makri |
| ||the ||dress ||the ||red ||the ||long|
| ||'the long red dress'|
In this talk, I show that the phenomenon is more common than previously thought, by describing Determiner Spreading in two African languages: Kipsigis (Nilotic; Kenya) and Runyankore-Rukiga (Bantu; Uganda). I argue that DS in both languages occurs with indirect modification adjectives (cf. Cinque 2010), and I analyze the pattern as follows: adjectives participating in DS are in a relative clause structure, which is analyzed as a D head with a CP complement (Kayne 1994 a.o.). The presence of multiple adjectives involves recursion of such structures, and, therefore, multiple D heads. My analysis of DS in Kipsigis and Runyankore-Rukiga has been proposed before for Greek (Alexiadou & Wilder 1998) and Maltese (Winchester 2019). This highlights that a relative clause structure is crucial for the understanding of the phenomenon in a wide range of languages. I will argue that even in languages analyzed in terms of definiteness agreement (e.g., Hebrew), there are indications that a relative clause analysis is on the right track.
Milan Rezac: Unsatisfiable Probes: Iterated Case after Partial Intervention
Theories of phi/case dependencies tend to limit iterativity in general (Chomsky 2000) or for NOM/ERG (Yip, Maling and Jackendoff 1987; Marantz 1991; Bittner and Hale 1996). Finnish mostly has the expected system where NOM does not iterate on arguments, e.g. ADV be NOM [INF ACC/*NOM]. However, NOM iterates in contexts identified here as those of the Person Case Constraint, e.g. OBLQ be NOM [INF NOM]. The contextual difference in NOM behavior is not straightforward in models with iterated phi/case-dependencies for phi-hierarchy interactions like the PCC (e.g. Béjar and Rezac 2009; Coon and Keine 2019; Deal 2015; Nevins 2007). The talk explores how a PCC context affects a phi-probe, turning it from one that can be valued, and thereby halts at the closest entailed match, to one that cannot, and may iteratively match. The key is sought in the range of known interveners in the PCC, suggesting intervention for person but not number, and variation in 3rd person agreement in the PCC, approached through parametrically separate/composite probes for person and number. The result is a contextually derived unsatisfiability and so iterativity of a probe born satisfiable and so terminal in other contexts.
Jochen Trommer: Parallel Across-The Board Changes of Tone
Tonal across-the board changes are delusively similar to iterative tone spreading since they typically show directionality effects and change a potentially unbounded number of tone bearing units. However, they also crucially differ from spreading in that they do not involve complete surface assimilation to overt triggers, and may occur even though the involved languages don’t have unbounded spreading elsewhere. Thus in Nandi (Creider 1989), strings of low-tone elements are raised to Mid when following High tones, and in Jumjum (Andersen 2004) utterance-final sequences of High tones are turned into Low, whereas both languages otherwise lack unbounded tone spreading. Since both processes happen across word boundaries, they cannot be assumed to apply to a single multiply linked tone unless this has been generated by phrase-level fusion process before the across-the-board change, which is problematic for a strictly parallel approach to phrase-level phonology. In this talk, I show that this problem can be solved in the Autosegmental Containment version of Optimality Theory (Trommer 2011) under the assumption that fusion and lowering happen in parallel and are triggered by the association of subtonal features (Snider 1999).
Rachel Walker: Locality and Iterativity in Jingulu Vowel Harmony
Jingulu, a language of North Central Australia, exhibits a vowel height harmony where high suffix vowels trigger raising of /a/ → /i/ in the root (Pensalfini 1997, 2002, Nevins 2004, 2010, Kalivoda 2012). Questions arise surrounding the nature of iterativity and locality in this pattern. On the one hand, Jingulu height harmony is iterative and local in the sense that it can affect unbounded sequences of /a/'s in contiguous syllables, e.g. /ngarrabaja-ji/ → [ngirribiji-ji] 'tell neg.impv'. However, underlying high root vowels neither trigger nor transmit vowel raising, e.g. /mamambiyaka-mi/ → [mamambiyiki-mi] 'soft veg.' This may seem to suggest that the process is long-distance rather than an iterative local process in the sense that a single suffix vowel appears to serve as a trigger for all raised vowels. In this study, I pursue an approach in which Jingulu height harmony is understood in terms of chained local identity relations. Following Kalivoda (2012), Jingulu height harmony is interpreted as a weak trigger pattern, where the root serves as a prominent licensor (Walker 2005, 2011, Kaplan 2008, 2011). A novel approach to prominence-based licensing is proposed using Agreement by Correspondence (Hansson 2001, Rose & Walker 2004). For Jingulu, constraints governing surface correspondence produce the effect of iterative harmony that terminates in a faithful high root vowel or an initial syllable. An alternative using relational correspondence is also discussed (Steriade 2012).
Philipp Weisser: Different Levels of Prosodic Dislocation
In this talk, I argue that prosodically driven dislocation can apply to elements of all sizes on the prosodic hierarchy. Starting from typical cases of infixation where a segment or a syllable is dislocated to a position inside the following phonological word, (Prince & Smolensky 1993, McCarthy & Prince 1993, Yu 2007), to cases of complete phonological words being dislocated to a position inside the following intonation phrase (Weisser 2019). Building on this typology, I discuss the observation that - unlike the dislocation of smaller units -
prosodic dislocation of larger categories seems to be able to apply iteratively, as it often has various landing sites available (see also Horwood 2008).
I then discuss various possibilities to account for this observation including (a) the assumption of two fundamentally different operations, (b) potential iterative reapplication of the dislocation operation and (c) the assumption that the phonological categories involved can be partially recursive within a given domain.
Eva Zimmermann: Non-iterative Iterative Reduplication
In this talk, I argue that instances of iterative or multiple reduplication (=the coexistence of more than one reduplicative morpheme in a word) must be derived as non-iterative sharing of underlying activity. This new account of reduplication is based on a redefinition of the phonological copying operation of fission as distribution of underlying activity. This proposal relies on the assumption of Gradient Symbolic Representations (Smolensky and Goldrick, 2016) that all phonological elements have a certain activity or strength. If copying is formally modeled as fission of underlying activity, a 'Copying- Weakening-Correlation' is predicted stating that every copy operation gradiently weakens all copies. It is shown that this correlation is empirically borne out in the typology of reduplication. For one, reduction symmetrically affects the reduplicant, the base, or both these constituents in different languages. And secondly, languages can show threshold effects where only multiple copying results in reduction, not single copying.
Crucially, it is shown that this redistribution of activity must apply in a single step if multiple reduplicative morphemes are present. This argument is based on a case study of Sikaiana, where an innermost reduplicant shows reduction effects in multiple reduplication contexts (Red2-Red1-Base). Crucially, Red2 copies from the base and not the adjacent Red1 - in an iterative copying model, Red1 was hence only copied once and should not show reduction effects that are bound to multiple copying. If copying is non-iterative distribution of underlying activity, on the other hand, both R1, R2 and the copied base are symmetrically weakened in a multiple reduplication context and the Sikaiana reduction falls out straightforwardly.
The workshop will take place in the GWZ building (= Geisteswissenschaftliches Zentrum) in Beethovenstr. 15. The room will be H1 5.16 (roughly: the westernmost room on the 5th/top floor). See also the linguistics department website for more details.
Show bigger map
The GWZ is close to the city centre (around 10-15min walking distance). The city has a convenient bike rental system (that operates in many cities in Germany and other countries: Nextbike) -- and bikes are definitely the way to get around in Leipzig! But there is also a good TRAM-system: the nearest TRAM station is Neues Rathaus or Münzgasse, LVB; for more information see the homepage of the Leipziger Verkehrsbetriebe.
Getting to Leipzig
Arrival by train
Leipzig has railway connections with almost all European cities. Hauptbahnhof Leipzig (Leipzig's main railway station) is located right in the city centre. Most hotels in and around the city center can be reached within 10-15 minutes.
Arrival by plane
The Halle-Leipzig International Airport (LEJ) is served by several major airlines. Deutsche Bahn InterCity (IC) and RegionalExpress (RE) trains, as well as local S-trains, run frequently between the Leipzig/Halle airport and Leipzig main station. The trip takes about 15 minutes. Tickets are available at vending machines on the platforms (be aware that the ticket for the faster IC train is also more expensive). More information about train connections can be found at the homepage Deutsche Bahn.
It might also be a convenient option to fly to Berlin (TXL or SXF). It takes less than 2 hours to get to Leipzig from both airports (there is a high-speed train between Leipzig main station and Berlin main station).
Call for Papers
Iterativity characterizes central phonological phenomena such as syllabification, stress assignment and vowel harmony, but is also crucial to major syntactic processes such as wh-movement (see e.g. Georgi 2017 and references cited there), and found in many morphological constructions (as in grand-grand-grand....mother). The motivation for this workshop is the observation that iterativity is a much more pervasive property of linguistic systems than is usually appreciated. Iterativity plays an important role in highly heterogeneous current debates, for example on the mechanics of syntactic Agree (Béjar & Rezac, 2009), and the evaluation of constraints in Agreement-by Correspondence approaches to phonological assimilation and dissimilation (Rose & Walker 2004, Bennett 2015). Importantly, recent theoretical developments also suggest that the classical notion of iterative rule application in phonology, which presupposes a built-in restriction to some form of strict structural locality is an epiphenomenon of more general mechanisms, either global constraint evaluation and independent constraints on structural coherence (Walker 2014), or a more general iterative optimization mechanism where locality restricts derivations and not structures as in Harmonic Serialism (Kimper 2012).
We invite abstracts on all aspects of grammatical iterativity, including, but not restricted to the following questions:
- How to resolve the tension between formal and descriptive iterativity?
We are especially interested in contributions on phenomena which are intuitively iterative, but might be better captured in a non-iterative way (as by parallel constraint evaluation, see e.g. Walker 2014), and conversely in iterative formalizations of data which are not prima-facie cases of iterativity (see Müller 2019 on affix order).
- Why are comparable processes iterative in one language/grammar, but non-iterative in another? Are there iterativity parameters (as for phonological rules in Archangeli & Pulleyblank 1994, or are iterative and non-iterative processes substantially different from each other (as has been claimed for vowel harmony vs. metaphony, see e.g. Kaplan 2008)?
- Processing and Computation: How is the iterativity of specific processes related to their complexity in human language processing and their computational complexity? (see e.g. McMullin & Chandlee 2018, Keine to appear)
- How does iterativity of single processes relate to more global grammatical iterativity? A case in point are grammatical formalisms where sets of procedures apply iteratively, as in Harmonic Serialism (McCarthy 2010, Müller 2019) or Stratal OT (Kiparsky 2015).
Abstracts should be maximally two pages, including data, references, and diagrams, in at least 11-point font, with one-inch (2,54 cm) margins. Submissions must be anonymous and are limited to 2 per author, at least one of which is co-authored. Only electronic submissions in pdf format will be accepted. Send your abstract via email to jtrommer [æt] uni-leipzig.de by October 15, 2019, 23:59 MET.
Archangeli, D. & Pulleyblank, D. (1994) Grounded Phonology. Cambridge MA: MIT Press.
Bennett, W. (2015) The Phonology of Consonants: Harmony, Dissimilation and Correspondence. Cambridge University Press.
Béjar, S. & Rezac, M. (2009) Cyclic Agree. Linguistic Inquiry 40(1):35-73.
Chen, M. (2004) Tone Sandhi Patterns across Chinese dialects. Cambridge University Press.
Chomsky, N. (1977) On Wh-Movement. In: P. Culicover, T. Wasow, and A. Akmajian (eds.) Formal Syntax. New York, Academic Press.
Dresher, B. E. & Nevins, A. (2017) Conditions on Iterative Rounding Harmony in Oroqen. Transactions of the Philological Society 115(3):365-394.
Georgi, D. (2017) Patterns of Movement Reflexes as the Result of the Order of Merge and Agree. Linguistic Inquiry 48(4):585–626.
Gibson, E. & T. Warren. (2004). Reading-Time Evidence for Intermediate Linguistic Structure in Long-Distance Dependencies. Syntax 7:55–78.
Kaplan, A. F. (2008) Noniterativity is an Emergent Property of Grammar. PhD thesis, UC Santa Cruz.
Kimper, W. (2012) Harmony Is Myopic: Reply To Walker 2010. Linguistic Inquiry 43(2):301-309.
Kiparsky, P. (2015) Stratal OT: A Synopsis and FAQ’s. In: Yuchau E. Hsiao and Lian-Hee Wee (ed.) Capturing Phonological Shades. Cambridge, Cambridge Scholars Publishing 2-44.
Keine, Stefan (to appear): Locality Domains in Syntax: Evidence from Sentence Processing. Syntax.
McCarthy, J. (2010) An Introduction to Harmonic Serialism. Language and Linguistics Compass 4(10):1001--1018.
Kushnir, Y. (2018) Lithuanian Pitch Accent. PhD thesis, Universität Leipzig.
Müller, G. (2019) Inflectional Morphology in Harmonic Serialism, Ms. Universität Leipzig.
Rose, S. & Walker, R. (2004) A Typology of Consonant Agreement as Correspondence. Language 80:475-531.
Smolensky, P. & Goldrick, M. (2016) Gradient Symbolic Representations in Grammar: The case of French Liaison. ROA 1286.
van Urk, C. & Richards N. (2015) Two Components of Long-Distance Extraction: Successive Cyclicity in Dinka. Linguistic Inquiry 46:113-155.
Vaux, B. (2008) Why the Phonological Component must be Serial and Rule-Based. In: B. Vaux & A. Nevins (eds.) Rules, Constraints and Phonological Phenomena. Oxford University Press, 20-61.
Walker, R. (2014) Nonlocal Trigger-Target Relations. Linguistic Inquiry 45(3):501-523.
Deadline for Abstracts: October 15, 2019
Notification of Acceptance: November 1, 2019
Workshop: December 2+3 2019
Institut für Linguistik
jtrommer [æt] uni-leipzig.de
Jochen Trommers Homepage