Joan Mascaró Altimiras
Phonologically (and syntactically and lexically) conditioned allomorphy
Phonologically conditioned allomorphy has been analyzed as an instance of The Emergence of The Unmarked (TETU). Allomorphs are listed in the lexicon with no contextual subcategorization, and the phonology chooses the allomorphs that yield a less marked structure, depending on the context in which they appear. I will analyze two specially difficult cases of phonologically conditioned allomorphy, Haitian definite suffix selection and Northeastern Central Catalan s -deletion.
In the first case allomorph selection seems to be governed by unnatural phonological conditions: the allomorph la appears after consonants, as in liv-la 'book-the', and the allomorph a appears after vowels, as in papa-a 'father-the'. I will show that once we allow partial ordering allomorphs in the lexicon (ordering reflecting relative markedness), natural alignment conditions derive the right results.
The second case regards s- deletion. Here "deletion" is subject two three heterogenous conditions: a phonological condition ( s must be final in a complex coda and followed by a consonant), a lexical condition ( s must be the plural morph), and a syntactic condition (the lexical element ending in s must be prenominal). Thus in bon- s vin-s blanc-s franceso-s 'good-pl wine-pl white-pl French-pl', the plural marker in prenominal bon- s doesn't appear, but plural markers in postnominal vin-s and blanc-s show up. Assume N is final within the DP and raising causes agreement with elements appearing to its right, but agreement with the rest takes place at PF. This forces postnominal agreement but leaves prenominal agreement subject to PF conditions. The bare root bon will be preferred to the number-inflected bon-s because it doesn't violate the marked structure CsC even if it violates (PF) Concord. Other cases of prenominal-postnominal asymmetry will be briefly discussed.
Omer Preminger
Basque Ling-Lunch Redux
Part II of the Basque Ling-Lunch series will begin with a recap of Episode One- attendance of previous talk will *not* be assumed! - where it was shown that apparent cases of Long-Distance Agreement (LDA) in dialectal Basque do not in fact constitute a case of true LDA (construed as agreement that spans across the boundaries of established locality domains). I provide evidence that the cases in question fall into one of two categories: either (i) the apparent LDA relation is comprised of two separate agreement relations, "stacked" on top of one another, each of which is perfectly well-behaved with respect to the relevant locality restrictions; or (ii) the agreement relation in question spans the boundaries of neither DP nor CP, and is thus typologically unexceptional.
In this brand new episode, I turn to the distinction between Agree (conceived of as a relation between a probing head and a goal) and clitic-doubling (conceived of as the generating of a clitic which is matched in phi-features with a full argument DP). Certain asymmetries in the reach of so-called LDA when targeting dative noun-phrases and targeting absolutive ones suggest that absolutive agreement is an instance of Agree proper, whereas the dative (and ergative) exponents on the auxiliary are the result of clitic-doubling. In the climactic finale, I present an independent diagnostic for distinguishing Agree from clitic-doubling: when so-called LDA fails to obtain, the agreement-bearing form of the auxiliary is obviously ruled out; the question is whether what shows up is default agreement on the corresponding exponent, or rather an auxiliary form that lacks the relevant exponent altogether. I show that precisely in those relations hypothesized here to be Agree relations, failure of the relation results in default agreement-whereas in those relations hypothesized here to be clitic-doubling, failure results in the wholesale absence of the relevant exponent.
Critics rave:
"Mildly disappointing."
- Morphology and Agreement Weekly
"Seriously, I'm gonna punch you in the face."
- an anonymous colleague
Thomas McFadden
DPs aren't licensed, they're selected (or not)
The premise behind abstract Case is that DPs are in some way defective and thus require explicit licensing to appear. Among other things, this is meant to explain the distribution of subject DPs -- overt and non-overt -- in certain familiar finite and non-finite clause-types. In this talk I consider data which militate against this explanation from a wider range of clause types in Irish, Portuguese, Tamil and other languages, in addition to English. I propose that matters are simpler if we view the distribution of different DP types in subject position the other way around -- overt DP subjects are generally available, but are explicitly blocked in certain clause types. Adopting and extending some recent ideas about the distribution of PRO, I argue that this can be modeled in terms of selection by (clausal) functional heads for (interpretable features which characterize) the specific DP types. Among other things, this allows a straightforward account of those clause types where both overt DPs and PRO are possible without any detectable difference in tense, agreement or other factors related to finiteness. Here there is simply no selectional restriction.
Edward Garrett
Impersonal Subjects Have No Taste
Sverre Johnsen
Binding in complements of perception verbs
Norwegian is one of the most discussed languages in the literature on reflexive binding, with its system of simple and complex reflexives seg vs. seg sjøl. This talk will present new Norwegian data showing not only a previously unknown pattern in Norwegian for long-distance binding of seg, but also a generalization for long-distance binding that has not been reported in any language before. The data will reveal that the reflexive seg is exceptionally allowed in complement clauses only if the clause is the complement of a perception verb.
Across languages, perception verbs exhibit a special behavior in a different domain, namely in terms of tense dependency. In languages without the phenomenon known as sequence of tense, a past tense in the complement clause of another past tense verb only allows a past-shifted reading. This is the case in Russian and Hebrew. In the complement clause of a perception verb, on the other hand, a past tense has a preferred simultaneous reading, meaning that its temporal interpretation fully depends on the tense of the matrix verb. In languages where 'sequence of tense' generally exists, such as English and Norwegian, the tense in complements of perception verbs still behaves differently from other complements in that it shows a greater dependency on the matrix clause tense.
I will adopt the common view of tense dependency as being caused by structural syntactic binding of tense (Enç 1987). With the analysis of reflexive binding within the minimalist framework developed in several papers by Reuland, I will show that interclausal reflexive binding can be a natural fall-out of interclausal tense dependency.
Conor Quinn
Applicative and antipassive: Algonquian transitive “stem-agreement” as differential object marking
The standard Algonquianist model of transitive morphosyntax has remained essentially unchanged since Bloomfield 1946 introduced a fundamentally descriptive set of categories and terms. The analysis informing these traditional categories is problematic for a number of distinct reasons. Here we offer an alternative approach, one that accounts for the observed properties of Algonquian transitivity in a more concise and less language-family-specific way, while simultaneously doing away with a number of longstanding problems introduced by the traditional analysis.
The specific new claim is this: contrary to standard Algonquianist analysis, the view that the terminal element of categorically transitive stems---the morphological element known as the transitive Final---agrees for the gender of the internal argument is unnecessary and untenable. What looks like stem-internal agreement for the grammatically [+animate] or [ animate] feature of the notional direct object argument---giving rise to contrast between Transitive Animate (TA) and Transitive Inanimate (TI) stems---is better identified as two respective feature-driven syntactic constructions, both of a much more cross-linguistically motivated kind: dative-accusative syncretism, in the first case, and antipassivization, in the second.
With this, the TA vs. TI contrast becomes an instance of grammaticalization of precisely the syntactic argument-prominence structures commonly triggered in the case of high-prominence ([+animate]) and low prominence ([-animate]) internal arguments respectively. In short, what appears to be stem-agreement is in fact differential object marking (Aissen 2003, inter alia).
This view of the system has not been immediately obvious because the two patterns manifest in Algonquian languages through a head-marking morphosyntax, rather than through the primarily dependent-marking strategies (i.e. nominal case- and adposition-marking) by which both are better known. The overall contribution of this new analysis is twofold: a de-exoticization of the Algonquian transitive system within the context of cross-linguistic comparison, and a removal of several problems of overgeneration and underprediction in the traditional account.
Enoch Aboh
Multiple copies and parallel chains
This paper discusses predicate fronting with doubling cross-linguistically and demonstrates that it is an instance of parallel chains in the sense of Chomsky (2005). Under this analysis, what superficially looks like a spell out of multiple copies within a single uniform chain turns out to be the expression of two simultaneous chains of which the heads only are spelled out. The analysis extends to subject intrusion in Dutch, as discussed in Barbiers and van Koppen (2006), and to auxiliary doubling in English child language.
Sabine Iatridou
Negative DPs and Scope Diminishment: Some Basic Patterns
Lasnik 1999 argued that Negative DPs like "no student" do not undergo reconstruction in A-chains. Lasnik drew the conclusion (from this and other data) that there is no reconstruction in A-chains at all. We disagree with this larger conclusion but do think that Lasnik was right for the particular case of NegDPs, We explore the significance of this constraint on reconstruction.
This work was presented at NELS 38 last fall.
Robert Ladd
Correlations between Interpopulation Differences in Two Human Genes (ASPM and Microcephalin) and the Distribution of Lexical and/or Grammatical Tone
We consider the relation between allele frequencies and linguistic typological features. Specifically, we focus on the derived haplogroups of the brain growth and development-related genes ASPM and Microcephalin, which show signs of natural selection and a marked geographic structure, and on linguistic tone, the use of voice pitch to convey lexical or grammatical distinctions. We hypothesize that there is a relationship between the population frequency of these two alleles and the presence of linguistic tone and test this hypothesis relative to a large database (983 alleles and 26 linguistic features in 49 populations), showing that it is not due to the usual explanatory factors represented by geography and history. The relationship between genetic and linguistic diversity in this case may be causal: certain alleles can bias language acquisition or processing and thereby influence the trajectory of language change through iterated cultural transmission.
Dong-Whee Yang
Phase-internal Scrambling and Edge Feature Movement
In this paper the notion of EF (edge feature) movement (Chomsky 2005) is characterized as pure internal merge, which is optional, hence induces D(=discourse) effects according to (1), leading to the chain condition of EF-movement (2):
Given this characterization of EF-movement, it is shown that not only phasal scramblings but also phase-internal ones are EF-movements, offering a unified account of long-distance and very short clause-internal scramblings. Furthermore, given that Agree and EF-movement are separable for an Agree-movement, i.e., need not occur together, it is shown why an optional Agree-movement may function as an EF-movement inducing D-effects, along with Agree in situ, like OS and Subject Raising in languages like Icelandic, Korean, etc. Thus, it is captured that all and only optional movements induce D-effects. This paper also offers an explanation for why EF-movements are not subject to the minimality-type constraints though subject to island-type constraints. Note that EF-movements are only subject to the architectural conditions of the minimalist theory, and I claim that island-type constraints are essentially those against violating the architectural conditions of the grammar like the PIC unlike the minimality-type constraints. This paper also shows how the EF-movement theory of (2) offers optimal accounts for problems like successive cyclic A?-movements (Bokovic 2007, Preminger 2007, Heck and M|ller. 2000) and the criterial freezing (Rizzi, 2004). This paper also suggests constraints on movements based on the chain condition of EF-movement (2), accounting for why idiom chunks may not optimally undergo EF-movement though corresponding non-idiom chunks may. Lastly, this paper claims that reconstruction and covert movement are the two sides of the same coin in the minimalist theory, offering a new analysis of the two phenomena, given the characterization of EF-movement in this paper.
References
Boskovic, Zeljko. 2007. On the locality and motivation of move and agree: an
even more minimal theory. Linguistic Inquiry 38:589-644.
Chomsky, Noam. 2001. Derivation by phase. In Ken Hale: A Life in Language, ed.
by M. Kenstowicz. 1-52. The MIT Press.
Chomsky, Noam. 2005. On phases. Ms., MIT.
Heck, Fabian and Gereon M|ller. 2000. Successive cyclicity, long-distance
superiority, and local optimization. WCCFL 19:218-231.
Preminger, Omer. 2007. Toxic syntax: yet another theory of syntactic movement.
Ms., MIT.
Rizz, Luigi. 2004. On the form of chains: criterial positions and ECP effects.
Ms.
Yang, Dong-Whee. 2008. On edge feature movement. Ms., MIT.
(downloadable)
Elena Benedicto, MIT/Purdue U.
H2-shift as a morphological device for Switch Reference in (HK)SL
Serial Verb Constructions (SVC) are present in a good number of languages of the world. Some of them combine SVCs with a system of Switch Reference, by which the inflectional system of the verb indicates whether the following subject has the same or different reference from its own subject. In this talk, I will contemplate the hypothesis that H2-Shift (that is, changing the dominant hand when signing a predicate in an SVC) acts as a morphological device for Switch Reference in Hong Kong Sign Language (HKSL). The use of H2-Shift in HKSL contrast with the behavior of H2 in ASL, where the complementary distribution occurs. I will consider potential sources for this parametric variation, among them, the different structural representation of SVCs in these languages. This will have consequences, in general, for the structure that we posit for SVCs and for the restrictions we impose on the complexity of the sub-clausal trees in an SVC.
This is work in progress, in collaboration with Dr. Gladys Tang, of the Center for Sign Linguistics and Deaf Studies at the Chinese University of Hong Kong, and her team.
Alya Asarina & Kirill Shklovsky
Optativity in English and Other Languages
As shown in Rifkin (2000), across many languages, a counterfactual (CF) conditional in combination with "only" can be used to express a wish. For example, in English, (1) indicates that the speaker wants to be hospitalized (pragmatically odd):
(1) If only I broke my leg, I would be hospitalized.
The standard accounts of the semantics of CF conditionals and "only" cannot be straightforwardly combined to derive the correct meaning, which presents a puzzle.
In this talk, we describe some properties of these optative constructions. We present generalizations made by Rifkin (2000), as well as novel observations. We show that:
- the "wish" part of the meaning is a presupposition, and not an entailment
- there are some cross-linguistic correlations between the availability of "if only" optatives and what other uses are available for "only"
- optatives cannot be embedded in negative environments
We conclude with some possible proposals for what the semantics of "if only" may be.
