Quantified numeral subjects in Russian may famously trigger plural or singular verb agreement. Generative accounts (Pesetsky (1982), Franks (1995), Bošković (2006)) tie the variation to Case and the DP/QP distinction. Corpus-based accounts (Revzin (1978), Corbett (2000), Robblee (1993)), in addition to precedence and definiteness/specificity, note a strong correlation between agreement choice and the animacy of a QNP subject. In this paper, I attempt to reconcile the original generalizations in both linguistic traditions by proposing an account whereby (i) the animacy condition on agreement is an argument structure effect (ii) the connection between Nom case and agreement with QNP subjects is captured by Case as accessibility condition for agreement (Marantz (1991), Bobaljik (2008), Baker (2010) Baker and Vinokurova (2011) but contra Chomsky (2000), (2001)) (iii) definiteness/specificity effects with agreement follow from Diesing’s (1992) Mapping Hypothesis and a locality condition on semantic agreement.

This corpus-based keyword analysis study investigated the letters to the shareholders from two commercial banks, Bank of America and Citigroup, over a three-year period from 2008, 2009, and 2010. The letters were compiled to facilitate a diachronic assessment of profit/loss reporting from two prominent institutions over a time period in which the recession commenced, peaked, and concluded. To conduct the analysis on the node texts, two sets of reference corpora were compiled. The first reference corpus set consisted of the letters to shareholders from eight consistently high-performing corporations not within the commercial banking industry for each of the three years; the second reference corpus set consisted of the letters from the 10 banking institutions that also appeared in the Fortune 500 listings for the three period. The corpus-based analysis revealed that in years of low performance companies create messages that assert a vision and forward a strategy for ensuring future success while also deflecting responsibility for past failure. In contrast, when companies perform well, the keyword lists display a clear tendency of the company and the author to accept praise and responsibility for high performance.

In the present study, a corpus of more than 18,000 online dating ads (downloaded from Craigslist.com, ~ 1.4 million words) is used to investigate differences in language use between men and women in the online dating context. Few studies have investigated gender differences in written texts, Newman, Groom et al. (2008), Mulac and Lundell (1994) and Koppel, Argamon et al. (2002) being the notable exceptions. These papers, however, differ remarkably in methodology and results. In the dataset studied here, regression analysis reveals marked differences the use of linguistic features such as emoticons or abbreviations. Writer gender and addressee gender emerge as predictors of variation.

In this research, we explore the linguistic structure of the Spanish of Heritage Speakers, those who have acquired Spanish as the home language in a minority language context (Iverson, 2010). We contribute to the discussion of the properties of Heritage Languages here by examining Preposition Stranding in Heritage Speakers versus native monolingual speakers of Spanish. We claim that the distinct behavior of Heritage Speakers of Spanish supports the claim that Heritage Languages may differ from native monolingual language in the narrow syntax, affecting uninterpretable features of the grammar.

Basic spatial and temporal meanings of the preposition after in Standard American English and Highland Scottish English are accounted for in this proposed semantics of the preposition. In addition, two seemingly aberrant meanings in Highland Scottish English and Lowlands Scots/ Lowlands Scottish English are discussed; the proposed semantics accounts for these meanings as well. The meaning of after is related to the meanings of first and subsequent, which are given definitions in terms of points of reference.

Isbukun, a major dialect of Bunun, is one of the Austronesian languages spoken in Taiwan. According to the Taiwan government statistics in 2009, Bunun had a population of 51,447, around half of which were registered as Isbukun. As Mandarin and Southern Min are the predominant languages, the Austronesian languages in Taiwan including Bunun are endangered. This study investigates word stress in Isbukun Bunun from the perspective of Optimality Theory. In particular, the stress patterns of monomorphemic and compound words, derivational and inflectional suffixed words, and words with clitics are explored. In Isbukun Bunun, a single quantity-sensitive trochee formed at the right edge of a word. Consequently, prefixation is generally irrelevant to footing, whereas suffixation closely interacts with stress placement. This paper presents two types of extrametricality along with quantitative adjustment that are found in suffixed words. Morphological extrametricality prevents inflectional suffixes and clitics from being footed and thereby prosodically distinguishes derivation and compounds from inflection and clitics. Positional extrametricality avoids forming a foot at the left edge of a root, which only emerges in derivational words, because inflectional words are required to fulfill morphological extrametricality first. In addition, since feet are constructed at the end of a word and derivational suffixes are allowed to be footed, quantitative changes take place in derivational suffixation: adjacent vowels with the same quality merge into one when two vowels come from different morphemes (i.e. the final segment of the stem and the initial segment of the suffix), and moras are deleted if otherwise the number of syllables in a word would increase. To account for the morphological extrametricality, a pair of output-output faithfulness constraints are used: a higher ranked OO-IDENT(stress)INF with an index referring to inflectional suffixes and clitics, and a lower ranked clone OO-IDENT(stress) without an index. The constraint ranking OO-IDENT(stress)INF >> ALL-FT-R >> OO-IDENT(stress) generates inflectional words with stress on the same syllable as in their stems, while derivational words follow general footing principles. As for the positional extrametricality, an anti-alignment constraint *ALIGN-L(Root, FT) is proposed, which concerns positions of feet only with respect to the root, rather than the stem that may be polymorphemic. The constraint ranking OO-IDENT(stress)INF, DEP-μ >> *ALIGN-L(Root, FT) >> *STRUC-σ confines the emergence of non-initiality in derivational words with roots not smaller than two moras.

In this paper, I examine the hitherto unnoticed issue of what syntactic categories mean may take as complements, analyzing the syntactic structure of sentences like What do you mean that I’m a liar? from the minimalist perspective. On the basis of the intuition that what starts out as the object of mean, I consider the mechanism of Case-licensing of wh-arguments. Noticing that under the current minimalist framework, object shift to SpecvP and vP-adjunction result in exactly the same configuration with respect to chain length, I consider how Case-licensing of wh-arguments in the complement clause of mean circumvent improper movement characterized in terms of the traditional A/A-distinction. I also discuss the Doubly Filled COMP Filter and the Superiority Condition in relation to the syntax of mean, both of which are shown to be consistent with the proposed analysis.

In this paper I propose and motivate a syntactic analysis of Hixkaryana (a Carib language spoken in the Amazon in Brazil), drawing on the extensive, linguistically-informed fieldwork of Desmond C. Derbyshire (1979, 1985, inter alia). Hixkaryana displays basic/unmarked Object Verb Subject (OVS) word order, which is found in very few languages of the world (Dryer 2008). There are three main components to the proposal presented here. I argue that the syntax of Hixkaryana involves (i) a marked hierarchy of agreement projections, AgrO over AgrS; (ii) movement of the subject to a high topic position; and (iii) fronting of the rest of the clause over the subject. This analysis accounts for a constellation of properties in Hixkaryana, including the surface order of constituents (OVSX, where X is an adjunct PP or AP), surface constituency (the object and verb form a constituent exclusive of the subject), verbal morphology (agreement is a prefix while all other inflectional affixes are suffixes), structural relations (the subject c-commands the object and obliques/adjuncts), the position of particles (which are either in second position or invariantly post-verbal), and exceptional OSV word order (triggered by the first person exclusive pronoun amna). OVS languages, like Hixkaryana, are important for syntactic theory because they likely have special insights to contribute, given how rare they are; however, OVS languages receive very little attention in the literature. This paper aims to call attention to OVS word order as a real linguistic phenomenon that must be accounted for in mainstream linguistic theory.

The goal of this paper is to show that internal arguments of verbs in Turkish do not uniformly occur in the complement position of the verb (contra e.g. Perlmutter 1978, 1989). We examine syntactic positions of bare arguments in Turkish on the basis of aspectual (Aktionsart) properties of VPs (e.g. Vendler 1967) and prosodic structure. Looking at where low adverbs appear, we propose that there are (at least) two different positions where bare internal arguments can occur in syntax – bare internal arguments of Turkish achievements occur in SpecVP while those of accomplishments occur in the complement position of V. This proposal is also supported by prosodic evidence.

This paper presents a formalization of view that auxiliary verbs such as be are in some sense default verbs. On the basis of languages in which auxiliaries arise only in certain combinations of inflectional categories (Latin, Kinande), it is argued that auxiliary be is not present in the syntax, but is instead a morphological strategy for realizing “stranded” inflectional features. A model of verbal inflection that implements this approach to auxiliaries is developed, providing a unified analysis of the auxiliary pattern found in languages of the Latin/Kinande type with the more familiar pattern of languages such as English.