From 2017 to 2018, I was a member of the AHRC-funded project Prominent Possessors at SOAS University of London and worked with Irina Nikolaeva (PI; at SOAS), Oliver Bond, and Greville Corbett (at the University of Surrey).
The project dealt with possessive constructions in which a noun phrase-internal possessor behaves as if it were external to the noun phrase, for example with respect to agreement or switch-reference.
This means that in certain languages, it is possibly to say something like My bike am fast, where the possessor my controls agreement on the predicate am. This contrasts with a more familiar phrase like My bike is fast, where the predicate is agrees with the third person subject bike.
We addressed the nature of this phenomenon in a number of publications and conference talks, some of which I describe here in more detail. See also my CV for my contributions, as well as the existing project websites at SOAS and Surrey.
“On adjoined possessors” (with Irina Nikolaeva). To appear in Linguistic Inquiry. DOI: 10.1162/ling_a_00370
In this squib, we discuss the behaviour of PIPs in Tundra Nenets. In contrast to regular lexical possessors, PIPs can control switch-reference, and bind possessive pronouns in other DPs in the clause. In addition, PIPs cannot co-occur with certain other DPs in the clause. These patterns are surprising for possessors inside a DP. We suggest that PIPs are actually not fully part of their host DP, but adjoined to it.
More technically, this adjoined position allows them to c-command out of the DP. This is in turn makes it possible to bind possessive pronouns they c-command and participate in switch-reference by controlling the PRO subject of certain converbial clauses. The co-occurrence restrictions follow from the assumption that PIPs are proximate and thus cannot co-occur with another proximate DP in the same clause.
Download the pre-print version.t.a.
“Possessive and non-identity relations in Turkic switch-reference” (with Irina Nikolaeva). To appear in Studies in Language.
We discuss switch-reference in a number of Turkic languages. Nearly all Turkic languages have the converb (a gerund-like non-finite verb form) ending in -(V)p which usually requires its subject to corefer with the subject of its superordinate clause. However, like in other languages, this same-subject requirement is sometimes overridden: in some languages, for example Turkish and Bashkir, this is possible when the subject of the converbial clause and the subject of the superordinate clause are in a possessive relation with each other (similar to what we found in the Glossa paper, see below).
In several other Turkic languages, however, the converb is also found with two disjoint subjects: we argue that this is only possible when the event expressed by the converbial clause and the event expressed by the superordinate clause are causally and/or temporally linked (inspired by Andrew McKenzie's work).
Finally, we sketch a diachronic pathway along which the Turkic languages could have developed based on the behaviour of -(V)p in Old Turkic.
You can download the pre-print version.2019
“Possessors in switch-reference” (with Irina Nikolaeva). Glossa: a journal of general linguistics, 4(1), 81. DOI: 10.5334/gjgl.865
In this paper, we discuss a type of non-canonical switch-reference, in which possessors seem to act as switch-reference pivots. Roughly, this means that in certain languages, it is possible to say Walking and walking, my shoes wore out to mean I was walking and walking and my shoes wore out, while this is impossible in other languages.
We discuss the cross-linguistic distribution of this phenomenon and propose two generalisations: some languages only allow nouns expressing part-whole relations to appear in such constructions (such as body parts like leg), while others allow alienably possessed nouns like shoe, too. When a language allows the latter, it seemingly always allows the former as well.
In addition, in languages which allow both types of nouns to participate in switch-reference, the possessive relation must be overtly coded as such, for example using a possessive pronoun, an overt possessive phrase, or a possessive suffix. Languages which only allow part-whole nouns in these constructions do not have to code possession overtly.
You can download the paper from Glossa.
“Towards a typology of prominent internal possessors” (with Irina Nikolaeva and Oliver Bond). In Prominent internal possessors, ed. by András Bárány, Oliver Bond, and Irina Nikolaeva. Oxford: Oxford University Press, 1–38.
In this introductory chapter to the OUP volume on Prominent Possessors, we discuss the empirical phenomena we refer to as prominent possessors, present a typology of these phenomena, and discuss some potential analyses of prominent internal possessors.
E-mail me for a copy!
“Proximate possessors” (with Irina Nikolaeva). In Prominent internal possessors, ed. by András Bárány, Oliver Bond, and Irina Nikolaeva. Oxford: Oxford University Press, 228–258.
In this paper, we focus on the Uralic language Tundra Nenets, in which a particular type of possessor can interact with switch-reference. Tundra Nenets distinguishes two types of lexical possessors: only one of these controls agreement on the possessed noun (like in Hungarian or Turkish, for example) and only this type of possessor can participate in switch-reference.
We argue that this is a consequence of the syntactic structure of the possessive phrase: agreeing possessors are high enough inside the possessive phrase to control embedded subjects.
We also discuss restrictions on the distribution of these possessors in terms of obviation, a grammatical reference-tracking mechanism that usually distinguishes different third person noun phrases from each other. We show that the distribution of prominent possessors in Tundra Nenets can be explained as a consequence of an obviation system in the language.
E-mail me for a copy!
Database of Prominent Internal Possessors (with Oliver Bond and Sandy Ritchie). University of Surrey. Access the database here.
In this online database, we will shortly make available the data we collected from the literature and through work with native speakers during the project. Each example is tagged with respect to the properties of the possessive relation, and the interaction of the possessor with the sentence, for example with respect to agreement or switch-reference.
The database has hundreds of examples from 52 languages that are fully glossed, can be searched and cited straight from the website.