WS5: Differential Internal Possessors

Workshop 5: The syntax of Differential Internal Possessors


Irina Nikolaeva (SOAS, University of London)
András Bárány (SOAS, University of London)
Oliver Bond (Surrey Morphology Group, University of Surrey)


Many languages have more than one possessive construction in which the possessor is internal to the same syntactic phrase as the possessum. This phenomenon is sometimes referred to as Differential Possessor Marking or Differential Possessor Expression (e.g. O’Connor 2003), by analogy with Differential Argument Marking. Although recent years have seen a growing interest in the study of Differential Argument Marking (Aissen 2003; de Hoop & de Swart 2008; Dalrymple & Nikolaeva 2011; Seržant & Witzlack-Makarevich 2016; among others), Differential Possessor Marking is by far less systematically investigated. Especially little is currently known about the syntactic effects differential internal possessors may have. This perspective will add a new dimension to the traditional typological studies of internal possessive constructions, which have mostly concentrated on the morphosyntactic encoding of their components or the (relational) semantics underlying their distribution.

The aim of the workshop is to bring some important issues regarding the cross-linguistic variation in the syntax of internal possessive constructions to the attention of the typological community. Without claiming to encompass the whole range of Differential Possessor Marking phenomena, the workshop will focus on two partially interrelated questions that appear to be typologically understudied.

First, the discussion will largely be on the behavioural syntactic properties of differential internal possessors. For example, in Turkish (Turkic) the possessor normally bears the genitive case and the possessed noun may host possessive agreement, although it is not obligatory. Constructions without agreement require discourse contexts which establish the possessor as a clearly identifiable referent, so that it cannot be indefinite or quantified. Unlike the possessor in constructions with possessive agreement, the possessor that does not trigger agreement is incompatible with restrictive relative clauses (Öztürk et al. 2015):

Ex. 1

This indicates a split in the syntactic behaviour of possessors within the possessive phrase.

Even more striking are cases where an internal possessor exhibits syntactic effects outside of its own phrase and participates in syntactic processes which typically target a phrasal head, such as predicate-argument agreement (e.g. in Maithili, Stump & Yadav 1988, see (2); Ngumpin-Yapa, Meakins & Nordlinger 2017; Chimane, Ritchie 2017) or switch-reference (Turkic, Bergelson & Kibrik 1995; Aleut, Bergsland & Dirks 1981; Tundra Nenets, Nikolaeva 2014; Uto-Aztecan, Hill 2016).

Ex. 2 

As in Turkish, in Tundra Nenets (Uralic) the lexical possessor ‘optionally’ triggers third-person agreement on the head noun. Possessors that trigger agreement within the subject noun phrase can control same-subject converbs, see (2). Thus, the agreeing possessor of the subject behaves in part like a clause-level element, even though no independent evidence exists that it is external to the possessive phrase.

Ex. 3 

These data are theoretically challenging, yet internal possessors that exhibit properties of a clause-level element have so far not been sufficiently recognized in linguistic typology.

Second, the workshop will focus on functional factors determining differential expression of internal possessors in their relationship to syntax. The alternative possessive constructions are usually specialized on the expression of possessive relations of a different semantic nature, as is observed in languages with possessive classifiers or an alienability opposition (Chappell & McGregor 1996), or they reflect a split in the lexicon and the inherent semantic properties of the possessor (e.g. lexical vs. pronominal possessors). The languages we want to focus on here differ from many of those in Chappell & McGregor (1996) in that we are specifically discussing internal possessors.

Some languages exhibit at least two internal possessive constructions for the expression of (roughly) identical possessive relations, with the choice between them being conditioned by subtle discourse-related factors which reference properties of the situation and context. Such alternations are found in English (Saxon vs. ‘of’-genitives; see a.o. O’Connor 2003), Hungarian (nominative vs. dative possessors; Szabolcsi 1994), and German, where possessors can be marked by dative case or a dative-like preposition (Weiß 2008). For some languages such cases raise the question to what extent the non-syntactic factors motivate possible difference in the syntactic behaviour of internal possessors. For instance, in Maithili agreeing possessors are associated with contrastive focus and prominence in terms of honorific degree. The Tundra Nenets example (3) falls under the large notion of non-canonical switch-reference marking (cf. Broadwell 1997; Camacho 2010; Keine 2013; de Sousa 2016), in which the possessor of the subject functions as a pivot of the same-subject relation. Since the possessor which triggers agreement is associated with higher level of discourse prominence (possibly some version of topicality or proximate status) than the non-agreeing possessor, this may suggest that the switch-reference mechanism in Tundra Nenets is not oriented exclusively towards the structural source (subjecthood).Non-canonical switch-reference marking has led various researchers to propose that more abstract semantic notions that pertain to the whole clause may lie at the centre of the switch-reference phenomenon (e.g. Stirling 1993; McKenzie 2012).Research on differential internal possessors in switch-reference systems will contribute to the discussion of how switch-reference should be modelled in theoretical terms.

Having these two large and novel research questions in mind, we aim to bring together for the first time researchers working on these topics and propose that the workshop should concentrate on the following more specific aspects:

  • syntactic behaviour of differential possessors within the possessive phrase
  • grammatical interaction between internal possessors and a larger syntactic domain;
  • discourse factors that affect the choice between alternative internal possessive constructions within one language to what extent the factors that determine differential coding of internal possessors are analogous to DAM?
  • recurrent cross-linguistic patterns and parameters of variation in discourse-conditioned differential internal possessors;
  • correlations between functional properties and syntactic prominence of internal possessors.

Selected references

  • Chappell, H & W McGregor (eds.). 1996. The grammar of inalienability. De Gruyter.
  • McKenzie, A. 2012. The role of contextual restriction in reference-tracking. UMass Amherst PhD diss.
  • Nikolaeva, I. 2014. A Grammar of Tundra Nenets. De Gruyter.
  • Öztürk, B et al. 2015. Possessive-free genitives in Turkish. Turcologica 103, 189-203. Harrassowitz.
  • Stump, G & R Yadav. 1988. Maithili verb agreement and the control agreement principle. CLS 24, 304-321.


Online user: 1