Coordination project operates as the SFB’s central administrative and scientific hub, supported by one FWF-funded staff member (Sarah Melker). Additionally, two PhD students are funded by the University of Graz for three years (Martina Barili and Nastaran Divani). Administrative responsibilities encompass organizing SFB events, handling employment procedures, coordinating gender-mainstreaming measures, ensuring open access for SFB publications and data, promoting equity and inclusion, and strengthening public relations through diverse channels. Scientifically, the project is tasked with integrating subproject outcomes and theory development, spotting and capitalizing on synergies within empirical studies, providing methodological and statistical support, monitoring scientific progress, managing publications, and laying the theoretical groundwork for the SFB’s next four-year phase.

The role of self in pronominal coreference Seeking for a better understanding of pronoun binding, the subproject tackles the question why cross-linguistically intensified pronouns (IntPros) tend to grammaticalize into reflexive anaphors. The hypothesis is that binding functions at the level of information structure roles: the topic binds the focus. The intensifier is the element that drives this process: a focal reflexive bound by the topic. While this is transparently the case in compositional IntPros, redundant and deficient interface operations lead both to the licensing of local rather than sentential topics as binders (see P06 for a simlar process in clitic doubling) and to the facultative interpretation of focal feature on the bindee, giving rise to vP and DP level reflexivization. Plausibly, the cycle closes with the full grammaticalization of IntPros via anaphors into reflexives, and the reemergence of fully interpreted IntPros. In result, IntPros as the strongest pronouns do not follow the general tendency of strength regarding binding that the stronger a pronoun, the larger the structural domain in which it must be free (Cardinaletti and Starke, 1999): although strong, IntPros contain a focal reflexive, which imposes local binding by the topic. This sets the choice of target language types to Malayalam with IntPros both for intensification and as anaphors, English with IntPros mainly as anaphors, requiring additional focal pronouns for intensification, German with simple anaphors and an intensifier that strengthens their reflexivity and BCMS with simple reflexives and fully compositional IntPros. The main hypotheses directly address binding, a hallmark of R and D, as a way how grammar exploits pragmatics to economize the expressed information.

Animacy features and the flexibility of pronouns investigates the mapping between morphosyntactic and semantic features of French and Spanish personal pronouns from the perspective of R and D. These pronouns are highly flexible expressions due to their minimal descriptive content and additional flexibility arises as their animacy features are not invariable (e.g., French strong pronouns are mostly used for human referents, but they can refer for inanimates under certain conditions (Heidinger, 2019)). The subproject assumes that strong and weak pronouns are specified differently for animacy (Cardinaletti and Starke, 1999, Dobrovie-Sorin, 1999), and scrutinizes how the pronouns’ animacy features are then manipulated via R and D to increase their flexibility and allow for noncanonical mappings (e.g., strong pronouns referring to inanimates). The subproject thereby investigates how R and D equip core grammar to deal with general cognitive pressure via additional flexibility; e.g., adaptations in pronoun choice triggered by requirements of anaphora resolution. The animacy features of French and Spanish personal pronouns (mostly in PPs) offer a window into the range and restrictions of such additional flexibility. The comparative perspective on French and Spanish allows the subproject to pursue yet another theoretical objective directly related to the SFB’s goals: The pronoun inventories of French and Spanish share the opposition between strong and weak pronouns. The languages differ, however, with respect to the availability of weak forms in PPs. This provides an ideal testing ground for how R and D work in minimally different pronoun systems.

Mismatches in binding and coreference focuses on two empirical domains—fake indexicals (and other bound pronouns) and pronouns showing covariance without binding. P03 investigates the morphological, syntactic and semantic properties of these types of pronouns and unites them in a model where syntax mediates between LF and PF and transfers features exhaustively or redundantly to the interfaces. Fake indexicals are shown to involve an R operation from syntax to LF, whereas non-bound pronouns employ R from syntax to PF. The two otherwise quite distinct phenomena are thus mirror images of each other in the framework of R/D of the SFB. The starting point of the subproject involves a detailed syntactic investigation of the structure of pronouns (e.g., the size and types of syntactic projections); the features involved in syntax, morphology, and semantics; any mismatches arising between the components; and the origin of features (binding, non-expressed structure). The initial hypotheses are that i) fake indexicals involve syntactic binding by the true indexical which is enabled by (syntactic) focus binding, and ii) that non-bound pronouns are underlyingly definite noun phrases in syntax which undergo a process of ellipsis. The subproject is directly relevant to all questions raised by the SFB and aims to contribute a model that allows the formalization of the R/D mechanisms. In addition to the theoretical focus of P03, it also involves an empirical component, a detailed survey of fake indexicals in Romance and Slavic. P03 is connected—empirically and/or theoretically—to most other subprojects and several collaborations have been envisaged already.

The (Non-)Deficiency and (Non-)Redundancy of Clitic Pronouns

Focusing on Clitic Doubling (CD), the subproject asks the foundational question of the ontology of the range between agreement, clitics and weak and strong pronouns and the deep nature of the categories included. The strong hypothesis is that these all instantiate the same grammatical class of items, which however differ from each other in the size of the underlying structure, as well as what features trigger the realization of various degrees of strength. Hypothesizing that the answer to the latter question is to be sought in Kiparsky’s 2008 D-hierarchy (Kallulli, 2016, 2018, 2019), and postulating universal realization of all the items included in the range, the subproject sets to model the strength of realization in terms of various types of D and R, including operations such as impoverishment, or rules of exponence. In particular, it is hypothesized that the responsible feature for all repeated realizations is [topic], but that in the process of grammaticalization it gets weakened. Ultimately doubling is generalized and driven by mere syntactic locality. What begins as topic-doubling in a certain position ends as agreement with the highest argument or universal clitic doubling. Depending on various factors, among which crucially the degree of grammaticalization and the syntactic position, languages display doubling of various degrees of strength, from different zero elements, to agreement, to clitics. The main empirical puzzles targeted by the subproject concern binding, reconstruction effects, resumption, the Person-Case-Constraint and (anti-) locality effects associated with clitic pronouns. In comparing languages with and without clitics or CD regarding the underlying semantic features informs Q1 and Q3, the way in which CD affects binding relations is relevant to Q2 and the analysis of LF and PF effects related to CD inform Q4.

Deficiency and redundancy in Iberoromance pronominal systems

The subproject investigates the variation of accusative clitic doubling and null direct objects in and across Spanish and Portuguese varieties. While the set of pronominal elements and the available surface configurations in these varieties overlap to a considerable extent, the differences between them is striking, challenging unified theories (Maddox, 2021). Moreover, many of the generalizations underlying unified theories require robust empirical verification. Building on the work of Di Tullio et al. (2019) on accusative clitic doubling in Rioplatense Spanish, the subproject addresses further research gaps regarding null objects in Spanish varieties and clitic doubling in Portuguese, such as the recently “discovered” system of accusative clitic doubling in the Mineiro variety of Brazilian Portuguese (Machado Rocha and Ramos, 2016). The subproject will use parallel experimental studies to achieve a highly comparable dataset for at least four varieties (Rioplatense and Peninsular Spanish, Mineiro Brazilian Portuguese and European Portuguese) and explore the “minimal pairs” of contrasts found in them as a window into how R and D play out in the respective systems. While the existence of almost opposite behavior in such closely related languages remains essentially a mystery both for conventional formal and functional accounts, the subproject will pursue the idea that the main hypothesis of the SFB provides (at least a partial) answer to this. If R and D are a means of the symbolic grammatical system to adapt to the stochastic cognitive environment, their general availability is a straightforward starting point for an explanation, which will be pursued by the subproject, thus directly contributing to Q1. Moreover, the empirical studies conducted in the subproject provide crucial data directly pertaining to Q3 and Q4.

Fading reference – the pragmaticalization of pronoun + verb investigates pronoun verb combinations (prVC) mainly in spoken Spanish and Italian focusing on their prototypical appearance as contentful predicates, their potentially grammaticalized form as discourse markers, but especially on the varying degree of fixation in intermediate, e.g. parenthetical forms. It maps out the discourse functions of prVCs built of verbs of cognition, communication and perception, and their dependence on the types of pronouns and their feature specification, the verb-type and -form, the presence of other functional elements as well as the syntactic positions involved. Basic grammatical operations in morphosyntax and morphophonology as well as at the semantics-pragmatic interface seem either not to consider features extant in the pronominal item (R) or to re-interpret them at the interface of LF against the discourse context as different, or additional, pragmatic features that originally are not specified as such (D). Therefore, this subproject investigates the influence of the feature specification of pronouns, in terms of R (fading referential features) and D (enrichment by pragmatic features related to the speech context), in the development of prVCs. The subproject contributes to the overall SFB goals as it takes a microscopic look on how the flexibility of R and D at the interfaces to PF and LF can be modeled theoretically in order to explain the change potential and repercussions of contextual linguistic and extralinguistic factors (including frequency in language use) on the grammatical system. Thereby, the subproject addresses three main questions of the SFB: Q1 by focusing on grammaticalization/pragmaticalisation processes, as one strategy to bridge mismatches between forms and communicative goals, Q2 by exploring the the syntactic consequences of these mismatches at the left periphery, and Q3 by examining the feature specifications of pronouns in prVCs.

Deficiency and Redundancy in Expressing Temporal Relations develops a formal semantic theory of tenses as pronouns and embeds it in a larger semantic model of temporal expressions, including temporal adverbials, quantificational adverbials, and aspect. Of particular interest is the temporal (under)specification of tense features on tense morphemes and their behavior when bound by other tenses, intensional verbs or temporal adverbs. The focus on theses aspects allows us to directly apply and tests the ideas about free and bound personal pronouns developed, in particular, in Subproject 02 and Subproject 03 in a different ontological domain, thereby testing the hypotheses that there are non-superficial commonalities between pronominal systems across domains. The investigation in the subproject is based comparatively on languages with articulated tense systems such as English and German, languages with articulated aspect systems such as Russian and other Slavic languages, and so-called tense-less languages, such as Mandarin Chinese. The subproject not only extends the empirical scope of the SFB to a new ontological domain, it also centrally addresses Q1, exploring equilibria between the range of meanings and a limited domain of available forms, it informs Q2 regarding binding in the temporal domain, and explores the role of syntactic structure (Q4) and feature specification (Q3) in formal semantics.