Artemis Alexiadou
Humboldt University, ZAS Berlin
“Building possessive adjectives”

In this talk, I will examine how possessive adjectives are built in Greek by comparing them to their English (e.g., moneyed, blue-eyed) and Spanish counterparts. Greek has two ways to form possessive adjectives, by employing participial morphology, similar to English, and via special affixes. Both processes are productive and provide evidence that possessive adjectives cannot be restricted to inalienable possession. As not much is known about these processes in Greek, I will explore the relationship between possessive adjectives and possessive constructions (see Myler 2016, Fábregas 2020), as well as the relationship between adjectival participles and possessive adjectives (Koontz-Garboden, Deo & Francez 2012 for English).

Bronwyn Bjorkman
Queen’s University
“The puzzle of (apparently) phonologically motivated empty morphs”

Empty morphs are stable units of form that occur without any associated meaning or function, the inverse of zero morphs. Some empty morphs seem to occur for phonological, rather than morphological, reasons, and some appear to be repurposed from an ordinary contentful use elsewhere in the langage. In Ndebele, for example, subminimal words can be augmented either by prefixing yi- or suffixing -an (Sibanda 2004: 113)—the former is plausibly a least-marked epenthetic syllable, but the latter resembles the applicative suffix, occurring without applicative meaning. Such patterns pose a puzzle for a theory like Distributed Morphology, just as in the more widely discussed case of (apparently) phonologically optimizing suppletive allomorphy: if phonology derivationally follows morphology, as it does in standard DM, morphological realization should not be constrained by phonological wellformedness. The additional puzzle posed by empty morphs, however, is how they come to be repurposed from existing morphological uses. This talk explores the options for accounting for such empty morphs within a theory like DM.

Matthew Hewett and Ruth Kramer
Georgetown University
“Subject agreement in Amharic complex verbs: the role of Kinyalolo’s Constraint”

In Amharic, certain verbal constructions include a main verb that bears subject agreement and a non-past auxiliary that also bears subject agreement (i.e. ‘V-Agr-AUX-Agr’). However, in certain φ-feature combinations, one of the agreement markers is absent (i.e. ‘V-Agr-AUX’ or ‘V-AUX-Agr’). Assuming that there is full agreement on V and AUX in the syntax, we argue that the disappearance of agreement arises in the postsyntax. Specifically, we propose a novel, exponent-based implementation of Kinyalolo’s Constraint (KC; Kinyalolo 1991, Carstens 2005); KC prohibits the insertion of identical exponents within a single M(orphological)-word at Vocabulary Insertion, acting as an anti-homophony constraint. Because Vocabulary Insertion operates from the bottom up (Bobaljik 2000), KC blocks insertion of auxiliary agreement when it is identical to main verb agreement (deriving ‘V-Agr-AUX’). This analysis also correctly predicts (i) that auxiliary agreement resurfaces when main verb agreement is null as a result of contextual allomorphy (yielding ‘V-AUX-Agr’) and (ii) that KC is inactive when V and AUX are not in the same M-word. We contend that this analysis is more successful than previous approaches (e.g. Diertani & Eilam 2010, Kramer To Appear) and that it can account for similar patterns in other Ethiosemitic languages (Bulakh 2014).

Maria Kouneli
Rutgers University
“Marked nominative is ergative in disguise”

Languages with marked nominative case systems display what looks like nominative – accusative alignment, but unlike regular nominative – accusative systems, the morphologically marked case form is the nominative (with accusative usually left unmarked). Marked nominative is a rare alignment system, but it is very common in Nilo-Saharan and Afro-Asiatic languages spoken in East Africa. Such systems pose a clear challenge to many existing theories of case: On the syntactic side, they are problematic for theories that treat nominative as the absence of case (Kornfilt & Preminger 2015 a.o.), while on the morphological side, they do not fit into analyses that treat nominative as the least marked case (Caha 2009 a.o.). In this talk, I argue that marked nominative (at least) in Nilotic languages is a type of ergative, and not nominative. What makes Nilotic special is that the sole argument of unaccusatives can be assigned ergative. I argue that this can be explained if: i) ergative (=“marked nominative”) is uniformly assigned to SpecVoiceP (Woolford 1997, Legate 2008 a.o.), and ii) the language has an ‘expletive’ Voice head (Alexiadou et al. 2015 a.o.), which requires a specifier but does not assign a thematic role to it; the internal argument of unaccusatives moves to SpecVoiceP to satisfy its EPP requirement and is assigned ergative.

Tom Leu
Université du Québec à Montréal
“Concurrent gender systems—with distributional affinities to classifiers: Evidence from German”

Fedden & Corbett (2017) write that “surely no one would want to claim that French had two concurrent gender systems.” Challenge accepted: I claim that French has two concurrent gender systems, though the talk will mainly be about German, addressing: (A) the non-categorical nature of noun classification systems; (B) the co-existence of more than one gender system within a single language; and (C) the epistemological advantage of a piece-based view on morphosyntax.

I will argue that (A) typical gender systems and typical classifier systems may share distributional properties, cross-cutting the traditional typological criteria (cf. Dixon, 1986); that (B) German has two such gender systems, which share distributional properties with noun classifiers and numeral classifiers, respectively; and (C) that on an inferential approach to exponence the discovery of those systems would be a lunar leap for an infant, whereas a piece-based view makes it a small step, at least for a linguist.

One consequence is that, if looked at the right way, concurrent systems may be much more common than is usually thought.

Carol Rose Little
University of Oklahoma
“Case competitors and the syntax of impersonals”

The realization of case on internal arguments with impersonal initiators can vary, as shown by the two examples in (1). In Irish (1a), the internal third person plural argument is realized in the accusative case; in Finnish (1b) the internal argument is nominative.

(1) a. Buaileadh aríst iad.
beat.PA.IMPRS again 3.PL.ACC
‘They were beaten again.’                             Irish (Stenson 1989: 384)

b. Vie-ttiin koira talo-on.
take-PST.IMPRS dog.NOM house-ILL
‘The dog was taken into the house.’            Finnish

Many theories of impersonal syntax argue for a null nominal in the initiator position. However, as shown by the examples above, this null initiator may or may not act as a case competitor in the sense of Baker (2015). This talk investigates the properties of case competitors in impersonal constructions, focusing on nominative-accusative aligned languages. I explore these facts through a dependent-case theory approach (e.g., Marantz 1991 et seq.).

Terje Lohndal, David Natvig, Michael T. Putnam, and Yvonne van Baal
Norwegian University of Science and Technology and UiT The Arctic University of Norway, University of Stavanger, Pennsylvania State University, and University of Stavanger
“Comparing frameworks: A case study of the syntax-morphology interface”

The conceptualization of features has remained a central theme in generative approaches to the syntax-morphology interface since the early 1990s. These have played an enhanced role in late-insertion approaches to morphosyntax, in which the establishment of feature-exponency relations is of fundamental importance. Multiple frameworks have been developed to model the syntax-morphology interface, opening the door to framework comparisons. In this paper we review the advantages and disadvantages of what we label direct and indirect approaches to this interface, namely Nanosyntax and Distributed Morphology, respectively. To illustrate the similarities and differences between these frameworks, we demonstrate how nominal suffixes are realized in varieties of Norwegian. We show that both direct and indirect approaches to the syntax-morphology interface are capable of modelling the alternations observed in the data. That is, they both achieve descriptive adequacy. However, they do so by virtue of rather different architectural assumptions, raising the question of how to compare and adjudicate their relative strengths and weaknesses.

Neil Myler
Boston University
“Imagining life without Rules of Exponence and the Elsewhere Condition”

The notion that there are Rules of Exponence, that they compete, and that their competition is regulated by the Elsewhere Condition, is a point of consensus that unites many approaches to morphology which otherwise differ significantly in their core tenets, forming the cornerstone of the family of theories dubbed “realizational” (Anderson 1992:132; Halle and Marantz 1993:123; Starke 2009:4; Stump 2001:22, 2016:50, amongst many others). Recent work in the Morphology as Syntax (MaS) framework seeks to undermine this cornerstone (Collins 2018, 2020; Kayne 2020; Collins and Kayne 2023). In its place, MaS proposes a non-realizational theory in which all allomorphy and morphotactics are captured using mechanisms which are familiar from syntactic theory, leaving no room for competition via the Elsewhere Condition. In this talk, I describe a fragment of a grammar for Latin noun declension which I have developed in the MaS framework. The fragment highlights a number of interesting properties of MaS, including at least one which I will suggest weighs heavily against it: eschewal of the Elsewhere Condition comes at a steep price in the amount of accidental homophony the account must assume.

David Natvig, Michael T. Putnam, and Emmeline Wilson
University of Stavanger, Pennsylvania State University, and Pennsylvania State University
“Umlaut as combined suffixation”

Umlaut in Germanic – both the historical process of conditioned back vowel fronting and the resultant synchronic morphological alternation – has been and remains a hotly debated phenomenon, both in terms of its historical development (Boutilier, 2019; Iverson & Salmons, 1996; Janda, 1998; Salmons, 1994, 2018) and its morphophonological representation in con- temporary languages (Iverson & Salmons, 2004; Trommer, 2021; Wiese, 1996, and many others). We focus in this talk on umlaut as a morphological aspect of German plurals. Specifically, we trace the structural development of umlaut across multiple noun classes from Old High German and New High German to the present, and compare those results to proposals of its contemporary morpho-phonological representation as a single concate- native process (Trommer, 2021; Wiese, 1996). We seek a better understanding of how the diachrony of a pattern with multiple, complex phonological sources coalesces into a single process. This, we argue, sheds light on the division of labor between syntax and phonology in morphological patterns (e.g., Bermúdez-Otero, 2012), as well as how these representations and relationships may change over time.

Heather Newell
Université du Québec à Montréal
“Readjusting the analysis of the irregular past tense in English”

English irregular verb conjugations (e.g. sing-sang-sung, take-took-taken) have been proposed to offer evidence that T0past and the verb root must undergo Vocabulary Insertion in the same cycle/phase, as well as requiring morpheme-specific Readjustment Rules to derive their surface forms (e.g. Embick 2010). Here I examine these arguments and offer an alternative analysis that requires neither: English irregular verbs are derived by the regular phonological grammar. This analysis requires an appeal to abstract, sometimes underspecified, autosegmental phonological representations for the relevant affixes (in the vein of Trommer’s (2021) analysis of German plural morphology). This reanalysis is couched within a discussion of how to distinguish phonological from morphological alternations, what evidence we have for the exact timing of spell-out (Chomsky’s PIC1 and PIC2) and whether and how phases are accessible after they have undergone spell-out.

Sylvia Schreiner
George Mason University
“Realizations of temporal semantics in Akuzipik”

Akuzipik (St. Lawrence Island Yupik; endangered, polysynthetic, Inuit-Yupik) belongs to a language family with members that have variously been claimed to be tenseless (e.g. many but not all varieties of Inuktitut; Kalaallisut, etc.). What descriptive literature on Akuzipik exists labels various morphemes as marking ‘tense’ and ‘aspect’ (as does the literature on its closest relative Central Alaskan Yup’ik), but the language’s temporal realizations are under-documented and the meanings and behavior of these morphemes are unclear. In this talk I discuss the results of fieldwork on the various realizations of temporal semantics that occur in Akuzipik, including their semantics and morphosyntactic properties. I investigate in particular the semantics of the unmarked form and the behavior of an apparent progressive morpheme. The (non-)obligatoriness of these morphemes informs our understanding of what it means for a language to realize tense and Tense.

Einar Freyr Sigurðsson and Jim Wood
The Árni Magnússon Institute for Icelandic Studies and Yale University
“Resolution by case syncretism in Icelandic passives”

Passivized ditransitives with reflexive indirect objects in Icelandic are subject to speaker variation: speakers either reject the construction entirely (Grammar D), accept it with nominative (Grammar A) or accusative (Grammar B), or accept it only when the noun happens to be syncretic between the two (Grammar C). In this talk, we propose that the structural problem is that the DP gets only a subset of its case features valued, and remains unvalued for the distinction between nominative and accusative. The case we focus on most is Grammar C, which is reminiscent of “Rescue by Syncretism” effects found in argument-sharing constructions, such as ATB-movement (Citko 2005, Asarina 2013, Bjorkman 2021). In this case, however, there is no argument sharing or source for multiple-case assignment. We propose that the undervaluation results in splitting the feature bundle into two separate feature bundles, one with nominative features and the other with accusative. We refer to this as Individuation, following Soares (2023). We show that this process looks like a certain kind of Fission (the “pre-Vocabulary Insertion” kind), in that the marked/conflicting features are split, and the irrelevant features are copied onto each bundle. The difference is that Fission creates two feature bundles, two loci for Vocabulary Insertion and two slots for separate Vocabulary Items to be realized. The mechanism in this case creates two feature bundles, two loci for Vocabulary Insertion, but only one slot for a single Vocabulary Item to be realized. We explore how this could be accounted for in the system of Vocabulary Insertion being developed by Soares (in progress).

For a long-form version of this abstract, click here.

Tran Truong
Pennsylvania State University
“Microvariable lexical stratification: The view from honorifics”

No abstract available; see slides linked to in schedule.