The Schedule: Day 1
All talks and coffee breaks at Room B01, Life Sciences and Engineering Building, 24 Cummington Mall, Boston MA 02215
All talks and coffee breaks at Room B01, Life Sciences and Engineering Building, 24 Cummington Mall, Boston MA 02215
Day 1
Registration and Breakfast 9:00 - 9:30
Session 1 Chair: Neil Myler
9:30 - 9:40
Neil Myler
This talk seeks to address morphologically conditioned variation in root semantics, which we can broadly call "root allosemy" (but we all know is secretly quite similar to "idiomaticity" in the syntactic sense). We analyze three phenomena in English that can be examined through this lens: 1) Overriding root semantics in complex word formation. Harley (2014) discusses this in the context of Latin borrowings such receive, deceive, and conceive, but we will approach native forms such as understand, undertake, and undergo. 2) Incomplete blocking of regular forms by suppletion suppletive forms (such as badder vs. worse, persons vs people, brothers vs brethren). Finally, 3) The causative-inchoative alternation in English, which is never marked productively and regularly but appears in a restricted set of irregular alternations (such as raise vs. rise, sit vs set, fall vs fell). Harley (2014 ) discusses these briefly to argue that teach and learn could be construed as root suppletion (and Siddiqi 2024 disagreed). We show that the constraint-based, modular architecture of LRFG (which we still maintain is a spinoff of Distributed Morphology) allows us to formally model these types of alternations while shedding new light on the phenomena.
Mouthing is a language contact phenomenon in which a signer may mouth, in part or in whole, a word from the ambient spoken language. Mouthing is distinct from mouth morphemes, which are oral movements that convey a wide variety of aspectual and adverbial information. In BSL linguistics, mouthing is referred to as spoken components and mouth morphemes as oral components (Sutton-Spence & Woll 1999). Oral components can be used for emphasis, to specify the intended meaning of a polysemous sign, and to facilitate comprehension when interacting with less fluent signers. Whether oral components (and non-manual markers more broadly) forms part of the phonological representation of a sign is a matter of ongoing debate, but this question cannot have the same answer in all signed languages. In BSL, there are a number of single manual letter signs, such as GEOGRAPHY and GEORGE, that are distinguishable only by mouthing. I argue that these are compounds of the manual component and the spoken component and not code-switching or code-blending. I draw analogies between this practice and CODA Speak (Berent 2012), cross-registral compounding in spoken languages, and the practice of 義訓 gikun ‘false readings’ in Japanese. It emerges that single manual letter signs have much to contribute to a theoretically informed study of cross-modal language contact.
Poster Session and Lunch
(111 Cummington Mall, Suite 138 and nearby rooms)
11:15 - 11:25
11:25 - 1:00
Lunch 1:00 - 2:00
Provided
2:00 - 2:10
Session 2 Chair: Nadeem Siddiqi
2:15 - 3:00
Ruth Kramer (joint work with Maša Bešlin)
Consider the following two puzzles. First, there is a well-known subset of male-denoting nouns in Bosnian-Croatian-Serbian (BCS) that have variable gender agreement: masculine in the singular, but either masculine or feminine in the plural; see (1)/(2). We refer to this subset as bishop nouns.
(1) Nov-i/*-a vladika je sazva-o/*-la sastanak. Masc Sg Agr
new-m.sg/*f.sg bishop aux called-m.sg/*f.sg meeting
‘The new bishop called a meeting.’
(2) Nov-i/-e vladike su sazval-i/-e sastanak. Masc/Fem Pl Agr
new-m.pl/f.pl bishops aux called-m.pl/f.pl meeting
‘The new bishops called a meeting.’
Second, it is controversial how to properly characterize grammatical gender features in languages like BCS, which have a three-way gender system. Some previous work has argued that each gender is identified by two binary gender features: [+/-masc] and [+/-fem] (e.g., Despić 2016). However, one part of the typology is unattested: there are said to be no nouns in these languages that are [+fem][+masc].
We address both of these puzzles simultaneously, contending that bishop nouns have [+fem][+masc] gender features. They therefore contain an overspecified gender feature bundle, which we argue must be reduced on agreement targets (Noyer 1997, Nevins 2011, Keine & Müller to appear, a.o.; see also Despić 2017 on BCS, though with a different feature specification and implementation). We propose that a markedness-targeting Impoverishment rule (Nevins 2011, Arregi & Nevins 2012) deletes the more marked [+fem] feature from a feature bundle that contains [+masc][+fem]; this generates the masculine singular agreement in (1).
As for the plural, languages generally make fewer morphological distinctions between genders in the plural, i.e., plural agreement contexts generally result in the reduction of gender features (Greenberg 1966, Bobaljik 2002, a.o.). Thus, we propose that the gender specification of plural bishop nouns triggers a common kind of markedness-triggered Impoverishment rule which deletes a gender feature due to the marked plural context. Since this rule is not triggered by the combination of gender features, either of the two gender features is deleted, generating the variable agreement in (2).
Our analysis is both simpler and empirically superior to existing alternatives. For example, we show that there is no need to postulate novel agreement-blocking mechanisms or non-standard arrangement of features on the nominal spine for BCS (contra Puškar 2018). Moreover, acknowledging the distinction between markedness-targeting vs. markedness-triggered Impoverishment rules gives us a handle on why the agreement is fixed in the singular and variable in the plural, and not the other way around. We also show how our analysis predicts that bishop nouns will be in declension class II, and also (unlike previous accounts) correctly predicts agreement resolution patterns with coordinated plural nouns. Finally, this new approach unites bishop nouns in BCS, previously viewed as perhaps uniquely recalcitrant, with the analysis of similar gender agreement patterns in other languages, e.g., Romanian and Guébie (Kramer & Sande 2023), lending support to Impoverishment approaches to changes in gender agreement from singular to plural.
Break 3:50 - 4:00
Session 3 Chair: Karlos Arregi
4:00 - 4:45
Mike Putnam (Joint work with Deborah Adeyeye, Andrew Hoffman, Alex Koch, Björn Kohnlein, David Natvig, Tran Truong, & Emmeline Wilsom)
While umlaut and other (related) non-concatenate alternations in standard German have been the focus of several recent analyses (e.g., Shwayder 2015, Embick & Shwayder 2018), detailed treatments of these phenomena in non-standard varieties have been less prominent (but see Köhnlein 2018). In this presentation, we take a closer look at umlaut (and closely related processes) in Gottscheerisch, a moribund Bavarian-based heritage variety spoken in Slovenia and in a handful of diasporic communities in the US (i.e., Queens, NY and Cleveland, OH; Tschnikel 1908). As we demonstrate, a comparison of the umlaut processes in Gottscheerish and standard German with respect to domains of (i) pluralization and (ii) diminutivization require deeper theoretical analysis and introspection.
Adopting the basic tenets of morphological decomposition found in Distributed Morphology, we outline the theoretical machinery responsible for untangling the relationship between pluralization and diminutivization in Gottscheerish (Ritter 1992, Harbour 2014, Wiltschko 2021, Alexiadou & Lohndal 2023). More explicitly, in Gottscheerish we observe a more diverse set of root vowel alternations when compared with standard German. In this presentation, we focus on four phenomena: First, in Gottscheerish, we have instances of roots that exhibit distinct vowels in plural and diminutive forms. This pattern contrasts with what we observe in standard German. Second, there are several morphologically complex diminutive forms that appear to block the application of umlaut. Third, Gottscheerish exhibits apparent disfixation in the form of a schwa-less plural for some singular forms ending in schwa, which reappear in both singular and diminutive forms.
Taken together, we develop the sketch of analysis of how these unique aspects of non-concatenate morphophonology in Gottscheerish can be captured in a formal analysis.
References
Alexiadou, Artemis & Terje Lohndal. 2023. Germanic diminutives: A case study of a gap in Norwegian. The Journal of Comparative Germanic Linguistics 26(2): 1-23.
Embick, David & Koby Shwayder. 2018. Deriving morphophonological (mis)applications. In: From sounds to structures: Beyond the Veil of Maya, eds. R. Petrosino, P. Cerrone, & H. van der Hulst (pp. 193-248). Berlin: Mouton de Gruyter.
Köhnlein, Björn. 2018. A morpheme-based to subtractive pluralization in German dialects. Phonology 35(4): 617-647.
Harbour, Daniel. 2014. Paucity, abundance, and the theory of number. Language 90: 617-647.
Ritter, Elizabeth. 1992. Cross-Linguistic evidence for Number Phrase. Canadian Journal of Linguistics 37: 197-218.
Shwayder, Koby. 2015. Words and subwords: Phonology in a piece-based syntactic morphology. PhD thesis, University of Pennsylvania.
Tschnikel, Hans. 1908. Grammatik der Gottscheer Mundart. Halle a.S.: Max Niemeyer.
Wiltschko, Martina. 2021. The syntax of number markers. In: The Oxford handbook of grammatical number. Eds., Patricia Cabredo Hofherr & Jenny Doetjes (pp. 164-196). Oxford: Oxford University Press.
The Basque verbal morphology shows the morpheme ordering template in (1):
(1) A-√-D-E
(2) a. n-au-zu b. d-i-da-zu
1sA-√-2sE 3sA-√-1sD-2sE
There are two second person singular pronouns in Basque: zu (formal) and hi (informal). Dative and Ergative (and allocutive) clitics of hi have a gender distinction; this is the only grammatical phenomenon that shows gender features in Basque.
Besides, Basque verbal morphology also presents a fourth person clitic: allocutive. The allocutive is a second person clitic, identical in its form to second person dative and ergative clitics, and is a non-thematic addressee marker, which appears when a second person is not an argument of the sentece.. It is obligatory when the speaker refers to the addressee as hi in all varieties, and it is also obligatory with the pronoun zu in the Zuberoan dialect. Allocutive clitics of hi also have gender features. (3) is an example of feminine hi allocutive: as the speaker refers to the addressee as hi (either overtly or not), a non-allocutive variant of the auxiliary is ruled out. (4) is the updated version of the template in (1).
(3) Mirenek ni (hirekin) ikusi n-au(-ø) / n-ai-n(-ø). 'Miren saw me [+ feminine addressee]'
Miren.E I.A (you.with) seen 1sA-√-All.F(-3sE) / 1sA-√-All.F(-3sE)
(4) A-√-D-All-E
Haddican (2018) observed that the presence of the allocutive clitic disrupts the morpheme ordering in the template in (4), and brought the contrast in (5) to out attention, where the position of the allocutive morpheme and the ergative are inverted:
(5) a. d-i-a-gu b. d-i-te-k
3sA-√-All.M-1pE 3sA-√-3pE-All.M
Haddican (2018) analyzed this contrast as a Person Effect: 1st person ergative clitics follow the allocutive clitic, whereas 3rd person ergative clitics precede it. His neoperformative analysis captured this effect as follows: 1st person clitics move to a Speaker head whereas 3rd person clitics move to T, where Speaker > T.
I propose that the contrast in (4) can be interpreted as a Gender-Plural Effect: clitics with a [+M/F] feature follow -(t)e-, which are dative or ergative plural markers resulting from plural fission (Arregi & Nevins, 2012). Thus, I suggest a morphological analysis, where postsyntactic rules target [+M/F] and [+pl] clitics and, under specific circumstances, yield a metathesis or doubling rule.
In order to test a better explanation of data, I analyzed microvariation data from Irizar (1991, 1992a, 1992b, 1997). More specifically, I analyzed the relative order of dative, allocutive and ergative clitics in whole auxiliary paradigms in 147 varieties of 5 dialects of Basque, close to the half of the Irizar corpus.
The results lead to the following conclusions:
(i) gender and plural features are relevant in clitic ordering patterns that contradict (4), which points to the necessity for the Gender-Plural Effect.
(ii) a strong version of the Person Effect is necessary to account for specific data, especially in the Lapurdian dialect and the varieties of the Sakana area.
(iii) the morphological analysis constitutes a more straightforward way to explain the clitic ordering diversity in the microvariation of the Basque verbal morphology
Dinner for Presenters and Organizers at Cornwalls
7:30pm (Kenmore Square, 644 Beacon Street, Boston MA 02215)