Guest talk: Elizabeth Bogal-Allbritten (University of Massachusetts, Amherst)

We are pleased to welcome Elizabeth Bogal-Allbritten, a Ph.D. candidate at the University of Massachusetts, Amherst. Her research is centered around linguistic fieldwork and the insights that less-studied languages can bring to semantic theory. Particular interests of hers are Navajo and belief, desire, modality, and modification across languages.

She will be giving a talk on Friday at 3 PM in SS 560A: "Decomposing attitudes of belief and desire." Afterwards, there will be a reception in the department lounge.

In the past 20 years, cross-linguistic investigation has challenged and informed our theoretical and typological understanding of a wide variety of semantic phenomena, including quantification, modality, and comparison. Other areas - including attitudes of belief and desire - have received comparatively little cross-linguistic attention, however, Since Hintikka (1982), attitudes of belief and desire (Alice thinks it will rain, Alice wants it to rain) have been fruitfully analyzed as modal statements. Authors subsequent to Hintikka have argued on the basis of data from English and related languages that attitude verbs (e.g. think, wish, want) are the source of modality in attitudes, where each verb expresses universal quantification over worlds consistent with, e.g., the attitude holder's beliefs or desires.

In this talk, I show that consideration of a more diverse set of languages makes it clear that while the now-standard 'neo-Hintikkan' picture may be satisfactory for English, it is not universally applicable. Navajo (Athapaskan) attitudes of belief and desire all contain the same verb, nízin. On the basis of novel fieldwork evidence, I demonstrate that nízin cannot be analyzed as a modal. Instead, I argue that nízin contributes the meaning common to both beliefs and desires: nízin denotes situations of mental attitudes. All modal meaning characteristic of beliefs and desires comes from modal operators (overt (1a,b) or covert (1c)) in the clauses that nízin embeds:

(1) a. Alice [nahodoołtį́į́ł shaʼshin] nízin.
Alice it.will.rain PRT-belief 3S.att(itude)
'Alice thinks it will rain.'

(2) b. Alice [nahodoołtį́į́ł laanaa] nízin.
Alice it.will.rain PRT-desire 3S.att
'Alice wishes for it to rain.'

(3) c. Alice [nahodoołtį́į́ł             ] nízin.
Alice it.will.rain Ø-belief / Ø-desire 3S.att
i. 'Alice thinks it will rain.'
ii. 'Alice wants it to rain.'

I propose that Navajo nízin-sentences present novel evidence for recent 'decompositional' theories by Kratzer (2006, 2013), Anand and Hacquard (2009), and Moulton (2009), who argue that modality can be severed from the lexical entries of English and German verbs of belief, perception (see), and communication (say). Navajo enriches the decompositional theory - and our understanding of attitudes more generally - by demonstrating how a language can construct attitudes of belief and desire from a light attitude verb and independently motivated modal operators: I will argue that all of the overt and covert modal operators invoked in (1) are also found in Navajo matrix clauses.

While it may be the case that all languages have the ability to communicate beliefs and desires, Navajo demonstrates that languages may differ significantly in the grammatical strategies they employ to communicate these meanings. I will argue, however, that admitting such diversity into our semantic theory of attitudes can shed light on the analyses of data from other languages, including parallel constructions in other Athapaskan languages and apparently 'ambiguous' verbs of communication in Romance and Hebrew: in each language, the contents of the embedded clause drive the interpretation of the clause as a whole.