Events

Comparing attitudes and modals

In this meeting, I will discuss my ongoing work on some puzzles that emerge from the interaction between attitude predicates/modals and comparatives. The first puzzle is a cross-linguistic difference between Mandarin and English in comparatives with attitude predicates. In English, belief ascriptions are comparable (1a), but knowledge ascriptions are not. In Mandarin, both can be graded with the morpheme geng, comparing the degrees of certainty assigned to propositions by attitude holders.

(1)   
a. He believes more than I do that the organization of the executive branch of the federal government matters a great deal.
b. # He knows more than I do that the organization of the executive branch of the federal government matters a great deal.

(2)   
Zhangsan bi Lisi geng juede/zhidao qinfen bushi weiyide da’an.
Zhangsan compared-to Lisi GENG think/know diligence isn’t the-only answer   
Zhangsn is more certain (of the fact) that diligence isn’t the only answer.

Second, geng can be used to compare modals, but not all of them. For example, yinggai “should” can be both an epistemic modal and a deontic modal, but only the deontic flavor is comparable:

(3)   
#Zhangsan meiyou buzaichangzhengming, suoyi ta bi Lisi geng yinggai shi xiongshou.
Zhangsan no alibi, so he compared-to Lisi GENG should at home
(intended) “Zhangsan has no alibi, so he is more likely to be the murder than Lisi.”

(4)   
Zhangsan shi laoshi, ta bi wo geng yinggai zunshou jilv.
Zhangsan is a teacher, he compared-to me GENG should follow the rules
“Zhangsan is a teacher, he needs to follow the rules more than me”

It seems that geng expresses a form of comparison that differs from English more. I will argue that this is not meta-linguistic comparison. Futhermore, I will also argue that the prior proposals about gradable modals and attitude predicates (Lassiter 2017, Koev 2018, Pasternak 2019), and other comparative morphemes (e.g. more in English, Wellwood 2015; eher in German, Herburger and Rubinstein 2018) all have some difficulty capturing the cross-linguistic differences. This work is ongoing, so comments and suggestions will be welcome.