Meaning Meeting - Will Zumchak / Scalar implicatures need scales
Meaning Meeting - Will Zumchak / Scalar implicatures need scales
On Cinco de Mayo, the Meaning Meeting has Will Zumchak, discussing the oddity of saying "Some Xs are Ys" when all Xs are Ys. Such oddity has often been explained by claiming that use of such sentences mandatorily triggers an inference that is obviously false: 'Not all Xs are Ys.' But Will pursues a different approach, whereby such sentences actually block implicatures that would make their use relevant. Will's abstract is below.
Magri (2009) discusses a contrast in felicity between sentences like “some lions are mammals” and “lions are mammals.” Common knowledge makes these two sentences equivalent, since presumably all lions are the same species. Magri notes, however, that assertion of the "some” version is significantly odder than its bare plural counterpart. I discuss some current accounts of this “Equivalency Oddness” (EO) and present some new data involving explicit questions not predicted by the existing proposals. I then consider an alternative account: unlike “Blind ᴇxʜ” approaches (Magri, 2009, 2011, 2017; Del Pinal, 2021; Bar-Lev, 2024), which assume scalar implicatures are mandatory and automatic, I assume contextual equivalencies actually block the generation of certain implicatures. This prevents EO sentences from directly addressing a (sometimes implicit) question, resulting in oddness.