ExQ: Exhaustiveness in embedded questions across languages

Principal investigators:
Prof. Dr. Edgar Onea
Universität Göttingen
Prof. Dr. Malte Zimmermann
Universität Potsdam

Funding period:
2017-2020; this project is a continuation of “ExCl: Exhaustivity in Cleft Sentences”

Project description:
Theoretical literature offers an ongoing debate about the interpretation of embedded questions, such as “Ms. Smith knows who solved the problem”. Under a strong exhaustive reading, the sentence suggests that Ms. Smith has to know the complete list of people who solved the problem, and moreover that this list is complete. Groenendijk and Stokhof (1984) argued that this is the only available reading, whereas other scholars (e.g. Guerzoni & Sharvit 2007, Uegaki 2015) point out the possible existence of weak or intermediate readings. Under a weak exhaustive interpretation, Ms. Smith would not need to know that the list is complete as long as she knows the complete list. The theoretical literature focuses strongly on the various ways in which the respective readings, and only the respective readings, can be derived. The main research question is whether the different readings are semantic in nature, or whether they result from pragmatic inferencing based on various factors such as the domain of quantification, at-issueness, utility etc.

The empirical basis of this theoretical discussion is less clear, however. There is only one experimental study (Cremers and Chemla 2016) that tries to elucidate the empirical situation, but this study is limited to two embedding verbs in English. Experimentally controlled data on embedded questions in other languages and with more verbs of embedding are not available.

The proposed project attempts to close this research gap by conducting a series of experiments on the interpretation of embedded questions in a wide range of unrelated languages (from Ulster English to Akan (Kwa, Niger-Congo)). The questions are embedded under various verbs (e.g. “know”, “surprise”, “predict”, “tell”) and tested in two independent experimental paradigms against all combinations of readings discussed in the literature: Experimental paradigm I employs a sentence-verification task, whereas experimental paradigm II employs a felicity-judgment task. The two experimental paradigms complement each other, and they will provide clear evidence as to which of the exhaustive readings of embedded questions are pragmatic in nature, and what contextual factors affect their derivation.

Since the exhaustivity of embedded questions is of fundamental importance for the theoretical discussion of what questions are (e.g. partitions or Hamblin-alternatives), the theoretical implications of the expected results of the project are far-reaching. They pertain to all areas in discourse semantics and pragmatics that rely on questions as semantic objects, including presupposition projection, exhaustification, derivation of scalar implicatures etc.