Is this an alternative question or a polar question? Yes.

Linguistics jokes are great. Here is one that was posted on Facebook by linguist Jonah Katz on September 26, 2024 (original source unknown).

The humorous effect here has many layers, and draws in part on an ambiguity that occurs in written English between a polar question containing “or” and an alternative question. A polar question is one that asks for a “yes” or “no” answer from the addressee, such as “Do you like cats?”. The question in panel 2 can have this reading, and happens to contain “or” – if the speaker on the left does either of the two things, the answer would be “yes”, otherwise the answer would be “no”. An alternative question also contains “or”, but on this reading, the question does not ask for a “yes” or “no” answer; rather, the answer to an alternative question is one of the two things connected by “or” (the disjuncts). Additionally, we usually think of alternative questions as presupposing that one of the disjuncts is true.

In this meme, the polar question interpretation is likely intended by the speaker on the right – an answer of “yes” to the question as a whole would provide an explanation for why the speaker on the left’s teeth are turning yellow. (As an aside: it’s interesting that the polar question interpretation is the most natural one here, because in many contexts this is not the case.) And unless the speakers know each other quite well, it might be presumptuous for the speaker on the right to use a question form (i.e., an alternative question) that presupposes that the speaker on the left either smokes or drinks coffee. But the speaker on the left has interpreted the question as an alternative question, and answers with one of the disjuncts: He drinks coffee. So part of the humorous effect here is that the speaker on the left has mis-interpreted the speaker on the right’s question.

This ambiguity only occurs in written English. In spoken English, prosody disambiguates between the two question types. Polar questions in (many varieties of) English occur with a final rise intonation, whereas alternative questions have a rise on the first disjunct (actually, on all the non-final disjuncts if there are more than two things being connected by “or”), and a fall on the final disjunct. This means that in a real life conversation, it would be very unlikely that the speaker on the left would mis-interpret the question in panel 2, since the speaker on the right would have used one of the intonation contours, thereby disambiguating their question.

The humorous effect of the meme further draws on the fact that there are actually three readings for the question in panel 2 – there are in fact two alternative question readings, depending on exactly how the prosodic contour is pronounced, i.e., what is in focus. This in turn has an effect on how we interpret what the disjuncts are. On one structure (a): Do you [[smoke] or [drink coffee]], the question is asking whether the speaker on the left smokes or whether he drinks coffee. With the other structure (b): Do you [[smoke or drink] coffee], the question is asking whether the speaker on the left smokes coffee or drinks it. The (b) interpretation is non-sensical given world knowledge (we don’t smoke coffee), but panel 3 makes clear that this is in fact the interpretation that the speaker on the left has given the question, adding a further layer to the humour conveyed here. His use of the pronoun “it” in panel 3 is crucial here (as pointed out by Seth Cable in a comment on the aforementioned Facebook post). The use of “it” seems only compatible with the (b)-reading here – if the speaker on the left had said “I drink coffee”, this would amount to an answer to the (a)-reading of the question, and would not be construed as a joke.

Leave a Comment

Filed under Blog post

Why Oh Why? — The Dutch Question Word Hoezo

This sub-project zooms in on why-questions, more specifically a special kind of why-question that occurs in Dutch. Why-questions are typically used to ask for a reason or explanation (for instance, Why was the train delayed?). But reasons and explanations can be of different kinds. Dutch has two separate question words that are both usually translated in English as why, but that differ from each other in interesting ways.

In addition to waarom (the run-of-the-mill counterpart of English why), Dutch also uses the more specialized question word hoezo. Questions with waarom are pretty much like English why-questions, but hoezo (which literally means ‘how so’) is used to ask why the previous speaker just said what they did. A typical example is the following mini-dialogue between two speakers (A and B) who are waiting for a train on the station platform:
Speaker A: The trein is vertraagd. (translation: ‘The train is delayed.’)
Speaker B: Hoezo vertraagd? (loose translation: ‘What do you mean, delayed?’)

By saying hoezo, B expresses skepticism about A’s statement and challenges them to give a reason for why they made that statement. This could for instance happen if B sees a train approaching in the distance and assumes that that’s the train they are waiting for. In other words, by saying hoezo, B does not ask A why the train is delayed, but rather why A said that.

Our working hypothesis is that hoezo-questions in Dutch are used by speakers to challenge the immediately preceding discourse move by their interlocutor. Hoezo is a pragmatic question word — it can best be characterized in terms of the conditions on its use rather than by describing its literal meaning. In this way, it shares some of the characteristics of what linguists call discourse particles: little words whose function it is to regulate or manage the conversation. Discourse particles are very common in many languages (including Dutch) but are often hard to translate adequately in English. We explore the possibility that hoezo is both a discourse particle and a question word at the same time. If that’s the case, this is evidence for the existence of a novel category of words that has not previously been recognized as such in the linguistic literature. We are currently testing our hypothesis by gathering examples of hoezo and waarom from a linguistic corpus of Dutch, and analyzing them in a theoretical model of the semantics and pragmatics of questions and dialogue. In the longer run, we plan to broaden our investigation by making comparisons to other ‘why’-type question words in closely related languages, such as English whywhat forhow come, and how so, and German warum and wieso.

by Hotze Rullmann and Sander Nederveen

Poster at LSA Meeting January 2024.

Leave a Comment

Filed under Blog post