IS THAT A RHETORICAL QUESTION?

:
A PRAGMATIC ANALYSIS
by

Jacklyn Ryan

A Dissertation Submitted in
Partial Fulfillment of the
Requirements for the Degree of

Doctor of Philosophy
in Linguistics

at

The University of Wisconsin-Milwaukee

August 2023
 ABSTRACT

IS THAT A RHETORICAL QUESTION?:

A PRAGMATIC ANALYSIS

by

Jacklyn Ryan

The University of Wisconsin-Milwaukee, 2023

Under the Supervision of Professor Nicholas Fleisher

There has been much work on the syntax, semantics, and pragmatics of questions. While

the argument herein is that rhetorical questions do not function like typical information-seeking

questions, it remains the case that they are, if nothing else, syntactically interrogative. This fact

is explored by examining different types of rhetorical questions through various lenses, including

question semantics, Gricean pragmatics, and Speech Act Theory. A pragmatic framework is pro-

posed to explain the effects that rhetorical questions have on the conversational scoreboard. Their

illocutionary force is also considered, as it, along with contextual factors, can affect how rhetorical

questions are interpreted. This paper offers a new definition of rhetorical questions as well as

providing an analysis of their pragmatic effects.

ii
©Copyright by Jacklyn Ryan, 2023
All Rights Reserved

iii
For Zooey and Von

iv
 TABLE OF CONTENTS

Page

ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii

DEDICATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv

TABLE OF CONTENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii

1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Don’t you know what a rhetorical question is‽ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.2 Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2. THE LANDSCAPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.1 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.1 Alternative and Polar Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.2 Wh-Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.3 Tag Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2 Pragmatics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.1 Gricean Pragmatics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.2 Speech Act Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3 The Conversational Scoreboard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

3. A TAXONOMY OF RHETORICAL QUESTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

3.1 Rhetorical counterquestions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

3.1.1 Who knows? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.2 Indirect questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.2.1 Rhetorical tag questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.2.2 Requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

4. A PRAGMATIC FRAMEWORK FOR RHETORICAL QUESTIONS . 50

v
 4.1 General framework for rhetorical questions in conversation . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.1.1 Components of the scoreboard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.1.2 Pragmatic principles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.2 Rhetorical suggestions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.3 Rhetorical counterquestions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

5. ODDS AND ENDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

5.1 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

vi
 LIST OF TABLES

TABLE Page

2.1 Partition for alternative question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2.2 Main Types of Tag Questions in English (Kimps 2007) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.3 Context State K4 (Farkas and Bruce (2009). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

2.4 Types and commitments (Farkas and Roelofsen 2017) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

2.5 Clause Types (Portner 2004) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

4.1 Can you pass the salt? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

4.2 Where do you think you’re going? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

4.3 Do you want to go to the Radiohead show? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

vii
 ACKNOWLEDGMENTS

My first year in the Linguistics department at UWM was a whirlwind. I was doing a lot

of things for the first time, including teaching. I was a TA for the entirety of my time in graduate

school, for which I am deeply thankful. During my second semester I was one of the teaching

assistants for Kelsie Pattillo, who was an excellent mentor and became a great friend. I learned a

lot from her about how to be a good teacher, a good student, and a good researcher.

My advisor, Nicholas Fleisher, was instrumental in my success for both my master’s degree

and my PhD. His immense knowledge and considerable patience were the two things that I needed

most in the writing of this dissertation. Peter van Elswyk’s course on Speech Act Theory, and

subsequent conversations, kicked my project in a direction that was really exciting to me and that it

wouldn’t have otherwise gone. Anne Pycha has an eye for detail and understands how to approach a

hypothesis and determine the best way to test it. This was very helpful to me over the course of my

graduate studies. Sandra Pucci’s perspective has always been refreshing. Her questions about how

linguistic constructions work cross-linguistically and cross-dialectally make me think and make

me want to find out. This group’s enthusiasm and expertise helped me to focus the broad topic of

“what’s going on with rhetorical questions?” to something more manageable and definitely more

interesting.

I’m not sure I would have pursued a graduate degree in Linguistics if it had not been for Erica

Benson and Lynsey Wolter. They recognized and nurtured my interest in semantics, pragmatics,

and sociolinguistics during my undergraduate studies at UWEC. I am forever grateful to them.

I could not have done this without my husband. He moved hundreds of miles three times

viii
so that I could achieve this goal. Our children came along for the ride. They are the reason I felt

strong enough to take the first step down this path, and they have been amazing and encouraging

throughout the years. They were both very young when this journey began. Von was not walking

yet, and Zooey was going to school for the first time when we moved to Shorewood, WI. Because

of this they both think of this village as their hometown. We have wonderful neighbors who have

helped tremendously with the kids when I needed some quiet time or when Mark and I were both

still at work when school got out. Shoutout to Morris Blvd!

I have an incredibly supportive family, but they have gone above and beyond the last couple

years. I’d like to thank my mom for unwavering moral support; my sister, Jess, for being my first

student and for believing in me always; my brothers, Jordan and Jared (yes all Jays), and my cuñada

Liz, for helping me take my mind off my work from time to time; my aunts, Karen and Karol, for

good conversations; my niblings, Cora, Nathan, Roark, and Franklin, for making me smile; my

in-laws, Barb and Dave, for being understanding when I moved their only grandchildren 600 miles

away; my grandpa Jerry, who was my biggest advocate; and my father, for all that he inspired.

Conversations with my friends Nikki, Steve and Jessica, Cassie and Adrian, Jason, and Sarah were

always welcome and sometimes quite helpful in working out a problem.

Thank you all for giving me the support that I needed to complete this journey.

ix
 1. INTRODUCTION

Rhetorical questions have been of interest for centuries. Around 1575, about a century after

the invention of the printing press, a reversed question mark (?) was introduced by printer Henry

Dunham. It was called the percontation point (shown in (1) and was invented as a way to punctuate

rhetorical questions (Houston, 2013).

(1) ?

Because of other printers’ use of italicized or boldfaced question marks to mimic the same punctu-

ation, the intention of the author to use this different question mark was eventually muddled to the

point of the extinction of the percontation point (Houston, 2013).

Almost 400 years later, another punctuation mark was introduced, the interrobang, shown

in (2). In part it was created to indicate a surprised or rhetorical question but it was also created

to eliminate the side-by-side use of the question mark and exclamation point separately (i.e., "!?").

This gained a bit more traction than the percontation point, but still lost favor by the early 1970s

(Houston, 2013).

(2) ‽

What this demonstrates is that rhetorical questions have been confusing for a while. There

have been times when people felt the need to make it clear that the question was not being asked in

order gain information but for some other reason. So what is a rhetorical question?

1
1.1 Don’t you know what a rhetorical question is‽

There are a number of questions that do not require answers in the form of a verbal response.

Some questions don’t require answers at all. Some are requests not for information but for some

action on the part of the addressee. All questions require something of the addressee in a cooperative

exchange, whether it be an answer, an affirmation, a rejection, or an action.

The types of questions that will be of interest here, rhetorical questions, have been examined

from numerous perspectives but there is not widespread agreement on their function or meaning.

This is in part because many rhetorical questions inherently indicate some degree of bias or speaker

commitment (Gunlogson 2002; Van Rooy & Safarova, 2003, Malamud & Stephenson, 2015; Farkas

& Roelofsen, 2017; inter alia).1

Additionally, rhetorical questions do not have a universal definition. There is general agree-

ment that rhetorical questions are not information-seeking in the traditional sense (Caponigro &

Sprouse 2007, Biezma & Rawlins 2017, Han 2002). In fact, there is a broad category of speech

acts that have the form but not the force or function of a question (e.g., requests, suggestions). I

will explore whether these should be considered rhetorical questions.

The term rhetorical question can be and has been used to refer to a large range of rhetor-

ical devices. They can be categorized according to their syntactic structure (i.e., as polar or wh-

questions), whether the answer is known to the speaker and/or other discourse participants, their

role in conversation, or their intended meaning. Some of this can be difficult to define and will be

a predominant aspect of the following discussion.

There is a type of rhetorical device that is used in written pieces and formally delivered
1
This is certainly the case in rhetorical questions with tags and likely the case in other structural types of
rhetorical questions. This will be a key component of the ensuing discussion.

2
speeches that looks like the type of rhetorical question discussed here. Below is an example taken

from a speech by President Obama:

(3) Are we a nation that tolerates the hypocrisy of a system where workers who pick our fruit

and make our beds never have a chance to get right with the law? Are we a nation that

accepts the cruelty of ripping children from their parents’ arms? Or are we a nation that

values families, and works to keep them together? (Obama White House)

This type of "rhetorical question" is deliberately designed as a way to introduce a topic or to engage

the audience in a written statement or speech. This is known as hypophora and is asked in order to

be immediately answered by the speaker (Oxford English Dictionary, n.d.).

There exist one or two other types of rhetorical questions that behave similarly in that they

are questions that the speaker is asking of themselves rather than their interlocutors. Percontatio is

the asking of an open-ended question, which in some cases do not have a known answer, like that

in (63) and (5) (Houston, 2013). Another type, interrogatio, attempts to confirm or deny a previous

argument, shown in (6) (Houston, 2013, p. 37).

(4) How do you solve a problem like Maria?

(5) How could I be so stupid?

(6) If you believe that climate change is not real, then why are the ice caps melting?

There is a monologue-like quality involved in these type of rhetorical devices, and therefore they

are considered to be different than other types of rhetorical questions, like those shown in (7).

(7) a. Is the Pope Catholic?

b. Could he be any louder?

3
 2
c. Who knows?

The yes/no questions, (7a) and (7b), have fewer potential answers and are therefore sim-

pler to interpret, even in rhetorical questions. The wh-questions are more complicated and will be

discussed in more detail below.

The most prevalent classification of rhetorical questions has been to define them as ques-

tions that have no answer, or as “a question asked only to produce an effect or make a statement,

rather than to elicit an answer or information” (Oxford English Dictionary, n.d.). On the broadest

definition, they are questions asked for any purpose other than to obtain information. Biezma and

Rawlins (2017) summarize a common perspective on rhetorical questions, shown below in (8a) and

(8b) (302). They include (8c) in their theory of rhetorical questions.

(8) a. Rhetorical questions don’t expect an answer.

b. Rhetorical questions have the feel of an assertion.

c. Rhetorical questions can optionally be answered.

Although this is one of the most recent definitions of rhetorical questions, it remains incom-

plete, even with the addition of (8c). Due in part to the indirect nature of rhetorical questions, an

additional component of rhetorical questions is that they can have the feel of a command. This will

be discussed below.

Just as the definition of what makes a question a rhetorical question is not unanimous,

previous research on rhetorical questions has defined their functions in a number of different ways.

Some of the ways in which they’ve been characterized are as redundant interrogatives (Rohde 2006),

as negative assertions (Han 2002), as retorts (Schaffer 2005), and as ordinary questions (Caponigro
2
In a context where the answer is unknown to all discourse participants, like whether there is life on Mars.

4
& Sprouse 2007).

Rohde (2006) considers data that show that rhetorical questions are not used to seek informa-

tion or to provide information. In order to be understood as such, this requires that the interlocutors

have a sufficient familiarity with each other. Because they don’t behave like ordinary questions or

as assertions, and because they are used to answer a previously asked question, Rohde classifies

them as redundant interrogatives. Farkas and Bruce (2010) also mention the redundant nature of

rhetorical questions.

Some previous work suggests that rhetorical questions are not questions at all. Han (2002)

proposes that rhetorical questions are interpreted as assertions of the opposite polarity. According

to her, when a wh-word is used in a rhetorical question, it denotes the bottom element in its de-

notational domain (214). That is to say, that in the case of rhetorical questions, a covert operator

(such as whether in the case of yes/no questions) is present in the structure and is interpreted as a

negation to the assertion rather than as a question. She does not offer a mechanism for how they

come to be understood as negative assertions.

On the other hand, an argument that rhetorical questions are semantically identical to ordi-

nary questions is offered by Caponigro and Sprouse (2007). They argue that an utterance’s inter-

pretation as a rhetorical question is at the pragmatic level. This is because they assume that both

the speaker and hearer know the answer to the rhetorical question (i.e., the location of the actual

world in the question’s partition of accessible worlds).

Biezma and Rawlins (2017) introduce speaker attitudes into the discussion and a distinction

between asking, which proposes to update the context, and questioning, which involves actually

changing what question is currently under discussion in the discourse.

This is done by adopting certain speaker presuppositions into the common ground. Those

5
presuppositions are validated by the acceptance of the rhetorical question as rhetorical. This is

similar to Schaffer’s description of their use as a retort (2005). This results in a strengthening of the

implicated assertion. She claims that rhetorical questions used as retorts are informal and “generate

indirect implicatures” (454). Schaffer suggests that speaker intentions are crucial in interpreting the

implicature or humor raised in uttering a rhetorical question.

Frank (1990) finds that defining rhetorical questions simply as questions which don’t elicit

a response is insufficient and suggests that their primary uses are as a persuasive device and as

a politeness strategy. She also attributes rhetorical questions’ effectiveness in those uses to indi-

rectness and implicature. Frank discusses the underlying representation of rhetorical questions as

containing a declarative or imperative performative verb.

Many of the theories mentioned above depend on the intentions of the speaker to not expect

a response, as well as to the hearer to not give one. This is problematic for theories that want to

attribute the meaning of rhetorical questions to the grammar (such as Han 2002). So what is a

rhetorical question? A preliminary definition ofrhetorical question is given in (9).

(9) A rhetorical question is any question that

a. does not require a linguistic response

b. does not seek information

1.2 Outline

The remaining chapters of this dissertation will address the following questions:

(10) What is a rhetorical question?

6
 (11) Are there different types of rhetorical question?

(12) What role(s) do rhetorical questions play in conversation and what are their effects?

(13) What is the illocutionary force of rhetorical questions?

This is done first by providing an overview of the literature in which this view of rhetorical questions

will be couched. Chapter 2 includes a description of previous research in the semantics and prag-

matics of questions, general theories of conversation analysis, as well as a discussion of Gricean

pragmatics and Speech Act Theory.

Chapter 3 explores different types of questions according to their use (counterquestions,

orders, unknown answers, etc.) and discusses their structure, meaning, and illocutionary force. A

revised definition of rhetorical question is offered and an analysis of the rhetorical questions that

adhere to this revised definition. A taxonomy of different types of rhetorical question is proposed.

Chapter 4 consists of a detailed framework for the pragmatic analysis of rhetorical questions.

Two pragmatic principles are defined and tested through the examination of several rhetorical ques-

tions within various discourse contexts.

Finally, Chapter 5 summarizes the project and offers a look into related phenomena that

were outside the scope of this project.

7
 2. THE LANDSCAPE

2.1 Questions

There has been much work on the syntax, semantics, and pragmatics of questions (Ham-

blin 1973; Karttunen 1977; Groenendijk & Stokhof, 1984; Roberts 2012; van Rooy & Safarova,

2003; inter alia). While the argument herein is that rhetorical questions do not function like typical

information-seeking questions, it remains the case that they are, if nothing else, syntactically inter-

rogative. There are two main types of interrogatives in the literature, wh- and alternative. Of these,

two subsets of alternative questions will be described: polar questions and tag questions. There

exist some theories of the semantics of questions that claim that all questions commit the speaker

to act (Lewis 1969; Roberts 2012; Condoravdi & Lauer, 2012). This, in addition to the pragmatic

role of questions, particularly as they affect the discourse commitments of interlocutors, will also

be considered in the analysis of rhetorical questions. This chapter will begin with a brief overview

of the semantics of questions, followed by theories on how they behave in conversation. Section 2

focuses on the effects of speech acts, with discussion of Gricean pragmatics and Speech Act Theory.

2.1.1 Alternative and Polar Questions

I will begin with a brief overview of the types and interpretations of polar and alternative

questions. Below is an example of a polar question and an alternative question, given with intonation

indicators.

(14) a. Polar question: Did Audrey dance? ↑

b. Alternative question: Did Audrey dance ↑ or sing? ↓

8
 Groenendijk and Stokhof (1984) propose a partition-based theory of questions. This ap-

proach represents questions as dividing the space of possible worlds (i.e., situations or states of

affairs) into mutually exclusive and exhaustive sets of possibilities, called "cells." Each cell corre-

sponds to a potential answer to the question.

The concept of "question under discussion" (QUD) plays a central role in many theories

on the semantics and pragmatics of questions. QUD refers to the current topic of conversation,

which helps structure the discourse and guide the flow of information exchange (Roberts 2012).

Conversational participants address the QUD by providing information that narrows down the cells

in the partition until a specific answer is reached.

In the context of alternative questions, Groenendijk and Stokhof’s theory treats them as a

series of individual yes/no questions. For example, consider the alternative question: "Do you want

tea or coffee?" This question can be viewed as two separate yes/no questions: "Do you want tea?"

and "Do you want coffee?" The QUD provides a framework to address each alternative separately

until the desired answer is found.

For a question like (14a) there are only two cells in the partition: the one in which Audrey

danced and the one in which she didn’t. There is a little more involved with alternative questions.

(14b) presupposes that Audrey did only one of the two conjuncts introduced in the question. The

partition for (14b) would look something like in Table 2.1, in which only w1 and w3 are possible.

2.1.2 Wh-Questions

Wh-questions differ from yes/no questions in the G&S framework because they create more

complex partitions. While yes/no questions divide the space of possible worlds into two cells (one

9
 Audrey danced Audrey sang
w1 ✓
w2 ✓ ✓
w3 ✓
w4

Table 2.1: Partition for alternative question

for "yes" and one for "no"), wh-questions create multiple cells corresponding to various potential

answers.

For example, consider the wh-question: "Where are you going on vacation?" This question

creates a partition with cells for each possible vacation destination, such as London, Barcelona,

Milan. The respondent’s answer will identify the specific cell (destination) that applies to their

situation.

Wh-questions differ from alternative and polar questions semantically and syntactically.

While alternative questions introduce a closed set of alternatives (e.g., yes or no), a wh-question

introduces a set of complete answers to the question (such as the one in (15b)). This is shown in

(16).

What the G&S semantics here does is create a partition, where two worlds will occupy the

same cell if they share a strongly exhaustive answer to the question, i.e. if they agree on who the

dancers are.

(15) a. Did Audrey dance?

b. Who danced?

(16) [[ Who danced?]] = λw1 λw2 .[(λx.x dancedw1 ) = (λx.x dancedw2 )]

Hamblin’s (1973) theory focuses on the idea that a question can be understood as a set of possible

10
answers. In other words, when someone asks a question, they are essentially requesting information

from a range of potential responses. For example, consider the question: "Who danced?" According

to Hamblin’s theory, this question can be represented as a set of all possible propositions of the form

"x danced", and the respondent’s answer will be one element from this set.

Hamblin’s proposal for the semantics of wh-questions is that the wh-item is interpreted as

an existential quantifier, the value for which can be selected from a contextually relevant set of

propositions (e.g., Audrey danced, Laura danced, Shelly danced).

(17) Q = λp.∃x[people@ (x) ∧ p =ˆdanced(x)]

= {ˆdanced(x) : x ∈ people@ }

(18) Q = λp.∃x[people@ (x) ∧ p =ˆdanced(x)]

= {ˆdanced (x) : x ∈ people@ }

= {danced(A)}, {danced(S)}, {danced(L)}

In addition to the semantic interpretation of questions, intonation is another aspect of sev-

eral recent studies of questions (Gunlogson 2002, Biezma and Rawlins 2012, Biezma and Rawlins

2016). Farkas and Roelofsen (2017) categorize the rising and falling intonation that occurs with

declaratives and interrogatives.

(19) a. Audrey danced. ↓ [falling declarative]

b. Audrey danced? ↑ [rising declarative]

c. Did Audrey dance? ↓ [falling polar interrogative]

d. Did Audrey dance? ↑ [rising polar interrogative]

e. Audrey danced, ↓, didn’t she ↓? [falling tag interrogative]

11
 f. Audrey danced, ↓, didn’t she ↑ ? [rising tag interrogative]

While rising and falling declaratives will appear in the following discussion, the star of the

analysis here will be questions, with particular attention to polar questions. Polar questions have

two basic syntactic structures. The canonical polar question, shown in (19d), has the structure of an

interrogative in English (i.e., T-to-C movement or subject-auxiliary inversion). The tag question,

shown in (19e), has a declarative anchor followed by a tag consisting of an auxiliary verb and a

pronoun.

2.1.3 Tag Questions

Tag questions present some interesting complications in major semantic theories of ques-

tions. There are several reasons given for this in the literature. One is that they occur mostly (or only)

in spoken language, making this a peripheral phenomenon. Another is that there is not agreement

on whether tag questions have the semantics of alternative or polar questions. Additionally, there

is not agreement on how the speaker’s intention in uttering the tag question affects the discourse

context. This also seems to be the case for rhetorical questions (discussed in the next section).

Types of tag questions have been widely discussed (Tomaselli & Gatt, 2015; Kimps, 2007;

Kimps et al., 2014; Tottie & Hoffmann, 2006; inter alia). Tag questions are generally categorized

into four main types in the literature. These are given below in Table 2.2, adapted from Kimps,

(2007, 273).

Kimps (2007) finds that the speaker uses tag questions to introduce implicit information

about their commitment towards the truth of the proposition. In this way, tag questions give attitu-

dinal information (i.e., they introduce bias, see examples (20-23)).

12
 Mood of Stem Polarity Example
Declarative Reverse + – You’re going, aren’t you?
Reverse – + You aren’t going, are you?
Same + + You’re going, are you?
Same – – You’re not going, aren’t you?
Interrogative Same + Are you going, are you?
Imperative Reverse + – Come here, won’t you
Reverse – + Don’t come here, will you?
Same + + Come here, will you?
Same – – Don’t come here, won’t you?
Exclamative Reverse + – What a game, wasn’t it?

Table 2.2: Main Types of Tag Questions in English (Kimps 2007)

Tottie and Hoffman (2006) synthesize pragmatic classifications of tag questions from pre-

vious studies on tag questions. They found that over 90% of the tag questions found in British and

American English corpora were from three categories, shown below.

(20) Confirmatory: the speaker wants confirmation

Ex: I don’t need a jacket, do I?

(21) Facilitative: speaker is sure of the truth but wants to involve the hearer

Ex: The answer is 47, isn’t it?

(22) Attitudinal: for emphasis, does not expect a reply

Ex: I don’t know where it is, do I?

There is one more category that is relevant to the present discussion. Algeo (2006) defined

peremptory tag questions as tags that “follows a statement of generally acknowledged . . . truth,”

that is “intended to . . . close off debate.” This is similar to the meaning of RQs such as Does the

sun set in the west?.

Tottie and Hoffman’s discussion focused on the sociolinguistic and dialectal distribution

13
of tag questions. Tags are used considerably more by British English speakers and women. One

interesting finding from their corpus study was the age distribution of tag question use in British

versus American English. The highest rate of use in the British corpora was in ages 25-34, but it

was not significantly more than other age groups (2006, p. 305). The American corpus, on the

other hand, had a much higher rate of tag question usage in speakers over the age of 60 years old.1

Van Rooy and Safarova (2003) also discuss the bias of tag questions with negative versus

positive anchors and that they can’t be used interchangeably without changing the effect of the

question on the discourse context.

(23) a. John is nice, isn’t he?

b. John isn’t such a bad guy, is he?

They argue that these examples demonstrate that the speaker intends to assert the proposi-

tions expressed by the anchor and therefore are biased toward that proposition.2 Interestingly, the

examples in (23) are labeled as rhetorical questions by van Rooy and Safarova (2003, p. 293). This

is because the examples in (23) are uttered by speakers who have the belief that John is nice in their

commitment set (more on this below) and are therefore not requesting that information.

2.2 Pragmatics

Investigating the pragmatics and illocutionary force of RQs and related constructions could

be a useful way to differentiate information-seeking questions (ISQs) from RQs, particularly given
1
Interestingly, the only tags which were consistently used more in the American corpus in Tottie and
Hoffman’s study were those with negative anchors and positive tags.
2
Here van Rooy and Safarova make a distinction between different levels of commitment, presumably
to demonstrate that a weak commitment to John’s niceness (in (23b)) indicates bias in the same way that a
strong commitment does (in (23a)). If the example in (23b) had been "John isn’t nice, is he?" it would have
the same level of commitment to the negation of John is nice.

14
the role of context in their interpretation.

2.2.1 Gricean Pragmatics

Grice’s Logic and Conversation (1975) introduced several innovative notions of the behav-

ior of cooperative communication. A major contribution of this seminal work was the introduction

and definition of conversational implicatures.

Conversational implicatures are used to convey some meaning other than the surface or

literal meaning. This secondary meaning comes about due to the speaker’s intentional violation

of what Grice (1968) refers to as the “Cooperative Principle.” This principle basically states that

a speaker’s contribution to the conversation should be appropriate (in its length, timing, direction,

etc.). In order to define this more clearly, Grice breaks it down into four categories: Quantity,

Quality, Relation, and Manner. These are assumed to be followed in a cooperative conversation.

(24) a. Quantity: Make your contribution as informative as is required. Do not say more than

is required.

b. Quality: Contribute only what you know to be true. Do not say false things. Do not

say things for which you lack evidence.

c. Relation: Make your contribution relevant.

d. Manner: Avoid obscurity and ambiguity. Be brief and orderly.

Speakers can deliberately flout (i.e., not follow) one or more of these maxims with an ut-

terance. Interestingly, it seems to be the case that RQs flout all of these maxims. Consider the

exchange in (25).

15
 (25) A: Do you want to watch Twin Peaks?

B:Does the sun set in the west?

A question as a response is by its very nature not providing enough information to answer

the previous question (Quantity). A question has no truth value (Quality). There is no obvious

connection between Twin Peaks and the sunset (Relation). In (25), B’s utterance is a question and

an answer. All RQs are ambiguous in that they have at least two interpretations (Manner). What

are the effects of flouting these maxims?

The category of Quantity puts forward that speakers should make their contributions as

informative as necessary but without being too informative. To illustrate this point, Grice gives the

example of a professor writing a letter of recommendation to a prospective employer for Mr. X.

The letter reads as follows:

(26) Dear Sir, Mr. X’s command of English is excellent, and his attendance at tutorials has been

regular.

In this case, the professor is intentionally violating (or flouting) the maxim of Quantity by failing

to be as informative as is required or expected of a recommendation letter. In doing so, the professor

is implicating that X has no other positive attributes.

The category of Quality concerns the truth of what-is-said, specifically that speakers will

avoid intentionally lying and saying things for which they have no evidence. Examples of when

these maxims are exploited can be found in the use of some linguistic devices such as irony,

metaphor, and hyperbole, for example:

(27) There are a million things I have to do today.

The utterance in (27) is not meant to be taken literally, but rather to implicate that the speaker has

16
a lot to do and probably feels overwhelmed by the quantity.

Grice’s position on the maxim of Relation is that it rarely results in an implicature (more on

this below), but that when it does, it looks something like this:

(28) A: Mrs. X is an old bag.

B: The weather has been quite delightful, hasn’t it?

In other words, speaker B has made a deliberate conversational move away from speaker A’s utter-

ance as a way to delegitimize it.

Finally, under the category of Manner are various maxims such as “avoid ambiguity” and

“be orderly”, among others. For example, imagine A has been arrested and calls B from jail to

inform B of the location of some piece of evidence. Speaker A says the following:

(29) It’s in the place where I put that thing that time.

In this case, speaker A is flouting the maxim of “avoid obscurity” in order to achieve the conversa-

tional goal of informing B while preventing others (e.g., law enforcement) from being able to work

out the intended implicatum.

There are cases, on the other hand, wherein it is unclear whether a maxim has been violated.

Grice claims that these are most often related to the maxim of Relation. An example that he gives

to illustrate this point is the following exchange between two friends about a mutual friend of theirs:

(30) A: Smith doesn’t seem to have a girlfriend these days.

B: He has been paying a lot of visits to New York lately.

If speaker B here is understood to be observing the maxim of Relation, it is assumed that they

believe the resulting implicature (in this case, either that Smith has a girlfriend in New York, or that

17
due to his frequent trips to New York, Smith is too busy for a girlfriend).3

Occasionally, one maxim might be violated in order to avoid violating another. Grice de-

scribes a scenario in which two friends are planning to visit C on their trip to France.

(31) A: Where does C live?

B: Somewhere in the South of France.

Speaker A is intentionally violating the maxim of Quantity to be as informative as required (this

information isn’t specific enough for the situation) in order to avoid violating the maxim of Quality

“don’t say that for which you lack adequate evidence.” In situations like this, there is no conversa-

tional implicature.

In addition to situations in which no conversational implicatures arise, Grice discusses cer-

tain features that conversational implicatures have that distinguish them from other types of linguis-

tic relations (such as presupposition and entailment).4 One of these features is that conversational

implicatures are explicitly cancellable. In other words, a speaker may continue their utterance

without contradiction, thereby cancelling the possible implicature. This is most easily illustrated

through the violation of the maxim of Relation. Imagine that John usually rides his bicycle to work

and his roommates have the following exchange:

(32) A: When will John get back?

B: The weather is terrible today but he’ll be home early.

3
The fact that there is more than one implicature here speaks to the feature of indeterminacy which is
mentioned below.
4
This differs from conventional implicatures, which do not follow these diagnostics. For example, a
speaker, in uttering (1a), is implicating a causal relationship between being an Englishman and being brave
without overtly stating such a relationship. A speaker cannot utter (1b) without contradicting themselves.
(1) a. He is an Englishman; he is, therefore, brave.
b. #He is an Englishman; he is, therefore, brave, but Englishmen are cowards.

18
The implicature that arises from the first clause of B’s utterance is that John could take longer than

usual to get home because of the storm; however, this implicature is effectively cancelled without

contradicting the second clause of what-is-said.

Another feature of conversational implicatures is that they remain intact even when the

words used to express the utterance (i.e., what-is-said) are different. Grice calls this detacha-

bility. For example, A and B were very close but recently had a significant falling-out. During

a conversation with C, A says, ‘B is a great friend,’ flouting the maxim of “do not say what you

believe to be false.” Speaker A can implicate that B is a terrible friend through a number of other

utterances, shown in (33).

(33) a. B is the best guy I’ve ever known.

b. I would recommend the friendship of B to anyone.

c. Friends don’t get any better than old B!

This example also illustrates that the truth of the conversational implicature is independent

of the truth-conditions of what-is-said. In other words, the utterance can be false and the implicature

true, as in (33), or what-is-said can be true while the implicature is false.

Because speakers are assumed to be following the Cooperative Principle and because con-

versational implicatures necessarily have a non-conventional meaning, they must be calculated by

the hearer. This causes them to be indeterminate to an extent (see (30) above).

A key component of these features (and some other implicatures more generally) is that

most of them are either dependent on the intentions of the speaker or on the hearer’s interpretation

of what-is-said. In other words, the speaker is relying on the hearer to be able to work out that a

conversational implicature is present based on a shared understanding of the following:

19
 (34) a. linguistic conventions

b. the context of the utterance

c. adherence to the Cooperative Principle

d. the speakers’ background knowledge

e. the common ground

2.2.2 Speech Act Theory

In How to Do Things with Words Austin proposed a theory of speech acts and the effects of

those acts. Austin’s Speech Act Theory deals with the responsibilities and reactions of conversa-

tional participants in their roles as speakers and hearers. Speech acts are divided into three types.

First is the locutionary act, which he defined as the act of saying something “in the normal sense”

(94). A sentence like (35), when uttered, can be taken as a mere act of saying (but crucially, not

with the intention to inform the hearer).

(35) It is raining.

It can also be taken as a warning or prediction, which results in the utterance having a certain effect

or force. When this occurs, the utterance becomes an illocutionary act. This is the speech act

that in saying it, the speaker is doing something.

The final speech act that Austin discusses is the perlocutionary act. This is performed by

the speaker when their utterance results in some action or reaction by the hearer. In Austin’s words,

the speaker is doing something by saying it. A speaker will have performed a perlocutionary act in

uttering (35) if the hearer of that utterance is alarmed by it and grabs their umbrella before leaving

20
the house. This is a perlocutionary effect. It is the carrying out or performing of an action by

the hearer upon hearing the speaker’s utterance.

This seems to suggest that any utterance that is understood by the hearer (i.e., any illocu-

tionary act) has the perlocutionary effect of causing understanding through changing the common

ground of the conversational participants. According to Searle and Vanderveken (1985), the act of

understanding is an illocutionary effect (12). In fact, there is some general disagreement on what

constitutes a perlocutionary act or effect (see Searle & Vanderveken 1985; Bach & Harnish 1979;

Kurzon 1998, for some discussion). This is likely due, in part, to the fact that Austin himself does

not offer the clearest definition in his seminal work on the subject.

Austin mentions in passing a scenario in which “one performative utterance is used as an

indirect means to perform another act” (130). To illustrate this, he gives the example of bidding in

clubs in a card game to inform (or implicate) that he has no diamonds. Austin labels both of these

acts as illocutionary and because no further context is given, this remains the case. However, if the

implicature (i.e., that the speaker has no diamonds) results in the hearer passing the bid, or bidding

in hearts instead of diamonds, then the speaker would have performed both an illocutionary act in

what was said, as well as a perlocutionary act in what was meant.

Although Grice (1975) doesn’t explicitly discuss the effects on the hearer of conversational

implicatures, he offers enough information for us to infer that these utterances can result in perlocu-

tionary effects. Take, for example, B’s utterances in (33). Assume that A convinced C that B is a

bad friend. This is an illocutionary act. However, if C stops talking to B and stops following B on

social media, then A will have performed a perlocutionary act through what was meant (i.e., that B

is a terrible friend) rather than what was said (that B is a great friend).

This is the case for most of the implicatures discussed above. An additional illustration

21
can be seen in (29) when exploitation of the category of Manner occurs. If it results in the hearer

working out the meaning of the deliberately obscure utterance (i.e., the object and location of ‘that

thing’, the evidence) and has the outcome of B retrieving said evidence (the perlocutionary effect),

then A has performed a perlocutionary act.

When successful, it is the conversational implicature that will result in a perlocutionary

effect, but this is not always the case. If the letter of recommendation written for Mr. X in (26)

resulted in his being hired for the philosophy job, then the professor’s implicature would have been

ignored, and it would be what was said (not what was meant) that was the perlocutionary act.5 On

the other hand, if Mr. X does not get hired as a result of the professor’s implicature that he is a bad

candidate for the job, then the perlocutionary act is what was meant, not what was said.

All of this suggests that a perlocutionary act does not need to be spoken in the traditional

sense. Austin spends some time describing the effects of nonlinguistic communication (such as

waving a big stick). The stick wielder can intimidate or cause fear and can intend to produce those

effects. Gesture can also contribute to the meaning of what-is-said. For example, if a speaker

shrugs their shoulders as they make a statement such as “The restaurant is three blocks away”, it can

indicate uncertainty about the truth of the statement.6 If conversational implicatures are understood

as being a form of nonlinguistic communication, then this could be evidence that other forms of

nonlinguistic communication can be perlocutionary acts and can result in perlocutionary effects.

Jeong and Potts (2016) designed an experiment to investigate illocutionary inferences, per-

locutionary effects, and the interactions that those have with the type of sentence uttered (e.g.,
5
Bach and Harnish (1979) limit the classification of perlocutionary effects to those that are intentional
only, so this would be considered an illocutionary effect, not perlocutionary. A position will not be taken
here as to whether a speaker’s intentions are necessary in order for an a perlocutionary act to have been
performed.
6
This can be accompanied by a rising intonation, which also indicates uncertainty.

22
declarative, interrogative, imperative). They did this in part through manipulating intonation pat-

terns in constructions that can be interpreted in more than one way (e.g., Can you close the window).

They state that there "seem to be conventional effects attached to specific type + tunes, guiding per-

locutionary effects relating to the listener’s view of the speaker’s social persona" (4). This study

suggests that there are important distinctions to be made between illocution and perlocution (and

that both remain relevant in examinations of speech acts).

Speech acts can be categorized into two broad classes: direct and indirect. Indirect speech

acts occur when the intended meaning of an utterance differs from its literal or surface meaning.

According to Searle, an indirect speech act involves a speaker conveying a particular illocutionary

force indirectly, rather than stating it explicitly. Illocutionary force refers to the intended function or

speech act performed by an utterance, such as making a request, giving an order, asking a question,

or expressing a command. In indirect speech acts, the illocutionary force is derived from the context

and the pragmatic interpretation of the utterance.

A direct speech act is an utterance whose illocutionary force (i.e., intended meaning) matches

its literal meaning. For example, if someone says "I promise to be there at 5 pm", their intended

meaning is conveyed directly through the words they use.

An indirect speech act, on the other hand, is an utterance where the intended meaning does

not match the literal meaning of the words used. Instead, the speaker relies on context, social norms,

or other linguistic cues to convey their intended meaning indirectly. Indirect speech acts involve

conveying illocutionary force indirectly, relying on context and pragmatic interpretation. Searle

and Vanderveken’s work provides a theoretical framework for understanding how indirect speech

acts function in communication and how they are interpreted within a social and linguistic context.

Indirect speech acts have several key features. One of the most important is the use of

23
language in ways that can be implied rather than stated directly. This can include the use of sarcasm,

irony, or other forms of figurative language. For example, if someone says "Oh great, now it’s

raining" when they really mean "I’m annoyed that it’s raining", they are using irony to convey their

true feelings indirectly.

Another feature of indirect speech acts is that they often rely on context to make sense. For

example, if someone says "Can you pass the salt?" during a meal, their intended meaning is likely a

request for salt. However, if they say the same thing during a job interview, their intended meaning

might be a test of the interviewee’s communication skills.

Finally, indirect speech acts can be used strategically to achieve specific goals, such as to

be polite, assertive, or persuasive. For example, if someone says "I was wondering if you could

help me with this project", they are using an indirect speech act to make a request in a polite and

deferential way.

Examples of indirect speech acts are numerous and can include:

(36) a. A boss saying "Would it be possible for you to work overtime this week?" when they

really mean "I need you to work overtime this week."

b. A parent asking a child "Do you know what time it is?" when they really mean "It’s

time for bed."

c. An employee saying "I was wondering if we could talk about my salary" when they

really mean "I want a raise."

Speech acts are further categorized according to their illocutionary force. For the most part,

questioning falls with the category of directives in Searle’s (1965) taxonomy of speech acts, shown

below (taken from Roberts, 2018, p. 322).

24
 (37) a. Assertives: Commit the speaker to the truth of a proposition: suggesting, putting for-

ward, swearing, boasting, concluding. No one makes a better cake than me.

b. Directives: Attempt to make the addressee perform an action: asking, ordering, re-

questing, inviting, advising, begging. Could you close the window?

c. Commissives: Commit the speaker to some future course of action: promising, plan-

ning, vowing, betting, opposing. I’m going to Paris tomorrow.

d. Expressives: Express how the speaker feels about a state of affairs: thanking, apolo-

gizing, welcoming, deploring. I am sorry that I lied to you.

e. Declarations: Change the state of the state of the world to bring it into conformity with

the propositional content: You are fired, I swear, I beg you, I hereby pronounce you man

and wife.

Roberts (2018) suggests that this taxonomy be amended to more accurately describe linguistic phe-

nomena, as shown in (38).

(38) a. assertion: an act of proposing an addition to the interlocutors’ C(ommon)G(round)

(Stalnaker 1979).7 If adopted, this addition would commit the interlocutors to accepting

that (and behaving as if) the world fits the words. Note that this is a weaker commitment

than belief, in keeping with Stalnaker’s (1979) characterization of the interlocutors’

Common Ground.

b. suggestion: an act of proposing that interlocutors adopt intentions to act in specific

7
The concept of common ground as defined by Stalnaker (from his 2002 paper Common Ground), is given
below:
(1) It is common ground that ϕ in a group if all members accept (for the purpose of the conversation)
that ϕ, and all believe that all accept that ϕ, and all believe that all believe that all accept that ϕ, etc.

25
 ways. There are two types of speech act proposals reflecting an essential distinction in

the types of goals interlocutors may propose in discourse:

i. direction: an act in which a speaker proposes to her addressee that he adopt a

particular intention to act in the world. This is the sort of speech act typically

performed with an imperative. It is a proposal to make the world fit the words.

ii. interrogation or question: an act of proposing that the interlocutors collectively

commit to collaborative inquiry, thus an act which would establish a direction for

the discourse itself. It is a proposal that the interlocutors endeavor to discover the

proper fit between world and words, thereby resolving the question.

Roberts’ taxonomy pays more attention to the role that speech acts play in conversation than

earlier analyses in Speech Act Theory did. Searle and Vanderveken (1985) attribute to the illocu-

tionary act various degrees that when built upon each other interact with the context in different

ways. This can be seen in certain RQs, shown in (39).

(39) a. A: Is Mary coming to the party?

B: Who knows?

b. A: I won at Scrabble again last night.

B: What else is new?

c. What’s the matter with kids these days?

The example in (39a) is similar to Rohde’s (2006) proposal that RQs are retorts. In this case

it can mean something like "I don’t know." In a scenario where it is well-known that Mary does

not attend parties, it could be a sarcastic way of answering "no." (39b) is an example of an RQ that

doesn’t expect a response. The fact that (39c) has more than one meaning could be due to a number

26
of factors. One is that this is a question that has (or expects) no particular answer (i.e., a non-null

answer RQ, see Rohde 2006). It could also be an assertion of the same polarity (contrary to Han

2002), meaning something like “There’s something the matter with kids these days.” These types

of RQs will be among those discussed in the next chapter.

2.3 The Conversational Scoreboard

There have been several analyses in the last 20 years of the role of discourse components

in conversation (Roberts 2012; Farkas & Bruce, 2010; Malamud & Stephenson, 2015; Farkas &

Bruce, 2017). Much of it has focused on polar questions. In this section, I will summarize a few of

those theories in addition to Portner’s (2004) analysis of imperatives’ contribution to the discourse

context. Components of these frameworks will be used to evaluate the contribution of RQs within a

conversation. The idea of the conversational scoreboard is related to David Lewis’s (1979) theory

of conversational scorekeeping, which suggests that participants in a conversation keep track of

various factors, such as presuppositions, implicatures, and common ground, in order to understand

and interpret each other’s utterances effectively.

Building on Roberts (2012) (to be discussed in greater detail below) and Gunlogson (2008)

among others, Farkas and Bruce (2010) discuss how questions and assertions affect the conversa-

tion differently. Their “discourse context” is comprised of three components: the discourse com-

mitments of the participants, the common ground, and the projected set. Discourse commitments

are those propositions that a participant has publicly committed to in the conversation. The com-

mon ground differs from discourse commitments in that it is the set of commonly held beliefs by

all discourse participants. The common ground is constantly changing as propositions are added

27
to the conversation, whereas discourse commitments remain largely the same. The projected set is

the superset of future common grounds proposed by a new proposition being added to the Table.8

The main objectives of a conversation according to Farkas and Bruce are to increase the

common ground and to reach a stable state. A conversation is unstable if an item is on the table

(i.e., an unresolved projected set). The projected set is the most intriguing and relevant of these

components because it is this that is affected most by the addition of questions to the Table. How-

ever, in the case of assertions, only one proposition is added to the projected set. Questions add the

set of potential answers to the question, which in the case of polar questions is p or ¬p. Below is

an example of the effects of a polar question in a conversation between two discourse participants

(Farkas & Bruce, 2009, p. 95).

(40) K4 ‘Is Sam home?’ was asked relative to the initial input context K1

Table 2.3: Context State K4 (Farkas and Bruce (2009)

A Table B
⟨ ’Sam is home’[I]; {p, ¬p} ⟩
Common Ground s1 Projected Set ps1 = {s1 ∪
{p}, s1 ∪ {¬p}}

The empty cells beneath A and B in the table represent the discourse commitments of the

participants with respect to the current issue on the Table. Because it’s a question, there are no

commitments.9 The most efficient way to reach a stable state after the introduction of a question is

by the addressee providing an answer to the question, thereby to the common ground. Although they
8
Farkas and Bruce’s table is analogous to Roberts’ question-under-discussion stack.
9
In the case of many RQs, these cells would not be empty, cf. (23).

28
don’t address it in their analysis, alternative questions should add all the relevant salient answers to

the question to the projected set.

Because RQs are rarely interpreted as questions in the traditional sense, it seems likely that

the effect on the projected set would be the implicated meaning (i.e., what-is-meant) of the RQ

(and would therefore contribute to something other than the QUD). Farkas and Bruce mention that

implicated content behaves differently than literal content but concern themselves with the literal

content only.

This is an interesting issue. I think it will depend on how mechanistically we think the F&B

model should interface with linguistic forms. If we take a more mechanistic view, then that might

open up some possibilities for explanation: could a speaker use an RQ to introduce absurdities into

the projected set, and thereby steer things toward a preferred destination?

Farkas and Roelofsen (2017) propose a similar definition of the discourse context, given in

(41) (255).

(41) A basic discourse context is a triple (participants, table, commitments) where:

a. participants is the set of discourse participants;

b. table is a stack of propositions, representing the proposals made so far;

c. commitments is a function that maps every participant x ∈ participants to a set of

possibilities, those possibilities that x is publicly committed to.

Farkas and Roelofsen define a basic convention of use in which the utterance of a declarative

or interrogative sentence affects the discourse context by (i) adding the proposition expressed by

the sentence to the table, and (ii) by adding the information content of the sentence to the discourse

commitments of the speaker (265).

29
 This is the case for falling declaratives and rising polar interrogatives because there is either

full commitment or no commitment toward the proposition. Special discourse effects come into play

in the case of rising tag interrogatives and falling polar interrogatives. In addition to the proposition

and informative content being added to the table and to the commitments of the speaker, a special

effect of the level of the speaker’s evidence toward one of the alternatives is added to the discourse

context.

Sentence type Type of commitment

Falling declarative full commitment to one alternative
Rising declaratives bias toward one alternative, but no full commitment
Tag interrogative bias toward one alternative, but no full commitment
Polar interrogative neutral

Table 2.4: Types and commitments (Farkas and Roelofsen 2017)

A crucial aspect to Farkas and Roelofsen’s framework is the evidence that discourse partic-

ipants signal to have in addition to their commitments. As shown Table 2.3, rising declaratives and

tag interrogatives are not neutral with respect to speaker bias, which indicates evidence toward one

of the propositions over the other.10 The examples below from Farkas and Roelofsen demonstrate

how their definitions of evidence and commitments affect the discourse context.

(42) Context: Belinda is going through a pile of job applications. Chris has not seen any of

them yet. Belinda hands Chris the application that she just finished reading, and tells him

to have a look at it. Chris to Belinda:

a. This is a good one? ↑

10
Bias can be signaled in other ways, such as the structure of the tag, that is, whether the anchor and tag
are positive or negative. This will be discussed below.

30
 b. # This is a good one, isn’t it? ↑11

(43) Context: Belinda and Chris are looking at a sunset together. Belinda to Chris:

a. # This is a beautiful sunset? ↑

b. This is a beautiful sunset, isn’t it?

The intonation of the tag in (43) does not affect its overall interpretation. However, if the

tag in (42b) has falling intonation it could indicate that Chris believes that Belinda has a reason for

handing him that particular application, namely that the application is good. The felicitousness of

the sentences in (42) and (43) is derived through the effects described in (44), the basic meaning

of which is that the proposition (i.e., the anchor) is added to the table and the speaker has a strong

commitment to the content of the anchor.

(44) Conventional discourse effects of falling tag interrogatives

When a discourse participant x utters a falling tag interrogative ϕ, expressing the proposition

ϕ = {α, α}↓, the discourse context is affected as follows:

1. Basic effect

• The proposition expressed by ϕ, {ϕ}, is added to the table.

• The informative content of ϕ, {ϕ} is added to commitments(x).

S

2. Special effect

• ⟨α[high]⟩ is added to evidence(x)

11
Note that this would be felicitous if the tag had the same polarity as the anchor (as in This is a good one,
is it?). Farkas and Roelofsen don’t discuss same-polarity tags but it is the focus of Malamud and Stephenson
(2015) discussed below.

31
 In (44), ⟨α[high]⟩ refers to the credence level of the speaker. This is defined as “the degree

to which the speaker believes the alternative itself to be more likely than its complement” (256).

(43b) is felicitous because Belinda believes it to be a beautiful sunset more than she believes it not

to be. On the other hand, (42b) is infelicitous because Chris has been presented with no evidence

to believe that the application that Belinda is handing him is good.

In applying this to an example like the question Can you pass the salt?, the speaker has

evidence that the addressee likely has the ability to pass the salt, in which case adding p or ¬p to

the discourse context would not resolve the question. They admit that these additional pragmatic

effects are not their concern; nevertheless, components of their framework can be used to add to

the understanding of these effects.

While Farkas and Bruce (2010) and Farkas and Roelofsen (2017) discuss tags that have

the opposite polarity of the anchor, Malamud and Stephenson (2015) break tag questions down

further and discuss both same-polarity (SP-tag) and reverse-polarity (RP-tag) tags. They argue

that the polarity of the tag affects the interpretation of the question. Their judgments all concern

information-seeking questions and therefore are different than they would be for RQs.

Malamud and Stephenson (2015) build on Farkas and Bruce (2010) with two interesting and

useful modifications. The first is their addition of a projected set of commitments. They argue that

this allows participants to make tentative commitments toward the proposition, which they refer

to as an “implicature of tentativeness” (Malamud & Stephenson, 2015, p. 288). The projected

commitment set also serves to allow speakers to guess the commitments of others, adding to the

hearer’s projected commitments.

In the examples below, an uninformed speaker is guessing what the addressee’s potential

commitments are, which demonstrates a tentative commitment to the proposition expressed by the

32
anchor in the cases with tags. RP-tags only add p to the common ground and are therefore infelici-

tous in cases where the addressee has no prior knowledge or experience regarding the proposition

(as shown in (45a) (where Malamud & Stephenson use a question mark it indicates rising intona-

tion).12

(45) Context: A and B are gossiping. A doesn’t know anything about B’s neighbor. B says,

blushing, ‘You’ve GOT to see this picture of my new neighbor!’

Without looking, A replies:

a. # A: He’s attractive, isn’t he?

OK
b. A: He’s attractive, is he?

OK
c. A: He’s attractive?

d. # A: He’s attractive.

(46) Context: A teacher (B) is quizzing a student (A) on state capitals. The teacher says: ‘What’s

the capital of New York?’ The student isn’t sure of the answer, but thinks it might be Albany.

The student says:

OK
a. A: It’s Albany, isn’t it?

b. # A: It’s Albany, is it?

OK
c. A: It’s Albany?

OK
d. A: It’s Albany.

The effects of RP-tags on the scoreboard are shown in (47). Malamud and Stephenson use

an asterisk to signify the projected sets.

12
(45a) can be understood as rhetorical with falling intonation.

33
 (47) A utters p with an RP-tag:

(Proposition q is already in the CG )

(i)
Previously After A’s move (uttering p with an RP-tag)
DCA {} {}
DCA* {{ }} {{p}}
DCB {} {}
DCB* {{ }} {{p}}
Table ⟨⟩ ⟨{p}⟩
CG {q} {q}
CG* {{q}} {{q, p}}

Portner (2004) claims that there are three universal clause types: declaratives, interroga-

tives, and imperatives. He approaches the contributions that these types make in a conversation

in terms of their discourse functions and their force. Portner argues that “force is in a one-to-one

correspondence with type” (236). Therefore, the type declarative has the force of an Assertion,

interrogative the force of Asking, and imperative the force of Requiring. On this theory of force,

there is not a distinction among the different subtypes of imperatives (e.g., orders, threats).

Portner’s analysis of conversation is in keeping with Roberts (2012), where assertions con-

tribute propositions to the Common Ground and questions add a set of propositions to the Question

Set. Imperatives add a property to the addressee’s To-Do List. The To-Do List is an individual

component whereas the other two are shared among the interlocutors. This is shown in Table 2.5.

This works due to a preference-ordering of possible worlds, with the most preferable one

influencing the interpretation of the imperative. These preferences are derived through the assump-

tion that participants in the conversation are adhering to the Cooperative Principle (discussed in

34
 Type Denotation Discourse Component Force
Declarative proposition (p) Common Ground Assertion
CG ∪ {p}
Interrogative set of propositions (q) Question Set Asking
set of sets of propositions QS ∪ {q}
Imperative property (P) To-Do List Function RequiringA
function from individuals to TDL(A) ∪ {P}
sets of properties

Table 2.5: Clause Types (Portner 2004)

Section 2.2.1). Portner (2004) gives the definitions in (48) and (49) for the implementation of the

To-Do List (p. 243).

(48) Partial Ordering of Worlds:

For any w1 , w2 ∈ CG, w1 <w2 iff for some P∈TDL(i), P(w2 )(i)=T and P(w1 )(i)=F, and
T

for all Q∈TDL(i), if Q(w1 )(i)=T, then Q(w2 )(i)=T.

(49) Agent’s commitment

For any agent i, the participants in the conversation mutually agree to deem i’s actions

rational and cooperative to the extent that those actions in any world w1 ∈ CG tend to
T

make it more likely that there is no w2 ∈ CG such that w1 <w2 .

T

Condoravdi and Lauer (2012) offer a more simplified version of this agent commitment to

avoid quantifying over (and thereby complicating) the common ground (55).

(50) An agent i is committed to act in such a way so as to make true as many propositions on

TDL(i) as possible.

They also discuss the relationship between illocutionary force and imperatives and offer an

analysis of the TDL that takes utterance force into consideration. Bach and Harnish (1979) further

35
developed the concept of illocutionary acts by distinguishing between illocutionary force and illo-

cutionary content. Illocutionary force is determined by the speaker’s intention, while illocutionary

content is derived from the linguistic form of the utterance. For example, consider the statement, "I

promise to help you." The illocutionary force is the act of promising, which reflects the speaker’s

intention to commit to future assistance. The illocutionary content consists of the information that

the speaker will help the listener. Both aspects contribute to the overall meaning and impact of the

utterance.

2.4 Summary

In this chapter much of the previous literature that bears on the present analysis has been

offered. This included a summary of major theories on the semantics of questions. Because this

is a predominantly pragmatic analysis, Grice’s Maxims of Conversation and Austin’s Speech Act

Theory are outlined. Finally, current takes on the role of questions in conversation and their effects

on the beliefs/commitments of participants within the discourse as well as on the common ground

are discussed. The next chapter will provide examples of how different types of questions behave

at the intersection of these various theories of interpretation.

36
 3. A TAXONOMY OF RHETORICAL QUESTIONS

There are two main characteristics of rhetorical questions that distinguish them from information-

seeking questions: that their answers are obvious (or unknowable) and that their illocutionary force

is not inquisitive. In this section different types of rhetorical questions are discussed according to

their structure, their intended meaning, and the contextually appropriate response to the question.

They will be categorized based on their type and illocutionary force.

Here is a preliminary definition of rhetorical question (given earlier in (9) and repeated

in (51). The criteria will be exemplified in the following discussion.

(51) A rhetorical question is any question that

a. does not require a linguistic response

b. does not seek information

3.1 Rhetorical counterquestions

One of the most widely employed uses of RQs is as an answer to a previously asked question.

These will be referred to as counterquestions. Interestingly, this is a term that has scarcely been

mentioned in the literature on questions. The Oxford English Dictionary defines a counterquestion

as “A question in reply to another question, a question asked by the person questioned.” While this

is not specific to RQs, it is one of the most common uses of this strategy. This is often done with

polar questions because they have the fewest possible answers (“yes” or “no”). Several examples

were discussed in the previous chapters, a few of which are shown in (52), and variations of which

will be discussed in more depth below.

37
 (52) a. Is the pope catholic?

b. Does a bear shit in the woods?

c. Does the sun set in the west?

d. Do pigs have wings?

The clearest and simplest example of this is a polar question used as a response to a polar

question.

In a context where B’s favorite band is Radiohead and A and B are good friends, the fol-

lowing exchange occurs.

(53) A: Do you want to go see Radiohead next month?

B: Does the sun set in the west?

This exchange is acceptable because the answer to B’s question is known by both discourse partic-

ipants. Additionally, both questions can be answered with the same response particle, in this case

“yes”. These work the same way in cases where the response particle is “no.” The same initial

question can be used to exemplify this, but in a context where B (somehow) dislikes Radiohead.

(54) A: Do you want to go see Radiohead next month?

B: Do pigs fly?

In some cases, an initial question which the addressee deems unnecessary (because the answer

should be known by the speaker), a RQ like those in (55) can be used.

(55) a. Do I look stupid to you?

b. Are you joking?

c. Did you have a brain tumor for breakfast?

38
 These types of RQ counterquestions have a more accusatory flavor than those in (52), which

affects their illocutionary force. This is likely due to the fact that there is a second person subject

in each of these examples (more on this below).

There are cases where polar RQs can be used to answer a different type of question. The

example below is a conversation among a group of high school girls. They are playing croquet in

a backyard and there is an obvious hierarchy in the group, where Heather Chandler is the leader

(Lehmann, 1988, 14:10).

(56) Heather Duke: So what are you gonna do Heather? Take the two shots or send me out?

Heather Chandler : Did you have a brain tumor for breakfast? First you ask if you can be

red, knowing that I’m always red. [hits Heather Duke’s ball out]

Heather Duke : Shit.

Heather Chandler : It’s your turn, Heather.

The first question (Heather Duke’s) is an alternative question which, crucially, cannot be

answered with "yes" or "no," unlike Heather Chandler’s question. The reason that this example

works is through figurative extension of the content of the RQ. In this case, the question of having

a brain tumor is intended to suggest that there are issues with Heather Duke’s brain that caused

her to ask the first question. In other words, Heather Chandler, through the use of the RQ “did

you have a brain tumor for breakfast” is implying that Heather Duke’s question was unnecessary.

The RQ is being used to mock Heather Duke. It also demonstrates the power dynamic between the

two girls. Most rhetorical counterquestions with different syntactic structures do not work. In this

case, it works by completely invalidating the first question (or indicating that it is informationally

defective).

39
 In most cases, though, exchanges in which the initial question is a different syntactic type

will not work.

(57) a. A: How many North American tour dates has Radiohead scheduled this year?

B: #Does a bear shit in the woods?

b. A: Do you want coffee or tea?

B: #/?Is the pope Catholic?1

c. A: At what temperature does water vaporize?

B: How old is the universe?

There are many counterquestions that are in the form of wh-questions. In fact, (57c) can

possibly be viewed as a successful discourse move if the approximate answer to both questions is

something like “a large number.” This type of RQ often takes the form of a polar question but not

always, as shown in (58). In the exchange below (taken from Game of Thrones), Tyrion and Catelyn

are talking about Petyr who is a well-known liar (Martin, 1996, p. 330).

(58) Catelyn: Why would Petyr lie to me?

Tyrion: Why does a bear shit in the woods? Because it is in his nature."

This is an interesting example because it demonstrates that questions that typically take the form of

polar question (“does a bear shit in the woods?” can have a similar effect as a wh-question. Here

the why-question seems to be expressly evoking the well-known polar RQ in order to imply that the

given answer is obvious. In most of these cases, the answer (i.e., "because it’s in his nature") goes

without saying, a hallmark of a rhetorical question.

1
There is a scenario in which this can be a felicitous answer. If the “or” in the initial question is understood as an
inclusive “or” and the RQ “Is the pope Catholic?” is understood to mean “yes” then the result of this exchange is that
B would like both coffee and tea.

40
 While this example from Schaffer (2005, p. 436) seems similar to (57c), it is more similar

to the brain tumor example above in (55c).

(59) A: How reliable is he?

B: How shallow is the ocean? How cold is the Sun?

The example in (59) suggests that he is not at all reliable because of the analogy between his

reliability and the depth of the ocean and the temperature of the sun. The two gradable adjectives

in B’s response are negatively oriented adjectives, which are not neutral with respect to the corre-

sponding positive form. "How hot is X?" doesn’t imply that X is hot; but "How cold is X?" typically

does imply that X is cold. So through the use of "shallow" and "cold" instead of "deep" and "hot",

the RQs force an identification between their own absurd answers (a kind of presupposition failure)

and A’s question that they are responding to (Rett, 2015). In her paper, Schaffer claims that this

example demonstrates that RQs are used to achieve a humorous effect. But it is also intended to

answer the question. Here the literal act is that of posing a question. The illocutionary act expresses

how the speaker feels about the first question, namely that it is a question with an obvious answer.2

The illocutionary effect is that of answering the question.

Sometimes wh-counterquestions are similar to echo questions in that they repeat the previ-

ously asked question but with reverse polarity. This is shown in (60) (Whedon et al. Solomon,

2009, 9:51).

(60) Echo: I can help you.

Topher: Why would you want to?

Echo: Why wouldn’t I?

2
More on illocutionary effect below.

41
 Topher: Did I just lose an argument to a doll?

(60) is a nice example because it shows the effect of the RQ (which is actually another

rhetorical question). In this scene, from the series Dollhouse, Echo has the mentality of a young

child, which is why Topher is surprised at her response. His reaction to the counterquestion is

evidence that the RQ is non-inquisitive. Another similar example is in (61), with the addressee’s

reaction to the RQ (Rosenberg, et al, & Lowney, 2023, 18:50).

(61) Dad: What should we pack? The throat roaster, obviously.

Son: Obviously, yeah.

Father: And the chainsaw.

Son: In what scenario would we need to take the chainsaw?

Father: In what scenario would we not take the chainsaw?

Son: Yeah, good point.

In both of these cases, the RQ is accepted as a successful counterpoint to the initial question

and that seems to be the intended meaning. Because of this they also have the illocutionary force

of an assertion, but of the opposite polarity. That is, each example has the intended meaning “it is

not the case that” and then the content of the proposition. In (60), “it is not the case that I wouldn’t

help you.” This is the argument that Han (2002) makes, namely that RQs should be understood as

assertions of the opposite polarity of the question.

Sometimes a rhetorical counterquestion can be a true echo question. For example, in this

exchange from the series LOST, Walt asks his father, Michael why Jin (another survivor on the

island) attacked him (Lieber et al., & Zinberg, 2004, 19:02).

(62) Walt: What did you do to him?

42
 Michael: What did I do to him? You tell me? I’ve been with you since we crashed. Have

you seen me do anything, to anyone? What kind of man do you think I am, anyway?

Upon seeing his father attacked by Jin, Walt asks “what did you do to him?” which both

expresses confusion about the situation and expects an answer. Michael echoes the question back

at Walt, but rhetorically, not expecting an answer. Michael’s utterance of the question has the

illocutionary force of incredulity. He goes on to ask a series of questions, the force of each becoming

more accusatory as he advances in his response.

In the example from The Sound of Music, several nuns are discussing a novitiate, Maria,

who does not behave in the way that is expected of a nun (Wise, 1965, 8:00).

(63) Nun 1: How do you solve a problem like Maria?

Nun 2: How do you catch a cloud and pin it down?

Nun 3: How do you keep a wave upon the sand?

In this example, it’s likely that all of the questions are unanswerable, or they have the same

partial answer of something like “you can’t” or “no one knows.” There are some questions that are

used to initiate thought on a subject by posing a question to which there is no known answer. This

is a commonly employed rhetorical device.

3.1.1 Who knows?

(64) What is the sound of one hand clapping?

The example in (64) is a classic Zen Buddhist koan, a paradoxical statement or question

meant to challenge the student’s understanding and lead to enlightenment. The question asks the

43
addressee to contemplate the nature of reality and the limitations of language and perception, not

to give a verbal response. This is a question for which there is no known answer.

There are other types of questions that are used in the same way, that is, to provoke thought.

For example, the questions in (65) are impossible to answer3 .

(65) a. What came first, the chicken or the egg?

b. Is there intelligent life in the universe?

c. How many roads must a man walk down before you call him a man?

The illocutionary force of the examples in (64) and (65) seems to be inquisitive, in other

words, not fitting with the working definition of rhetorical question given in the previous chapter.

This differs from the cases where the answer is a negative quantifier. These are often used as rhetor-

ical counterquestions when the initial question is either unanswerable or deemed by the addressee

as being informationally defective or not worth answering. In (66) the answer “who knows” or

“who cares” can be given in each case.

(66) a. How many licks does it take to get to the center of a Tootsie Pop? (not worth answering)

b. How many stars are in the sky? (unanswerable)

The illocutionary force of “who knows?” essentially makes the initial question invalid. The

answer to this question is negatively quantified but not all such RQs have this effect.

(67) Who doesn’t want to go to Spain?

(68) Who likes homework?

(69) What’s the point?

3
That is, impossible to answer at the time of the utterance. I do not assume that the question will always be impos-
sible to answer.

44
 When these are used as rhetorical counterquestions they generally have the force of dis-

agreement or defeat.

3.2 Indirect questions

As described in the previous chapter, indirect speech acts are those which express two dif-

ferent illocutionary forces with the same utterance. Therefore it should not be controversial to say

that rhetorical questions are a type of indirect speech act. What might be, however, is the proposal

that all indirect speech acts that are in the form of a question are rhetorical questions. One type

of indirect question is tag questions.

3.2.1 Rhetorical tag questions

As discussed in the previous chapter, there are several different types of tag questions. Many

do not expect a verbal response because their answer is the same as the anchor. The examples below

break down a few different uses of tags and their intended meanings.

(70) "You wouldn’t want to miss out on this, would you?" - This tag question is used to persuade

the addressee to do something. The speaker is implying that the addressee would regret it

if they didn’t do the thing that is being asked of them.

(71) "That’s not true, is it?" - This tag question is used to challenge a statement that has been

made. The speaker is implying that the statement is false and that the addressee knows it.

(72) "I’m sure you understand, don’t you?" - This tag question is used to make a demand or

request. The speaker is implying that the addressee has no choice but to comply.

(73) "You’re not going to tell anyone, are you?" - This tag question is used to make a threat. The

45
 speaker is implying that the addressee will suffer consequences if they tell anyone about

something.

(74) "You can do it, can’t you?" - This tag question is used to make a request seem less demand-

ing. The speaker is implying that the addressee should do what is being asked of them

because it is not a big deal.

While these have the same syntactic structure and intended meaning, they are considerably

different illocutionary acts. For example, (70) has the illocutionary force of persuasion, (73) has

the illocutionary force of a threat, and (74) is a request.

The dialogue shown in (75) has two different types of RQs. In this scene, a villain, Dr.

Hamsterviel, and his sidekick Gantu were just caught in a lie by their enemies. The first RQ in

(75) is a tag question. This dialogue nicely illustrates the ambiguity of some RQs in conversation.

This question is being asked with sarcasm, as their plan was not successful. It is possible that

Gantu believes he should attempt to answer the question due to the fact that Dr. Hamsterviel is his

boss and not answering the question would be akin to disobeying an order. The second RQ in the

exchange is a negative polar question. With this RQ, Dr. Hamsterviel intends to deride Gantu for

not interpreting the first RQ correctly (Craig & Gannaway, 2006, 40:45).

(75) Dr. Hamsterviel: I think that went very well, don’t you?

Gantu: Actually sir I think—

Dr. Hamsterviel: I didn’t ask what you think!

Gantu: Actually you did I—

Dr. Hamsterviel: It was a rhetorical question! Don’t you know what a rhetorical question

is?

46
 Gantu: Yes sir! I believe it’s a—

Dr. Hamsterviel: No no! That was a rhetorical question too!

Tags do not always have the form of an auxiliary verb and a pronoun. Sometimes they

consist of a single variable, such as right or huh. Regardless of the structure of the tag, they seem

to behave the same way. In (76), Tony Stark (a.k.a. Iron Man) has seen video footage of Spiderman

and has figured out his identity, Peter Parker. Tony goes to Peter’s apartment to recruit him. The

following exchange occurs (Russo & Russo, 2016, 78:37):

(76) Tony: (showing Peter a video of Spiderman) Quick question of the rhetorical variety: that’s

you, right?

Peter: Um, no. What do you – what do you mean?

Tony: Yeah, look at you go.

The illocutionary force in this example is to accuse. In general, tag questions are good ex-

amples of these different types of illocutionary force due to their relatively straightforward intended

meaning, as discussed in 2.1.3.

3.2.2 Requests

Another type of indirect speech act is requests. This is discussed extensively in the literature,

some common examples shown in (77).

(77) a. Can you pass the salt?

b. Are you going to pick that up?

c. Where do you think you’re going?

47
 For example, "can you pass the salt?" usually means "give me the salt" although in some

cases it can be asking whether the addressee is capable of passing the salt (in which case it is an

information-seeking question and not rhetorical). In some cases there are two different rhetorical

meanings for the same question. For example, the RQ "are you kidding me?" can have the force

of excitement (in a context where you just told me that Radiohead is putting on a free show in my

neighborhood) versus having the force of scolding in a context where a child has just come in from

outside and tracked mud throughout the entire first floor of the house. You might also be able to

argue that these have the same force (that of incredulity). This will be discussed in more detail in

the next chapter.

3.3 Summary

This chapter detailed the different types of RQs. They are defined according to their struc-

ture, their role in the conversation, and their intended meaning. There are two main types: rhetorical

counterquestions and rhetorical suggestions. These are the main focus of the following chapter. The

speaker can convey multiple different illocutionary forces in one RQ. This is typically what marks

an indirect speech act.

This chapter also provided a clearer picture of what the definition of a rhetorical question

is. A revised definition is shown in (104).

(78) A rhetorical question is any question that

a. does not require a linguistic response

b. does not seek information

c. has (at least) two interpretations

48
d. does not have the illocutionary force of asking

49
 4. A PRAGMATIC FRAMEWORK FOR RHETORICAL QUESTIONS

Rhetorical questions play a number of different roles in conversation. As discussed in pre-

vious chapters, there are two types that only occur in a discourse context. The first type is used

to answer a previously asked question, called rhetorical counterquestions. The other, rhetorical

suggestions, are used to elicit a nonverbal response from the addressee, either through indirectly

requesting some action (“can you pass the salt?”) or indirectly commanding some action (“where

do you think you’re going?”). In this chapter I propose a mechanism by which RQs are evaluated

within a conversational context.

4.1 General framework for rhetorical questions in conversation

As discussed in Chapter 2, the view of the semantics of questions taken here will be similar

to what many current theories take (e.g., pieces from Hamblin, 1973; Groenendijk & Stokhof, 1984;

Ginzburg, 1992) where a question denotes a set of alternatives. This creates a partition over a set of

worlds. Choosing among these alternatives can offer a partial or complete answer to the question.

A definition of answerhood is shown below (adapted from Roberts, 2012, p. 11).

(79) a. A partial answer to a question q is a proposition which contextually entails the evalua-

tion—either true or false—of at least one alternative

b. A complete answer to a question q is a proposition which contextually entails an eval-

uation for each alternative in q-alt(q)

There are several aspects of the conversational scoreboard that must be intact in order for

RQs to be appropriately interpreted in a discourse context. This section will provide definitions of

50
the relevant components of the framework. Additionally, two pragmatic principles will be proposed

that allow for the interpretation of rhetorical questions, not as information-seeking questions, but

as rhetorical questions as defined in the previous chapter.

4.1.1 Components of the scoreboard

I propose that when a discourse participant utters a question, all relevant alternatives are

added to the Projected Common Ground (defined in (81)). If an appropriate proposition as defined

in (79) is proffered by the addressee (because they believe it to be true), then it exists in the Projected

Commitment Set of the addressee, defined in (82). If the proposition completely or partially answers

the question, it will be added to the Table where, if accepted, it will become part of the Common

Ground. This is the case for regular information-seeking questions and will not be discussed in any

more detail. If, however, one of the alternatives introduced by the question is already present in the

current Common Ground, then the proposition is eliminated from the Projected Common Ground

and added to both the speaker’s and the addressee’s Projected Commitment Set.

(80) Common Ground:1 It is common ground that ϕ in a group if all members accept (for the

purpose of the conversation) that ϕ, and all believe that all accept that ϕ, and all believe that

all believe that all accept that ϕ, etc.

The Common Ground is a set of commonly-held beliefs among the discourse participants.

This includes general facts that are known (or assumed to be known) by everyone, as well as oc-

casionally the feelings and opinions of other members in the group, dependent on their degree of

intimacy. Propositions introduced during the conversation that have not yet been accepted are in
1
This definition of Common Ground is from Stalnaker, 2002, p. 716.

51
the Projected Common Ground.

(81) The Projected Common Ground is a set of potential Common Grounds where the propo-

sitions are held until they can be resolved through being added to the Table, the Common

Ground, or the To-Do List.

When propositions are added to the Projected Common Ground, they can also be added

to the Projected Commitments. This is not typically the case for questions, as they do not add

information to the discourse context, but rather introduce a set of alternative propositions. However,

in the case of RQs, the speaker’s intention in uttering the question (i.e., the preferred alternative) is

added to the project commitment set, defined in (82).

(82) The Projected Commitment Set is a collection of sets of possible commitments, including

what was introduced by the most recent utterance, conversational implicatures, speaker

intention, and other interlocutors’ possible commitments.

Malamud and Stephenson (2015) define their projected commitment set as “the expected

next stage of the conversation” (p. 288). In addition to speaker intentions and proposed commit-

ments, this is where the speaker’s beliefs about the likelihood of one alternative over another (i.e.,

({α, α}) are located.

Rhetorical suggestions cause the preferred alternative to be added to the addressee’s To-Do

List. I adopt Condoravdi and Lauer’s (2012) simplification of Portner’s (2004) To-Do List as a set of

propositions rather than a property or function. This remains compatible with preference-ordering

which commits an agent to act in a way that makes the proposition true.

(83) To-Do List

For any agent i, the participants in the conversation mutually agree to deem i’s actions

52
 rational and cooperative to the extent that those actions tend to make it more likely that the

largest subset of propositions on the TDL(i) becomes true; and

An agent i is committed to act in such a way as to make true as many propositions on the

TDL(i) as possible.2

4.1.2 Pragmatic principles

There are two instances considered here wherein the speaker has either knowledge of or

a preference for one alternative over another when uttering a question. The Pragmatic Principles

shown in (84) and (85) provide a mechanism for contextual interpretation of rhetorical questions.

(84) evidence: When a discourse participant A utters the polar question q?, the alternatives q

and ¬q enter the Projected Common Ground. If a preference for an alternative, α or α,

is present in the speaker’s Projected Commitment Set, then the proposition matching the

preferred alternative (q or ¬q) will be added to the addressee’s Projected Commitment Set

3
and their To-Do List.

This is the case for indirect questions and other types of rhetorical questions that have the

illocutionary force of requesting or ordering. This will be discussed in more detail in Section 4.2.

If an utterance causes a proposition to be added to the addressee’s Projected Commitment

Set, then the addressee becomes responsible for resolving it (i.e., adding it to their discourse com-

mitments). This is usually done by the addressee’s assumption that the speaker of the RQ was

adhering to the Cooperative Principle and, in this case, the maxim of Quality. Because the speaker

assumes that their interlocutors are being informative, the answer to the RQ is compared to the

potential answers of the first question.

2
This includes taking action in order to make the propositions true.
3
This is also the case for alternative and wh-questions.

53
 (85) redundancy: When a discourse participant A utters the polar question q?, the alternatives

q or ¬ q are added to the Projected Common Ground. If one of the alternatives is present

in the Common Ground when q? is uttered, then there is a redundant proposition. This

results in:

i the evaluation of the polarity features of the redundant proposition in the addressee’s

Projected Commitment Set

ii the corresponding response particle (yes or no) is added to the speaker’s and the

addressee’s Discourse Commitments

There are two types of polarity features as defined by Roelofsen and Farkas (2015, p. 384).

These are shown in (86).

(86) Absolute polarity features: [+], [–]

Relative polarity features: [agree], [reverse]4

The absolute polarity features [+] and [-] presuppose that their prejacent expresses a propo-

sition containing a single possibility, which is highlighted and has positive or negative polarity

(Roelofsen & Farkas, 2015). This is what is added to the Projected Commitment Set while the

prejacent is elided, resulting in the response particle’s acceptance as a valid discourse move.

Redundancy is manifested in a variety of ways across the world’s languages. The redun-

dancy I am proposing differs from other types of redundancy in that the redundant feature is in

the Common Ground rather than as part of the utterance, but it has a similar function. Wit and

colleagues (1999) discuss two types of redundancy. Grammatical redundancy encompasses those
4
These will be set aside for the remainder of the discussion here, although their inclusion in future work
on other types of questions that are not strictly information-seeking would be worthwhile.

54
syntactic or morphological features that are required by a language’s grammar. An example of this

is double negation which is obligatory in Spanish, for example (shown below in (87).

(87) No sabes nada.

Not you know nothing.

You don’t know anything.

The type of redundancy that is more relevant to this analysis is contextual redundancy, which is

the repetition of information. One of the main functions of redundancy is that it is often used for

emphasis. In Spanish, for example, personal pronouns are repeated to add emphasis, shown below

in(88).

(88) A mi me gusta el café.

To me to me enjoy the coffee.

I LIKE coffee.

It results in adding emphasis to the utterance as well as providing an answer. In this way re-

dundancy is quite an efficient way to update the conversational scoreboard – it both answers the

question and provides information about the speaker’s attitude toward the first question.

In the case of rhetorical suggestions, the proposition is added to the To-Do List where, in

the case of polar RQs, the prejacent clause is made true through the addressee’s actions, and in all

5
other cases the preferred alternative is made true through the addressee’s actions.

5
This will work in cases like “where do you think you’re going?” where the preferred alternative is present
in the speaker’s PCS.

55
4.2 Rhetorical suggestions

An important feature of this type of RQ is that, like imperatives, they have a second-person

subject (although imperatives almost always have a null second-person subject (Zanuttini et al.

2012; Portner, 2007; Condoravdi & Lauer, 2012)). This is why utterances like (89) are ungram-

matical.

(89) #Mary kiss you!

This is interesting for the discussion of RQs that require the addressee to act. Below are

some examples with an overt second-person subject.

(90) Can you pass the salt?

(91) Are you going to pick that up?

(92) You’re not going to let him get away with that, are you?

(93) Where do you think you’re going?

Maybe the most commonly referenced example of an indirect question (considered to be a

rhetorical question here) is that in (90). This is a nice example because it clearly has two interpre-

tations. One possible interpretation of the question is the regular information-seeking one, where

one of the alternatives is true of the addressee’s actual ability to pass the salt. The other more com-

mon interpretation is as a request for the addressee, not to give a verbal answer, but to take some

action, namely that of handing over the salt. The mechanism for this interpretation is evidence as

discussed above.

This depends on contextual evidence in order to achieve the desired result. When the ques-

tion p?=can you pass the salt is uttered, the alternatives p and ¬p enter the Projected Common

56
Ground. If the speaker of the question p? has in their Projected Commitment Set, a preference for

the alternative p (represented in the table by α) then the proposition p is added to the addressee’s

(i.e., B’s) To-Do List. This is shown in Table 4.1.

(i) (i)
Previously After A’s move After B’s acceptance
DCA {} {} {{ p, α}}
PCSA {{}} {{p,α}} {{ }}
DCB {} {} { p}
PCSB {{ }} {{ }} {}
Table ⟨⟩ ⟨{p?}⟩ ⟨⟩
CG {} {} {p}
PCG {{ }} {{ p,¬p}} {{ }}
TDLA {{ }} {{ }} {{ }}
TDLB {{ }} {{p}} {{ }}

Table 4.1: Can you pass the salt?

After B’s acceptance of the preferred alternative and their resolution of the proposition on

the TDL (by passing the salt), p? is removed from the Table, p is removed from the TDLB , and the

PCG, and added to the Common Ground.

Another example of this type of rhetorical question is in (91). It can be posed in a number

of different scenarios with the same result. Likewise, there are a number of scenarios where the

questions shown in (94) can have a different syntactic structure and be used to mean different things

but still have the same result.6 Below are a handful of indirect questions relating to (91).7 This type

of rhetorical question has the illocutionary force of a command (whereas the previous example had
6
Although the illocutionary force can be different – but the pragmatic principles evidence and redun-
dancy do not rely on the type of illocutionary force.
7
Granted, some of these because of their syntactic structure much more closely resemble commands than
questions, but as discussed earlier, tag questions are considered a type of indirect question.

57
the illocutionary force of a request). Otherwise this is treated in almost exactly the same way.

(94) a. Are you going to pick that up?

b. You’re going to pick that up, aren’t you?

c. Pick that up, will you?!

d. Can you pick that up?

(95) Below are two contexts for which any of the questions in (94) is appropriate.

(i) Context: A mother and child are making cookies. The child drops the bag of choco-

late chips on the floor and the mother says: You’re going to pick that up, aren’t you?

(ii) Context: A father and son are talking about their fantasy football teams. The father

has been thinking about picking up the Green Bay defense for a while and the son

thinks that he should just do it. The son says: Are you going to pick that up?

In these cases the scoreboard would look identical to Table 4.2 and most other polar ques-

tions of the type discussed here would as well. This is the case regardless of the syntactic structure

of the question (i.e., the tag question “You’re going to pick that up, aren’t you?” behaves the same

way as "Can you pick that up?").

The situation is a bit different for wh-questions. Take an example like (93), repeated in (96).

(96) Where do you think you’re going?

This, like most examples in this discussion, could be interpreted as an information-seeking

question. In scenarios where it isn’t, it usually is meant to be understood as something like the

examples in (97). Crucially, these all have the illocutionary force of an order.

58
 (97) x = You are going there8

y = You are not going there

z = You are not going anywhere

There are a few things that make this question more difficult to interpret. One is the fact

that there are bound to be more contextually relevant alternative propositions to the wh-question

than there are in polar questions. This will be a bit tricky to represent fully, so for the discussion

here let’s assume that the context is such that when A utters the question (96), the only alternatives

introduced by the question are those in (97).

The table below could represent a discourse situation in which A is B’s teacher. B gets up

to leave the classroom and A utters w? = where do you think you’re going? with the illocutionary

force of ordering B to stay put.

(i) (i)
Previously After A’s move After B’s acceptance
DCA {} {} {{ z, α}}
PCSA {{}} {{z,α}} {{ }}
DCB {} {} { z}
PCSB {{ }} {{ }} {}
Table ⟨⟩ ⟨{w?}⟩ ⟨⟩
CG {} {} {z}
PCG {{ }} {{ x,y, z}} {{ }}
TDLA {{ }} {{ }} {{ }}
TDLB {{ }} {{z}} {{ }}

Table 4.2: Where do you think you’re going?

Evidence is able to accommodate differences between syntactic structures and illocution-

8
Here there is contextually determined.

59
ary forces and therefore can be applied to a wide variety of rhetorical suggestions (i.e., indirect

questions). A couple examples have been shown in this section to illustrate its versatility in the

evaluation of this type of rhetorical question.

4.3 Rhetorical counterquestions

One of the most widely employed uses of a rhetorical question is as an answer to a previously

asked question. Several examples were discussed in the previous section, similar examples are

shown below.

(98) A: Do you want to go to Radiohead’s next show?

B: Does the sun set in the west?

(99) A: Are the Denver Nuggets going to win the NBA Championship?

B: Does a bear shit in the woods?

(100) A: Is New York City the most populous city in the U.S.?

B: Is the pope Catholic?

(101) A: Do you like celery?9

B: Do pigs have wings?

One of the more interesting things about examples (98–101) is that some of them have

contextually-dependent answers (like (98), where it depends on the musical tastes of the addressee

and the degree to which the interlocutors are familiar with each other’s preferences. While others do

not, like (100), where the answer to the question is a fact about the world. However, the framework
9
This example functions in the same way as the other questions but here the answer is "no" to the first
question rather than "yes" as in the other cases.

60
proposed here should be able to accommodate all of these different types.

Imagine a scenario where two good friends are talking about the summer’s concert schedule.

A asks B the question p? = Do you want to go to the Radiohead show? This adds p and ¬p to the

PCG. Because of the intimacy of the discourse participants, A should know B’s music preferences.10

Therefore B responds with a polar rhetorical question of the type shown above, q? = Does the

sun set in the west? This adds q and ¬q to the PCG. The alternative q is already present in the

Common Ground (being a fact about the world), thereby triggering redundancy. This results in

the polarity of the RQ (i.e., B’s utterance, or q) being evaluated in the Projected Commitment Set.

The polarity particle ("yes" in this case) is then added to the Discourse Commitments of the speaker

and addressee. The effects of this on the conversational scoreboard are shown in Table 4.3.

(102) p? = Do you want to go to the Radiohead show?

(103) q? = Does the sun set in the west?

(i) (ii) iii

Previously After A’s move After B’s move After A’s Acceptance
DCA {} {} { p} {p}
PCSA {{ }} {{p}} {{ p[+]}} {{ }}
DCB {} {} {} {p}
PCSB {{ }} {{p}} {{ p [+]}} {}
Table ⟨⟩ ⟨{p?}⟩ ⟨ p?, q? ⟩ ⟨⟩
CG {q} {q} {q} {{ p, q }}
PCG {} {{p,¬p}} {{p, ¬p, q!}} {}
TDLA {{ }} {{ }} {{ }} {{ }}
TDLB {{ }} {{ }} {{ }} {{ }}

Table 4.3: Do you want to go to the Radiohead show?

10
This suggests that this question is informationally defective, or that it was an unnecessary question.

61
 This example is illustrative of how polar rhetorical counterquestions work in general. Rhetor-

ical counterquestions in the form of wh-questions are different, however. Because Redundancy

relies on the polarity particle associated with the RQ, this analysis cannot, by its nature, generalize

to other types of rhetorical counterquestions. It is possible that this framework could be modified

to accommodate different syntactic types of counterquestions in future work.

4.4 Summary

In this chapter a pragmatic framework was proposed to explain the effects that rhetorical

questions have on the conversational scoreboard. Their illocutionary force was also considered, as

it, along with contextual factors, can affect how rhetorical questions with the same structure are

interpreted. It was determined that not all types of rhetorical question can be accommodated with

this framework, but it does work to explain the most common types of rhetorical counterquestions

(polar RQs) as well as different types of rhetorical suggestions.

62
 5. ODDS AND ENDS

5.1 Discussion

I have provided a description of the different roles that questions play in a conversational

context. Some of these are considered to be rhetorical questions based on the definition provided

at the end of Chapter 3 and exemplified in Chapter 4.

I proposed a revised definition for rhetorical questions that includes some types of questions

that were typically not considered by previous definitions of rhetorical questions, namely, indirect

questions. It also excludes questions that are unanswerable, which are by their nature included in

other theories of rhetorical questions.

(104) A rhetorical question is any question that

a. does not require a linguistic response

b. does not seek information

c. has (at least) two interpretations

d. does not have the illocutionary force of asking

5.2 Future Work

There is an additional type of alternative question that was omitted from the discussion above

but that could have interesting implications for the proposal. The distinction between canonical

polar questions and alternative or not questions has been discussed in previous literature (van Rooy

& Safarova 2003; Biezma & Rawlins, 2012). The pragmatic difference between the examples in

63
(105) suggest a different semantic structure.

(105) a. Do you want to play chess? ↑

b. Do you want to play chess, or not? ↓

Van Rooy and Safarova (2003) discuss or not questions as questions that don’t invite an

answer. For example, in the case of an invitation, adding or not would be considered impolite

(295).

(106) a. Do you want something to drink?

b. Do you want something to drink, or not?

They also define its use in certain cases as rhetorical (295).

(107) a. Are you crazy?

b. Are you crazy, or not?

Van Rooy and Safarova define this type of alternative question as having a weak presuppo-

sition for both p and ¬p, compared to other polar questions that evoke different presuppositions.

This is similar to Farkas and Roelofsen’s (2017) concept of evidence and credence level.

A more intriguing alternative question to consider that hasn’t gotten as much attention in

previous research is or what. Some examples of its use are shown below.

(108) a. Do you want to play chess, or what?

b. Didn’t you hear the phone ring, or what?

c. Can you pass the salt, or what?

I propose, similar to Biezma and Rawlins (2015), that these are also examples of alternative

questions rather than polar questions. While they are very similar to the alternative or not questions

64
in form, they seem to have both a different semantic interpretation as well as a different pragmatic

function. One piece of evidence for this is that or what works in cases where the host clause is

negative, whereas it seems that or not doesn’t. Compare (109) with (108b).

(109) ??Didn’t you hear the phone ring, or not?

This is problematic for existing theories of alternative questions because it doesn’t seem to

have any acceptable responses. Therefore, it doesn’t behave like an alternative or not question. Or

what also doesn’t seem to be like a canonical alternative question in that giving one of the disjuncts

as an answer is only questionably felicitous, as shown in (110). However, it does seem to work like

a rhetorical question by the definition proposed in (??).

(110) Do you want to play chess or what?

a. ?(I want to play) chess.

b. ?what

c. #both

d. #neither

e. ?I want to play cribbage.

f. Yes

g. No

Biezma and Rawlins (2016) discuss or what as an answer to an imperative (in a two-

participant exchange), which could suggest that the host clause in an or what question is actually

more like an imperative than like a regular interrogative (or declarative) clause. Or what questions

are not compatible with those imperative tag questions shown in Table 1, but the possible effects on

65
the discourse context by an imperative are the same as what seems to be introduced by an or what

question. These contributions are shown in (111) (Biezma & Rawlins, 2017, p. 8).

(111) What are the alternatives for the future? / What will happen?

Another example of the rhetorical nature of or what questions is in their use as exclamatives

as shown in (112). While outside the scope of this paper, it would be beneficial in the study of

rhetorical questions to investigate what role exclamatives such as these play.

(112) a. Was that a game, or what?

b. Does she know how to win, or what?

c. Can she serve, or what?

Or what seems to also demand an answer in a way that variant tag questions do not. In other

words, they appear to have a stronger force. This is problematic for the interpretation of Portner’s

(2004) claim that force and type have a one-one correspondence. For example:

(113) a. Are you going to move, or what?

You’re going to move, aren’t you?

Meaning: Move!

b. Are you going to pick that up, or what?

You’re going to pick that up, aren’t you?

Meaning: Pick it up!

Another comparison to be made between or what questions and other question types in-

volves intonation. Unlike the examples described in Table 1 (most of which have rising intonation),

66
or what tags almost always have falling tone. Does this mean they pattern with the falling tag ques-

tions Farkas and Roelofsen (2017) proposed? Because this indicates a high credence value, this

seems to work with their analysis.

Intonation was not included in the framework proposed in Section 4, although it was men-

tioned throughout the paper. In considering various other clause types in the future (e.g., rising

declaratives), it would be necessary to include a mechanism that can manage the prosodic structure

in addition to the syntax and semantics.

The sociopragmatics of rhetorical questions was not addressed in this project - that is, the

possible social and communicative intentions of the speaker when using RQs. Is social meaning

encoded in the grammar and/or in the use of intonation patterns in the utterance of RQs?

Early sociolinguistic research focused on the relationship between language and demograph-

ics such as gender, socioeconomic status, and geographic region. For example, pioneering studies

by William Labov (1966) and Peter Trudgill (1974) described differences in the pronunciation of

the phoneme /ô/ and the associated social significance of producing one variant over the other. The

studies found production of /ô/ increased or decreased according to the socioeconomic status of the

speaker/addressee, with the prestigious variant being used the majority of the time by the upper

class (in Labov’s study, /ô/ was produced more at the higher-end department store than at the others

in the study) and gradually moving toward the less prestigious variety in the lower classes in the

studies.

Many linguists have continued to observe this phenomenon, notably Campbell-Kibler (2007)

and Podesva (2011). Campbell-Kibler investigated perceptions of variants of (ING). She found that

speakers who produced the alveolar variant /In/ were considered to be less educated, less articulate,

and less likely to be wealthy (Campbell-Kibler 2007). Use of the /Iŋ/ variant was more likely to

67
be viewed as "gay" than /In/. Podesva has extensively studied the role of speech (predominantly

phonemes) in the presentation of a gay identity. He has looked at the production and perception of

vowels (e.g., the California Vowel Shift), /s/, and /t/, among other variables. He found that variants

of these phonemes have been become a social index that signals a gay identity.

Recent research has expanded from looking at a single phoneme to a morpheme (Beltrama

et al., 2021), or a lexical item (Acton, 2021). In their investigation of intensifiers, Beltrama and

Casasanto (2017) suggest that social indices like totally or -issimo have a similar mapping for their

semantic and sociolinguistic features. Acton (2014) suggests that the use of a determiner in an

expression like The Americans like coffee is marked due to its redundancy. This results in the

speaker signalling a social relationship with a group, oftentimes distancing themselves from it.

Research on the encoding of social meaning in the grammar of constructions is harder to

find. Moore and Podesva (2009) analyze style and indexicality by looking at the use of tag questions.

They propose an alternative framework for examining social meaning that differs from previous

approaches. Style, they claim, "is not only about how to render an utterance; it is about whether

to utter it, and when" (Moore & Podesva, 2009, p. 480) and they treat content as a component of

style. Applying this definition of style (taken from Eckert, 2008) to the examination of the social

meaning of RQs should yield some interesting results.

Another area of interest in sociolinguistic research has been the role of intonation patterns

on the presentation of identity and attitude/emotion (Podesva, 2011; Podesva et al., 2015). The re-

lationship between rhetorical questions and intonation patterns is an interesting one. Much research

has been done on questions and intonation (e.g., Gunlogson 2002, 2008; Farkas & Roelofsen, 2017

- discussed above), and rhetorical questions have come up in some of those discussions (Neitsch,

2019; Caponigro & Sprouse, 2007) but only insofar as to determine whether the question was in-

68
tended to be understood as rhetorical. This introduces a host of interesting questions:

i What role does intonation play in determining/intending a question as rhetorical?

ii How does the use of intonation in rhetorical questions interact with linguistic social cues?

iii Is there a sociolinguistic style that is more or less likely to use rhetorical questions and how

does intonation figure into that likelihood?

iv How is speaker attitude and/or emotion conveyed through the use of RQs?

A potentially useful place to find data on RQs are corpora of spoken language, such as the

Corpus of Contemporary American English. Searching the utterances preceding and following

questions like those in (114) could provide insight into why and when RQs are used.

(114) a. Who knows?

b. Who cares?

c. Is it, though?

In addition to corpora, social media is another resource for conversational linguistic data.

Ranganath and colleagues (2018) identify RQs on Twitter by determining author intention based

on hashtags following the tweets in which the RQ occurs. While they focus more on their algorithm

than where the information fits in the larger linguistic literature, it introduces some interesting data

and methods for collection. While not a user myself, social media seems like a good place to

investigate possible patterns of RQs usage in the different communities.

There is still much more to be explored in the area of rhetorical questions as well as other

linguistic phenomena whose interpretation relies heavily on pragmatic and contextual factors.

69
 REFERENCES

Acton, E. K. (2014). Pragmatics and the social meaning of determiners. Dissertation. Stanford

University.

Acton, E. K. (2021). Pragmatics and the third wave: The social meaning of definites. Social

meaning and linguistic variation: Theorizing the third wave, 105.

Austin, J. L. (2007). How to Do Things with Words: Edited by James O. Urmson and Marina Sbisa.

2nd ed. Harvard University Press.

Bach, K., & Harnish, R. M. (1979). Communication and speech acts. MIT Press.

Beltrama, A., Staum Casasanto, L. (2017). Totally tall sounds totally younger: Intensification at

the socio-semantics interface. Journal of sociolinguistics, 21(2), 154-182.

Beltrama, A., Casasanto, L. S., Hall-Lew, L., Moore, E., Podesva, R. (2021). The social meaning

of semantic properties. Social meaning and linguistic variation: Theorizing the third wave,

80-104.

Biezma, M., & Rawlins, K. (2012). Responding to alternative and polar questions. Linguistics and

philosophy, 35(5), 361-406.

Biezma, M., & Rawlins, K. (2015). Alternative questions. Language and linguistics compass,

9(11), 450-468.

Biezma, M., & Rawlins, K. (2016). Or what?: Challenging the speaker. Proceedings of NELS 46th,

1, 93-106.

Biezma, M., & Rawlins, K. (2017, December). Rhetorical questions: Severing asking from ques-

tioning. Semantics and linguistic theory, 27, 302-322).

Campbell-Kibler, K. (2007). Accent,(ING), and the social logic of listener perceptions. American

70
 speech, 82(1), 32-64.

Caponigro, I., & Sprouse, J. (2007). Rhetorical questions as questions. In Proceedings of Sinn und

Bedeutung 11, 121-133.

Condoravdi, C., & Lauer, S. (2012). Imperatives: Meaning and illocutionary force. Empirical

issues in syntax and semantics, 9, 37-58.

Craig, T., & Gannaway, B. (2006). Leroy & Stitch [Motion picture]. United States: Walt Disney

Studios Motion Pictures.

Eckert, P. (2008). Variation and the indexical field 1. Journal of sociolinguistics, 12(4), 453-476.

Farkas, D. F., & Bruce, K. B. (2010). On reacting to assertions and polar questions. Journal of

semantics, 27(1), 81-118.

Farkas, D. F., & Roelofsen, F. (2017). Division of labor in the interpretation of declaratives and

interrogatives. Journal of semantics, 34(2), 237-289.

Frank, Jane. 1990. "You call that a rhetorical question?: Forms and functions of rhetorical questions

in conversation." Journal of pragmatics 14, no. 5: 723-738.

Ginzburg, J. (1992). Questions, queries and facts: A semantics and pragmatics for interrogatives.

stanford university.

Grice, H. P. (1975). Logic and Conversation. In P. Cole, & J. L. Morgan. (Eds.), Syntax and

semantics, Vol. 3, Speech Acts (pp. 41-58). Academic Press.

Groenendijk, J., & Stokhof, M. (1984). Semantic analysis of wh-complements. Linguistics and

philosophy, 5(2), pp. 175-233.

Gunlogson, C. (2002, September). Declarative questions. Semantics and linguistic theory (Vol.

12, pp. 124-143).

Gunlogson, C. (2008). A question of commitment. Belgian journal of linguistics, 22(1), 101-136.

71
Hamblin, C. L. (1973). Questions in Montague English. Foundations of Language, Vol. 10, No. 1

(May, 1973), pp. 41-53.

Han, Chung-hye. 2002. "Interpreting interrogatives as rhetorical questions." Lingua 112, no. 3:

201-229.

Houston, K. (2013). Shady Characters: The secret life of punctuation, symbols, & other typograph-

ical marks. New York, USA: W.W. Norton & Company, Inc.

Huddleston, R. (1994). The contrast between interrogatives and questions. Journal of linguistics,

30(2), 411-439.

Kimps, D. (2007). Declarative constant polarity tag questions: A data-driven analysis of their form,

meaning and attitudinal uses. Journal of pragmatics, 39(2), 270-291.

Kimps, D., Davidse, K., & Cornillie, B. (2014). A speech function analysis of tag questions in

British English spontaneous dialogue. Journal of pragmatics, 66, 64-85.

Kurzon, D. (1998). The speech act status of incitement: Perlocutionary acts revisited. Journal of

pragmatics, 29(5), 571-596.

Labov, W. (1966). Hypercorrection by the lower middle class as a factor in linguistic change.

Sociolinguistics, 84-113.

Lehmann, M. (Director). (1988). Heathers [Motion Picture]. United States: New World Pictures.

Lewis, D. (1969). Convention: A philosophical study. Harvard Universtiy Press.

Lewis, D. (1979). Scorekeeping in a language game. Journal of philosophical logic 8, 339–359.

Lieber, J., Abrams, J. J., & Lindelof, D. (Writers), Zinberg, M. (Director). (27 October 2004).

House of the rising sun [Television series episode]. In Lost. Los Angeles, CA: American

Broadcasting Company.

Malamud, S. A., & Stephenson, T. (2015). Three ways to avoid commitments: Declarative force

72
 modifiers in the conversational scoreboard. Journal of semantics, 32(2), 275-311.

Martin, G. R. R., (1996). Game of thrones. Bantam Books.

Moore, E., & Podesva, R. (2009). Style, indexicality, and the social meaning of tag questions.

Language in society, 38(4), 447-485.

Neitsch, J. (2019). Who cares about context and attitude?: Prosodic variation in the production and

perception of rhetorical questions in German (Doctoral dissertation).

Obama, B. https://obamawhitehouse.archives.gov/the-press-office/2014/11/20/remarks-President-

address-nation-immigration

Oxford English Dictionary. (n.d.). Rhetoric. In OED.com dictionary. Retrieved June 1, 2023, from

https://www.oed.com/view/Entry/165178.

Oxford English Dictionary. (n.d.). Rhetorical. In OED.com dictionary. Retrieved June 1, 2023,

from https://www.oed.com/view/Entry/165181.

Podesva, R. J. (2011). Salience and the social meaning of declarative contours: Three case studies

of gay professionals. Journal of English linguistics, 39(3), 233-264.

Podesva, R. J., Reynolds, J., Callier, P., & Baptiste, J. (2015). Constraints on the social meaning

of released/t: A production and perception study of US politicians. Language variation and

change, 27(1), 59-87.

Portner, Paul. 1997. "The semantics of mood, complementation, and conversational force." Natural

language semantics 5, no. 2: 167-212.

Portner, P. (2004, September). The semantics of imperatives within a theory of clause types. In

Semantics and linguistic theory (Vol. 14, pp. 235-252).

Portner, P. (2007). Imperatives and modals. In Natural language semantics, 15, 351-383.

Ranganath, S., Hu, X., Tang, J., Wang, S., & Liu, H. (2018). Understanding and identifying rhetor-

73
 ical questions in social media. ACM transactions on intelligent systems and technology (TIST),

9(2), 1-22.

Rett, J. (2015). The semantics of evaluativity.Oxford University Press.

Roberts, C. (2012). Information structure: Towards an integrated formal theory of pragmatics. In

Semantics and pragmatics, 5, 6-1.

Roberts, C. (2018). Speech acts in discourse context. In Fogal, D., Harris, D. W., & Moss, M.

(Eds.) New work on speech acts (pp. 317-359). Oxford University Press.

Roelofsen, F., & Farkas, D. F. (2015). Polarity particle responses as a window onto the interpreta-

tion of questions and assertions. Language, 359-414.

Rohde, Hannah. 2006. "Rhetorical questions as redundant interrogatives." San Diego linguistic

papers, 2: 134-168.

Rosenberg, A., Stikeman, C., & Walsh, M. O. (Writers), Lowney D. (Director). (10 May 2023).

Deerfest: Part one [Television series episode]. In The big door prize. Los Angeles, CA: Apple

TV Plus.

Russo, J., & Russo, A. (2016). Captain America: Civil War [Motion picture]. United States: Walt

Disney Studios Motion Pictures.

Schaffer, Deborah. 2005. "Can rhetorical questions function as retorts?: Is the Pope Catholic?."

Journal of pragmatics 37, no. 4: 433-460.

Searle, J. R., & Vanderveken, D. (1985). Foundations of illocutionary logic. CUP Archive.

Tomaselli, M. V., & Gatt, A. (2015). Italian tag questions and their conversational functions. Jour-

nal of pragmatics, 84, 54-82.

Tottie, G., & Hoffmann, S. (2006). Tag questions in British and American English. Journal of

English linguistics, 34(4), 283-311.

74
Trudgill, P. (1974). Linguistic change and diffusion: Description and explanation in sociolinguistic

dialect geography. Language in society, 3(2), 215-246.

Van Rooy, R., & Safarova, M. (2003, August). On polar questions. Semantics and linguistic theory

(Vol. 13, pp. 292-309).

Whedon, J., Chambliss, A., & Bellomo, T. (Writers), Solomon, D. (Director). (10 April 2009).

A spy in the house of love [Television series episode]. In Dollhouse. Los Angeles, CA: Fox

Network.

Wise, R. (1965). The Sound of Music [Motion picture]. United States: Twentieth Century Fox.

Zanuttini, R., Pak, M., & Portner, P. (2012). A syntactic analysis of interpretive restrictions on

imperative, promissive, and exhortative subjects. Natural language & linguistic theory, 30,

1231-1274.

75
 ProQuest Number: 30531310

INFORMATION TO ALL USERS

The quality and completeness of this reproduction is dependent on the quality
and completeness of the copy made available to ProQuest.

Distributed by ProQuest LLC ( 2023 ).

Copyright of the Dissertation is held by the Author unless otherwise noted.

This work may be used in accordance with the terms of the Creative Commons license
or other rights statement, as indicated in the copyright statement or in the metadata
associated with this work. Unless otherwise specified in the copyright statement
or the metadata, all rights are reserved by the copyright holder.

This work is protected against unauthorized copying under Title 17,

United States Code and other applicable copyright laws.

Microform Edition where available © ProQuest LLC. No reproduction or digitization

of the Microform Edition is authorized without permission of ProQuest LLC.

ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346 USA