The Semantics and Pragmatics of Honorification Register and Social Meaning Elin Mccready Full Chapter

The Semantics and Pragmatics of
Honorification: Register and Social

Meaning Elin Mccready
Visit to download the full and correct content document:
https://ebookmass.com/product/the-semantics-and-pragmatics-of-honorification-regist
er-and-social-meaning-elin-mccready/
The Semantics and Pragmatics
of Honorification
OX F OR D STU DIES IN SEMANTICS AND PR AGMATICS
General Editors
Chris Barker, New York University, and Chris Kennedy, University of Chicago
recently published in the series

3
Weak Island Semantics
Márta Abrusán
4
Reliability in Pragmatics
E. McCready
5
Numerically Qualified Expressions
Chris Cummins
6
Use-Conditional Meaning
Studies in Multidimensional Semantics
Daniel Gutzmann
7
Gradability in Natural Language
Logical and Grammatical Foundations
Heather Burnett
8
Subjectivity and Perspective in Truth-Theoretic Semantics
Peter Lasersohn
9
The Semantics of Evidentials
Sarah E. Murray
10
Graded Modality
Qualitative and Quantitative Perspectives
Daniel Lassiter
11
The Semantics and Pragmatics of Honorification
Register and Social Meaning
Elin McCready
12
The Meaning of ‘More’
Alexis Wellwood
in preparation
Comparing Comparison Constructions
M. Ryan Bochnak
Meaning over Time
The Foundations of Systematic Semantic Change
Ashwini Deo
Plural Reference
Friederike Moltmann
A History of Formal Semantics
Barbara Partee
The Semantics and
Pragmatics of
Honorification
Register and Social Meaning
E L I N MCC R E A DY
1
3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Elin McCready 2019
The moral rights of the author have been asserted
First Edition published in 2019
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2019937076
ISBN 978–0–19–882136–6 (hbk.)
ISBN 978–0–19–882137–3 (pbk.)
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
General Preface
Oxford Studies in Semantics and Pragmatics provides a platform for original research on
meaning in natural language within contemporary semantics and pragmatics. Authors
are encouraged to present their work in the context of past and present lines of inquiry
and in a manner accessible to semanticists and pragmatists in linguistics, philosophy
and cognitive science, as well as to professional linguists in related subfields such as
syntax and lexicology. They are also asked to ground argument in numerous examples
from English and, where possible, from a variety of other languages.
This is a companion series to Oxford Surveys in Semantics and Pragmatics, which
provides critical overviews of the major approaches to research topics of current
interest, a discussion of their relative value, and an assessment of what degree of
consensus exists about any one of them. The Studies series equally seeks to put empir-
ical puzzle and theoretical debate into comprehensible perspective, but its authors
generally develop and defend the approach and line of argument which they find most
convincing and productive. The series offers researchers in linguistics and related
areas—including syntax, cognitive science, computer science, and philosophy—a
means of disseminating their findings to potential readers throughout the world.
In this volume, Elin McCready investigates the semantics and pragmatics of hon-
orifics: linguistic expressions which are conventionally associated with the expression
of politeness, respect, or formality. Honorifics are found in a large and typologi-
cally diverse array of languages, and come in various morphosyntactic forms, from
sentence-level particles to word-level morphology to free-standing lexical items. And
although it is clear that such expressions are conventionally associated with the expres-
sion of politeness, deference, formality, and respect, and are used both to express and to
form the kinds of social relations that trade in such concepts, it is less clear exactly how
this association should be captured, and the extent to which it stems from the semantics
of honorific language vs. the choices speakers make about whether or not to use
such language vs. language-independent characteristics of the relevant social relations
themselves. In The Semantics and Pragmatics of Honorification, McCready establishes
a much-needed linguistic basis for exploring these issues, by carefully distinguishing
distinct classes of honorifics based on both distribution and contributions to meaning,
and then providing a precise formal semantics and pragmatics that accords both
with their grammatical/compositional properties and with their particular semantic
and pragmatic properties, which place them in the expressive dimension. This work
represents the most comprehensive analysis of honorifics in the formal semantic and
pragmatic literature to date, and provides a foundation for future work geared towards
deepening out understanding of the relation between compositional meaning and
social meaning.
Acknowledgments
This book arose from a paper written for PACLIC in 2014 (McCready, 2014b). This
paper was one of those which springs into the mind already fully formed and seems to
write itself, but was limited to the notion of register and (to a lesser extent) the proper
way to think about pronouns. Looking more carefully at issues around pronominals
and role honorifics, it became clear that there was a great deal more going on than
showed itself at first, and that a lot of it was also relevant to other issues in the general
area of social meaning and semantics/pragmatics: slurring, gender biases, issues of
subordination and injustice, among others. It turned out to take a book-length work
to address them all, even to the limited degree that this book manages; in general,
this work probably raises as many questions as it answers. In addition to the analyses
worked out here, I gesture at many directions for further research; I plan to pursue
some of them myself, and I hope others will find them intriguing as well.
Thanks for discussion, suggestions or other speech acts (this joke stolen from
Jakub Szymanik) to Nicholas Asher, David Beaver, Daisuke Bekki, Heather Bur-
nett, Chris Davis, Patrick Elliott, Michael Erlewine, Robert Henderson, Magdalena
Kaufmann, Lily Kobayashi, Chung-min Lee, Midori Morita, Hiroki Nomoto, David
Oshima, Pittayawat Pittayaporn, Paul Portner, Yasutada Sudo, Shoichi Takahashi, Yuki
Takubo, Upsorn Tawilapakul, Grégoire Winterstein, Akitaka Yamada, audiences at
Texas Linguistics Society, LENLS, FAJL, WAFL, ICAL, GLOW in Asia, PACLIC, ICL,
Cornell University, University of Delaware, Chulalongkorn University, Mie University,
NUS, and ZAS Berlin, and to anyone else I have forgotten at the moment of writing this
set of acknowledgments (which is likely a large class of people: sorry everyone). Thanks
also to the Japan Society for the Promotion of Science for supporting the project (via
JSPS Kiban C Grants #25370441 and #16K02640).
Thanks to Julia Steer and the anonymous reviewers who have seen this project in
the variety of forms it has taken, and to Chris Barker and Chris Kennedy for their
work as series editors. Thanks especially to Chris Kennedy for extensive and very useful
comments, and to Judith Tonhauser for recommending that I expand the paper which
was the original seed of this book into a monograph; this version is far better than the
previous one(s) in a number of ways, and it wouldn’t have occurred to me to make a
book out of it without your suggestion.
Thanks to Midori, Colin, Kai, and Tyler too, and everyone else I love.
List of Abbreviations
acc accusative
antihon antihonorific
arg argument
benef benefactive
CI conventional implicature
cond conditional
cop copula
dat dative
DP determiner phrase
DS discourse segment
evid evidential
exc exclamative
fem feminine
form formal
gen genitive
hon honorific
inf infinitive
LFG Lexical-Functional Grammar
LP “linking particle”
masc masculine
MP Maximize Presupposition
neg negative/negation
nom nominative
NP noun phrase
obj object
pln plain
pol polite(ness marker)
pres present tense
prog progressive
pst past tense
pt particle
q question
SDRT Segmented Discourse Representation Theory
subj subject
top topic
T/V tu/vous
utt utterance
1
Introduction
This book is about the semantics and pragmatics of honorifics. Honorifics are lexical
items or morphological units which have the expression of politeness or formality
as one primary aspect of their meaning. They are found widely across languages
(see Agha, 1994 for a useful survey), and have received extensive attention in lin-
guistics, both from formal and informal perspectives. The interest linguists take in
honorifics stems partly from the fact that they are common and partly from the fact
that they play a crucial role in anchoring linguistic agents in social hierarchies and
relationships. The aim of this book is to show how this anchoring works via formal
techniques, and to argue that it is an instance of a more general class of expressions
that let speakers situate themselves and others within society and its structures. As
such, it will turn out that the phenomenon of honorification is also highly relevant for
the work of philosophers interested in social facts, especially those interested in how
social facts and language interact.
It should be noted at the outset that there is a vast amount of work on topics related
to honorification within sociolinguistics and anthropological linguistics: the use of
honorifics, how politeness is expressed, how honorification and politeness relates to
social structure and hierarchies, and so on. The range of empirical data at our disposal is
extremely rich. Still (as elsewhere in linguistics) the dialogue between sociolinguistics
and formal semantics/pragmatics has been relatively sparse until extremely recent
times; consequently, the data needed to see how honorifics and related expressions
have been analyzed is not always available. One goal of this book is therefore to try to
bridge this gap, or begin to; hopefully, the framework presented in what follows will
allow the researcher interested in semantics and pragmatics to make a more direct use
of this literature.
Most of the existing work on honorifics, which is either done within semantics and
pragmatics or speaks directly to those domains, has focused on three general topics.
First, from a formal semantic perspective, researchers have been concerned with the
way in which semantic composition with honorific expressions takes place, and with
the kinds of denotations which they have; some main results of these investigations
will be summarized later in the book.1 A second line of research is found within the
sociolinguistic tradition (and also within discourse analysis), and looks at ways in
which speakers use politeness expressions to indicate aspects of their social identities
1 Work on syntactic aspects of honorification is closely related (Niinuma, 2003; Miyagawa, 2017), but since
morphological affixes with honorific meanings will not be my primary focus here, I will not focus too much on
these issues; see Chapter 5 for discussion.
The Semantics and Pragmatics of Honorification. First edition. Elin McCready

© Elin McCready 2019. First published in 2019 by Oxford University Press.
2 1 introduction
and further their general social goals (Brown and Levinson, 1987; Watts, 2003). Finally,
there is a tradition which attempts to situate the use of politeness, including honorifics,
within a general theory of rational linguistic behavior; this work begins with Brown
and Levinson (1987) and continues to game-theoretic accounts like that of van Rooy
(2003). This book lies in the first strand of research.
Given the amount of research done in this area, it is no surprise that significant
results have been obtained. However, a problematic feature of the literature is that the
three domains of research mentioned above do not engage extensively with each other.
Research on honorific meanings tends not to consider observations made within dis-
course analysis; game-theoretic accounts try to predict rational honorific use without
proposing a truly adequate formal semantics for honorific content. A theory which
can bring the various aspects of politeness together seems necessary, especially given
the current interest in honorification in formal circles, and further is essential for the
automatic generation of appropriate speech in computational pragmatics. The aim of
the present work is to propose a semantics which is capable of making predictions
about the felicity conditions and discourse effects of honorific content and thus can
serve as a foundation for such a theory; modeling substantial sociolinguistic observa-
tions (excluding some general discussion of the social role of honorification and related
domains such as slurs which are tied directly to the semantics I will propose) and tying
the result to game-theoretic calculation is left for later stages of the current project.
1.1 Honorifics: definitions and examples
Before entering properly into the semantic analysis of honorifics, some points must
be clarified to delimit the domain of inquiry of this book. In particular, there are two
issues which must be addressed. First, what exactly counts as an honorific? There are
a wide range of expressions in natural language which might be thought of as having
an honorific character, or which sometimes function to mark or perform politeness.
But not all of these fall into the class of honorifics proper. The first task, therefore, is to
indicate what I take to fall under the category of honorific expressions for the purposes
of the present analysis. The second question is closely related to the first, and is (to some
degree) difficult to separate from it: what is the relationship between honorification and
politeness? That is, are honorifics necessarily used to indicate politeness, and is their
use necessarily polite? Is a full theory of politeness and polite behavior required for a
theory of honorification? Answering these questions is the goal of the remainder of this
chapter. For the first, I will define honorifics as expressions which have as their main
function (in a sense to be defined shortly) the expression of formality or informality.
For the second, I will claim that their meanings are distinct from politeness and polite
behavior in a way that allows the two to be profitably teased apart. As I explore these
issues, I will also give a kind of preview of some of the empirical content of the book.
Here is a standard definition of honorific expressions from the literature: they are
those expressions which perform the linguistic marking of “honorification: relation-
ships involving social status, respect or deference between communicative interactants
1.1 honorifics: definitions and examples 3
(Agha, 1994).” This definition is compelling, but remains somewhat underspecified.

On it, the range of expressions that potentially count as honorific is vast. Many, or
even most, expressions seem to come with implications for the speaker’s beliefs about
social status, the formality of the context and the respect she deems it advisable to
pay to the other conversational participants. Let us consider a few examples, moving
from the uncontroversial instances of honorification to less clear or obvious cases. I will
indicate the dividing line between the sorts of expressions this book will treat and those
it will not when the point of demarcation is reached, though the line will be revisited
in Chapter 8 from the perspective of the formal theory I will propose.
The examples in (1.1) contain expressions which are uncontroversially honorific.
The Thai sentence in (1.1a) contains the honorific particle khá, which is used by people
presenting as female in polite speech.2 (1.1b) is a Japanese sentence containing the
honorific suffix -mas-. Both of these honorifics are usable only in unembedded clauses,
and are therefore classifiable as root phenomena: they exemplify a class of honorific I
here call utterance honorifics, which always reference the situation of utterance. Their
analysis is the subject of Chapter 4.
(1.1) a. foň dtòk khá

rain fall pol.pt
‘It’s raining’ + the speaker is being polite and presenting as female
b. ame-ga fut-tei-mas-u
rain-nom fall-prog-hon-pres
‘It’s raining’ + the speaker is being polite
The next set of examples are instances of what I will call argument honorifics: hon-
orific expressions (here verbs) which target sentential arguments for honorification.
The individual toward whom politeness is expressed therefore doesn’t need to be a
conversational participant. These are also clear instances of honorific expressions,
and these are the object of the most study in formal linguistics, especially syntax.3
(1.2a) shows a suppletive honorific form meaning both ‘come’ and an expression of
formality toward the denotation of the DP which serves as the sentential subject; (1.2b)
is an instance of honorific verbal morphology which indicates formality toward the
denotation of the object DP. Forms like this are the subject of Chapter 5.
(1.2) a. sensei-wa ashita irassharu yoo desu

professor-top tomorrow come.hon evid cop.hon
‘The professor is apparently coming tomorrow’ + the speaker is being
respectful toward the professor
b. Taroo-kun-ga sensei-o o-tasuke-shi-ta
Taro-hon.inf-nom teacher-acc hon-help-do-pst
‘Taro helped the teacher’ + the speaker is showing respect for the teacher
2 Questions of gender and gender presentation will be returned to in Chapters 7 and 8.

3 I will not gloss the utterance honorific aspects of these sentences in these particular examples in the interest of
readability. See subsequent chapters for more adequate treatments of the other honorific elements (e.g. utterance
honorific suffixes and titles).
4 1 introduction
With the next set of items, we see forms which are less clearly purely honorific.
While the above forms indicate formality or honorification as their core meaning,
or indeed their only meaning, the honorification expressed by terms of address such
as professor, the Japanese sensei or the Thai aacaan (both meaning ‘teacher’) seem to
have honorification as a kind of side effect of their core meaning, which is to indicate
the profession of the addressee (or other individual named, in the case of examples
like (1.3a). These contrast with cases like the Thai khun ‘Mx.,’⁴ an honorific modifier
of nominals indicating respect for the individual denoted by the nominal, where the
content of the modifier is rather bare excluding the honorific content, as with the
Japanese suffix -san. I call examples of the first type derived honorifics and treat them
in Chapter 6. Forms like -san and khun are treated in Chapter 5, as a special kind of
argument honorific, partly for reasons having to do with the way in which they carry
out their honorific functions. Details on the parallel with more standard argument
honorifics can be found there.
(1.3) a. Professor Mendoza is not here today.
b. Yamada-sensei-ga kita
teacher-nom came
‘Professor Yamada came.’
c. Khun Somsak maa thîinîi
Mx. Somsak came here
‘Mx. Somsak came here’ + the speaker indicates respect for Somsak
In (1.4) we see the final set of items which are more or less uncontroversially deemed
honorific: pronouns which are specified for (in)formality, among other aspects of their
interpretation. The simplest kind of system is exemplified by many European lan-
guages, such as French and German, in which two distinct second person pronominal
forms are found, one used in formal speech and one casually (1.4a). But more complex
systems are common in the world’s languages, for instance in many languages of Asia;
this book focuses on Japanese and Thai, in which a wide variety of first person pronouns
with different implications for formality, gender specification, and other presentational
aspects of persona can be found, as with the Japanese atashi ‘I (informal, feminine)’ and
boku ‘I (semiformal, masculine), shown in (1.4b,c).
But these languages are interesting beyond just their wide variety of first person
pronouns. Both Thai and Japanese also have a large array of second person pronouns,
each also with its own shades of meaning, for example the Thai examples in (1.4d,e):
here khun is polite (unsurprisingly, since it is the pronominal version of the honorific
prefix above), and mung is extremely rude. The availability of both first and second
forms in different registers makes available the possibility of combining forms in a
variety of different patterns, which is exploited differently (or disallowed entirely)
⁴ ‘Mx.’ is a nongendered honorific term. I use it here instead of ‘Mr.’ or ‘M(r)s.’ as khun, like the Japanese -san,
indicates nothing about the gender of its referent.
1.1 honorifics: definitions and examples 5
by different languages. All this will be detailed further in Chapter 7, where I show
that Japanese and Thai make use of different strategies for the lexical introduction of
honorific content in their pronominal systems: Thai via direct register specification,
and Japanese via the introduction of speaker commitments about social behavior.
(1.4) a. French tu/vous, German du/sie

b. atashi-wa iku yo
I.inf.fem-top go pt
‘I’m going’ + speaker is presenting in a feminine manner in an informal
context
c. boku-wa iku yo
I.semif.masc-top go pt
‘I’m going’ + speaker is presenting in a masculine manner in a not too formal
context
d. khun yàak bpai?
you.pol want go
‘You want to go?’ + speaker is being polite to the addressee
e. mung yàak bpai?
you.antihon want go
‘You want to go?’ + speaker is being rude to the addressee
The forms already discussed comprise the bulk of the empirical domain of this book.
Let me mention now some linguistic phenomena which have an honorific character, in
the sense that they have implications for the degree of formality the speaker thinks it
expedient to speak with and consequently for both the discourse context and the social
relations between the interlocutors, but this meaning is entirely incidental to their main
function. The remaining forms here thus do not form part of the empirical domain of
this book in any substantial sense, though I believe the analysis to be developed can
easily be applied to them, as will be discussed in Chapter 8.
The first type are stylistic variants of the kind in (1.5). This sort of contrast has
been studied extensively in variationist sociolinguistics (e.g. Eckert, 1989) in terms of
the indexing of various aspects of speaker personae and self-presentation (as recently
formalized by Burnett, 2017): for instance, fully articulating the final sound in the verb
in (1.5a) implicates a careful, professional or even pedantic character, while simplifying
it, as in (1.5b), produces a friendly and casual, but possibly not very professionally
component impression. However, these kinds of variants also have implications that
have an honorific feel. Specifically, while the form in (1.5a) would be appropriately
used in a formal setting, that in (1.5b) might not be; this kind of fact can be tied to
the sort of analysis I will present in Chapter 7 of honorific pronouns. Chapter 8 will
be extensively concerned with the relationship between honorific and other kinds of
social meaning; examples of this type will be returned to there, though they are not,
properly speaking, honorifics.
6 1 introduction
(1.5) a. I was cooking up some dinner.

b. I was cookin’ up some dinner.
The last kind of case I want to mention is that of distinct lexica or registers. The
term ‘register’ will be used throughout this book in a different sense, but here is meant
to refer to the kind of difference observed in (1.6): cases where two or more words
exist with the same or at least extremely similar denotations, but where one is much
more formal or specific than the other. In English, such register distinctions are found
in e.g. scientific discourse, as in (1.6a); but in other languages such as Javanese, they
are much more extensive, in a way that is naturally deemed honorific in the standard
respect. Javanese in particular will be discussed in Chapter 4, where the three distinct
registers—krama, madya and ngoko—available in that language will be treated as a kind
of utterance honorific, as exemplified by the terms in (1.6b), both of which mean ‘rice’
but which are associated with different registers.
(1.6) a. cat, feline
b. sega, sekul
The book therefore runs the gamut of expressions that might be considered honorific
in character, starting with the uncontroversial pure lexical honorifics, moving to
‘impure’ derived honorification, and concluding with expressions that often aren’t
thought of as honorific at all. This breadth is intended. The aim of this book is twofold.
Its first goal is to give a theoretical framework capable of accounting for the meanings
of honorific expressions, their appropriate use, and their effects (at some level of
idealization and abstraction, as usual in the formal side of linguistic theory). But its
second goal is to consider honorifics in the broader context of social meanings in
general: this area is as yet underdeveloped in a formal sense, but interest is rapidly
growing, and, given the clearly social nature of honorific meanings and effects, it seems
appropriate to think about them in the context of general effects related to register and
social relations. Doing so is the purpose of the final chapter of the book.
With that last paragraph this book might begin to look as if it runs the risk of
attempting to give a theory of everything (in a linguistic sense). Despite possible
appearances, I do wish to avoid doing so. In particular, I want to consider honorifics
in isolation as much as possible, in that I would like to restrict attention to the lexical
meanings of honorifics and to their effects on formality in particular (in a sense defined
in more detail in subsequent chapters). Doing so requires considering only a subset of
their pragmatic effects. I want specifically to avoid the need for a full theory of honorific
use; justifying this choice requires showing that the analysis of honorifics is possible
without giving a full theory of politeness. I turn to this task in Section 1.2.
1.2 Honorifics and politeness
Can one give a theory of honorification without giving a theory of politeness? If the
answer is negative, the whole idea of giving a theory of honorification quickly appears
1.2 honorifics and politeness 7
to be very difficult, and possibly even impossible in a single book. There are few
candidates for a formal theory of politeness within linguistics (Brown and Levinson,
1987 being the main exception, though its limitations are well known; there are also
game-theoretic approaches to politeness such as Quinley, 2012; van Rooy, 2003 which
I will not discuss in detail here). The whole project of properly formalizing politeness
appears extremely difficult, because it requires at least (i) a theory of what counts as
a polite or impolite act, which in turn requires a theory of norms of social behavior,
(ii) a way to analyze speaker motivations for being polite or impolite, because politeness
is in part an intentional notion (or so I would argue), and (iii) a theory of strategic
behavior surrounding both conventional and nonconventional content. A theory of
politeness then starts to look like a full theory of human behavior. Each element
of such a theory is extremely complex; if a theory of politeness is a prerequisite for
a theory of honorification, the whole project is daunting.
Fortunately, politeness and honorification can be separated. This can be shown in
two ways: first, by showing that honorification does not necessarily indicate politeness,
and, second, by showing that politeness does not require honorification. If these claims
are correct, then it is possible to separate politeness from honorification, and a full
theory of politeness is not necessary for an analysis of honorifics.
The first thing to do is to show that the use of honorifics does not necessarily indicate
politeness. I will do so in two ways: first, by showing that it is possible to use honorifics
in an impolite way, and, second, to call into question the idea that honorific meaning
implicates politeness in any way, based on claims found in the literature.
Ordinarily, one thinks of honorifics as indicating politeness: using an honorific
means that the speaker is being polite. But this conclusion is too quick. Minegishi
Cook (2011) presents a study of argument honorifics which shows that they are not
used primarily to indicate politeness (as opposed to something like please), but rather
to show the speaker’s placement in a social hierarchy; this kind of usage will be returned
to in Chapter 8, where the relationship between honorification and social meaning will
be further discussed. This shows that the two notions are at least separable. Further, it
is possible to use honorifics in a directly impolite or offensive manner; consider the
Japanese phenomenon of inginburei ‘(hypocritical courtesy),’ which refers to the use
of excessive honorifics to be rude, which would be highly unexpected if honorifics
always correlate with politeness.⁵ Thus, the use of honorifics does not always indicate
politeness.
It remains to show that the other direction also does not hold: politeness doesn’t
require honorification any more than honorification always indicates politeness. This
is trivial, especially if one counts positive politeness (i.e. the indication of solidarity):
choosing to avoid honorifics already has connotations of positive politeness in many
⁵ I should note that I will not address this phenomenon further in this book, because it seems to fall into the
category of strategic uses of honorification; intuitively, given that honorifics appropriate to the current context
of speech should be used, various pragmatic effects will arise from purposely using honorific patterns which
fall outside of those parameters. The exact way in which an impression of rudeness arises does indeed require a
theory of politeness and, likely, how politeness interacts with broadly Gricean considerations. I leave this domain
for future work.
8 1 introduction
contexts, so it follows that honorifics are not required for politeness. The clearest
exemplar of this case is the Thai politeness particles discussed in the previous section:
the omission of khá(p) indicates that the situation is informal, not that the speaker
is being rude. But the same point can be made even without considering specific
linguistic items. One can be polite without even speaking, via gesture, posture, or even
less symbolic cues such as the way one chooses to behave. There is no sense in which
the use of honorifics is required for politeness.
The upshot of this discussion is that honorifics and politeness can be teased apart
quite straightforwardly. This is, in some sense, not a surprise: honorifics are a tool for
indicating politeness, and so should not be indistinguishable from politeness, just as
hammers are distinct from carpentry, though they are used to perform it. In this book,
therefore, I will forego tying honorification and politeness together, focusing instead
on understanding why the hammer can do what it does. Still, Chapter 8 will include
some discussion of the relation between honorification, politeness, and the strategic
use of language, and the conclusion will return to this question as well in the context
of game-theoretic analysis.
With this background in place, we are now ready to move into the meat of the book:
the formal analysis of honorifics. I will start by providing a framework in Chapters 2
and 3, and then turn to empirical analysis in the remainder of the book.
2
Honorification as expressive
One obvious initial question that has to be addressed for any theory of honorification is
the type of meaning that honorifics introduce. The standard toolkit makes four options
available: at-issue or truth-conditional content, conversational implicature, presuppo-
sition, and expressive content (as distinct from conventional implicature). Examining
the options makes it clear that honorification is best viewed as expressive.
This conclusion is not only my own. Twenty years ago, Kaplan (1999) wrote, in the
seminal (though still unpublished) paper on expressives which kicked off the current
wave of research on formal properties of expressive content, the following:
Many languages contain a distinction between “formal” and “familiar” second person
pronouns […] It can hardly be doubted that this distinction belongs to the semantics
of the pronoun, and within semantics, not to the semantics of reference, but to the
expressive side of meaning. (Kaplan, 1999: 26–27)
The idea of honorification as expressive, then, was already present at the early stages
of this work. Subsequently, there has been substantial recent research in this area,
all of which appears to take honorifics to introduce expressive meanings (Potts and
Kawahara, 2004; Sells and Kim, 2007; Horn, 2007; McCready, 2010b).1 The main
reasons for thinking so are that honorific meanings are not affected by denial, do not
interact with operators like negation, and appear to resist nonexpressive paraphras-
ing. This chapter will summarize some existing discussion of both expressives and
honorifics, and provide additional data showing that honorifics show the properties
of expressive content. However, as we will see, some of the standardly accepted criteria
for expressivity don’t seem to apply to honorifics; it turns out, however, that they
also fail to apply to certain other items which can be taken to be expressive. This
observation leads to the conclusion that not all expressive items behave identically,
which is perhaps not a surprise. The final part of the chapter briefly considers and
rejects the other possibilities for honorific meanings: at-issue content, presupposition,
and conversational implicature.
2.1 Properties of expressives
Potts (2007) provides the following six criteria for expressive items.
1 Some aspects of this proposal are anticipated by Pollard and Sag (1994), as pointed out by a reviewer.

10 2 honorification as expressive
(2.1) Properties of expressives:

a. Independence: Expressive content contributes to a separate dimension of
meaning.
b. Nondisplaceability: Expressives predicate something of the utterance
situation.
c. Perspective dependence: Expressive content is evaluated from a particular
perspective (often the speaker’s).
d. Descriptive ineffability: Speakers are never fully satisfied when they para-
phrase expressive content using nonexpressive terms.
e. Immediacy: Expressives achieve their intended effect by being uttered.
f. Repeatability: Repeating an expressive strengthens its content; it is not
redundant.
Examining these properties will help to understand the intuitive notion of expressivity.
Our main test case will be the expressive adjective fucking. For the purposes of the
present discussion (and following Potts), I will characterize it as indicating that the
speaker is highly emotionally affected by the object denoted by the term which fucking
is predicated of; a more sophisticated view can be found in McCready (2012b).
The first property, Independence/(2.1a), is often taken as canonical for expressives
and also the closely related conventional implicatures. Consider the following two
examples.
(2.2) a. I didn’t see a brown dog.

b. I didn’t see a fucking dog.
(2.2a) indicates that the speaker didn’t see a dog which satisfies the property of being
brown. The sentence is made true even if the speaker has seen a dog which wasn’t
brown; thus it is sufficient for the adjectival content to be false in order for the
sentence to be true. The content of the adjective thus falls in the scope of negation.
Compare this situation with what is found in (2.2b): here, if the speaker saw a dog the
sentence naturally comes out false, regardless of the speaker’s attitude toward the dog.
If the speaker isn’t in the requisite excited emotional state, the sentence is inappropriate
rather than false, a notion explicated by Kaplan (1999) by taking expressives to
introduce use-conditions rather than truth-conditions. Empirically this means that,
semantically speaking, the content of the expressive adjective is not in the scope of
negation. Another way to put this is that the content of the adjective invariably projects
out of the scope of negation; it is independent of the operator. Presupposition is similar,
of course, but admits for the possibility of binding in universal constructions such as
conditionals (Kartunnen-style filters, Karttunen, 1974); this point will be returned to
in §2.4, where I argue against treating honorific meanings as presuppositional partly
on this basis.
2.1 properties of expressives 11
The same holds for other kinds of semantic operators. Uttering (2.3a) indicates that
the speaker believes that he might see a dog, without reference to what kind of dog it
is; (2.3b) doesn’t ask whether the hearer saw a dog which she had strong feelings about,
and the speaker of (2.3c) commits to calling the addressee if he sees a dog, regardless
of whether he finds it emotionally affecting.
(2.3) a. I might see a fucking dog.
b. Did you see a fucking dog?
c. Fine! If I see a fucking dog, I’ll be sure to call you right away.
The immunity of expressive items to semantic operators extends to operations at
the level of speech acts. Consider (truth-directed) denials, as in (2.4): here, again,
the denial is unable to ‘target’ the content of the expressive. B’s utterance questions
whether A actually saw a dog, not whether A had the relevant attitude. Note that this
‘undeniability’ is not a function of content, though one might think that it is simply
impossible to deny an individual’s private attitudes, due to something like privileged
access (Mitchell, 1986): A’s utterance in (2.5) means something roughly similar to the
way I have characterized the meaning of fucking (on a negative interpretation), but B
is still able to deny it successfully.
(2.4) A. I saw a fucking dog in the park this morning.
B. That’s not true.
(2.5) A. I’m upset that I saw a dog this morning in the park.
B. That’s not true.
The independence property actually goes further. Potts et al. (2009) show that
expressive content fails to participate in many semantic operations ordinarily under-
stood as involving the ‘copying’ of content from one point to another, such as ellipsis
and anaphora: considering the case of one-anaphora, in (2.6), B need not have the
excited attitude A expresses for anaphoric reference to succeed, while in (2.7), the
property of being zebra-striped must hold of the object B saw as well.
(2.6) A: Yesterday I saw a fucking zebra in the park.
B: I saw one too!
(2.7) A: Yesterday I saw a zebra-striped dog in the park.
B: I saw one too!
The second criterion, nondisplaceability, references the ‘displaceability’ criterion of
Hockett (1960). Hockett defines a number of properties of human language; displace-
ability refers to the ability to talk about objects distant in time or space (and of course
modality, cf. Schlenker, 2006). Expressives lack this ability, at least according to Potts,
who takes them to apply to the situation in which they are uttered. This criterion in
fact follows from Independence at least without additional assumptions or machinery;
if semantic operators cannot apply to expressive content, then, because displacement

is the result of the action of such operators, it is expected that expressive content is not
displaceable.2 For examples of how this property works, consider (2.8): if expressive
content was displaceable, (2.8a) should have a reading on which the speaker is upset
every time she pours wine, and (2.8b) should be able to indicate that the speaker was
unhappy yesterday; but these readings do not exist.
(2.8) a. Every time I pour wine, the damn bottle drips. (Potts, 2007), attributed to
Florian Schwarz
b. Yesterday I dropped my fucking wallet.
Still, there is some wiggle room here, which can be brought out by the following
question: What are the truth conditions of expressive adjectives like fucking? The
question does not yet have an undisputed answer; but it is plain that it is not even
sensible without a notion of perspective. How can something objectively qualify as
e.g. damn (on an expressive interpretation)? This observation motivates the notion
of perspective dependence (2.1c): all expressives are interpreted from a particular
viewpoint.3 This appears to be universally true. It also lies at the root of criticisms that
have been leveled at Independence, and, by extension, Nondisplaceability.
Independence—(2.1a)—is usually taken to be canonical for expressives, but it has
been called into question on the basis of attitudes. Potts cites the example (2.9),
attributing it to Angelika Kratzer; here, plainly, the attitude is not that of the speaker,
for (s)he presumably wants to marry Webster, implying that (s)he doesn’t think he is a
bastard, while it is all too likely that this is his/her father’s attitude.
(2.9) My father screamed he would never allow me to marry that bastard Webster.
Potts suggests treating this example in terms of mixed quotation (see also Harris and
Potts, 2010). The idea is that it is possible to shift the perspective from which the
expressive is used. In this case, the perspective picked up is that of the father, in the
manner of quotation; in other circumstances, it can be that of some other contextually
salient individual. This is a powerful mechanism, as noted by Amaral et al. (2008); the
precise constraints on its application remain to be specified, though Harris and Potts
(2010) make significant progress in this direction. This is something we won’t have
much occasion to consider in the analysis of honorifics, though, so I will mostly put it
aside. It mainly reappears in the notion of register choice found in the formal treatment
in Chapter 3: there, it is possible for particular honorifics to be associated with different
individuals, both honored and honoring, though speech act participants are targeted
by default.
2 The same holds for contexts in which the relevant operator is not overt, as with for example the generic
passages of Carlson and Spejewski (1997).
3 The relation between the perspective-dependence found in expressives and that in more ‘vanilla’ truth-
conditional perspectival expressions like deictics (Oshima, 2006), various kinds of adjectives (Lasersohn, 2005;
MacFarlane, 2016), and indexicals (Kaplan, 1989) remains a sadly understudied area.
2.1 properties of expressives 13
These two properties of expressives mean essentially that they can’t be embedded
under operators and they are interpreted in the utterance situation. We will see in the
coming sections that these properties do in fact hold for honorifics, as does the next to
be discussed.
The property of immediacy is closely related to this notion of nondisplaceability.
This property says that the use of an expressive item is enough for it to achieve its
effect; thus, again, negating it or otherwise embedding it has no effect. But more is
at issue here. It is a commonplace that speech acts present proposals for changing
the commitments of conversational participants: imperatives can introduce hearer
commitments for particular actions, questions commitments to provide answers, and
so on (e.g. Portner, 2007). Assertions on this view are proposals to make changes in
the common ground, or in the private information states of conversational agents.
Expressives can then be viewed as inducing changes in the common ground without
the mediation of a proposal: in a way, they are proposals that must be accepted by
virtue of their utterance. Potts compares them to performative speech acts, which
also introduce commitments by virtue of their use. It’s clear that this is a quality
that honorifics have as well, though this claim will be substantiated by empirical data
in the following sections: merely by using an honorific, one does indeed express an
honorification of the relevant individual.
The final two properties are either controversial, or clearly not applicable to all
varieties of expressive. Descriptive ineffability is a kind of metalinguistic property;
according to it, expressives cannot be satisfactorily paraphrased or translated without
recourse to other expressives, and indeed perhaps do not admit proper paraphrases at
all. In other words, they are not ‘effable’ via descriptive content. This seems true: it is
not easy to find a way to determine whether a proposed translation of an expressive
like fucking or even hello genuinely captures the whole range of possible uses, because
the metric of comparison may not be clear. For instance, take the Japanese exclamation
chikusho ‘beasts,’ often used to translate sentences like the following.
(2.10) a. Damn!
b. Shit!
Is this a good translation? It does seem that it is fairly appropriate in terms of how
the two terms are used, but the two sentences in (2.10) don’t mean exactly the same,
so some content is being lost: it is in fact quite unclear how to consistently map the
degree of ‘emotional intensity’ (or whatever it is that is being expressed by these terms)
from language to language in the absence of an external metric. More generally, Geurts
(2007) observes that even terms like green may not have satisfactory paraphrases either,
and neither may any lexical item at all. This property thus seems to need some addi-
tional argumentative basis, some of which is given by McCready (2014a), who argues
that ineffability is the precise basis of the unavailability of translations in certain cases
and which opens the door to genuine differences in cross-linguistic expressive power.
Still, unarguably, to the degree that such items as expressive adjectives, exclamations,
or interjections lack satisfactory paraphrases, honorifics do as well.
Finally, Repeatibility is obviously a special case: though repetition of expressive

adjectives like fucking clearly does heighten the emotive impression given by the
adjective, this is definitely not the case of items treated as expressive across the board.
Many other expressives, like Kaplan’s (1999) oops, don’t seem to be strengthened by
repetition—one instance of oops seems to be no weaker than oops! oops! oops!—and
other terms such as hello or good morning may not admit repetition at all on pain of
infelicity; even if they do, it’s not clear what a strengthening effect would even amount
to in such cases. It seems likely that a rather specific kind of gradability is needed for
repeatability to hold. Thus, this last property is certainly not universal for expressives.
For the honorific case, I believe that repeatability is true for some kinds of honorifics,
but not all; this discussion is complex and will be elaborated on in the following
chapters, but I will take honorifics to separate into two distinct types with respect to
this phenomenon. The first type is associated with a particular register or ‘highness’ of
speech, as with utterance honorifics; there, repetition serves to incrementally pull the
register toward the register picked out by the honorific. For the second type, the pull
to a register happens via inferential mechanisms, but this pull happens as a side effect
of the use of the honorific itself, which tags other aspects of social reality and identity
relevant to social status and familiarity. In such cases, since the tie between honorific
and register is less direct, the effects of repetition are more malleable. This issue will be
returned to in Chapter 8.
What properties must we then consider when trying to determine whether
honorifics are expressive? The first thing is that they must introduce a kind of
not-at-issue meaning; this follows from the Independence constraint, and is shared
with other meaning types such as presupposition and conversational implicature, a
point stressed by Simons et al. (2011); Tonhauser et al. (2013). But expressives differ
from these other meaning types (and from the otherwise rather similar conventional
implicature) in having meanings that aren’t even in principle truth-conditional. While
presuppositions (for example) have meanings expressible in terms of truth, it is not
appropriate to talk about expressive meanings in these terms. I take this to be the
intuition underlying the idea of ineffability above. We then need to determine whether
honorifics are paraphrasable in the proper manner. I will argue that they aren’t, and
that honorifics do in fact fit the meaning profile of expressives.
2.2 Honorification as expressive: initial data
In this section, I want to consider the above criteria for expressivity with respect to
honorific meanings. We will see that the meanings of honorifics do indeed satisfy
these criteria, so it is reasonable to treat honorific meanings as expressive, as is largely
done in the recent literature on the topic. Alternatives will be considered and rejected
later in the chapter. The data in this section will mostly be drawn from Japanese, with
secondary data from Thai and French.
2.2 honorification as expressive: initial data 15
Let’s begin with Independence. Above we saw two phenomena related to this
property: the fact that expressive content can’t be targeted by denial, and its indepen-
dence from semantic operators.⁴ Consider first the Japanese dialogue in (2.11).
(2.11) A. Ame-ga futtei-masu
Rain-nom falling-hon
‘It’s raining’ (and the speaker is being polite)
B. Sore-wa nai yo
That-top not pt
‘That’s not true’ = ‘It’s not raining’ ≠ ‘You’re not being polite’
B’s utterance cannot be construed as denying the honorific content of A’s utterance;
this is expected if this content is expressive, or at least not truth-conditional (cf.
Potts, 2005 a.m.o). For a second piece of evidence, consider what happens when hon-
orifics are placed under negation. The first example is from Japanese, and the second
example from Thai; both involve a pronoun appearing in the scope of negation, where
the Japanese pronoun is highly formal and the Thai pronoun is associated with informal
contexts.
(2.12) a. watakushi-wa itta to iu koto-wa nai
I.Formal-top went C say thing-top not.exist
‘It’s not that I went’ (and the speaker is self-presenting formally)
b. chaň mây chɔɔp ̂ khun
I.fem.mid not like you.mid/high
‘I don’t like you’ (and the speaker is being somewhat informal)
In the above, we see that the honorific content of the pronouns projects past the
negation, as in fact do their gender specifications, an observation that will become
important in Chapter 7, where the semantics and pragmatics of honorific pronouns
are analyzed.
What about other kinds of contexts, such as the scope of epistemic modals, ques-
tions, or conditionals? We find the same behavior here: the content of honorifics fails
to remain within the scope of these operators.
(2.13) Iwabuchi-sama-ga kuru kamoshirenai
Iwabuchi-hon-nom come maybe
‘Mx. Iwabuchi might come’ + ‘the speaker is being highly respectful to
Mx. Iwabuchi’
(2.14) Iwabuchi-sama-ga ki-masu ka
Iwabuchi-hon-nom come-hon q
‘Is Mx. Iwabuchi coming?’ + ‘the speaker is being highly respectful to
Mx. Iwabuchi’
⁴ Non-truth-oriented denials can however challenge expressive content, though not perhaps directly deny it.
See von Fintel (2004) for the closely related case of presupposition.
(2.15) a. moshi Iwabuchi-sama-ga ki-tara kono ii sake-o

if Iwabuchi-hon-nom come-cond this good sake-acc
das-oo
put.out-hort
‘If Mx. Iwabuchi comes, let’s put out this good sake.’ + ‘the speaker is being
highly respectful to Mx. Iwabuchi’
b. daiji-na kaigi dat-tara Iwabuchi-sama-mo kuru
important-cop meeting cop-cond Iwabuchi-hon-also come
hazu-da
must-cop
‘If it’s an important meeting, Mx. Iwabuchi will surely come too.’ + ‘the
speaker is being highly respectful to Mx. Iwabuchi’
Finally, consider the unavailability of expressive content to ‘copying’ operations
such as anaphora and ellipsis. Suppose that A is a company employee and B is a
member of the board of the same rank as Mx. Fukuda. In the following discourse,
B’s utterance does not imply any special respect toward Mx. Fukuda, indicating that
honorific content does not participate in this ellipsis. Similarly, it is often noted in
the literature (e.g. by Sells and Kim, 2007 for Korean) that honorific content is often
eliminated when quoting or glossing speech in TV news programs, which again shows
that it is peripheral to the main content.
(2.16) A. saki Fukuda-sama-o omenikakari-mashi-ta
a.moment.ago Fukuda-hon-acc see.hon-hon-pst
‘A moment ago I saw Mx. Fukuda’ + ‘the speaker is being extremely
respectful to Mx. Fukuda’
B. boku-mo
me-too
‘I did too’
On the basis of evidence of this kind, together with the much more extensive
evidence in the previous literature, we can conclude that honorific content has the
Independence property.
The second property we should consider, Nondisplaceability, states that the effect of
the expressive item holds at the speech time, not at other ‘displaced’ temporal points,
places, or worlds; thus, if a sentence contains an honorific, it should express the current
attitude of the speaker, not that of the speaker (or some other individual) at a distinct
time (place, world). I claimed above that this property follows from Independence, but
let us see how the honorific data plays out here.
(2.17) koko ni kuru toki-wa maikai doko-ka-no kaisha-no
here to come time-top every.time where-q-gen company-gen
erai hito-ni o-ai-shimasu
high.grade person-dat hon-meet-do.hon
‘Every time I come here I meet some high-ranked person from some company
or other.’ + ‘The speaker is currently indicating respect for these people’
In this first example, we see that, while the speaker is indeed indicating respect for each
individual (s)he encountered previously, this respect is taken to hold at the speech time,
rather than at the time at which (s)he met those people. This is entirely parallel to the
interpretation of damn in (2.8a) above.
Consider next the property of perspective dependence. While it is slightly inde-
terminate (and controversial within the literature) exactly whose perspectives can be
adopted for the interpretation of a particular expressive, just as with other perspectival
expressions (see e.g. McCready, 2007a for more on this issue), it is at least clear that
honorifics are relational, in that they involve the way particular agents take themselves
to exist within social structures. Ordinarily the agent whose social relationships
are expressed is the speaker, just as with other expressives (something modeled in
Chapter 3 by assuming speaker perspective as a default); thus we can conclude that
honorifics have the property of perspective dependence as well. The property of
immediacy is similar. By uttering an honorific, the speaker performatively indicates
her recognition of the social status of the honored individual, or of the formality of
the situation. This effect is neither cancelable nor displaceable, in precisely the way
expected given this criterion.
(2.18) a. kyoo-wa samui desu ne

today-top cold hon pt
‘Today is cold, isn’t it?’ + ‘the situation is relatively formal’
b. #betsuni ima reigitadasiku hanasu hitsuyoo aru to
particularly now politely talk need cop copc
omotteinai kedo ne
don’t.think though pt
‘But I don’t especially think I need to be polite though’
Thus honorifics easily satisfy these two criteria, further supporting a view of them as
introducing expressive content.
The next property we must consider is ineffability. Can honorifics be paraphrased
in terms of other, nonhonorific content? It doesn’t seem that there is an adequate
paraphrase available. Potts and Kawahara (2004) simply write: “Speakers are never fully
satisfied when they paraphrase honorifics.” (p. 258) Anecdotally, this certainly seems
to be true: in my own experience, asking for paraphrases to native speakers has yielded
many things like the following, none of which either satisfies the speakers themselves or
gives much of an intuitive understanding of the meaning of the constructions, because
each admits immediate counterexamples. I think it is fair to conclude that (keeping
in mind the worries of Geurts, 2007 about the degree to which ineffability is specific
to expressives) honorifics are just as ineffable as other uncontroversial instances of
expressive content.
(2.19) Yamada-san-ga irasshaimashita

Yamada-Mx-nom came.hon
‘Mx. Yamada came’ + H
(2.20) Possible candidates for H:

a. The speaker respects Mx. Yamada.
b. Mx. Yamada deserves respect.
c. The speaker feels positively about Mx. Yamada.
d. Mx. Yamada has a high rank.
Finally, we can turn to repeatability. This property refers to the strengthening of the
effects of expressive items with their repetition. It might seem that this property doesn’t
hold of honorifics: one key property of honorifics in all languages is that they should
be used when it is appropriate to do so, usually in every sentence of a conversation or
discourse. In what sense does repeating the honorific make the ‘honoring’ it expresses
stronger? Intuitively, using (say) vous in every sentence when addressing someone in
French doesn’t indicate a higher degree of respect than using it once. Thus at first glance
honorifics might seem to fail to possess the repeatability property.
But I think this conclusion involves a view of repeatability that is too simple.
In examples with expressive adjectives, repeating the adjective does indeed lead to a
stronger meaning; it also seems that this repetition is more or less invariable, in that
the repetition consistently makes the attitude expressed seem stronger. Chris Kennedy
(p.c.) points out that the same is true of other gradable adjectives, as in (2.21), which
suggests that we might well be seeing a property of (certain) adjectives as opposed to a
property of expressive content.
(2.21) It’s a wide wide world.
But other kinds of expressive meanings might well involve other kinds of effects with
repetition, as already indicated in the previous section. In particular, it seems likely to
me that the strengthening quality of adjectives like damn or fucking is the result of the
fact that the scale with which they are associated has no upper bound: nothing bars
us from having ever stronger emotions with respect to some object or situation, so it
becomes possible for the adjectives to push the emotive quality ever higher.
It should be observed that this idea is not captured in the model of Potts (2007)
and subsequent authors in this tradition, for whom the scale is associated with the
interval [−1, 1] and subintervals thereof which are picked out by particular adjectives
(or other expressions). A more adequate model might allow for an unbounded scale, or,
alternatively, for the adjective to induce changes in the extremity of the values expressed
rather than directly referencing subintervals. I will not push this point further in the
present work, but I believe it connects closely to the differences one finds in the effect
of repeating honorifics and expressive adjectives.
Let us see how this idea can be spelled out. Suppose that, as in Potts (2007), we
make use of the interval [−1, 1] for the emotive adjectives. According to Potts, these
adjectives provide specifications of speaker attitudes toward the individual (or event)
of which they are predicated (see Chapter 3 for some further detail); specifically, given
the existence of expressions aIb which indicate that a has the attitude described by the
interval I to b, they provide either (a) initial specifications of that interval or (b) serve to
‘narrow’ the possible attitudes of a toward b by altering I to a new interval I′ such that
I′ ⊆ I. Effectively, the adjectives on the Potts view work to narrow down the intervals
which serve to mark attitudes.
But what if we use a less static system? Suppose that, instead of simply restrict-
ing existing attitudes or introducing new ones, expressive adjectives dynamically ‘pull’
existing attitudes toward the intervals they introduce. Consider, for instance, fucking
on its negative interpretation. Let’s suppose that it targets the interval [−1, −.7],
corresponding to a strongly negative attitude. The basic idea now is that the effect of
using the adjective is to move the attitude of the speaker toward the predicated object
in the direction of this interval. The same basic idea will appear in a different form
in Chapter 3, where it will play a key role in the analysis of honorifics I will propose.
Here is a simple implementation of the idea I have in mind for emotives. (2.22) is an
auxiliary definition allowing us to pick out the least and greatest elements of an interval;
the dynamic definition in (2.23) moves existing attitudes in the direction of the interval
expressed by the emotive expressive adjective.
(2.22) a. min([i, j]) = i
b. max([i, j]) = j
(2.23) Dynamics of emotive expressives (preliminary)
I if I ⊆ Exp
I[(Exp)]H = I′ , where I′ = { min(I)+min(Exp)) max(I)+max(Exp) .
[ , ] else
2 2
The main interest of this definition for present purposes is how it interacts with
the notion of repeatability. For emotive adjectives like fucking, which include extreme
elements of the scale (in this particular case, one of the endpoints of the interval
[−1, 1] which forms the domain of the emotive meanings in this theory), repeating the
adjective will continue to pull the attitude expressed in the direction of the endpoint,
giving a strengthening effect to each repetition. But this is not the only kind of
strengthening available; rather, it depends entirely on the fact that the lexical meaning
happens to contain one of the endpoints of the interval corresponding to the domain.⁵
Honorifics (on the view I will propose) directly reference registers of various kinds:
formal, informal, casual, and so on. These registers are also modeled using (like the
work of Potts on expressive adjectives) subintervals of [0, 1], but, crucially, ordinarily
not intervals which contain endpoints. Given this, dynamic definitions of the kind
above won’t necessarily give a ‘more polite’ or more honorific meaning as a result of
repetition, but rather a general trend in interpretation in which the politeness expressed
comes to coincide with the subinterval picked out by the honorific being repeated. With
sufficient repetitions of a single honorific in the absence of other factors, the discourse
register will eventually be identical to the subinterval expressed by the honorific, in
a way familiar from other kinds of continuous models. I want to suggest that on this
more nuanced understanding of repeatability, honorific meanings are just as clearly
⁵ For this analysis to go through, it is necessary to carefully think about what should happen with the
denotations of emotive expressives milder than fucking, such as damn or even darn; this question seems somewhat
complex, and I won’t address it here as my main purpose is to bring out the sense in which honorifics are also
strengthening.
strengthening as those of emotive expressives. Though they don’t necessarily result in

a stronger meaning in the sense of one that’s more honorific, the meaning is more
strongly correlating with the meaning of the honorific itself.
The upshot of the above discussion is that honorific meanings satisfy the definitional
conditions placed on expressives by Potts (2007) and subsequent authors. I conclude
that they are indeed properly treated as introducing expressive content.
2.3 Other languages and other data
The previous section gave extensive data from Japanese, concluding that honorifics are
expressive in nature. This discussion comes with two caveats. The first is that it was
limited to utterance and argument honorifics; role honorifics and honorific pronouns
were left aside. The reason for this will come clearer in later chapters, but, briefly stated,
is that I take the honorific content of such cases (in Japanese) to involve an inferential
process which is somewhat too complex to lay out at this point in the discussion. This
situation therefore will be addressed in later chapters. The second caveat is that data was
drawn exclusively from Japanese. This point will be rectified now, with the introduction
of data from Korean, Thai and other languages.
My strategy will be to roughly duplicate the tests shown above for Japanese utterance
and argument honorifics, though for some of the languages I will consider I will
be working from published sources, so running the whole battery of tests will be
impractical. I will here focus on the independence property, using (where possible)
the embedding, ellipsis, and denial tests; the other properties either relate closely to
independence, as with nondisplaceability, or are relatively conceptual, as with ineffa-
bility. I will mention or discuss this data where appropriate as well. My presentation of
this data will be nonexhaustive, but I hope the reader will be convinced, especially in
conjunction with the Japanese data above, that an expressive treatment of honorifics is
the right way to go.
We have already seen a bit of Thai data relating to the independence test, so let me
just provide a bit more. Consider the following case, in which a suppletive honorific
is embedded in a conditional construction. Here we see that the honorific meaning
carried by the suppletive survives outside the conditional, as expected if honorifics are
expressive and thus independent of at-issue semantic operators.
(2.24) a. thaa mii somtam kɔɔ khruu rápbpràthaan mang
if exists papaya.salad then teacher eat.hon probably
‘If there is somtam, the teacher will probably eat it’ + ‘the speaker is being
respectful to the teacher’
b. thaa khruu chɔɔp ̂ rápbpràthaan somtam kɔɔ chan
if teacher likes eat.hon papaya.salad then I
ja sɯ:́
will buy
‘If the teacher likes somtam I will buy some’ + ‘the speaker is being
respectful to the teacher’
2.4 other possibilities 21
The same point can be made by considering the simplest available case of honorific
systems, that of the T/V system commonly found in European languages. Consider
the case of French, which has both formal and informal second person pronouns.
For the second person singular, French has the tu and the formal vous. Embedding
these pronouns doesn’t affect the honorific level of the sentence in which they appear;
instead, the honorific level indicated by the pronominal is carries over to the whole
sentence.
(2.25) a. ce n’est pas que tu n’es pas une bonne
it is.not neg that you.inf neg.cop not a good
personne
person
‘It’s not that you’re not a good person’ + ‘the speaker is treating the hearer
informally’
b. ce n’est pas que vous n’est pas une bonne
it is.not neg that you.hon neg.cop not a good
personne
person
‘It’s not that you’re not a good person’ + ‘the speaker is treating the hearer
formally’
2.4 Other possibilities
The data adduced in this chapter led to the conclusion that honorific content should
be viewed as expressive. But there are other possibilities. The aim of this final section
is to consider, and dismiss, these possibilities in turn.
The first is the simplest to put aside. Could we simply view honorific content as
truth-conditional? No: it bears few or none of the hallmarks of truth-conditional
content. Truth-conditional content lacks the independence property: embedding such
content under a semantic operator places it into the scope of that operator, something
we have already seen does not hold for honorifics. Further, honorifics don’t affect
truth conditions themselves: adding an honorific to a sentence doesn’t change the
circumstances under which the sentence is true or false, merely those under which it is
appropriate, or simply eliminates some effect of the use of the honorific. For instance,
the sentence (2.26a) would never be true in a circumstance where (2.26b) was false,
though the latter would be inappropriate in most circumstances given standard social
norms. Taking honorific content to be truth-conditional is essentially a nonstarter.
(2.26) a. hanzaisha-wa ki-ta
criminal-top came-pst
‘The criminal came’
b. hanzaisha-wa irasshai-mashi-ta
criminal-top came.hon-utt.hon-pst
‘The criminal came’ + ‘the speaker honors the criminal’
Can it then be thought of as conversationally implicated? Such a view seems

implausible: honorific content is directly connected to the use of honorifics themselves,
rather than to the ‘larger’ properties of discourse contexts or speaker cooperative
behavior which give rise to conversational implicatures (Grice, 1975; McCready, 2015).
Meanings which arise from the choice of particular lexical items, and invariably
arise in association with those items, are properly viewed as conventional, which
is precisely what conversational implicatures are not. Further, since conversational
implicatures are not conventional, they can be canceled in further discourse, whereas
conventionally expressed meanings cannot be. Honorific meanings are not cancelable
either. These mismatches suggest again that conversational implicature is not a viable
candidate for the meanings of honorific items.⁶
We are left with a choice between presupposition and expressivity. One might
first wonder exactly what the difference between the two consists in. Folk wisdom
has the two categories completely separate, but recent research raises two questions:
(i) whether the two are fully distinct, or just points on a continuum (cf. Simons
et al., 2011; Tonhauser et al., 2013), and (ii) whether there is a distinct category
of conventional implicature as opposed to expressive content. I will assume that
the answer to these two questions is negative for present purposes: whether or not
expressive content can be viewed as closely related to presupposition, it shows different
projection behavior in general, and the best way (if any) to draw the boundary line
between conventional implicature and expressive content is orthogonal to the goals of
this book. Thus, in order to determine whether honorification is expressive, we should
examine its projection behavior.
What exactly is the difference in projection behavior between presupposition and
expressive content? The clearest difference lies in the degree to which the content is
‘trappable’ or ‘bindable’: does it always project in conditional or universal construc-
tions? Consider the following three examples of presupposition. (2.27a) presupposes
that Alicia has a daughter (due to the possessive noun phrase); (2.27b), where the
possessive is in the antecedent of a conditional also does, as does (2.27c), where
it is in the consequent. So far this is all basically the same as what one finds with
expressive content. The interesting case is (2.27d), where the presupposition appears
in a conditional consequent and its content is entailed by the antecedent of the
conditional. In this case, the content of the presupposition does not project. This differs
from expressive content, which is always basically immune to being trapped in this
manner.
(2.27) a. Alicia’s daughter is a doctor.
99K Alicia has a daughter
⁶ Caveat: later in the book I will claim that some honorific meanings do indeed arise via reasoning about speaker
intention in conjunction with social facts. However, there, the meanings conveyed by the (so-called) honorific
items are not solely honorific in nature, instead expressing aspects of social role and speaker presentation. These
meanings are in no sense cancelable.
2.4 other possibilities 23
b. If Alicia’s daughter is a doctor, Alicia probably isn’t too concerned about her
future.
c. (I don’t know that much about Alicia, but) if Alicia is a helicopter parent,
then Alicia’s daughter is probably a doctor.
d. If Alicia has a daughter, then Alicia’s daughter is probably a doctor.
99K
/ Alicia has a daughter
The question now is whether honorifics pattern with presuppositions or with expres-
sives in this area. Unfortunately, this question cannot be directly posed because the
necessary conditions for the tests are not met. Consider the ineffability property. This
property states that expressive content cannot be paraphrased using only nonexpres-
sive content. If so, there is no way to construct an example like (2.27d): any attempted
paraphrase will fail, so the expressive content in the consequent will necessarily project
as it won’t be trapped. Using an expressive paraphrase won’t help as the result will
simply project in the way expected from (2.27c), where there is nothing to bind the
content in the antecedent. The direct approach won’t work. But perhaps there’s a less
direct strategy that can be applied. The appropriate use of honorifics relies on particular
social relations as reflected in the context of use, a view that will be formalized in the
next chapter and thereon. Perhaps, by constructing a context in which an honorific
is inappropriate but setting up conditions under which it would be in a conditional
antecedent, we can test whether projection occurs.
Thus, consider the following case. Here, the context is informal (postparty cleanup),
and the relationship between the conversational participants is also informal; but,
if the situation described in the conditional antecedent were true, the employee
addressed would be extremely socially distant, so using the suppletive honorific in the
consequent would be appropriate. This is an instance of what we might call situational
trapping: a context is created to trap the honorific content.
(2.28) Context: a company function where employees are forbidden to drink leftover
beverages, which are stored for subsequent functions. Two new employees
have been instructed to clean up; both want to open a bottle of champagne.
a. omae-ga shachoo dat-tara kore-o meshiagaru
you.inf-nom president cop-cond this-acc drink.hon
deshoo
would.hon
‘If you were president, you surely would drink this’ + ‘If you were president,
I would be honoring you’ (intended)
But does it work? No. My informants judge this example infelicitous or at least as
performing a far higher degree of honorification than necessary, yielding a jokey or
sarcastic effect. The situation in the antecedent fails to trap the honorific content;
if situational trapping is indeed a technique which properly duplicates for honorific
content that shown for presuppositions in (2.27d), this example shows that honorific
content projects in the way expected for expressives, not as with presuppositions. I take
this to be further evidence for an expressive treatment.
2.5 Conclusion
This chapter has argued for an expressive treatment of honorifics. I began with a critical
discussion of the characteristics of expressive content proposed by Potts (2007), after
which I showed how they apply to honorific content in Japanese. After exhibiting some
data from other languages, I argued against other possible treatments in terms of other
kinds of content such as presupposition. The conclusion of the whole discussion is that
treating honorific content as expressive is the best empirical option. In the remainder
of this book, I will build a theoretical account of how honorification works which takes
an expressive treatment as a starting point.
3
A theory of register for honorification
The previous chapter showed that honorific meanings should be regarded as

expressive, by examining how they behave with respect to the properties of
expressives proposed by Potts (2007), which were also critically discussed. This
conclusion, as also discussed there, is in line with the vast majority of the literature
on the topic. However, despite the extensive cross-linguistic support for this
conclusion, most of this work does not attempt to seriously propose denotations for
honorific meanings, instead using dummy expressions like 𝜆x[honor(s, x)] to indicate
honorification, and showing how these expressions play out in composition, which
is the focus of most work on expressives (here s denotes the agent of the utterance).
This particular example is taken from McCready, 2010b; Watanabe et al., 2014), but
it exemplifies something rather ubiquitious. The only two exceptions are Potts and
Kawahara (2004), a discussion of which will occupy the first section of this chapter,
and Portner et al. (2018), which I will discuss after presenting my own theory.1 The
former work is the starting point for the full semantics of honorifics I will present.
Given the empirical picture presented in the previous chapters, we can now consider
what is needed for a formal semantic analysis of the honorifics. First, many languages
have honorific forms which reference the current discourse context, specifically the
relationships which hold between the various contextual agents. This means that any
semantics for honorifics must provide a model of a discourse context which makes
available the requisite formality relationships and relativizes them to agents; further,
given that honorific use can evolve over a discourse, it is necessary to make whatever
contexts are introduced dynamic in a way that tracks patterns of honorific use. Second,
cross-linguistically we find honorific forms which reference gender and societal role;
a semantics for these honorifics must make such reference available as well, with
the proper sort of expressive meaning. However, we find linguistic differences in
certain patterns of use: for example, as we will see in Chapter 7, Thai allows only joint
use of pronouns in the same or contiguous levels of formality, but in Japanese pronouns
from distant levels can be used together. The theory must make allowances for the
existence of these kinds of facts as well.
The remainder of the chapter will be devoted to providing a semantics and prag-
matics for the honorific expressions discussed in this section, as situated in a general
model of honorification and politeness in general.
1 I became aware of this work quite late in the process of preparing this book.

26 3 a theory of register for honorification
3.1 Previous formal analyses of honorifics
We begin the discussion with a consideration of the work of Potts and Kawahara
(2004). Potts and Kawahara provide a compositional, type-theoretic semantics
for a limited domain of honorifics set within a Potts-style compositional seman-
tics for expressive items, essentially a system of composition rules together with a set
of dedicated semantic types for different sorts of content (see Potts, 2005; McCready,
2010b; Gutzmann, 2015 for details of such systems, and Section 3.3 for an overview).
The analysis begins with the proposal of a new expressive type 𝜀, which denotes
relations between individuals and attitudes. These attitudes are expressed by real-
numbered intervals, I ⊆ [−1, 1], which indicate positive (0 <) and negative (< 0)
attitudes in the obvious way. These intervals correspond to attitudes which relate two
individuals, and thus have the form aIb.
(3.1) a. dt[−1, −.9]bo ‘Donald Trump feels very negatively about Barack Obama’
b. dt[.6, .9]rd ‘Donald Trump feels quite positively about Rodrigo Duterte’
c. em[−1, −.9]dt ‘The author of this book feels very negatively about Donald
Trump’
These intervals are used to model the meanings of both honorifics and expressive
adjectives like damn and fucking.
(3.2) a. [[damn]] = 𝜆x.s[−.7, −.3]x ∶ ⟨e, t⟩c
b. [[fucking]] = 𝜆x.s[−.8, −1]x ∶ ⟨e, t⟩c
The combinatorics of the 𝜀-types follow the usual Pottsian rules for composition,
which ensures that they are independent of semantic operators; for details of the rules,
see the works cited above and also §3.3.
Potts and Kawahara provide the sample denotation in (3.3) for a Japanese subject
honorific. Subject honorifics are taken to denote functions from individuals to expres-
sive types, and to state that the speaker s has a highly positive attitude toward x, as
indicated by the closeness of the interval to 1, and by its specificity.
(3.3) [[SH]] = 𝜆x.s[0.8, 1]x ∶ ⟨e, 𝜀⟩
Unfortunately, this view cannot be quite right. On this semantics, emotive attitudes and
honorification are conflated, so that the subject honorific has a meaning close to the
positive interpretation of damn (or even the stronger fucking). McCready (2012b) is an
extensive exploration of the underspecification of such emotive expressive adjectives;
according to this work, the interpretation of the emotive is fixed by interpreter
reasoning about the speaker’s likely communicative intention on the basis of world
knowledge, including metalinguistic knowledge about how language is used and how
communication proceeds. On the positive interpretation of a strong emotive expressive
like fucking, it could be assigned a denotation which is the inverse of (3.2b), which is, of
course, precisely the denotation in (3.3), and simply says that the speaker has a strongly
positive attitude toward the object of which the adjective is predicated.
3.1 previous formal analyses of honorifics 27
But honorifics and emotive adjectives clearly are semantically distinct. Speakers can
use politeness markers without having any kind of emotive attitude at all, or even when
they have a negative one; it matters only that they are in a context which specifies formal
usage with respect to the ‘honored’ argument. Such contexts can, of course, involve
genuine emotive attitudes (and in certain circumstances positive implicatures do seem
to arise, depending on the honorific and the context), but honorification can also be
emotionally neutral, as when one uses honorifics to refer to a universally despised boss
in a formal setting.
The waters are muddied somewhat by pragmatic factors: using an honorific with
respect to someone present in the utterance situation can implicate positive attitudes,
mainly due to the expectation that if someone treats another as socially superior in
some sense, that person is worthy of respect, which in turn has a positive emotive
quality. Such implicatures can be controlled for by considering uses of honorifics which
are stated with respect to third parties, as with the Japanese argument honorifics.
An example like (3.4) does not indicate a positive attitude toward the teacher, but only
marks his or her status.2
(3.4) Tanaka-sensei-ga irasshai-mashi-ta
T-teacher-nom come.subjhon-pol-pst
‘Professor Tanaka came. + the speaker honors Prof. Tanaka’
Definitions of the kind in (3.3) have the drawback of only indicating an attitude
toward a specific individual. The facts about honorification we have seen are a bit
more complex: they seem to jointly indicate the speaker’s social level with respect
to a particular individual, and also to indicate the speaker’s assumptions about the
formality of the context of speech, a point which will be discussed further in the next
section, where a theory of honorific register is proposed which makes reference to
multiple dimensions of social position and distance.
Finally, when honorifics are used, they change the context; the speaker indicates a
particular level of formality (perhaps with respect to some individual, as in (3.3) above).
This point is neglected by Potts and Kawahara (2004), but Potts (2007), which is partly
an extension of the Potts and Kawahara approach, models it by assuming that discourse
contexts contain a set cI of indices of the form aIb, as above. This set can be updated
by a newly introduced index aIb in two ways: (i) if cI does not contain any index of the
form aI′ b, then c′I = cI ∪ {aIb}, and (ii) if it does contain such an index of the form aI′ b,
then aIb replaces aI′ b, where it is also required that I ⊆ I′ .
′
(3.5) c𝜀 ≈Ia,b c′𝜀 iff c𝜀 and c′𝜀 differ at most in that
a. aIb ∈ c′𝜀 ; and
b. if c𝜀 contains an expressive index aIb, where I ≠ I′ , then aIb ∉ c′𝜀 and I ⊑ I′ .
(Potts, 2007;11)
2 Thanks to Ryan Hearn and Will Starr for discussion here.

This last clause is problematic in that it certainly seems possible to indicate altered
attitudes as opposed to simply further specifying existing ones, something ruled out
by the requirement for inclusion. Indeed, that requirement eliminates the possibility
of change in attitudes over time, something which is key in the pragmatics of honori-
fication; consider the shift from formal to informal forms in languages with the T/V
distinction, for example (e.g. Brown and Gilman, 1960). Another difference from the
Potts–Kawahara treatment is that Potts (2007) now analyzes the indices as introducing
primitive relations, as opposed to always being emotive (implying that they can be
incompatible with one another). This is a move forward, but since the content or source
of these relations is never spelled out, it is hard to be fully satisfied with the analysis.
However, a way to model the dynamic character of honorification is certainly necessary
for a complete picture, so this part of the analysis is also in need of improvement,
though it provides a solid foundation.
A fully adequate semantics for honorifics and politeness markers must satisfy the
following criteria. Given the force of the above arguments that honorific meaning is
expressive, the proposed meanings must be expressive in nature, both in denotation
and in terms of the means by which they compose with other content; they must,
of course, also yield the intuitively correct meanings. Further, the result of semantic
composition must be able to support analysis of the rational use of honorifics and
politeness markers in communication. The proposals of Potts and Kawahara (2004)
and subsequent related work do not appear to fully satisfy these criteria, for they equate
honorific content with emotive attitudes, which results in the wrong intuitive sincerity
conditions for honorific use.3 Still, the notion of scales of politeness and the general
notion of expressivity at play seem highly useful; I will take them as a starting point for
my proposal, which is given in Section 3.2.
3.2 A formalism for register
To give a semantics for honorifics it is first necessary to decide the domain of meanings
over which they operate, and the kinds of effects which they have. My strategy here will
be to make use of observations from the literature on politeness and honorification,
some drawn from the discussion in Chapter 2, modeling them in a formal setting that
carries forward the best aspects of the work of Potts and Kawahara.
Iwasaki and Ingkaphirom Horie (1995) propose that politeness behavior in Thai
operates along three dimensions: psychological distance, social distance, and formality.
Psychological distance is the perceived interpersonal closeness of the discourse partic-
ipants; for instance, friends have a higher degree of closeness than work colleagues in
the general case. Social distance is determined by the societal roles of the participants;
3 It also seems to give wrong results when input to game-theoretic analysis, for strategic use of honorification
should not depend on the expression of emotivity; since such uses are not the focus of this book, I will not push
this point further here, but see Chapter 8 for more discussion of this aspect of the use of honorifics.
3.2 a formalism for register 29
there is a smaller social distance between a sanitation worker and a plumber than
between a plumber and the president of a large corporation. Formality is determined
by the situation of utterance together with the purposes and topic of the conversation;
a funeral is more formal than a drinking party, but even the latter may require formal
speech if I plan to ask you to lend me a substantial sum of money while we drink. I will
adopt this view of how politeness levels are determined here. The three dimensions are
obviously not completely independent (for instance, psychological distance will often
be partly a function of social distance), but for the purposes of the present work I will
treat them separately. The exact manner in which they interact is an empirical question
too complex to address here.
These considerations prompt the use of denotations for honorific expressions which
reference these three dimensions. I will thus take the domain associated with the
semantics of honorifics to be the set of real-numbered subintervals of [0, 1], which
is itself determined by a 3-tuple of such intervals ⟨P, S, F⟩, where each element is
associated with a range of politeness. Each element of these tuples corresponds to
one of the aspects of politeness determination discussed in the previous paragraph:
psychological distance, social distance, and formality.
(3.6) Base domains for politeness
𝒟𝜀 =df {⟨P, S, F⟩ | X ⊆ [0, 1] for X ∈ {P, S, F}}
This follows Potts (2007) in that it makes use of real-valued intervals, but differs in three
respects: (i) I assume a multidimensional domain for honorifics rather than a single
real-numbered interval, (ii) this domain, while a real-numbered interval as in Potts’s
work, inhabits the space between 0 and 1, as I take it that it does not make sense to have
a negative degree of (e.g.) social distance, and, of course, (iii) the conceptual basis of the
interval is different, as Potts takes honorifics to reference emotivity but I take them to
reference psychological distance, social distance, and formality. These three differences
entail that honorific denotations are distinct from what is found in the emotive domain
of e.g. expressive adjectives, which was shown to be desirable in the previous section.
The key points here are (i) and (iii). The empirical claim is that the three factors
represented here are the factors relevant for determining the use of honorifics, and the
social hierarchies that are (sometimes) relevant to their selection; but, since the basis
of these domains is multidimensional, more dimensions can be added if they prove to
be empirically necessary.
To analyze register, I will make use of the notion of discourse context. In semantics
and pragmatics, contexts are often taken to be sets of worlds or other elements, as
with the sets of attitudes utilized by Potts (2007) and discussed above. For honorifics,
I will take contexts to simply indicate the formality of the current discourse situation.
Situations can be distinguished in terms of formality at an extremely fine-grained level,
so they should be analyzed using continuous techniques; I take this to mean that they
too should be viewed as subintervals of [0, 1]. The exact range of a given context
is determined by the three factors mentioned above. So contexts 𝒞 have the form
⟨P, S, F⟩, each a subinterval of [0, 1], where higher intervals are associated with more
polite contexts, and ‘narrower’ intervals with more specific requirements for politeness
behavior. Update is carried out with respect to registers R, which are also subintervals
of [0, 1], derived from contexts as follows. The functions min and max pick out the
lower and upper bounds of intervals, respectively; the definition of discourse contexts
makes use of the projection functions 1, 2, and 3, which pick out the corresponding
elements of n-tuples and will also be used extensively below.
(3.7) a. min([i, j]) = i
b. max([i, j]) = j
(3.8) Discourse contexts for honorification
min(1(𝒞))+min(2(𝒞))+min(3(𝒞)) max(1(𝒞))+max(2(𝒞))+max(3(𝒞))
ℛ𝜀 =df [ , ].
3 3
The relationships between and weighting of the three factors is an empirical question.
I will simply assume that the three together determine a range of appropriate speech,
as it does not seem to be the case that honorifics directly reference these factors in
general but rather to the general register derived from all of them. This view could of
course be wrong, however; my use of a particular honorific may relate in principle to
psychological or social distance, or to the formality of the speech situation. In this case
the formalism makes it possible to make reference to the particular elements P, S or F;
it seems worth making this possibility available.⁴ Equating the current context with the
current politeness register yields a notion of ‘global register’ ℛ.
(3.9) Global register
ℛ =df ℛ𝜀 .
Thus the appropriate level of formality for a discourse context is derived (indirectly)
from the interpersonal and social distances of a context and its formality, and is itself
a subinterval of [0, 1].
With the above, the discourse context specifies an interval corresponding to a
formality level. But how should this tie to the use of the honorifics themselves? In
the previous discussion, expressions with honorific content—particles and pronouns—
were separated into three general levels of politeness: low, mid, and high. If one wants to
reference these levels directly, it is possible to define intervals corresponding to them,
as follows.
⁴ One might worry that this complexity is unnecessary, and that we might as well just have a single unanalyzed
interval determined by a black box referencing the mentioned factors. I do not go this route for two reasons: first, I
think it is a positive feature of the analysis to make the underpinnings of the model part of the theory, even if they
don’t play a significant role in the empirical analysis, both for reasons of transparency and also because (the second
reason) I think it remains to be seen whether the individual elements P, S, F don’t do any empirical work. This
book can analyze only a small subset of the world’s honorifics: but perhaps there are honorifics which reference
only one or the other factor, in which case they must be made available. Further, I think there’s interesting work
to be done in determining how discourse situations and contexts influence the weighting of the three factors in
determining global register. At a funeral, presumably the context outweighs personal relationships, at least when
others can overhear; but if no one can hear, perhaps the opposite holds. This kind of question might or might not
be linguistic, but to the extent that it has an influence on language use it should be representable in the model.
(3.10) a. High ⊆ [.6, 1)

b. Mid ⊆ [.3, .7]
c. Low ⊆ [0, .4]
Note that the categories overlap: High and Mid share [.6, .7] and Mid and Low share
[.3, .4]. The reason is that these forms are compatible in Thai: it is possible to use High
and Mid forms together, and the same is true for Low and Mid forms. However, doing
so indicates a relatively specific degree of formality. The use of Mid and High forms
together means that, while the speaker does not take the context to be an extremely
formal one, it is still relatively formal. This suggests that honorific use ought to be tied
closely to speaker assumptions about the nature of the discourse context, which appears
correct.Why not just use these categories and put aside the more complex formalism
that underlies them? The answer lies in how honorifics act on the discourse context.
To analyze these effects, a discrete system will be insufficient.
Now we are ready to consider the denotations and discourse effects of the honorifics
themselves. I will take honorifics to denote subintervals of ℛ, higher intervals for more
formal expressions, and lower intervals for less formal ones. The context will determine
whether a given expression is appropriate or not. Since these denotations are expressive,
appropriateness cannot be stated in terms of truth, but rather must involve conditions
of use. I follow Gutzmann (2012) in taking use-conditional judgements to involve two
values, ‘√’ and ‘×,’ indicating appropriateness and inappropriateness respectively.
(3.11) Appropriateness for honorifics
√ if Hon(DS) ∩ ℛ ≠ ∅
Utter(DS) in 𝒞 = {
× else
The above says that an utterance of a given discourse segment is honorific-appropriate
if its honorific level is compatible with the global register.⁵ This seems right, but
requires the derivation of the discourse segment’s honorific level. Recall that the
use of multiple honorific expressions in a discourse segment gives a different result
from using a single one; this means that honorific levels must be fairly nuanced,
but still derivable from the honorific levels of the expressions involved. However,
since denotations are expressive, we need not worry about interactions with semantic
operators (Potts, 2007). Thus it will be sufficient to take the average of all expressions
used in the sentence, with the proviso that their denotations also be compatible (in
order to rule out illicit combinations). This last condition serves to implement the
observation made for Thai by Iwasaki and Ingkaphirom Horie (1995), according to
whom high- and low-level items cannot be used together, though combinations of
high- and mid-level items are possible, as are combinations of and mid- and low-
level items. This is predicted in the present theory, as only adjacent speech levels have
⁵ I make use of discourse segments here rather than sentences mainly because sentences can be of arbitrary
length, which opens the possibility of having extremely long sentences; in such cases, (3.12) can give strange
results. Using discourse segments instead (which correspond to clauses in the case of long sentences) eliminates
the issue. Thanks to Benjamin Bruening and Satoshi Tomioka for discussion here.
nonempty intersections. (3.12) defines the honorific level of a discourse segment with
n honorifics.
(3.12) Honorific level of a discourse segment
min(1)+⋯+min(n) max(1)+⋯+max(n)
Hon(DS) = [ , ] if Hon1 ∩ ⋯ ∩ Honn ≠ ∅, else
n n
undefined.
The above seems a reasonable characterization of how the appropriateness of a given
honorific will be determined. If the context is formal, use of an extremely informal
pronoun will be inappropriate; in the context of casual speech among friends over
drinks, extremely formal pronouns will sound very unnatural. More detail will be
provided in later sections in conjunction with the semantics of particular honorific
items in Thai and Japanese.
This proposal also is able to account for changes in honorific use over the lifespan
of a conversation or long-term social interaction. It is well known that, in many
social situations, one tends to begin speaking formally and then move to informal
speech. This is reflected in the use of honorifics: often, formal pronouns and other
markers are initially used, and then at some point speakers jointly move to the use of
informal markers.⁶ In the present context, it corresponds quite simply to a change in
the parameters determining 𝒞: for example, as the measure of interpersonal distance
becomes smaller, a corresponding diminishment of the value of ℛ occurs, given
sufficiently low values for formality and social distance (i.e. a context which does not
automatically specify formal speech). Honorific use thus depends on external social,
parameters in the expected manner.
Several issues remain. First, while ordinarily changes in speech level are determined
by the external context (or so the model above has it), it is also the case that the use
of honorifics can impact the formality level of the discourse continuation. Specifically,
there are points at which it is obvious that the speech level should be changed; but
sometimes the use of an informal form causes a switch to an informal level, although if
the informal form had not been used, the level would not have changed. This is a kind
of performative effect and should be captured by the semantics. However, at present
the semantics simply assumes that the level of the honorifics is checked against the
context, and makes no provision for honorific-induced context change.
In the present theory, this observation can be made more concrete. Suppose that a
discourse segment DS with politeness level Hon(DS) is used in context 𝒞 determining
register ℛ. Then two cases arise. In the first, Hon(DS) ∩ ℛ ≠ ∅. In such a situation, DS
is deemed appropriate. The discussion so far has focused on case 1. In case 2, Hon(DS)∩
ℛ = ∅. Here, use of DS is inappropriate. But the use of DS can also serve as a proposal
to modify the context to one in which DS would be appropriate after all. In essence,
⁶ This situation has been analyzed by McCready et al. (2013) for the binary tu–vous distinction on second
person pronouns common in European languages, and for a Japanese honorific pronouns by Asher and McCready
(2013), using the tools of infinitely repeated games and topological analysis of strategy complexity. I will return
to these issues in Chapter 8.
the use of DS aims to move ℛ upward or downward in a way that makes Hon(DS) an
appropriate honorific level.
How should this process be modeled in the formal theory? One option is to allow
honorifics to modify the context directly and dynamically via their use. For instance,
a use of the Thai politeness particle khráp could be taken to preemptively change
the context to a formal one, irrespective of what it was formerly.⁷ However, this
view would seem to obviate the analysis so far, in that the definition in (3.11) would
become obsolete; since the use of khráp would change the context to one in which
khráp was appropriate, we no longer have any means to model inappropriate use of
honorific elements.⁸ Instead of allowing such extreme changes, I will model honorifics
as proposals to change the context in an incremental manner, if they were originally
incompatible.
The basic idea is to take honorifics to, as before, denote subintervals of [0, 1],
which are checked for compatibility with the register currently specified by the context.
However, the performative character of honorifics functions as a proposal to change
the register to one compatible with the honorific level. Thus, use of a formal particle
like khráp proposes raising the level of formality, and a particle associated with casual
speech like wóoy functions as a proposal to lower the register. But this register shift
cannot be completely unrestricted, as discussed in the previous paragraph. It should
be tied to the current formality of the context. I propose the following shift, where
ℛ[(DS)]H signifies ‘honorific update’ of the current register with the honorific content
of a discourse segment, ℛ′ is the register arrived at after such update.
(3.13) Dynamic registers
Let 𝒞 = ⟨P, S, F⟩. Then:
𝒞[(DS)]H = 𝒞′ , where
𝒞 if 𝒞 ⊆ Hon(DS)
𝒞′ = { ,
⟨P′ , S′ , F′ ⟩ otherwise
min(P)+min(Hon(DS)) max(P)+max(Hon(DS))
P′ = [ , ]
2 2
′ min(S)+min(Hon(DS)) max(S) +max(Hon(DS))
where S = [ , ].
2 2
min(F)+min(Hon(DS)) max(F)+max(Hon(DS))
F′ = [ , ]
2 2
This formula simply averages the honorific content of each element of the current
discourse segment with the corresponding element of the current context (i.e. the
current register) unless the honorific content is less specific than the current context.
Note that this generalizes the proposal of Potts (2007), who allows only restriction
to subintervals (with an emotive interpretation, of course). In case of change, both
previous context and current utterance are given equal say in the ultimate register.
This is the simplest option, which can of course be weighted as required by empirical
observation. Note that this is a proposal, which can be rejected by the hearer, just as
⁷ As we will see shortly, this is more or less the strategy adopted by Portner et al. (2018).
⁸ Of course, external constraints could be placed on the update mechanism, but this seems inelegant.
with other update operations (Stalnaker, 1978; McCready, 2015). The result of this
operation is used to check the appropriateness of an utterance via (3.11). Many detailed
derivations will be provided in later chapters, but for a simple and abstract example,
consider an utterance of a sentence including a single honorific, say the Japanese
utterance honorific desu, which (for the sake of example) can be taken to mark a level
of [.6, 1] in a relatively formal context, say a communication by a new employee to an
immediate supervisor, which is associated with a register of [.7, 1]. The formula above
will yield a new context, each element of which is the result of averaging the relevant
element of the discourse segment and the original context. Mapping this new context
to a register yields the new register ℛ′ = [.65, 1].
One lacuna in the discussion of the dynamic effect of honorifics so far is the fact
that, in certain cases, they can in fact function to change the context in more abrupt
ways. Consider the fraught change from using formal pronouns to informal pronouns
in European languages with a T/V distinction, or the choice to use or avoid honorifics
in Japanese or Thai. In certain cases, the context will be compatible with both; but in
others, a speaker can simply hazard a change to informal speech, hoping that it will be
acceptable to the addressee. Sometimes this gamble will fail, and the proposal to move
to an informal register will be rejected, but sometimes it will succeed.
The analysis so far only considers incremental change in honorific use and how the
choice of honorifics incrementally and gradually changes the context; ‘inappropriate’
speech levels are defined above as those which lack an overlap with the current register.
This accounts for a wide range of cases, and indeed quite generally for ‘safe’ honorific
use. But the performative character of honorifics allows for nonincremental change
as well. One can sometimes ‘float’ an honorific which is far more informal than
safely allowed as an attempt to make the discourse more casual (and consequently
try to deepen the personal relationship between the participants, as reflected in the
corresponding change in P), or use one which is more formal than would usually
be judged appropriate, for example to signal a ‘stepping back’ from the discourse
participants or discomfort with the discouse content. These pragmatic effects should
be addressable in the formal theory, or at least able to be grounded there. To account
for this, I propose the ‘reset rule’ in (3.14).
(3.14) Reset rule
𝒞[(DS)]H = Hon(DS) if ℛ𝒞 ∩ Hon(DS) = ∅.
‘If a speaker uses an honorific with no overlap with the current register, the
register is reset to that denoted by the honorific.’
Of course, this is merely a proposal by the speaker, and one which can be rejected by the
addressee, just as in any other case of dynamic update. But it allows us to understand
the occasionally abrupt changes in register induced by the combination of incautious
speakers and liberal addressees.⁹
⁹ More subtle takes on this operation are possible: we could, for instance, allow ‘abrupt updates’ to take the
average of the current register and the incompatible honorific level. I will not present these possible alternatives
in this book.
The reader will have noticed that this proposal, while it succeeds in capturing
the intuitive effects of honorifics (though further complications will be introduced
later), still fails at this point in adequately capturing their usage. The reason is that
the context only makes available a single register for honorific use. This predicts that
all conversational participants will use the same level of formality, given that every
participant is working with the same context; i.e., only reciprocal registers are properly
modeled. But clearly this prediction does not accord with the empirical facts: it is
extremely common for individuals to speak in different levels of formality when there
is a power differential among the conversational participants. Different agents must
therefore be associated with different levels of formality. Every conversation should
therefore make use of at least two distinct contextual representations, something
already expected from formal pragmatic work on context (Gunlogson, 2003). Still
more will be required when conversations involve more agents; ultimately, contexts
as described here must be lifted to context sets, where each agent is associated with a
distinct context, and such contexts represent each agent separately.
Carrying this task out is rather straightforward, given the discussion so far. Potts and
Kawahara (2004) actually do something quite similar to what is needed by associating
honorific meanings with triples aIb, an attitudinal relation of agent a to individual b.
Contexts for them are then sets of such tuples. Which member of the context is picked
out by a particular use of an honorific appears to be lexically specified; for instance,
the lexical meaning of a subject honorific given in (3.3) indicates that the speaker’s
attitude is at issue. Here, I would like to generalize this idea and claim that honorifics
are set to the register relating speaker and hearer(s) by default. For the subject case, this
parallels closely observations that have been made in the linguistics and philosophy
literature about other kinds of ‘subjective’ predicates such as predicates of personal
taste, epistemic modals and the like (e.g. DeRose, 1991; Lasersohn, 2005; Egan, 2006;
Weatherson and Egan, 2011); for the addressee case, the move is similar, though it does
not seem to have direct parallels.1⁰
For the necessary formal base, we can allow each agent to be associated with a set of
registers of the kind proposed above, except now given the form ℛ⟨a,b⟩ and construed
as measures of a’s sense of her social relation with respect to b; each register in this
set will further be indexed to another (contextually relevant) individual. This gives the
following construction:11
1⁰ It is well known for epistemic modals that it is difficult to determine who is supposed to be included in
the relevant relation in a given discourse context, but similar problems arise for honorifics. This is clear in the
hearer case. In a two-person dialogue, plainly the hearer is being addressed, but as the number of participants
rise, it may no longer be obvious whose social roles are being considered during a given utterance. This problem
has been discussed extensively in philosophy in the context of worrying about whose information is relevant in
determining the appropriateness of the use of epistemic modals. For present purposes, I will just assume that
honorifics relate to individuals, and leave the group case with its attendant complications for another occasion,
though a fully adequate analysis must address it, for honorific use with respect to groups can differ greatly from
what might be used for their component individuals. Thanks to Miloje Despic and Will Starr for discussion here.
11 A reviewer asks if this approach results in the availability of exclusively symmetric readings: if e.g. ℛ⟨a,b⟩ =
[.2, .5], then is it required that ℛ⟨b,a⟩ = [.5, .2], as with simple distance measures, where if d(a, b) = .3,
then d(b, a) = .3 as well? I think not; here, the relations are specified individually, and are no more required
Another random document with
no related content on Scribd:
The State, here, amongst ourselves, had, throughout the whole of
this middle period, been asserting that it had a domain in which it
was supreme; that the Church had usurped a great part of this
domain, and was still endeavouring to extend its usurpations; and
that there could be no peace till the whole of this usurped ground
had been recovered. At last the State became sufficiently
enlightened and strong to establish its supremacy in the domain it
claimed; and to estop the Church from its usurpations. This was a
great gain. The work, however, was very far from having been
completed. What was done, though much, was in truth only a
beginning. What further was required was that the State should
forthwith address itself to the discharge of the high and fruitful duties
that belonged to the position it had assumed. But the fact was that it
did not yet fully and clearly perceive either what had become its own
sphere, rights, and duties, or what had become the sphere, rights,
and duties of the Church. Some, indeed, of the conceptions it formed
on these points were entirely erroneous, as both the teaching of
History—now better understood—and the inconveniences, the evils,
and the necessities of our present condition have since
demonstrated. The correction of these errors is a very important part
of the task of the present generation. The unsettled character of the
actual relation of the State and of the Church to each other, and the
resultant uneasiness and tenderness felt by each, and the way in
which, by these causes, each is at present crippled for much good it
might be doing, are to be attributed to these errors. These are
matters in which History is our only guide and interpreter. A
knowledge of the origin, nature, aims, and fortunes of this long
conflict in past times, enables us to understand its present position,
and to foresee its future course. We are at a certain point in a chain
of events: and nothing throws light on the events that are coming
except the events that have been now evolved.
When ideas, through their having been traditional for many
generations, have got a strong hold on men’s thoughts and feelings,
it is impossible to break away from them, and in some matters to
face in the very opposite direction, at a moment. Ideas grow, and
decay: they are not subject to instantaneous transformations, like the
figures in a kaleidoscope. This explains the partial acquiescence by
the State in the theory that the Church was only the State acting in
another capacity: as it were a committee of the whole House for
some politically necessary objects; and with an authority that must
be maintained. There was merely a colourable amount of truth in
this. Practically, and relatively to the condition society had reached, it
was a mistake; and one that was unworkable in every particular. The
Church, whatever might have been the case in the early stages of
society, is not now the State in another capacity. It has ceased to
have now any directly political objects. It has no authority in the
sense in which the State has: the authority of the State being such
as can be enforced by pains and penalties, and by physical
constraints; whereas the authority of the Church is only that of moral
and of intellectual truth—as much as, and no more than, it claimed
eighteen hundred years ago. In this matter its present advantages
are that it has not to contend for existence against hostile
established religions, and a consequently hostile tone of morality and
of society; for what is now generally recognized, in the moral order,
is precisely its own principles.
The logical and practical issue of this mistake was the
mischievous conclusion that the teaching of all morality, including
that which is necessary for the order and well-being of modern
societies, must be left exclusively to the Church; and that the State
must confine its own action to the repression of crime, and to the
protection of person and of property; and this only by the way of
punishment. Now each of these two propositions has, in a certain
sense, and from a certain point of view, though not those belonging
to these times, enough plausibility to enable a kind of defence of it to
be set up; but, at the same time, each contains such an amount of
real falsity to the existing circumstances and conditions of society, as
to issue in incalculable mischief both to the State and to the Church;
both in what it has caused, and is causing, to be done, and in what it
has hindered, and is hindering, from being done.
This was a mistake which assigned to the Church work, which
what have now become its constitution, its real objects, and the
means and forces at its disposal, incapacitate it from doing; and
which led the State to abdicate what is now its highest, and really
paramount, function. It put both the Church and the State in a wrong
position, and on a wrong path. It enfeebled, depraved, and shackled
both. It brought them into inevitable conflict with each other. It made
them both aim at what could never be more than very imperfectly
attained by the means they were respectively endeavouring to
employ. Its results were confusion, anarchy, and failure. Hence came
about the neglect by the State of national education. And hence the
claims of the Church to educate the nation. Hence the fierce
contradictions to these claims, expressed in a blind demand, as if
that were the only way of effectually contradicting them, for secular
education, that is to say, for the exclusion of morality from education,
and its limitation to an acquaintance with the instruments of
knowledge, plus a little physical instruction. This would make things
far worse than they are at present. It would be prohibiting the
acquisition, by those who are now the depositories of power, of the
knowledge and sentiments requisite for its right use. It would be
creating, and setting at work, in the midst of us, the most efficient
machinery imaginable for the general demoralization of the
community. It would be going some way towards transforming the
commonwealth into an aggregation of wild beasts, but of wild beasts
possessed of knowledge and reason. The concession of this by the
State would be the renunciation of its first and most imperative duty.
Hence, in short, all the imbroglio and the evils of the present
situation of this great question; and all the misunderstandings and
hot conflicts between those on the one hand, whom logic, working
with wrong data, has made secularists, but to the exclusion of
secular morality, the chief point of all, and, on the other hand, those
whose fealty to what is highest and best, and should be supreme in
man’s nature, even when regarded only as a political animal, has
obliged them to enrol themselves as supporters of (I am afraid we
must say internecine) denominational teaching in the education of
the people. It is obvious that, as it is the duty of the State to regard
the community as a single family, and to endeavour to bring its
members to act harmoniously together, it would be better, both
theoretically and practically, to exclude the inculcation of these
differences from the Schools of the State: that, if it must come, would
come with less evil from the denominations themselves.
But truth, reason, right, and History must in the end triumph. It is
the duty of the State, and we rigidly exact from it the performance of
it, to punish and repress crime: it must, therefore, be its duty, but this
we will not allow it to perform, to teach that kind of morality which
manifestly has a tendency to prevent the commission of crime. The
evil is done when the crime has been committed: à fortiori, then, it is
better to prevent than to punish it. It is the duty of the State, and we
energetically insist on its being discharged effectually, to protect
person and property: à fortiori, then, it must be its duty to teach that
morality which shall dispose men to respect the rights of person and
of property. It is the duty of the State to do what it can, within its own
sphere, to promote the well-being of its members; we may presume,
then, that it is its duty to teach that morality which shall have a
tendency, above every thing else the State can do, to secure this
great object. How can it be argued that the State does rightly and
wisely in neglecting the one means which stands first in the order of
nature, and which is emphatically the most efficient, for bringing
about its great paramount object? To deny that the means for doing
this duty are within its sphere, is to deny that it has any duty at all,
except that of punishing. Possibly such means may not be within the
sphere, as some define it, of the political Economist. But, though a
Statesman ought to be a political Economist, he ought to be
something besides. And it may be very bad political Economy to
allow in these days the mass of the people to be vicious. This may,
in the highest degree, be destructive of wealth. But, at all events,
what the Statesman has to lay his measures for is the well-being of
the community, of which wealth is only one ingredient; and which,
too, may be so distributed, and so used, and productive of such
effects and influences, looking at the community generally, as on the
whole not to promote its well-being. At all events, man, even when
regarded in his social capacity exclusively, does not live either by, or
for, bread alone.
The present condition of society is never to be lost sight of. And
the two most prominent elements of its present condition are the
general diffusion, throughout all classes, of political power, which
almost means that the decision of political questions has been
entrusted to the most ignorant and uninstructed, because they are
the most numerous, part of the community; and the fact that every
member of the community is now required to think, and to act, and to
take charge of, and to provide for himself. Here are two reasons,
which have made it as much the duty of the State to teach, as to
repress, and to punish; for knowledge, and this means pre-eminently
moral knowledge, has become quite as necessary to it for self-
preservation. Though, indeed, punishment is a mode of teaching,
and the policeman and the magistrate are a kind of teachers; but it is
as unreasonable, as suicidal, to have recourse to no other mode of
teaching, and to no other kind of teachers.
I think, then, that none but unstatesmanlike Economists will deny
that it is the duty of the State to see to the education of the whole
people. The Egyptian Priest, and the Hebrew Prophet, never made,
nor could have made, a mistake of this kind; to their apprehension
the right training of the people was the paramount duty of a
Government—the very purpose and object for which it existed. This
must, amongst ourselves, be given mainly in schools established
everywhere. We have now at last got so far as to attempt their
general establishment. The schools, however, are only machinery;
and the great question is, what kind of work this machinery is to do?
and the State will not discharge properly its duty in this all-important
matter, if it does not take care that the schools shall teach the
morality indispensably required, under existing conditions, for the
well-being of society. This morality means the principles of Justice,
Truth, Temperance, Honesty, Manliness, Forbearance, Considerate
Kindliness, Industry, Thrift, Foresight, Responsibility. These are
political and social, and perhaps also economical, necessities of
modern communities. They are now the first great wants of society.
Speaking generally, they can be taught to the masses of the people,
and to the whole people, best, and, in fact, only by the State. Every
one, I think, must be ready to acknowledge, that if the State, during
the last fifty years, had seen to their having been taught, so far as
schools and early training could have taught them, to the population
of this country, we should be in a widely different position—all the
difference being on the right side—from that in which we are at this
day.
It is just because the State has made, at best, only half-hearted
attempts to do any part of this work, and has even at times loudly
proclaimed that it saw that it was not its duty to undertake it, that is to
say that it was its duty to renounce its most important duty, that that
part of the community in which the moral instinct predominates, has
turned to Church organizations, and called upon them to undertake
it. And this is a reason why many of this class have been attracted to
that particular branch of the Church which advances, most loudly,
the most unqualified claims to the superintendence of the whole
domain of morality, not making any distinction between that which is
social, civil, and political, and that which belongs to the higher
sphere of the spiritual life. Had the State seen its duty in this great
matter, and endeavoured to act up to it, nothing of this kind would, or
could, have occurred. On the contrary: the wisest and best part of
the community would have supported it in carrying out what it had
undertaken, with their whole heart and soul.
Of course it is a mistake to look to the Church for this kind of work.
Neither the Church of Rome, nor any other Church, either in this, or
in any other, country, has the means necessary for enforcing this
kind of teaching, or even for bringing it home, generally, to the bulk
of the population, that is to say to the very part of it which most
needs it. Nor under any conjuncture of circumstances, which can be
imagined as possible, will they have the means for doing it. And
even further, if the powers necessary for the purpose could be
conferred upon them, it would be putting them in a false position to
call upon them to undertake this mundane, political work. Besides
that, the false positions into which events and circumstances have
already, more or less, brought all Churches, have so damaged their
credit with large proportions of the population, in all the foremost
nations of the world, as that their teaching of this kind would not,
generally, be received, would even be strenuously resisted; and it
would still further weaken them, were they to attempt to teach these
things for these purposes. It would bring them before the world as
mere instruments of national police—a position that is now so utterly
and glaringly at discord with the purpose and idea of a Church, that
its assumption would go a long way towards obscuring altogether in
men’s minds that purpose, and that idea; far too much in that
direction having been done already. We know how disastrous an
effect the assumption, to some extent, of this position has had, in
this and other countries, on some branches of the Church. This is
true now, and will continue to be so, till the Church shall have
become an organization in which all of us, laity as well as clergy,
women as well as men, who shall be animated by the desire for the
higher moral and spiritual life, shall find ready for us places and
work; and until, in this matter, the first effort amongst us shall not be
to secure this-world power, and social and political position, which
must always be accompanied by separations and antagonisms, and
is demoralizing, and destructive of the very idea of a Church; but to
reform and improve, and to lift above the world; an effort which is
actively and fruitfully moral, and of the very essence of the work of a
Church. This is truly spiritual work.
Taking things, then, as they are, any Church would be but a bad
and inefficient teacher of the political, we may even call it the
secular, kind of morality we are now thinking about. While every one
can see that, as it is an affair of the State, and comes within its
sphere, and is useful for its purposes; and as it is the duty, and the
interest, of the State to teach it; and as the State has, and alone has,
the power of teaching it, it might be well and properly taught by the
State. But it may also be remarked that no Church can afford to give
to this work of the State the first place in its thoughts and efforts.
Every branch of the Church, from the greatest down to the least,
must be occupied, primarily, by its own necessities. Self-preservation
is the first law of nature, in the case of Churches as well as of every
thing else that has life. The first care, therefore, as things now are, of
every Church must be to maintain and enforce its own system; and,
as part of the same effort, to weaken those whose systems are
opposed to its own. This, however disguised, must be a main object
with all of them. That it is so, is very disastrous for Churches; still it is
a necessity of their present position. And the efforts that arise out of
this necessity can, at the best, be only non-moral: in truth, one
cannot but think that they must generally be demoralizing, and even
immoral: at all events, they can only be made at the expense of the
higher morality, which is the true domain of the Church. But, however
much this point may be controverted, the other is an obvious fact,
and incontrovertible, that no Church has the power of teaching to the
community, and this is especially true of the most numerous and
least instructed part of the community, that morality which is now
necessary for the well-being of political societies. In this matter there
is a wide difference between past and present times. Formerly this
teaching, however desirable it might have been, was not
indispensable under the old restrictive and paternal systems of
society. All that has now passed away. We have drifted from those
moorings, and out of those harbours. Our population has been
agglomerated into large masses; and these masses have been put
into a position to exercise the power which resides in numbers.
Every one, too, is now called upon, and this is a most important
element in the consideration of what ought to be done, to take care
of himself. No class is now put in charge of another class. The moral
training, therefore, which these conditions require has become the
paramount object and first duty of the State; and, one way or
another, perhaps the highest personal mundane interest of every
member of the community; and all would do well to demand from the
State the discharge of this duty.
That the State should awake to a sense of its duty in this matter,
and act up to that awakened sense, would be no encroachment on
the domain of the Church. In so doing, indeed, it would set free, and
strengthen, the Church for its own proper work. The State cannot do
the work of the Church, any more than the Church can do the work
of the State. Each has now, distinctly, marked out for it its own
sphere, its own aims, its own rights, and its own duties. The world is
rapidly advancing to a correct understanding of all this. Each should,
properly, by attending to and doing its own work, help the other.
Each is necessary to the other. The morality the State has charge of
is that which, obviously, contributes to the right ordering and
prosperity of the commonwealth generally, and of its members
individually. It is such as can be expounded, and made intelligible to
all and acceptable to many. Much of it too can be enforced on all.
Not, of course, in the old Egyptian fashion, but in a fashion which is
in accord with the conditions of modern societies.
There can be few things more mistaken and ridiculous than to
urge that the Master of a School, because he is a layman, cannot
teach such morality as the State requires for its own maintenance,
and for the well-being of its members. He is just as capable as the
Minister of Religion, or as any body else, of learning his own proper
work. The point that really needs to be seen clearly is that the proper
work of the State School Master, and of the Minister of Religion, so
differ, as that each is incapable of teaching fully and rightly what
ought to be taught by the other. The Minister of Religion puts himself
quite in a false position, and contradicts the idea of his office, when
he undertakes the work of the State; and the School Master goes out
of his way, and passes beyond the work of the State, when he enters
on the ground of the Minister of Religion. From the time that civil
societies existed, or that men had come to act from a sense of duty,
all well disposed Fathers of families, not excluding Masters of
Schools, have deemed themselves qualified to teach, and have
taught, with more or less success, to their children such ethics as
they themselves had attained to a knowledge of, and thought
desirable. Let any one refer to the duties I just now enumerated, as
socially and politically necessary in these days; and, when he has
considered what they are, will he be disposed to assert that a man of
ordinary intelligence, the business of whose life it is to teach, whose
attention has been particularly directed to this subject, and who has
studied it with the knowledge that he must teach it, will, after all, be
unable to teach it? Or would any teacher, with that list in his hand,
say that it never would be in his power to give lessons on each of the
heads it contains; and to see that the practice of the pupils
corresponded with what he taught? If the Clergy could do this, why
not the Masters of Schools? The fact, however, is that the Clergy
cannot, and that the Masters of Schools can.
Nothing else that is taught in Schools can be taught so naturally,
so easily, and so surely. Almost everything that occurs, or that is
done, supplies ground for a lesson on the subject. In nothing else
that we have to teach do we find a foundation laid for our teaching
already, as it is here, in the instinctive moral sentiments which have,
some how or other, come to be, or, if not, which may be made to be,
a part of the pupil’s nature. The discipline, too, of life here again aids
the teacher in a manner, which is not the case in anything else he
has to teach. The Ethics the State requires may be taught, as the
occasion in any, and each, case will suggest to the teacher, either
practically, or dogmatically, or scientifically; either with a reference at
the moment to the principle of utility, or to the voice of conscience, or
to experience. Lessons of this kind may also be set forth in Parables,
or illustrative stories: a large proportion of the reading lessons now
used in Schools have this aim. Nor would there be many who would
object to reference being made, in the teaching of the State School-
Master, to the Religious ground, that is to say, to the future life:
though of course it is manifest that this would belong rather to the
teaching of the Church and of the Minister of Religion. Practically,
however, that is with respect to the substance and form of the virtues
taught, there would be no antagonism between the two: for even with
respect to Charity, which Religion elevates above Justice, the
layman would still have something to say in the same sense, for he
would show that the kindliness, and consideration for others, he
taught supplemented and went beyond Justice. Indeed, what
antagonism could there be, seeing that our ideas of the several
virtues, wherever they differ from what Aristotle or Cicero would have
taught, are what our Religion has made them to all of us alike? The
chief difference, indeed, I can make out would be a very small one,
for it would be the importance the lay-teacher would have to assign
to industry and thrift, secondary virtues of which popular Religion
does not take much notice: an oversight which, of course, arises out
of popular misapprehensions, such, for instance, as those we are all
familiar with in respect of the purpose and character of the present
life, of the meaning of faith, and of the teaching of Jesus Christ on
the subject of Divine interposition in the current affairs of life.
But, however, this little difference, though indeed it happens to be
one that must ultimately disappear, for it arises out of a
misconception, will help us to understand the difference between the
morality the State requires and that which the Church presents to us.
The former is limited to what is useful politically and socially, and for
mundane purposes; while that of which the Church has charge
(there being ultimately no real contradiction between the two)
consists of the same principles, only purified, elevated, and rendered
more fruitful by the action of higher motives. It is that which is in
thought perfect; the morality of the kingdom of God, that is of those
who have been brought to understand that they have a citizenship
which is not of this world, and whose conversation is above. It is that
morality which is cast in the mould of the ideas we endeavour to
form of the moral attributes of the Deity; or rather the application of
that to our own present condition: its members endeavour to form
God within themselves. This cannot be enforced. The idea of
constraint contradicts its nature. Its motives are found in men’s
spontaneously engendered conceptions of moral perfection; and in
the hope of a future life, which alone can supply a stage and
conditions suitable for the complete realization of such conceptions.
The rights of the Church are those of humanity to complete freedom
in its effort to advance and purify its ideal of the moral and spiritual
life. This has been its work from the beginning, though in the early
stages of society it embraced the State, and has subsequently often,
during the struggles of the State to establish its independence, been
in conflict with it: sometimes one, sometimes the other, sometimes
both having been in the wrong: all this History explains. Its true
position is to be in advance of the State. It elaborates and diffuses
that interpretation of man’s nature, and position, and of the
knowledge man has attained to, those conceptions of virtue and that
morality which the State, following in the wake of the Church, adopts
in its own degree and fashion, and makes in such degree and
fashion the aims and principles of its legislation. Every virtue,
however elementary and indispensable, according to our ideas,
might once have been beyond the power and the ken of the State.
We can imagine such a condition of things, as that, during its
continuance the State would have been unable to enforce and
inculcate the principles of common honesty, and even of
responsibility. It may once have been so here, just as it is still, to this
day, in Dahomey. Scientifically, the condition of Dahomey is as much
a part of the subject as the condition of England. The question is,
what has brought about the difference? The answer is the Church—
the Church that was in Egypt, that was in Israel, that was in Greece,
that was in Rome, that was in the forests of Germany, that has been,
and is, amongst ourselves. The Church has, all along, been going
before and shaping, little by little and step by step, higher and clearer
conceptions of right, and of duty, and of life; and the State has
followed, little by little and step by step, accepting and adopting what
the Church had made possible for it. Its position has generally been,
and ex rerum naturâ it must be so, behind the Church. This is seen
distinctly in the early days of Christianity. The Church was then
working out, and diffusing, much that the State afterwards
recognized and acted upon. This is their true relation to each other. It
is not merely that the nation, organized for its immediate mundane
wants, is the State, and that humanity, organized for the needs of its
higher life, is the Church; but that, besides this, in the progress of
society and of humanity, each is indispensable to the other.
Universal History tells us this: and from universal History, in a matter
of this kind, there is no appeal. And what universal History tells us
the History, as far as it goes, of the two famous buildings before us
confirms.
And now we must take off our thoughts from the two great
organizations of society, whose action and interaction have all along
been at work in shaping our political, social, and moral growth, and
making us what we are, symbols of which, in the two buildings
before us, we have been looking upon, and must turn our thoughts to
the great million-peopled city itself, of the existence of which we are
reminded, at the spot where we have taken our stand, chiefly by a
few lordly mansions, glimpses of which we catch, here and there,
through the trees. What variety of life is stirring within its widely
differing regions! How much energy and power, and how much
waste of power, and neglect of opportunity, are there! What
principles are struggling into existence! What principles are dying
out! What a conflict of principles is going on! We shall think not only
of the lordly mansions environing the parks that are spread out
before us, but equally of the commercial city on the banks of the
river, and of the moiling and toiling, the rough and gin-drinking
myriads of the manufacturing quarters of this world-capital. We shall,
in our thoughts, set by the side of what is refined, and intellectual,
and energetic, what is frivolous and enfeebled, what is rough, and
degraded, and vicious. We shall become sensible of the
uncertainties, as well as of the power, of the great intellectual and
moral organism that is at work all around us.
How much is there that is good and hopeful in all classes, and how
much in all that is evil, and evil enough almost to cause
despondency! How vast and complex is the whole! Your thought
enables you to understand that the railway and the telegraph have
made the city in which you are standing the centre of English
business and life, in a manner that was impossible formerly; and
more than that, for the ocean steamers and electric cables have
made it the centre of the business of the world. How does the
imagination, when stirred by the suggestions of the scene, picture to
itself the fashion in which are peopled the decks and saloons of the
great steamships that are hurrying, outward and homeward, on all
seas and oceans, to carry out the plans that have been originated
and matured here! You think, too, of the countless messages that
are flashing to and fro, beneath those seas and oceans, every
moment, for the same purpose. Here is the heart of the world. The
life-sustaining blood, in the form of human thought, and which carries
along in itself the elements of construction as well as of life, is ever
going forth from this heart, and coming back to it again. How many
tens of thousands of steam-engines, in as many mines and factories,
are throbbing and working to supply the wants, and maintain the
wealth, of this manifold Babylon we have built. Of this wealth we see
an exhibition here every day; for this is the spot for the daily parade
of one of its braveries. How have the corn-fields and meadows of
this island been solicited year by year to yield more and more, and
how widely have Australian and African wildernesses been peopled
with flocks and herds, for the enlargement of this wealth. This has on
its surface only a material aspect. It is true that its first and most
obvious result is to give wealth, and the enjoyment of wealth; and
that neither of these are necessarily and in themselves good: for if
wealth lead only to the self-bounded fruition of wealth it is
deadening, corrupting, and degrading: and of this there is in the city
around you much. But, however, this is not all its effect. It has given
to many minds culture and leisure, which they have devoted to
advancing the intellectual wealth of man; and it has produced many
who have devoted themselves, according to the light that was within
them, and prompted by the noblest impulses of our nature, to the
improvement of the moral condition of those with whom they come in
contact. Which of the two preponderate, the good or the bad effect of
the sum of all that is going on, we need not attempt to estimate here.
But to whichever side the balance may incline at the present
moment, we believe that the bad will perish, as it has done in past
times, and that the good only will survive—for only what is good and
true is eternal.
And now we turn from the many who are wealthy to the greater
many who are poor, and are carrying on a painful struggle for bare
existence, in this vast assemblage of humanity: and here, too, we
find mingled with what there is of good much that is evil. Here, as
with the wealthy, are aims that are unwise, springing from misleading
instincts which society has, carelessly and ignorantly, allowed to be
formed in its bosom, and which tend in the individual to unhappiness
and degradation, and in society itself to disorder and subversion.
All this must be taken in by the mind in order that the scene before
us may be rightly understood. We could not interpret the scenes of
old Egypt till we had formed some conception of what old Egypt was,
and we must endeavour to do the same for our corresponding
English scene. It is in this way only that the study and understanding
of old Egypt can be of any use to us. It is only when we understand
both that we are in a position to ask the question whether old Egypt
has anything to teach us.
It tells us that the aims of society must be moral; and that the
morality required can, within certain limits, be created and shaped,
and made instinctive, where society itself honestly wishes and
intelligently endeavours to do it. But as we look upon old Egypt we
see that the morality we need is not precisely what they imagined
and established, and that we are precluded from attempting to
establish what we want in the fashion of old Egypt. Theirs was a
system of constraint, ours must be a system of freedom. Theirs was
a system that concentrated its highest advantages on a few, ours
must be a system that opens its advantages to all. We must present
what we have to offer in such a form that men will voluntarily accept
it for themselves and for their children, and allow it to shape them. If
we see distinctly what we have to do, and the conditions under which
we have to do it, this will be in itself the achievement of half our
work. Their method was to devise a system, in strict conformity to
the conditions of the problem as it then stood, and place it as a yoke
upon society. They could do that: we cannot. Our method must be
accepted freely by society, and by the individual. We, too, must
devise a system in strict conformity to the conditions of the problem
as it now stands; and it must be such as approves itself to the
understanding and the conscience of the men of these times. The
successful fulfilment of the first requirement will, probably, include
the second.
Egypt, Israel, Greece, Rome, each did the work that had been
allotted to it. What we have to do is not to repeat what any one of
them did. That, indeed, we could not do; and, if we could, it would be
of no use to us. Imitations at all times, but more particularly when
circumstances differ, are worthless and disorganizing. And yet what
each of them did was necessary for us. The work we have to do now
is a great advance upon theirs, and is to be done under very different
conditions from theirs, but is so connected with theirs that we cannot
dispense with their foundations, or with the principles they worked
with. We need them all, but we must use them in the way our work
requires. When men came to build with stone, they did not abandon
all the principles of construction they had worked out for themselves
during the time they had built with wood. Those principles were right
as far as they went. They were not all bad, and worthless, and
inapplicable to the new material and its grander possibilities. What
had to be done was to incorporate the new principles that were
needed with those from among the old that would still be
serviceable. The purpose and object of building, whatever the
materials might be, continued one and the same. And so, now that
we have come to use glass and iron largely in architecture, the same
process is again repeated. Some new principles may be introduced,
but we do not discard all the old ones. Just so is it with the social
fabric.
The great and governing differences in our case are that what we
have to do is to be done for all, and that this is accompanied with the
condition of not partial, but universal freedom. It never was so with
any of the old peoples. And though our work is new in some of its
conditions, and such as, in its reach and variety, was never dreamt
of by the four great teacher nations of antiquity, there is no more
reason for our failing in it than there was for their failing in theirs.
That it is to be done is, in some sort, proof that it may be done.
Indeed, there is apparently more reason for our success than there
was for theirs. We have their experience; and in the principles of
universal freedom, and universal justice, we have more to commend
what ought to be done now to men’s hearts and understandings then
they had. Freedom, knowledge, truth, justice, goodness; these must
be our aims, our means, our statecraft, our religion. We do not go off
the old tracks. They all converge into our path. And so we find that
we are advancing, having history for guide, through new conditions,
into a richer and better life, placed within the reach of an ever
increasing proportion of the community.
The greatest, perhaps, of the advantages that will be found in our
wealth is that it will enable us to confer on every member of the
community such knowledge and such training as shall have an
hopeful, perhaps a preponderant, tendency towards making
instinctive, at all events in the minds of the greater number, a rational
use of the freedom they already possess, and the love and practice
of truth, justice, and goodness. Though, indeed, when we look at the
educational efforts of Saxony, of Switzerland, and of New England,
we are almost brought to fear that this great and necessary work will
be undertaken more readily and intelligently, and done sooner and
better, among people, who have less of the material means for
carrying it out than ourselves. In saying this, I do not at all mean that
we should confine our efforts merely to what they have done, for
they have, to a great extent, omitted that morality which I consider
the main point of all; but that we should be much better than we are,
if we had done as much as they, with their very inferior means, have
already accomplished.
In Egypt submission and order; in Israel, though labouring under
most cruel disadvantage, during its better days belief in and devotion
to right, and during its latter days the determination to maintain at
any cost its morality and religion; at Athens the appreciation of
intellectual culture; in the Roman Empire, by the mere working of its
system, the idea of the supremacy of the law, and the sentiment of
the brotherhood of mankind—were made instinctive. Why should we
despair of doing as much for what we need? Our task, indeed,
though so much grander, and promising so much more fruit than
theirs, does not appear as hard as theirs. If it be beyond our powers,
then modern society is but a fermenting mass of disorder and
corruption. It cannot be so, however; for if it were, then the long
course of History would now have to be reversed. All the progress of
the Past, and all its hard-won achievements, would prove without
purpose; and there would remain for us only to despair of truth, of
right, of religion, and of humanity itself.
FOOTNOTES
[1] This was written in 1871. It was in the following year, that is,
in the interval between the first and the second edition of this
work, that the Livingstone-search Commissioner of the ‘New York
Herald’ found the great African explorer.
[2] Some, I am aware, are disposed to answer the question of
this Chapter by ascribing to the Egyptians a Turanian origin. The
following appear to be the steps in the process, by which they
endeavour to reach this conclusion. There was, in remote times,
on the banks of the Euphrates, a Priest Class, which, on the
supposition that in its sacred and literary language, there are
some traces of the early Turanian form of speech, might have had
a Turanian origin. (Though, indeed, a Priest Class is rather an
eastern Aryan, or even a Semitic, than a Turanian phenomenon.)
This Priest Class, thus conceivably Turanian, might, conceivably,
have had some ethnological connexion with the Priest Caste of
Egypt. (There is, however, nothing to lead us to suppose that its
antiquity was as great as that of the Priest Caste of Egypt.)
Therefore the Egyptians might have had a Turanian origin. To put
the argument abstractedly: We may imagine two presumable
possibilities; the first of which possesses little probability, and the
second still less; and then by the juxta-position of the two reach a
desired conclusion. In other words, some degree of probability will
be the product of the multiplication of the non-probability of a first
assumption by the improbability of a second. This is the form of
argument by which probability is inferred from the accumulation of
improbabilities.
Of course, there is no saying what discoveries the future may
have in store; but, in the present state of knowledge, it seems an
unlikely supposition that Arts, Science, Law, Philosophy and
Religion were, aboriginally, Turanian.
[3] It is a curious fact that the inhabitants of the Lake-villages of
Switzerland cultivated, in the prehistoric period, as may be seen
in the Zurich collection of objects from the sites of these villages,
the same variety of wheat—that which we call Mummy, or hen-
and-chickens wheat—as the old Egyptians. Did the first
immigrants into Europe, of whom we may suppose that we have
some historical traces, for the Etruscans may have been, and the
Laps, Finns, and Basques may still be, surviving fragments of
their settlements, bring with them this variety of wheat at the
same time that another swarm from the same Central Asian hive
were taking it with them to the Valley of the Nile.
[4] I am led to propound this conjecture from a desire to render
intelligible what Herodotus says of their hair and skin; for we
know, both from the old paintings and from the existing mummies,
that the true Egyptian’s skin was not black, and that there was no
kink in his hair. It is impossible then to take his statement as it
stands; and I can imagine no other way of correcting it.
The difficulty here I conceive to be of just the reverse kind to
that which meets us in his statement, that the circumference of
Lake Mœris was 450 miles; and which, therefore, in the chapter
on the Faioum, I endeavoured to render intelligible by just the
reverse process, that is to say, by suggesting that, while we
suppose he is speaking of the Lake only, he is really speaking of
the whole of a vast system of artificial irrigation, of which the lake
was the main part. Here he is speaking of a part of the Egyptian
population, only he puts what he says in such a way that we
suppose that he is speaking of the whole of it.
I will take the opportunity of this note to propound an
explanation of Homer’s having sent Jupiter, and all the gods, to
Oceanus, to feast, for twelve days, with the irreproachable
Ethiopians. We immediately ask, Why with the Ethiopians? Why
are they irreproachable? What have they got to do with Oceanus?
Why to feast? Why for so long a period? Why all the gods? The
light, in which things are viewed in this book enables us to see an
answer to each of these questions.
Homer, we know, was acquainted with the magnificence of
Thebes. In his time, and for many centuries before, the
Phœnicians had, through commercial intercourse, been closely
connected with the Greeks; having, during the whole of that time,
been an autonomous dependency, or dependent ally, of the
Egyptians, who, in going to and from their head-quarters on the
Euphrates, had kept open a line of communication through
Phœnicia. The Phœnicians, therefore, must have had a great
deal to tell the Greeks about the marvellous greatness of Egypt,
the chief ingredient in which was the magnificence of Thebes.
There was plenty of time for all this to be thoroughly talked over.
Sethos and Rameses, the great Theban builders, had preceded
Homer’s day by four or five centuries. And, as such things never
lose in telling, Homer’s contemporaries must have had no very
inadequate—we now know that they could hardly have had
exaggerated—conceptions of the temples and wealth of Thebes.
He mentions the great amount of its military population; its
hundred gates, which, as no traces of walls of fortification for the
city have been found, meant, probably, the propylons of the
temples; and its vast wealth. He knew probably that Egypt
consisted of an Upper and of a Lower Egypt, and that the
inhabitants of the Upper country were darker, and that in the
extreme south, as then understood, the complexion became quite
black; and so, to distinguish them from the maritime Egyptians, he
calls them Ethiopians. He uses the same word as an epithet of
dark objects, as of wine and bronze. And here among these
Ethiopians was the wondrous Thebes. When the Phœnicians had
told the inquisitive Greeks of its mighty temples, and of its
incalculable wealth, they must have described its commerce, the
source, to a very considerable extent, of its greatness. For
centuries it had been the emporium of the trade of India, Arabia,
and Africa. This, and its position in the supposed extreme south,
to Homer’s mind, connected it with the outer, world-surrounding
ocean. What was told to him, and to his contemporaries, of the
tides and monsoons of the Indian Ocean, suggested to them, and
most aptly, only the idea of a stream. They heard of tides on the
Atlantic also; hence his mighty stream of circum-ambient ocean.
As to the trade of Thebes, all international wholesale trade in
those times, and in that part of the world, was carried on in the
courts and sacred enclosures of temples. The greatness of the
temples was, in some measure, an indication of the greatness of
the trade. The great festivals were, in substance, only great fairs.
Trade was then under the guardianship of Religion. Society was
not yet sufficiently organized for the protection of trade: for such a
purpose the civil power could hardly as yet be said to exist.
Religion alone had either the wisdom, or the power, to enforce fair
dealing, or to ward off violence. At the season, therefore, that the
great annual caravans arrived from the interior, and the easterly
monsoons wafted the merchandise and products of Arabia and
India to Egypt, to be bartered for those of Africa (and the
caravans were doubtless so arranged as that their arrival
synchronized with that of the ocean-borne traffic), there were
great processions and feasts at the temples. Religion then put on
its most imposing aspect. We have now only to recall the number

The Semantics and Pragmatics of Honorification Register and Social Meaning Elin Mccready Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Semantics and Pragmatics of Honorification Register and Social Meaning Elin Mccready Full Chapter

Uploaded by

Copyright:

Available Formats

The Semantics and Pragmatics of

Honorification: Register and Social

recently published in the series

The Semantics and Pragmatics of Honorification. First edition. Elin McCready

1.1 Honorifics: definitions and examples

(Agha, 1994).” This definition is compelling, but remains somewhat underspecified.

(1.1) a. foň dtòk khá

(1.2) a. sensei-wa ashita irassharu yoo desu

2 Questions of gender and gender presentation will be returned to in Chapters 7 and 8.

(1.4) a. French tu/vous, German du/sie

(1.5) a. I was cooking up some dinner.

1.2 Honorifics and politeness

2.1 Properties of expressives

The Semantics and Pragmatics of Honorification. First edition. Elin McCready

(2.1) Properties of expressives:

(2.2) a. I didn’t see a brown dog.

if semantic operators cannot apply to expressive content, then, because displacement

Finally, Repeatibility is obviously a special case: though repetition of expressive

2.2 Honorification as expressive: initial data

(2.15) a. moshi Iwabuchi-sama-ga ki-tara kono ii sake-o

(2.18) a. kyoo-wa samui desu ne

(2.19) Yamada-san-ga irasshaimashita

(2.20) Possible candidates for H:

strengthening as those of emotive expressives. Though they don’t necessarily result in

2.3 Other languages and other data

2.4 Other possibilities

Can it then be thought of as conversationally implicated? Such a view seems

The previous chapter showed that honorific meanings should be regarded as

The Semantics and Pragmatics of Honorification. First edition. Elin McCready

3.1 Previous formal analyses of honorifics

2 Thanks to Ryan Hearn and Will Starr for discussion here.

3.2 A formalism for register

(3.10) a. High ⊆ [.6, 1)

You might also like