Publishing Partner: Cambridge University Press CUP Extra Publisher Login
amazon logo
More Info

New from Oxford University Press!


May I Quote You on That?

By Stephen Spector

A guide to English grammar and usage for the twenty-first century, pairing grammar rules with interesting and humorous quotations from American popular culture.

New from Cambridge University Press!


The Cambridge Handbook of Endangered Languages

Edited By Peter K. Austin and Julia Sallabank

This book "examines the reasons behind the dramatic loss of linguistic diversity, why it matters, and what can be done to document and support endangered languages."

Review of  Agency and Impersonality

Reviewer: Heiko Narrog
Book Title: Agency and Impersonality
Book Author: Mutsumi Yamamoto
Publisher: John Benjamins
Linguistic Field(s): Semantics
Subject Language(s): English
Issue Number: 18.402

Discuss this Review
Help on Posting
AUTHOR: Yamamoto, Mutsumi
TITLE: Agency and Impersonality.
SUBTITLE: Their Linguistic and Cultural Manifestations.
SERIES: Studies in Language Companion Series 78
PUBLISHER: John Benjamins
YEAR: 2006

Heiko Narrog, Tohoku University, Japan.

If one has the opportunity to deal with two or more widely different
cultures and languages on a more than superficial level, there is a good
chance that one comes across systematically different patterns in culture
and language that seem to correspond to each other. The enterprise to pin
down such correspondences between language and culture scientifically, that
is, beyond mere impressions and intuitions, has been known under the label
'linguistic relativity'. The author of the volume under review firmly
places herself in this tradition of linguistic relativity when she compares
the concept of agency and its realization in Japanese and English language
and thought.

It was already pointed out by Ikegami Yoshihiko (1991) and others that
English seems to emphasize agency while Japanese suppresses or backgrounds
it. Yamamoto builds on this observation and can thus be seen in a line of
research descending from Ikegami, a scholar who has been extremely
influential within Japan. At the same time the book is a direct
continuation of her previous book that appeared in the same series, on
''Animacy and Reference'' (Yamamoto 1999). In a preview of my evaluation, I
would say that, although the concrete linguistic analysis is not always
convincing, the author is successful overall in raising awareness about
systematically differing patterns of expression through which agency is
obfuscated in Japanese, while it is highlighted in English. Concerning the
connection to culture, or habitual thought, however, the book is less
persuasive, simply because of the failure to provide any hard evidence
(data) that would support this connection.


The rather slim volume (130 pages plus notes, references and an index)
consists of five chapters. After a 10-pages introduction, chapter 2 (11-37)
discusses the concept of agency in some detail. It starts out with the
treatment of the concept in philosophy, going back to Aristotle, and goes
on to discuss the relationship of agency with related concepts such as
intentionality, awareness of action responsibility, causation, and animacy.
Yamamoto acknowledges the relevance of all these notions for the concept of
agency, but ultimately sees intentionality at its core, and animacy as
inseparable from it, since ''[o]nly animate beings can be agents in a normal
sense'' (29).

Chapter 2 also introduces the concept of ''mind-style'' which consists of the
''preoccupations, prejudices, perspectives and values which may bias an
individual's 'world-view''' (19).
''Mind-style'' and ''world-view'' are the key terms in Yamamoto's concept of
linguistic relativity. Language patterns reflect mind-styles, and may also
influence mind-styles in a mutual relationship. Mind-style, therefore, as
Yamamoto uses it, seems to refer less saliently to the individual than
collectively to a language or culture (in comparison) as a whole.

Chapter 3 (39-69), the core of the book, deals with the linguistic
treatment of agency. Section 3.2 discusses agency in linguistic analysis,
referring to Fillmore, Cruse, and other scholars who have been concerned
with semantic roles and related notions, before settling on Dik's (1989)
Functional Grammar approach to the concept of ''agent''. Section 3.3 presents
data on agents in Japanese texts and their English translations and vice
versa. The data show that Japanese has an overwhelmingly strong tendency
towards either ellipsis of the agent or choosing verbal expressions that do
not require an agent. According to the author, it can thus be said that in
contrast to English, which has an ''actor-action pattern'' (65), Japanese
backgrounds actors into 'nothingness' and prefers 'impersonal' expression
of events (68).

Chapter 4 (71-117) tries to forge links between the linguistic data and
''mind-styles'' or ''world-views''. In the first part of the chapter, Yamamoto
introduces the cognitive linguistic concepts of ''figure'' and ''ground'' and
argues that in Japanese not only agents tend to be obfuscated but even
figures in general, which is ''[a]n ultimate demonstration of impersonality
in the Japanese language'' (76). It is then argued that in Japanese, if
agents are not elided altogether, 'de-agentivisation' takes place through
'positionalisation' of persons, or the replacement of individual agency by
group or collective agency (78-90). In section 4.3., the author revisits
linguistic relativity, boldly stating her support for Whorfian linguistic
relativism, in contrast to many modern linguists who, while pursuing a
relativist research agenda as well, tend to distance themselves from
linguistic-cultural relativism in its earlier forms. Finally, in section
4.4., the author makes the connection between language, mind-styles and
cultural norms, arguing that the linguistic obfuscation of agent correlates
to impersonalisation, de-agentivisation and dehumanisation of human agents
in Japanese society and to collectivism, as opposed to individualism in
English-speaking societies.

Chapter 5 (119-130) is devoted to issues that are seen as peripheral to the
discussion of the present book, namely agency and rhetorical style in
Japanese and English literary works, and impersonality in languages other
than English and Japanese, hinting at the possibility of future typological


This book refers not only to linguistics but also to philosophy and
literary studies. From my limited view as a linguist, it is in some
respects very attractive and in other respects very weak. To start out with
the positive features, I found this book always interesting to read, due to
the author's wit and entertaining style of writing, which is truly
admirable. The discussion of the concept of agency in chapters 2 and 3 is
comprehensive and useful. It is somewhat regrettable that the author
introduces and discusses various concepts of agency, but in the end lets
them stand as they are without a new synthesis, but the review of existing
concepts is of undeniable value. The ensuing exposition of the data in
chapters 3 and 4 is, to the reviewer's knowledge, more thorough than
anything written on the topic before. The author shows in detail how agents
are backgrounded in Japanese and highlighted in English. However, it should
be critically remarked here that the same (rather limited amount of) data
were already presented in Yamamoto's earlier book in the same series
(Yamamoto 1999: 164f), and that data-wise this book offers nothing new; the
focus of discussion has simply shifted from animacy and reference to
agency. The link between backgrounding agency in language and in society,
as presented in chapter 4, may be intuitively plausible to anyone familiar
with Japanese culture, but, as mentioned briefly above, it is established
exclusively on the intuitions of the author about Japanese society and
thought, and some general books on it, and is not supported by any data.

Here we come to the critical points. In the new wave of research
establishing links between language and thought (e.g. Lucy 1992, Levinson
2003, Gentner & Goldin-Meadow 2003), the general consensus is that
different linguistic structures as such are no conclusive evidence for
linguistic relativity. One needs corresponding data from non-linguistic
experiments that establish correspondences between language and general
cognition. As a result, most contemporary research on linguistic relativity
is being carried out within the frameworks of, or with the methods of,
psycholinguistics and anthropology. Yamamoto, in contrast, relies solely on
language data. Her intuitive observations on the link of linguistic
structures to 'mind-styles' are certainly interesting to read, and may have
their own value as ideas, but from a scientific point of view they are

Another serious problem lies with the linguistic analysis. Can ellipsis of
arguments (pro-drop), which is common in Japanese, be really equated with
'impersonalisation' or even 'nothingness'? Certainly, this is a possible
conclusion, but it seems to me that the author has jumped to this
conclusion too readily without seriously considering alternatives. Within
the discourse grammar of a language, ellipsis must be primarily seen as
part of the referent identification and reference-tracking system. This is
an important aspect of ellipsis not taken into account at all by Yamamoto.
In her book, sentences from coherent texts are presented in isolation as
examples of 'impersonality' and 'nothingness' although it is precisely
their position in discourse and context that prompts the ellipsis. Usually,
pro-drop languages dispose of various devices to indicate referents,
saliently including agents and other first arguments. In an excellent study
on the topic by Nariyama (2003), the author shows the system of
predicate-marking morphemic and syntactic devices that are employed in
Japanese, including ellipsis as a meaningful part of the system, and not as
'nothingness'. Also, consider the following example of 'impersonalisation'
from Yamamoto (61):

(1) ''Ø Sonna fuu ni o-kangae-ninatte wa dame yo...''
(You:Nom) such way in think-HON TOP no:good COP
''Oh! You mustn't think that...''

The point for Yamamoto is that while the English sentence (from an Agatha
Christie novel) has the personal pronoun you, the Japanese translation has
nothing in its place, and thus is ''impersonal'' and ''somewhat incomplete''.
However, to my mind, the Japanese sentence is not impersonal at all. While
it does not contain a pronoun, the honorific marking on the verb (one of
the devices pointed out by Nariyama) very directly indicates the
interlocutor as the referent (and agent). This makes the sentence in fact
highly personal.

Another very interesting point made by Nariyama is that the whole
reference-tracking system in Japanese is saliently centered on the
interlocutors, especially the speaker. In this sense, far from being
impersonal or event/situation-centered, ''the organization of Japanese
sentences has a propensity to be egocentric and anthropocentric'' (Nariyama
2003: 264).

Different languages have different systems of segmenting events and marking
referents. Are these all due to different mind-sets and world-views? I
suppose it is easier to jump to this conclusion if only two languages such
as English and Japanese are compared, which then appear as two polar
opposites, than if many languages are taken into account. For example, some
languages in West Africa (and presumably others elsewhere as well) tend to
split up events into more sub-events than English, and at the same time
require obligatory subject pronoun clitics for each verb. Also, they may
not allow the frequent subject ellipsis in complex sentences that English
has (e.g. they may have something like 'I went (to) I buy food' instead of
'I went to buy food'). That is, in the equivalent a sentence where you
might find one or two agentive subject pronouns in English, and possibly
none in Japanese, you might find three or four in that language. Does that
mean that speakers of such languages have a super-agentive world-view?

In conclusion, I believe that ellipsis should either be analyzed more
carefully or omitted entirely from the discussion of agentivity and
impersonality. This leaves us with those cases where Japanese chooses
different, intransitive expressions from English, and those are still
interesting enough. One of the merits of Yamamoto's book lies in a thorough
discussion of such cases, but I feel that this merit ends up being
diminished by the lack of a clear distinction between such cases and cases
of ellipsis.

One more point where I found that Yamamoto unfortunately undermines the
thrust of her own argument is the fact that the author initially included
English scientific texts in her study but later decided to exclude them as
it turned out that they abounded in agentless passives (59). From a human
point of view, it is understandable if the author did not want to include
data that would undermine her clear-cut picture of agent-oriented English
vs. event-oriented Japanese. However, it seems to me that an opportunity
was missed to point out that Japanese and English are possibly not
monolithic blocks but that various modes of discourse exist and can be
employed by speakers for different purposes. Overall, Yamamoto was perhaps
too eager to present a strong contrast between the two languages and
cultures while in fact a more balanced approach would have made the
impression of higher credibility. It should also not go unnoticed that
while Japanese language and culture may indeed be highly homogenous
(although many social scientists question this assumption), English is
nowadays spoken as a first language or bilingually in many countries with
cultures that are quite different from the UK or (Anglo-Saxon) North
America. Do all these different people and cultures share the same, or a
similar world-view?

Overall, the biggest problem with this book is that it is short on
substance. It contains no new data, and crucially, no data at all that
would serve to establish the purported relationship between language and
culture/thought. The analysis of the language data is in many respects only
cursory and conspicuously biased towards the desired conclusion. On the
other hand, I find the topic quite inspiring in the witty way it is exposed
by the author. Even if the analysis and the results may not be convincing,
a book like this, which manages to raise interesting research questions in
an interesting manner, can justly be called a success.


Gentner, D., & Goldin-Meadow, S. (eds.). (2003). Language in Mind: Advances
in the Study of Language and Cognition. Cambridge: MIT Press.

Ikegami Y. (ed.) (1991) The Empire of Signs. Semiotic Essays on Japanese
Culture. Amsterdam: John Benjamins.

Levinson, S.C. (2003) Space in Language and Cognition. Explorations in
Cognitive Diversity. Cambridge: Cambridge University Press.

Lucy, J. (1992) Grammatical Categories and Cognition. Cambridge: Cambridge
University Press.

Nariyama, S. (2003) Ellipsis and Reference Tracking in Japanese. Amsterdam:

Yamamoto, M. (1999) Animacy and Reference : A Cognitive Approach to Corpus
Linguistics. Amsterdam: Benjamins.

Heiko Narrog is an associate professor at Tohoku University, Japan. His
research interests include historical linguistics, syntax and semantics,
modality, linguistic typology, and the Japanese language.