A Multifaceted View of the Concept of Amae : Reconsidering the Indigenous Japanese Concept of Relatedness

Abstract

The indigenous Japanese concept of amae has provoked interest from scholars across disciplines. Many have provided their own version of defining amae without much attempt to synthesize it into a demonstrative definition. Non‐Japanese scholars have attempted to understand the concept through their own interpretations, which has often led to confusions and erroneous conclusions. The present paper analyzes the concept of amae, focusing on its everyday use with illustrative evidence to provide contextual meanings of varied amae phenomena. A multifaceted view to approach amae is introduced. The new approach proposes to consider amae in different contexts and in three developmental phases, with evidence to support that amae is, in fact, different in each category. Future amae research is discussed.

Full text

Fax +
41 61 306 12 34
E-Mail karger@karger.ch
www.karger.com
Ó2004 S. Karger AG, Basel
0018–716X/04/0471–0001$21.00/0
Accessible online at:
www.karger.com/hde
Kazuko Y. Behrens, Graduate School of Education
University of California Berkeley
4511 Tolman Hall, Berkeley, CA 94720 (USA)
Tel. +1 510 642 4201, Fax +1 510 642 3555
E-Mail kbehrens@uclink4.berkeley.edu
Paper
Human Development 2004;47:1–27
DOI: 10.1159/000075366
A Multifaceted View of the Concept of
Amae
: Reconsidering the Indigenous
Japanese Concept of Relatedness
Kazuko Y. Behrens
University of California, Berkeley, Calif., USA
Key Words
Amae
` Attachment ` Japan ` Relatedness
Abstract
The indigenous Japanese concept of
amae
has provoked interest from
scholars across disciplines. Many have provided their own version of defining
amae
without much attempt to synthesize it into a demonstrative definition.
Non-Japanese scholars have attempted to understand the concept through their
own interpretations, which has often led to confusions and erroneous conclu-
sions. The present paper analyzes the concept of
amae
, focusing on its every-
day use with illustrative evidence to provide contextual meanings of varied
amae
phenomena. A multifaceted view to approach
amae
is introduced. The
new approach proposes to consider
amae
in different contexts and in three de-
velopmental phases, with evidence to support that
amae
is, in fact, different in
each category. Future
amae
research is discussed.
Over the past few decades the concept of amae, known as the indigenous Japa-
nese concept of relatedness, has invited debates among scholars in various disci-
plines. Unfortunately, these debates have resulted in few agreements regarding a
consensus for a definition and description of the function of amae. Amae has pro-
voked much interest not only because this concept is often believed to play a key
role in understanding Japanese psyche, but also because it appears to be similar to
another major concept of relatedness, such as that found in attachment theory. At-
tachment theory has been proven to be one of the most influential theories of relat-
edness based on the evolutionary perspective. Attachment is a species-specific,
innate behavioral system that guides the young to stay close to their mothers to
Copyright © 2004 S. Karger AG, Basel

2
Human Development 2004;47:1–27
Behrens
assure survival, and thus all children attach to their caregivers [Bretherton, 1992].
Both attachment and amae phenomena are believed to first emerge in infants after 6
months of age, and the first experience is likely to occur in mother-child relation-
ships [Bowlby, 1958; Doi, 1973]. Yet attachment and amae as conceptualized dif-
fer in important ways, which I will discuss later in this paper. Nevertheless, van
IJzendoorn and Sagi [1999], major cross-cultural attachment researchers, consider
Japan a real challenge, partly because the concept of amae has an implicit emphasis
on psychological dependence, whereas autonomy is one of the key elements that
represents security in attachment theory. Rothbaum, Weisz, Pott, Miyake, and Mo-
relli [2000] have challenged the universality of attachment theory and selected the
concept of amae in supporting their own view of attachment and dependence.
Rothbaum et al. claim that typical amae behaviors, which are supposed to be plau-
sible and acceptable to Japanese mothers, will be considered as type C (‘insecure-
ambivalent’), based on the standard attachment classification system
1
: this ques-
tions the validity of attachment theory as an exclusively Western-biased theory.
What characterizes ‘typical’ amae behavior, however, needs to be re-examined.
What Is
Amae
?
When conceptualized, amae represents a cluster of behaviors, an emotional or
internal state, and a philosophical construct for Japanese people that can be viewed
either positively or negatively, depending on what is deemed appropriate with re-
spect to maturity or degree of social intimacy. Amae, sometimes translated simply
as ‘dependence’, is a phenomenon with multiple aspects or meanings, including
(a) one’s desire to be intimately close to another person or to be basked in the
warmth of the other [e.g., Doi, 1973]; (b) to act playfully, like a baby [e.g., Take-
tomo, 1986]; (c) to be lenient toward one another or have an indulgent relationship
[e.g., Lebra, 1976]; (d) to importune somebody [e.g., Okonogi, 1992], and (e) to
presume upon another person’s goodwill or to take advantage of them [e.g.,
Mitchell, 1976]. Amae is viewed as an interpersonal process and thus the phenome-
non can be observed only in interactions [e.g., Kumagai & Kumagai, 1986; Maruta,
1992]. Amae, in my view, always consists of some expectation or assumption on
the part of the amae doer of being understood and accepted, whether it is for pure
affection or instrumental needs, either within intimate or non-intimate relationships.
Further, amae is always relational and often involves the desire to be accepted for
asking for something that one is perfectly capable of doing oneself. Therefore de-
pendency on others (for something one is incapable of doing), I argue, does not
constitute amae. Very briefly, amae can be described as the presumption on others
to be indulgent and accepting [G. DeVos, pers. comm., February 4, 2003].
To date, however, there have been little systematic investigations of amae,
both empirically and theoretically. The goal of this paper is to depart from the way
1
Ainsworth, Bell, and Slayon [1971] recognized three major developmental patterns of attach-
ment – A, B, and C – based on the laboratory procedure known as the ‘Strange Situation.’ Group-B
children are regarded as secure, whereas Group-A children are classified as insecure-avoidant and
Group-C children are classified as insecure-ambivalent/anxious.

Amae Reconsidered 3 Human Development 2004;47:1–27
researchers and scholars have viewed amae, and to provide a new framework to
approach this indigenous Japanese concept of relatedness. I intend to show that
amae can show itself in many different forms and functions, depending on the con-
text in which amae is exhibited. It is possible, however, to understand and work
with amae if we treat amae seperately in each context, thus applying the multidi-
mensional approach that I propose together with recommendations for future re-
search.
In this paper, I first present existing definitions of the concept of amae, begin-
ning with a brief linguistic explanation. I then review debates regarding the concept
of amae to show variations in interpretations. Second, I review some amae research
to provide possible empirical evidence. Third, I propose a new, multifaceted ap-
proach to viewing the concept of amae by analyzing it in various contexts, thus
showing that it has different qualities and developmental phases. Here I provide
evidence to illustrate amae in various contexts, demonstrating what amae means to
Japanese people today in its everyday use. Finally, I discuss how this new approach
to amae can help to promote other research in Japan, attachment research in par-
ticular. I attempt to clarify the similarities and differences between the concept of
amae and attachment theory, suggesting how both can possibly work together.
The evidence I present here is by no means complete or sufficient to cover all
possible amae phenomena. Nonetheless, it provides a contextual framework to get a
sense of what amae is, and the evidence shows how varied theoretical definitions
are actualized. To show how the word amae is, in fact, used by native Japanese, I
quote directly from the participants’ responses to the amae question. I have left
amae unchanged in the text to show how the participants use it. I should also add
that, due to a lack of data, I am not discussing the father’s role in amae interactions.
Future research needs to include the role of the father in Japanese family relation-
ships.
Amae
– A Traditional View
Linguistic Background
The word amae is a noun and shares the same root with the adjective amai.
The adjective amai has two basic meanings: ‘sweet/sugary’ and ‘not strict/
generous’ or ‘seeing through rose-colored glasses/being naïve’, the latter with
rather negative connotations. The transitive form of the verb amaeru is amayakasu,
which refers to an act of letting someone amaeru. In order for an act of amae to be
completed, these two complementary roles – amaeru or amae-taker and amayakasu
or amae-giver – are necessary [Lebra, 1976]. Amae can be induced in this amaeru-
amayakasu interaction [Kumagai & Kumagai, 1986]. Further, mutually satisfying
amae can be achieved only when one who amaeru and one who amayakasu are
both in agreement [Taketomo, 1986]. It is believed that in an affective mother-child
relationship, the child is likely to be the one who actively amaeru and the mother is
likely to be the one who amayakasu, letting or accepting the child amaeru more or
less passively [Doi, 1973]. The mother can also take an active role in amayakasu
behavior, for example, by encouraging and inviting child to be physically close to
her and be cuddled [Lebra, 1976].

4
Human Development 2004;47:1–27
Behrens
Doi’s Definition of Amae
A Japanese psychiatrist, Doi, focused on the concept of amae as something
peculiar to the Japanese; he published his first book on amae, entitled ‘Amae no
Koozo’ (The Structure of Amae) in 1971. The English translation, entitled ‘The
Anatomy of Dependence’, was published in 1973. Doi has been accused of not pro-
viding a clear definition of amae, thus leaving the concept vague and ambiguous.
Doi [1973] regards the mother-child relationship as the prototype of the amae rela-
tionships throughout a person’s lifetime. The prototype of amae, then, is the in-
fant’s desire to be close to his mother after realizing that she has a separate exis-
tence from himself. Doi claims that amae fosters ‘a sense of oneness between
mother and child’ (p. 75), similar to Mahler’s [1968] concept of symbiosis. Doi
defines the amae mentality as ‘the attempt to deny the fact of separation that is such
an inseparable part of human existence and to obliterate the pain of separation’
(p. 75). Taking Balint’s [1966] idea of ‘passive love’ and ‘active love,’ Doi also
characterizes amae as ‘passive love’. According to Doi, such craving for closeness
should be a universal phenomenon, as love should be a primary need of all human
beings, as Balint claimed. The very existence of the word amae in Japanese, how-
ever, has enabled Doi to single out this psychological phenomenon. Doi does also
recognize deviated amae behaviors as the root of many pathological behaviors that
are prompted by the inability to express amae, or when amae was not granted.
Alternative View of Amae
Taketomo [1986] claims that Doi misses a metacommunicational feature in
amae, namely that it has a message for both interactants – the one who amaeru and
the one who amayakasu. The message may be ‘this interaction is undertaken with
a mutual agreement that it may deviate from certain ordinary rules of behavior’
(p. 535). Taketomo agrees with Doi on infantile amae being prototypical, however,
he focuses more on the amae phenomena in childhood and adulthood, beyond in-
fancy. He claims that a child’s ‘amae entails a playful interaction with a parental
(especially maternal) figure in which social pressures on the child to behave age-
appropriately are temporarily relaxed’ (p. 532). According to Taketomo, when a
child playfully mimics an infant’s behavior, the child is not really becoming an
infant, as Doi proposes. Instead, by such behaviour the child is trying to communi-
cate with his mother that he wants to amaeru and wants her to amayakasu him. If
the mother does not agree with or recognize the child’s message, there will be no
amae. Taketomo further presents adult manifestations of amae in the coquetry
situation, in which a woman behaves playfully, like a child who is acting as an in-
fant – in other words, a mimicry of a mimicry of the infant’s amae behavior. In
romantic relationships, both partners are likely to mutually agree with ‘the tempo-
rary lifting of the code of ordinary, or “proper”, mature behavior’ (p. 532) for an
amaeru-amayakasu interaction to occur.

Amae Reconsidered 5 Human Development 2004;47:1–27
Amae Symposium
A symposium on ‘Attachment and Dependency’ was held in Kyoto, Japan, in
1990. Its treatment of amae appeared in a designated section of the Infant Mental
Health Journal [Osofsky, 1992]. Four well-known psychiatrists, three of whom are
Japanese, presented their views on amae, with varied interpretations. Maruta’s
[1992] critical view on Doi’s claim is that Doi only focuses on the child who
amaeru, disregarding the role of the mother who amayakasu, without which amae
cannot exit. Maruta argues that either the infant or the mother can initiate the amae
process ‘in search of renewed intimacy or a sense of security’ (p. 15). Maruta also
briefly touches on amae among adults in everyday social situations, which Doi
barely discusses, such as relationships with friends, relatives and colleagues. The
key to successful amae, when it is pleasurable to both interactants, is to carry on in
a ‘mutually comfortable’ manner. Maruta adds ‘learning what is “mutually com-
fortable” is a significant part of growing up and surviving in Japan’ (p. 16).
Okonogi [1992] focuses on the child’s initiating role of amaeru through cer-
tain expressions and behaviors to elicit an adult’s role of amayakasu. Okonogi
points out that a child never actually says that he wants to amaeru, but rather goes
through a series of expressions, often highly manipulatively, to achieve his goal,
which is to let the adult amayakasu him. Okonogi argues that it is the adult who
actually uses the word amae to refer to the particular behavior of the child. He then
classifies amae into secure amae and anxious amae, simulating the concept of at-
tachment. For example, he argues that just as in a case of anxious attachment, a
child who clings excessively can be said to be exhibiting anxious amae. Okonogi
argues that many people often view amae negatively and some mothers firmly be-
lieve in not amayakasu-ing in order to discipline the child.
Watanabe [1992] claims that when a mother amayakasu her child, thus fulfill-
ing the child’s needs, it can be secure and beneficial to the child. However, it can
be insecure and harmful if the mother amayakasu her child to ‘fulfill the mother’s
egoistic need’ (p. 28). Watanabe’s view of amae assumes that the mother’s role is
that of an initiator, contrasting Okonogi’s emphasis on the child’s active role in
initiating the amae process. Watanabe distinguishes amae in childhood from amae
in adulthood and claims that in adulthood, social expectations and obligations need
to be met to fulfill amae.
Emde [1992], the only non-Japanese participant in this symposium, does rec-
ognize the need for operationalizing and clarifying the concept and observable be-
haviors to make it useful for clinicians. He also proposes possible amae research on
individual and gender differences within the mother-child relationship (i.e., mother-
son amae relationship vs. mother-daughter amae relationship).
Summary of Amae as a Concept
Amae has always been a part of life for Japanese people. Yet Doi was the first
to recognize its significance in affecting the way people think and behave in a cul-
tural context. Since his first introduction of amae as an indigenous Japanese con-
cept, Doi has provoked many scholarly reactions, as we have seen above. Johnson
[1993] has dedicated an entire volume to provide the most comprehensive review

6
Human Development 2004;47:1–27
Behrens
of the concept of amae. Johnson’s thorough review of amae seems to have con-
veyed the depth and complexity of amae on the anthropological, philosophical and
psychological levels. Johnson provides a summary of his argument with the follow-
ing nine points (pp. 200–207):
1 Amae is part of a larger domain of dependent and interdependent relationships.
2 Amae is identified by Doi as a basic desire, motive, or ‘drive’.
3 Amae constitutes a Japanese cultural expression of indulgent dependency; ver-
sions of this universal dynamism take on different forms in different cultures.
4 Amae-like dependency and interdependency are observable throughout the life-
span.
5 Amae is not a simple or unitary phenomenon; it may be examined at several
different levels of subjective and behavioral experience.
6 Doi’s theory of amae does not specify sharp continuities and discontinuities in
regard to indulgent dependency.
7 Amae can be productively applied to psychotherapeutic and clinical encounters.
8 Amae may be examined through a study of language and culturally specific ter-
minologies.
9 Doi’s social-philosophic commentary concerning amae is a useful contribution
to the understanding of modal Japanese personality.
Johnson’s extensive work is undoubtedly helpful in interpreting amae on a theo-
retical level. What is missing in this and other work emerging from the debate on
amae is empirical data or evidence of what amae means to Japanese people today
on a practical level. Next, I review empirical amae research.
Amae
Research
Amae Research in Japan
Although the concept of amae has interested many scholars abroad, its com-
plexity has led to inconsistent interpretations. A study of amae outside of Japan is
not readily available, making it difficult to gain greater understanding of the con-
cept. Empirical research on amae in Japan appears to be still surprisingly scarce.
Most amae studies conducted in Japan are based on questionnaires rather than
direct observational or experimental studies. To conduct their study, Yamaguchi et
al. [1998] applied Taketomo’s [1986] definition of ‘a mutually agreed-upon sus-
pension of certain ordinary restraints on behavior’ (p. 541) by rephrasing it as
‘inappropriate behavior with the expectation of being accepted’ (my translation).
They asked subjects to judge whether twenty scenarios of inappropriate behaviors
were amae or not. The subjects were presented with these scenarios under three
conditions and asked whether or not these descriptions reflected amae: (1) purely
inappropriate behavior, (2) inappropriate behavior with no expectation of accep-
tance and (3) inappropriate behavior with expectation of acceptance. For example,
one scenario depicted a high-school student with a part-time job at a coffee shop,
even though this is strictly prohibited at his high school. The first condition simply
stated that he was violating a rule. The second stated that he expected to be ex-
pelled from his school if the school found out, and the third condition was that he
expected to be forgiven even if the school found out. Yamaguchi et al. [1998]

Amae Reconsidered 7 Human Development 2004;47:1–27
found that less than 28% (n = 376) regarded ‘inappropriate behavior’ alone without
expectation as amae, whereas more than 83% regarded inappropriate behavior with
expectation to be forgiven as amae. Thus, ‘expectation’ of acceptance of inappro-
priate behavior played a key role in identifying amae. The second follow-up study
also confirmed the role of an expectation of acceptance of inappropriate behavior
through a word association task.
Ohsako and Takahashi [1994] conducted a study in two parts: the first part
examined how amae may relate to gender and birthorder, the second investigated
the effects of amae in hypothetical interpersonal conflict situations. Ohsako and
Takahashi found that female students scored generally higher in terms of a degree
of amae expressed in romantic relationships than did male students. As for a
birthorder effect in sibling relationships, the secondborns expressed more amae
than did the firstborns. In the second part of the study, college students were asked
to rate emotion scales and conflict solution strategies scales in six hypothetical in-
terpersonal conflict situations. Those who were significantly more likely to feel
negative emotions and apply various strategies to attempt to resolve the conflicts
were grouped into the high-amae, the others into the low-amae group. This study
shows that high- and low-amae groups co-exist, thus not all Japanese people ex-
press amae in the same way or to the same degree. The findings also indicate that
amae may be implicitly applied in resolving interpersonal conflicts.
Unlike Ohsako and Takahashi [1994] who treat amae more as a disposition or
trait, Kim and Yamaguchi [1995] approach amae as an episode involving two inter-
actants: one to engage in amaeru and another to let amaeru. They gave a compre-
hensive questionnaire to 847 subjects in four age groups (junior-high, high school,
college, and adults) to examine their responses to open-ended questions. The ques-
tionnaire covers situations when respondents amaeru and let amaeru. For example,
they were asked to whom, when, and how they amaeru as well as about feelings
when they could not amaeru. They were also asked who, why, when, and how they
let people amaeru as well as about feelings when the person does not amaeru to
them. Lastly, they were asked about word associations with regards to amae, emo-
tions, positive aspects of amaeru and letting amaeru, negative aspects of amaeru
and letting amaeru, who they thought was good or poor at amaeru, whom they felt
difficult or comfortable to amaeru, and what was acceptable or unacceptable amae.
The findings show that the majority of the students’ group listed ‘mother’ as some-
one to whom they amaeru most, although ‘friends’ came a close second, especially
among older students, whereas the adult group listed ‘spouse’ most frequently, fol-
lowed by ‘mother’ and ‘friends.’ In the situation when respondents let amaeru,
students listed ‘friends’ at the top, whereas the adults’ choice was more varied:
‘son’, ‘daughter’, or ‘spouse’ [see Kim & Yamaguchi, 1995, for a full review]. One
of the intriguing findings is that many reported to feel ‘lonely and sad’ when they
could not amaeru and when the person does not amaeru to them, indicating the
vital role that amae plays in maintaining the psychological well-being of Japanese
people. Kim and Yamaguchi [1995] also indicate that the older the respondents are,
the more complex views about amae they are likely to have.

8
Human Development 2004;47:1–27
Behrens
Other Amae Research
Three concepts, amae, attachment, and dependency, were compared by non-
Japanese researchers to illuminate the interrelations between them and the role they
play in what is deemed desirable children’s behavior in a two-part study
[Vereijken, Riksen-Walraven, & Van Lieshout, 1997]. In the first part of the study,
Vereijken et al. first asked US experts in the field to hypothetically describe a se-
cure and a dependent child, to construct an ‘attachment security criterion sort’ and
a ‘dependency criterion sort’ using the Attachment Q-Sort (AQS). The AQS con-
sists of 90 small cards that describe infants’ and young children’s behavior in the
home setting. Japanese experts in the field were asked to describe the concept of
amae, also using the AQS. Additionally, Vereijken et al. used an existing Japanese
AQS desirability criterion sort that depicts the ideal child according to Japanese
mothers. In the second part of the study, Japanese children were observed at 14 and
24 months and rated on the AQS. Vereijken et al. found that children’s amae be-
haviors correlated with their dependency behaviors, although neither amae nor de-
pendency were related to security. Furthermore, they found that Japanese mothers
view the behavioral construct of a securely attached child as desirable, whereas
neither the behavioral construct of dependency nor of amae was considered desir-
able – in stark contrast to Rothbaum et al.’s [2000] claim. Whether mothers who do
not desire hypothetical amae behaviors will not allow their children amaeru in real
life is, however, unknown. Although this study is helpful in demonstrating interre-
lations between amae, dependency, and attachment security and how Japanese
mothers view these behaviors in the global sense, use of the AQS limits a list of
behaviors to a framework of attachment to describe amae behaviors: it is not likely
to cover the full range of amae behaviors.
Morsbach and Tyler [1986] presented examples of amae, mostly from Japa-
nese literature. Their attempt to ‘convey some of the “flavour” which amae has in
Japanese social interaction’ (p. 291) is highly plausible and effective when seen in
the literary context. However, their selected excerpts appear to remain solely within
Doi’s theoretical framework without exploring possibilities for alternative amae
interpretations. It would also be more helpful if those examples were introduced
and grouped in some meaningful way so that readers can perhaps recognize some
amae phenomena and cross-reference them. My goal is to do just that.
Next I attempt to reorganize and differentiate amae by context. I hope to dem-
onstrate why viewing amae separately in each context and considering its develop-
mental aspect may better clarify this phenomenon. Here I provide narratives to il-
lustrate amae in different contexts through the eyes of Japanese people today. I first
show how I obtained evidence and how the separate categories of amae emerged.
Amae
Reconsidered
Sources of Evidence
Sapporo Mothers. In this study, data were analyzed from a larger study in
which 40 mothers from Sapporo, a large city on the northern island of Japan, partici-
pated. To consider possible social class differences, half the sample was recruited

Amae Reconsidered 9 Human Development 2004;47:1–27
from the parents of a low-tuition-fee preschool (serving mainly working-class fami-
lies) and half from a high-tuition-fee preschool (serving middle to upper-middle-
class families). This study investigated Japanese mothers’ general parenting prac-
tices in relation to their parenting self-efficacy and the relationship with their chil-
dren. It was comprised of open-ended interview questions followed by a question-
naire on parenting efficacy [for results, see Holloway & Behrens, 2002]. Interview
questions included three amae-related questions that were not published with the
findings of the study, but which I present here to provide evidence. The specific
amae-related questions were, (a) What are the situations when [child] wants to do
amae?; (b) Are there any particular situations when you want [child] to do amae?,
and (c) Are there any particular situations when you want [child] to do less amae?
Mothers of the two preschools did not show any notable differences in their re-
sponses to these amae questions. A majority of mothers responded that their child is
likely to exhibit amae most when he/she is sleepy or tired, whereas some mothers
stated that their child would come to them for amae after having been scolded by
another adult or having had a bad day at school. Many mothers stated that they
would want their child’s amae when they have free time or feel they spent too much
time with the child’s siblings, whereas some felt that their child should come to
them for amae more often and tell them more about what he or she had experienced
at school and so on. A majority of mothers also stated that they would not welcome
their child’s amae when they are tired or too busy. Some mothers wished that their
child be more autonomous and not be clinging to them when they felt that he or she
should be with his/her peers. These responses indicate that amae is not a static event
that is always accepted or rejected (even within the same interactive partners), and
that childrens’ amae behaviors are diverse and contextual.
San Francisco Bay Area Japanese Interviewees. Another set of evidence was
provided by 10 native Japanese interviewees who are mostly professionals and cur-
rently reside in California’s San Francisco Bay Area. A particular context was not
provided, as the purpose was to learn what kind of episode native Japanese would
spontaneously relate in response to the word amae. The stories were recorded dur-
ing general conversations covering a variety of topics, not as part of a formal study
to be presented elsewhere. They responded to the question, ‘Have you observed or
experienced amae lately?’
Japanese Interviewees in Japan. Another set of evidence was provided by 13
native Japanese interviewed when visiting several cities in Japan. Stories were
again recorded while the participants were engaged in general conversations, not as
part of a formal study to be presented elsewhere. To assure spontaneity, no particu-
lar context was provided. They responded to the questions, ‘What is your most re-
cent amae experience?’, and ‘Have you observed or experienced amae lately?’
Amae Categories
Separate categories of amae emerged, suggesting both theoretical and empiri-
cal considerations. I found it necessary to differentiate between amae situations
because each amae discussed by different theorists is contextually different from

10
Human Development 2004;47:1–27
Behrens
the others. In addition, examples of amae experience provided by native Japanese
who reside in the San Francisco Bay Area as well as by those who live in Japan
have presented yet more dimensions of amae that have not been elaborated by the
theorists. Watanabe’s [1992] claim that amae in adulthood represents part of the
complex social system and Maruta’s [1992] emphasis on ‘mutually comfortable’
amae as a key to successful socialization in Japan indicate that amae has an aspect
of ‘learning’ that evolves in the course of development. However, one’s ‘intuitive’
aspect of amae is also likely to remain, involving possibly different interactive part-
ners in different contexts. Thus I believe that maturation does not necessarily corre-
spond to the emergence of different categories of amae. Instead, the expansion of
the social network that comes with maturation will lead to different amae experi-
ences. Each amae category I introduce here represents a distinct amae experience
(with a distinct quality and in a distinct context) that average Japanese people are
likely to encounter in their lives.
First, I consider the amae that Doi [1973] and Taketomo [1986] focus on are
likely to serve similar purposes. In my view, Doi’s amae and Taketomo’s amae
involve different interactants only by age and relationship. Doi’s focus is on
mother-infant amae, which he claims to be the prototype, and is mainly affective.
Taketomo, on the other hand, focuses on a child’s amae when an older child acts
like an infant, being playful to enjoy affective interactions. Taketomo further dis-
cusses an adult’s amae, when, for example, a woman behaves coquettishly toward a
man to enhance mutual closeness. These three amae situations – mother-infant
amae, a child’s infantile amae, and a woman’s coquettish amae toward a man –
share the same basic desire: to enjoy affective interactions. Therefore I argue that
this particular category of amae can be observed in any developmental phase in
different expressive and interactive behaviors. Here I propose to differentiate this
category of amae (Amae I) into three separate developmental phases, roughly corre-
sponding to the three situations presented above: combining Doi’s and Taketomo’s
views, Amae I can thus be defined as ‘affective’ situations in infancy, childhood,
and adulthood.
Second, while reviewing the interview responses to the amae questions pro-
vided by the Japanese mothers, I found that there were two distinct characteristics
of amae that these mothers discussed. One is mostly affective in nature and seem-
ingly enjoyed by both mother and child; it is perhaps used to reassure their inti-
macy, as described above as Amae I – ‘Affective’ in childhood. The other kind of
amae that the mothers described involves a child’s helpless behavior often occur-
ring when a child is tired or sleepy. There is an attempt to get the mother to do
things that the child can normally do on his/her own, and this behavior is not al-
ways welcome. I believe differentiating between these two kinds of amae is useful
not because one portrays more positive feature of amae than the other, but rather
because these two serve different purposes. These two kinds of amae are not mutu-
ally exclusive and may appear in separate contexts within the same relationship.
This Amae II – ‘Manipulative’ in childhood is also visible in adulthood, with an
added complexity I will explain below.
Third, when I began compiling examples of amae by asking native Japanese
who live in the San Francisco Bay Area, I learned that the stories they told were
often about amae behaviors of people in non-intimate relationships. I again identi-
fied three distinct characteristics of amae, and all represent different contexts with

Amae Reconsidered 11 Human Development 2004;47:1–27
different relational meanings. One interviewee presented an amae experience re-
flecting the relationship with her close friend, which she regarded as mutually
agreed and reciprocal in nature. This Amae III – ‘Reciprocal’ in adulthood repre-
sents conscious, mutually accepted behaviors that serve mainly instrumental pur-
poses between close friends. Although I have no example to illustrate this type of
amae in childhood, I speculate that Amae III – ‘Reciprocal’ also exists in child-
hood, as children spend increasingly more time with peers, gradually replacing time
with their parents.
Many told amae experiences at work, in which the interviewees were rather
forced to accept unreasonable demands from their bosses or from their subordi-
nates, which they perceived as forms of amae. Others told their experiences of the
times when they felt they were taken advantage of by distant acquaintances with
whom they had no obligatory relationship, which they also regarded as amae. Due
to their connection to social status and the inherent complexities, these last two
categories of amae are not likely to be observed in intimate relationships or in chil-
drens’ behaviors. Thus Amae IV – ‘Obligatory’ and Amae V – ‘Presumptive’ in
Table 1. The five amae categories
Noninstrumental Instrumental
Amae I –
Affective
Amae II –
Manipulative
Amae III –
Reciprocal
Amae IV–
Obligatory
Amae V –
Presumptive
Infancy
Motivation
Behavior
Relationship
Interactant(s)
desire for physical and
emotional closeness,
oneness
snuggling, seeking to
be held
intimate, affective,
close
parents (mothers)
N/A
N/A
N/A
N/A
Childhood
Motivation
Behavior
Relationship
Interactant(s)
desire for physical and
emotional closeness,
fun
snuggling, seeking to
sit on lap
intimate, affective,
close
parents (mothers)
get their way, benign
manipulation
clingy, act helplessly,
temper tantrum
intimate, close
parents (mothers)
desire for emotional
closeness, to recip-
rocate favors
act desperate, deal
making
close, trusting
school peers, friends
N/A
N/A
Adulthood
Motivation
Behavior
Relationship
Interactant(s)
desire for physical and
emotional closeness,
fun
playful, childish,
coquettish
intimate, affective,
close
romantic partners
get their way, benign
manipulation
act helplessly, selfishly,
with little cues
intimate, close
married couples
desire for emotional
closeness, to recip-
rocate favors
act desperate, deal
making
close, trusting
peers, friends
take advantage of/
abuse power/
control situations
excessive, unrea-
sonable demands
nonintimate,
unequal status
boss, client,
subordinate
presume upon
one’s good will
socially inappro-
priate, no enryo
nonintimate, non-
hierarchical
distant acquain-
tance

12
Human Development 2004;47:1–27
Behrens
adulthood emerged. Moreover, reviewing more examples of amae from interview-
ees in Japan enabled me to reexamine each category of amae and to further expand
on its meanings and functions.
Lastly, I recognized that these five categories of amae that I identified above
can be further divided into two larger domains, instrumental and noninstrumental.
Amae I – Affective, naturally represents amae behaviors with noninstrumental mo-
tives, whereas the other four categories of amae share some instrumental motives in
either intimate or nonintimate relationships. Thus, the amae categories that I intro-
duce here have been derived in part theoretically and in part empirically. The labels
that I attach to each category of amae – Affective, Manipulative, Reciprocal,
Obligatory, Presumptive – do not necessarily represent a fixed sequence, stage, or
level. They more or less indicate some ‘accrual’ in terms of complexity as well as
experiencing and learning different relational meanings. Nonetheless, Amae II–V,
under the instrumental domain, appear to show a rough gradient in distance of rela-
tionship of amae interactants (i.e., Amae II within the family, Amae III among
friends, Amae IV between co-workers and Amae V between strangers). Below, I
examine the five categories of amae in three developmental phases, providing evi-
dence where it is available. I present a chart that summarizes these five contextual
forms of amae in a conceptual framework, which may, I hope, serve as a helpful
guide for the readers (see table 1).
To assure the amae category system that I established here is valid and reli-
able, a native Japanese bilingual research assistant reviewed forty transcripts for
responses to the amae questions. Reliability coding reached 92% agreement. The
reliability coder also categorized eleven stories from ten San Francisco Bay Area
interviewees, which were randomly presented, with 100% agreement. For twenty
stories from thirteen interviewees in Japan, there was 90% agreement.
Amae in Infancy – Noninstrumental
Amae I – Affective. Infants’ ability to influence their caregivers through proto-
conversations or various forms of nonverbal communications is well documented
[e.g. Tomasello, 1999; Reddy, Hay, Murray, & Trevarthen, 1997]. Infants’ amae
behaviors, however, are expected to be largely affective in nature, as Doi [1973]
stresses. In the US, mother-infant interactions have been observed and their affec-
tive speech and behaviors have been analyzed in depth for the significance of in-
fants’ overall development [e.g., Stern, 1977; Stern, 1985]. In cross-cultural stud-
ies, Japanese mothers used ‘affect-salient’ speech toward their infants more than
‘information-salient’ speech, whereas the reverse was true for American mother-
infant dyads [Bornstein, Tal, Rahn, Galperin, Pechuex, Lamour, Toda, Azuma,
Ogino, & Tamis-LeMonda, 1992; Morikawa, Shand, & Kosawa, 1988; Toda,
Fogel, & Kawai, 1990]. I argue that such affect-focused speech by mothers can
facilitate amae behaviors that the mothers attempt to elicit in their young infants. In
those studies, Japanese mothers may have tried to show their infants that they are
the same and attempt to be ‘one’ with the infant by mimicking his vocalization in
order to maintain the infants’ focus on themselves rather than on external stimuli.
Doi [1973] claims that the first amae behavior is welcomed as a happy occa-
sion to celebrate the infant’s healthy growth, but at the same time it may represent a

Amae Reconsidered 13 Human Development 2004;47:1–27
bit of sadness to the mother, realizing that the infant now recognizes itself as a
separate entity from his mother. Doi claims that the prototype of the amae relation-
ship can be observed in the mother-infant relationship. Amae behaviors do vary,
however, even in this developmental stage. Not all Japanese infants engage in amae
behaviors, as would be expected. The following example was given by a mother in
our sample who recalled that her 6-year-old son did not engage in amae behavior
even when he was an infant:
Since he was small [he didn’t amaeru much] ... When you leave a child at his grand-
parents’ house to do errands and things, a normal child is likely to cry, missing the parent
even at his grandma’s place, you know? ... But he never was like that at all. So, my mother
got worried and asked me, ‘Is he okay?’ [laughs]. She was worried. So, in those situations,
it’d be troublesome if he actually did cry, but I also kind of wished that he could at least
whine or something’ (mother of 6-year-old male).
Amae in Childhood – Noninstrumental
Amae I – Affective. Normally developing young children still enjoy closeness
to their mothers, sometimes by even mimicking infants’ behaviors. Many children
are likely to desire to be cuddled or sit on their mother’s lap. I argue that this kind
of amae behavior in children is qualitatively the same as the one in infancy. Take-
tomo’s [1986] claim that amae refers to a child who playfully acts like an infant fits
into this category. The following examples illustrate such amae behaviors:
Uh, he doesn’t do baby talk anymore, but if I say his name like this ‘[name of the child
in a sweet voice]’, then he says ‘babu babu’ [a sound a baby makes]. He’s joking but comes
to me, crawling, that’s also a joke, but ... (mother of 4-year-old male).
When I smile at her or when I show my sweet face, she comes toward me or clings to
me ... When I praise her, she becomes like a baby [laughs] (mother of 6-year-old female).
I argue that a child’s basic desire to be close to his/her mother does not neces-
sarily end in early childhood. One interviewee in Japan gave the following example
to describe her son and daughter, both in their early teens:
Once in a while, my son pretends to sit on my lap; if I tried to hug him he’d run away. I
think he wants to amaeru ... My daughter, I don’t see her too often these days, but still she
suddenly comes next to me and pushes her body against me. I think this is amae (Tokyo
female).
Amae in Childhood – Instrumental
Amae II – Manipulative. As children get older, they become more capable of
taking the perspectives of others. Developmental theorists claim that young chil-
dren acquire the basic principles of folk physics and folk psychology by the age of
4 or 5 years, either through simulation or through a tacit theory [e.g., Gopnik &
Wellman, 1992; Harris, 1992]. At the same time they may also engage in manipu-
lative behaviors, understanding what behaviors may get others’ attention. Children

14
Human Development 2004;47:1–27
Behrens
by 4 and 5 have developed a ‘representational model of the mind’, and the mind
becomes more ‘intentional’ [Gopnik & Wellman, 1992]. These children may en-
gage in amae behaviors, pretending to be helpless, demanding extra care and atten-
tion from adults in general and from their mothers in particular. Infants’ sensory-
motor actions are known to exhibit clear intentions with specific goals [e.g.,
Tomasello, 1999]. Lewis and Ramsay [1999], however, discuss different levels of
intention, adaptive, knowledge, and conscious. Children’s intentions, acting to ma-
nipulate the interactive partner through amae, I argue, refer to conscious intensions,
the highest level of intensions according to Lewis and Ramsay. Okonogi’s [1992]
claim that a child takes an active role in the amae process through a series of ex-
pressions and behaviors, some of which are manipulative, fits into this category.
Japanese preschool children engage in amae behaviors usually when they are
tired, sleepy, or when they go to bed, exhibiting helpless behaviors with an attempt
to get their mothers attend to them. Sixty percent of mothers in our sample stated
that their preschool children amaeru either when they are sleepy or tired, illustrated
as follows:
… when he goes to sleep, we still have to sleep together, that’s something he hasn’t
even attempted to change. We still hold hands and talk about things and this is the only way
he can go to sleep ... (mother of 4-year-old male).
When she wakes up in the morning and when she goes to sleep, she still acts like a
[helpless] baby (mother of 4-year-old female).
Forty-three percent of the mothers in our sample who have more than one
child stated that their preschool children come to them to amaeru when their sib-
lings are amaeru-ing to demand their share of the mother’s attention or to compete
with their siblings, not necessarily because they genuinely want to amaeru, illus-
trated as follows:
Well, when the younger one comes to me to amaeru, then she comes to me also. She
seems to be saying, ‘me too’. She does that often (mother of 5-year-old female).
Well, sometimes we can’t help paying more attention to the younger one. Then, at
those times, he will come with a look on his face of, ‘how about me?’ (mother of 5-year-old
male).
Some mothers may simply relent to their children’s demands. Not all Japanese
mothers, however, automatically welcome such amae behaviors and they may wish
their children to be more autonomous, often when the mothers are faced with other
demanding tasks or are simply physically tired themselves. The following examples
illustrate these situations when amae behaviors are not welcomed:
Of course when I’m tired, I get irritated easily. So, at those times, I beg my children,
‘Mommy is very tired now and I’m in a bad mood. So, please do not come any closer right
now’ [laughs] (mother of 6-year-old male).
… [in the mornings], I have to get my husband ready to go, prepare [child] for
Youchien, and wake the little one up. So in the middle of those things, if [child] comes to
me and says, ‘I can’t do it unless you give me a hug’ or something like that, I lose my tem-
per (mother of 5-year-old female).

Amae Reconsidered 15 Human Development 2004;47:1–27
Some mothers appear to have a more or less clear idea of what amae is accept-
able or unacceptable. These views support Taketomo’s [1986] argument and Ma-
ruta’s [1992] claim that both interactants (mother and child) must mutually agree
for the amaeru desire to be satisfied. The following statements illustrate such
views:
When I know it’s not the time for amaeru, then I won’t [let her]. I don’t like to spoil
her and always indulge or pity her for just about anything, like always saying ‘poor thing’ ...
So, I can tell when I watch her, you know? If such and such a thing has happened and that’s
why she wants to mentally amaeru, then I hug and let her [amaeru] (mother of 5-year-old
female).
When everyone is playing with friends, I don’t want her to cling to Mommy but I want
her to play with them ...When we meet other adults, I want her to be able to greet them prop-
erly without clinging to Mommy (mother of 4-year-old female).
When Vereijken et al. [1997] reported that Japanese mothers did not desire
amaeru children, it is possible that the list of amae behaviors included mostly those
from the Amae II category. Such amae behaviors can be observed in older children.
They may exhibit amae behaviors toward parents with less purely affective motives
but more in relation to financial needs. Children as old as high-school or even col-
lege age do approach their parents in helpless and desperate manners when they
want new clothes, shoes, or cars. In Kim and Yamaguchi’s [1995] study, respon-
dents (over 80% were students) reported that they amaeru when they want money
or objects, which they listed as one of the top five reasons to amaeru. I regard these
amae behaviors as qualitatively equal to younger children behaving helplessly or
throwing tantrums to get what they want. In Japan, parents often provide children
with allowances to purchase whatever they want as long as they maintain a certain
level of academic achievement [Iwao, 1993]. Through this form of ‘financial’
amae, Japanese adolescents and young adults may be more dependent on their par-
ents than their counterparts in the US. The following examples illustrate such amae
behaviors in young adults in Japan. I insert the narratives in this section because
they are still assuming the child role within the family:
Well, I’m just not financially independent yet (Tokyo male).
I’m over 30 now but still live with my parents. I keep telling myself that I should leave
home and be independent, but I am doing amae because of the convenience of having
‘home’ (Tokyo female).
Amae III – Reciprocal. When children begin to form new relationships with
their peers, some amae behaviors can be observed in those relationships and may
substitute amae behaviors with their parents. Good friends expect to do favors for
each other, whether it means sharing their favorite toys, borrowing and lending
notes, or taking each other’s side when there is a fight. Older children may know
the limit and reciprocal nature of amae, which differs from Amae I – ‘Affective’
and Amae II – ‘Manipulative’, which are performed mostly within intimate family
relationships. Negotiations between children or mutual expectations that emerge
among peers, I argue, constitute this category of amae. Research has shown that
true mutuality or reciprocity exists in early communications between infants and

16
Human Development 2004;47:1–27
Behrens
interactive partners [e.g., Brazelton, Koslowski, & Main, 1974; Reddy et al., 1997].
However, reciprocal amae is neither cyclical nor a simultaneous event. Execution
of reciprocal amae is not necessarily immediate and requires negotiation and accep-
tance between peers. Older children may view this kind of amae behavior more
objectively and fairly, knowing that excessive and unidirectional amareru behav-
iors will not lead to popularity with their peers. At the same time, they know that
permitting others’ excessive amae behaviors or amayakasu peers excessively may
also be viewed as weakness. In Kim and Yamaguchi’s [1995] study, the majority of
students indicated that they let their friends amaeru more than they let their family
members do. The top five reasons for their acceptance of such amae include ‘aite
ga suki dakara’ (because [I] like them) and ‘amaeru no wa otagaisama da-
kara’ (because amaeru is [a] reciprocal [act]; my translations).
Amae in Adulthood – Noninstrumental
Amae I – Affective. Adult men and women in romantic relationships do engage
in amae behaviors, acting a little helplessly or being coquettish and thus enjoying
other people’s care and attention. Their simple desire for closeness within the lov-
ing relationship is similar to an infant’s desire to enjoy intimate closeness with his
mother (except for sexual gratification). Such behaviors appear to be common
across many cultures. For example, American adults in affective relationships often
call each other ‘baby’ or use nicknames that are typically addressed to young chil-
dren. They are not truly ‘regressing’ to the infant state per se but behave in a way
that may induce others to increase closeness and intimacy. Taketomo [1986] dis-
cusses amae in adult situations in which women behave coquettishly toward men.
According to Taketomo, women are motivated to amaeru to enhance mutual close-
ness, for fun, or to be released from reality as well as from superego pressure. Men,
on the other hand, feel the gratification of caring for a need or to achieve greater
self-esteem and proceed to amayakasu (see p. 542). Although Taketomo’s focus
was more on the direction from women toward men, I expect this interaction to be
bi-directional, with men amaeru-ing women to enjoy playful interactions as well.
Amae in Adulthood – Instrumental
Amae II – Manipulative. In contrast to Taketomo’s [1986] emphasis on
women’s amaeru behaviors toward men, I argue that there is another intimate situa-
tion in which amaeru behaviors are perhaps overrepresented in men toward women.
Sometimes adult men amaeru toward women for instrumental rather than affective
needs. Iwao [1993] argues that the rather intense mother-son relationship in Japan
may lead boys, when they grow up and get married, to expect ‘their wives to take
over their mother’s role’ (p. 151). Lebra [1994] also elaborates this point:
The Japanese husband expects his wife to be an overall caregiver for him, including
body care, as if he saw a mother substitute in his wife. The wife is amused to call her hus-
band her ‘oldest child’, who demands the greatest attention and nurturance’ (p. 267).

Amae Reconsidered 17 Human Development 2004;47:1–27
One mother from our sample, in the other section of the interview, complained
about the lack of support from her husband due to his amae behavior:
I only have a sister, so when the first child was born, I wondered whether all men were
like this ... No comforting, just requesting [me] to cook what he wants to eat, I wasn’t in the
condition [to do things like that] ... He is busy, but maybe he was also doing amae, using the
‘busyness’ of his occupation as an excuse (mother of 6-year-old female).
This kind of behavior, expecting things will get done without their involvement
and expecting their wives to understand what they want and comply with even un-
stated requests at times, constitutes amae with rather manipulative motivations. I
argue that this is similar to young children acting age-inappropriately and helplessly
for their mothers to come and attend to their needs because they know this manipu-
lative strategy often works. Thus, Japanese wives, in such a status-sensitive society,
sometimes have no choice but to accept their husbands’ behaviors or to amayakasu
them as part of their roles, either willingly or reluctantly. Lebra argues that such
childlike behaviors exhibited by the husband signify his ‘machoness’, representing
‘an extreme gender hierarchy of male master and female servant’ (p. 267).
This kind of amae behavior is not exclusive to men. Married women can also
take advantage of their husbands’ good nature or manipulate them into situations
where husbands are given little chance to contest. The following examples illustrate
such amae behaviors:
I don’t ask him to get me things or do something, but I often expect him to accept [the
situations] when things are not done. Well, it’s more like I just decide on my own ‘ah, this is
good enough’ and expect him to say ‘OK’ (Osaka female).
My husband and I are going on a trip to [places] to visit friends I met at a language
school ... My husband doesn’t know any of them, but I’m telling myself, ‘I’m sure he’ll
enjoy my reunion with them, too’ – this could be my amae too (Osaka female).
Amae III – Reciprocal. Much like amae in peer relations in childhood, amae
exists between close friends or co-workers in adulthood. They allow each other a
roughly equal degree of amae. One-sided, excessive, or abusive amae behaviors
should not be tolerated for too long, because when that happens, it may cost a
friendship. Close friends may let each other borrow some personal items, money or
cars, or may arrange occasional dates for each other and play go-between or media-
tor; expectations of any such request being accepted are regarded as amae.
These behaviors also seem to be observable in many other cultures, although
they are not called amae. For example, popular American comedy dramas such as
‘Seinfeld’, ‘Friends’, or ‘Will & Grace’ appear to depict close friends who exhibit
this type of amae within a less defined boundary between themselves and their
friends. Such mutual expectations, that basically any behavior is acceptable in close
friendships built on trust without romantic feelings, are regarded as amae in Japan.
One of the interviewees from the San Francisco Bay Area sample discussed such a
relationship:
My friend and I are very close but we are so different in every aspect ... I have a key to
her apartment and to her car. When I need a car, I usually just use her car. She often invites
me for dinner ... and I always amaeru for that. But she also rings me when she needs some

18
Human Development 2004;47:1–27
Behrens
assembly work. Also, when her car needs some work done, she expects me to take it to the
mechanic to get it fixed ... I’d say we amaeru each other pretty fairly ... (San Francisco Bay
Area female).
One interviewee in Japan discussed this kind of amae when it was pushed over
the limit:
Our high school alumni members have been frequently exchanging e-mails lately. One
member spent his own time to create a CD-R for reunion pictures. Then, one member started
to say, ‘How about including more pictures of old friends’ get-togethers at various places?’,
but then others said: ‘The one who made the CD-R spent his own time to do it, don’t do
amae, if you think it’s a good idea, do it yourself.’ I realized that we have a limit to our
amae between classmates’ (Osaka female).
Amae in Adulthood in Nonintimate Relationships
Next I present the last two categories of adult amae, which occur in noninti-
mate relationships. The first category in the nonintimate relationships, Amae IV –
‘Obligatory’, concerns amae in unequal relationships with respect to social status,
social roles, or gender. This kind of amae is commonly observed by Japanese,
sometimes with resentment. The last category, Amae V – ‘Presumptive’, represents
another amae behavior, which is also often viewed negatively. Okonogi [1992], in
fact, claims that the word amae is often used with resentment, referring to desires
for ‘preferential treatment beyond the other person’s tolerance level’ (p. 21).
Amae IV – Obligatory. In nonintimate, unequal power relationships (such as a
boss and a subordinate or a businessperson and a client), excessive demands can be
made by the one in power toward the one in the weaker position. Expectations for
the acceptance of such unreasonable demands also constitute amae of this category.
Although such behaviors are certainly observable in other societies, they are
viewed extremely negatively. Excessive and highly visible cases may run the risk
of a lawsuit (especially in the US). In Japan, although awareness of harassment and
discrimination issues has been rising, such behavior appears to be often underre-
ported. It is felt to be easier to accept it as part of people’s roles in a given situation
[Iwao, 1993]. The following narratives illustrate this kind of amae behavior. A
Japanese boss expects his Japanese subordinates to accept his unreasonable de-
mands on the grounds of ‘being Japanese’, assuming that Japanese people do not
like confrontations and that Japanese people understand each other without explicit
explanations:
I work at a Japanese company ... Because we work a lot with Japanese people and I’m
the only one who speaks Japanese, I often get twice the workload as the American employ-
ees, who basically do the same job. ... The difference is clearly observable, but still my boss
comes to me with some more work, even if the particular task doesn’t require any Japanese,
simply because it’s easy for him to ask in Japanese! I’d say that’s amae ...’ (San Francisco
Bay Area female).
I work for a Japanese corporation. I was hired locally just like any other local Ameri-
can. When the time for a review comes, ... he [my boss] would say something like ‘you’re
Japanese, too, and you should know better’, basically hesitating to give me the same salary
increase as other American co-workers. Giving me a reason like ‘you’re Japanese, too’ is

Amae Reconsidered 19 Human Development 2004;47:1–27
definitely amae and he is expecting me to accept whatever he says, leaving me with little
option’ (San Francisco Bay Area female).
Iwao discusses the symbolic meaning of serving tea not just in the light of it
being a service but also as a lubricant in human relations. She presents contrasting
views that some Japanese women are apparently not bothered by a task of serving
tea to maintain office harmony, whereas some are clearly opposed to the expecta-
tion of the task as a woman’s job. One interviewee equates such expectation as a
form of amae as illustrated below:
I used to work for a Japanese financial institution. I was one of the ‘rare’ Japanese
female AVPs. Hired locally, my colleagues were mostly American men, but my immediate
supervisor was a Japanese man. One day, ... my boss had a Japanese visitor and he came
over to me and asked me to serve tea. I looked at him in shock but he said, ‘It looks better if
a Japanese woman serves tea, you understand, don’t you?’, expecting that I would agree. ...
That’s definitely amae ...’ (San Francisco Bay Area female).
The following example was given by a Japanese interviewee in Japan as an
observer of amae behavior of this category:
Recently, there was an incident at the Country Dance Club which led to it breaking up.
I understand the reason was the instructor who, being the eldest and the leader, really abused
his amae on the club members (Osaka female).
In unequal power relationships, amaeru behaviors in the reverse direction –
from subordinates toward superiors – can be also exercised, with the subordinates
expecting their desires to be understood and accepted without being explicitly
stated, as in the following example:
Students (college) often come and ask me about jobs, things that they should be asking
administrative people about, or whatever ... Students do no ask me directly, ‘please teacher,
do so and so’, but say something like ‘oh, gees, I don’t know’ or ‘ah, whaaat?’, acting in a
way that they want me to do something (Sapporo male).
An act of amayakasu can also be initiated by a superior toward his favorite
subordinate or simply to avoid a foreseeable conflict with a subordinate who is
viewed as a potential threat or trouble. The example above pointed out Japanese
bosses’ amayakasu behaviors toward American employees by letting them get
away with less work while asking the Japanese to do more work. This Japanese
boss’s behavior – amayakasu of Americans and amaeru on Japanese – is also
pointed out by the other interviewee whose Japanese boss’s unwillingness to give
her a salary increase while he meets American employees’ demand for a raise. Both
interviewees commented that these Japanese bosses are afraid of American employ-
ees because of the commonly practiced work-related lawsuits in the US, and also
because of the language barrier. Another interpretation may be that these Japanese
bosses may regard Americans as ‘children’ who do not know better, simply because
they are not Japanese, leading to amayakasu behaviors.
Kashima and Callan [1994] discuss amae in relation to other Japanese con-
cepts such as on (indebtedness) and gimu (obligation) in the Japanese workplace.

20
Human Development 2004;47:1–27
Behrens
They claim that if Japanese workers’ desires to amaeru to their superiors are satis-
fied, the subordinates feel on and gimu, which in turn motivate them to work hard
in a paternalistic work environment. However, with a recent drastic change in the
working conditions in Japan due to recession, it is not certain how much the amae-
on-gimu system still holds true as lifetime employment or other benefits, including
guaranteed salary increments and promotions based on seniority, can no longer be
fully expected.
The next two narratives illustrate a relationship between a businessperson and
a client where the client has more control in the situation and expects the Japanese
service provider to follow his unreasonable request:
I’m a travel agent ... One time, I had arranged a rent-a-car for a Japanese businessman
and basically, my job was done. But he realized that he had left his personal belongings in
the car when he returned the car. So, he called me and expected me to trace the stuff ...
That’s, I’d say, amae (San Francisco Bay Area female).
I used to work as a case manager for Japanese patients with [disease] ... We always
explain the support and service we provide as clearly as possible ... Too often, when we
follow up, we find that those patients haven’t finished the paper work or other tasks that the
patients are responsible for. Those patients expect me to do absolutely everything ... I con-
sider that amae (San Francisco Bay Area male).
Amae V – Presumptive. The final category of amae is a set of behaviors seen
largely among adults; it is nonaffective and involves no hierarchical relationship or
even no relationship at all in some cases. This kind of amae is sometimes intricately
intertwined with other subtle behaviors. Amae can be exercised tactfully under the
disguise of another Japanese concept, enryo, defined as ‘self-restraint’ [Lebra,
1976] or ‘ritualized hesitation or deference’ [Johnson, 1993]. For example, if a
woman is offered a delicious piece of cake to go with coffee, she is supposed to say
‘oh, no thanks’ at first to exhibit enryo. By refusing the cake, she is actually hoping
and expecting that the host will offer the cake at least a few more times. She will
eventually have the cake without having to make herself appear vulgar by jumping
at the cake the minute it is offered. This expectation is indeed one form of amae. In
other words, one expresses enryo to display socially desired behaviors, but at the
same time one counts on the other person’s understanding of what one ‘really de-
sires’ beneath this behavior, thus practicing amae toward people in nonintimate rela-
tionships. Such subtle interactions understandably appear very vague and confusing
to non-Japanese. If one expresses a true desire from the very beginning, it may seem
straightforward and easy to be understood. Using these seemingly contrasting con-
cepts together appears to be part of Japanese socialization among mature members
of the society. The following narrative demonstrates such case of amae:
I had this woman staying in my house for a week or so, a visitor from Japan. ... Well,
this woman was very modest at first, doing enryo a lot ... she said she’d call for a cab to go
to her appointments downtown. I told her that I’d give her a ride since it’s on the way to
some place I’m going. She was still doing enryo at first but then accepted my offer. Well,
once she came out of the enryo mode, she then became pretty demanding. ... She started
telling me where she needed to go next and basically, I had to spend much time driving her
around. This is amae, at first disguised as enryo … (San Francisco Bay Area female).

Amae Reconsidered 21 Human Development 2004;47:1–27
As a general rule, Japanese are expected to show some hesitation, known as
enryo, before they exhibit amae behaviors, except within very close relationships.
The following narrative demonstrates how some Japanese people bypass enryo and
attempt to engage in almost forceful amae with an individual whom they personally
do not know well:
One day, I got a call, all of sudden, from someone whom I barely knew, saying that she
had just arrived at the SFO airport. Well, I was like ‘So?’ Without her asking me if I could
pick her up, she just remained silent, waiting for me to say something. Of course, she ex-
pected me to say that I’m coming to the airport to pick her up. That’s amae (San Francisco
Bay Area female).
One interviewee described the situation below, when she was placed in an
awkward position because of amae by someone she hardly knew:
I’m a professional [occupation] ... One day, I was asked by my acquaintance’s wife if I
could help her translating a tax form or something. ... I only charged a modest fee as a favor
to my acquaintance, ‘Just for once’, I thought. One day, I got a call from this woman again
and she asked me if I could translate a letter. … she insisted to bring it to me herself. Well, it
turned out that the letter she wrote was a love letter for someone who she’s having an affair
with! It was definitely amae because somehow, even though she knew that I knew her hus-
band, she expected that I wouldn’t tell him about her affair. … (San Francisco Bay Area
female).
The final example below demonstrates how some Japanese adults engage in
socially inappropriate behaviors but expect to be accepted because they are
‘temporarily’ released from their social obligations, following Takatomo’s [1986]
basic principle. These people exhibit amae and target a nonspecific audience:
Once I was hired as one of the interpreters for a big group of Japanese businessmen ...
At one of the restaurants, as soon as we walked in, they started yelling loudly for beer and ...
It was really embarrassing, because somehow they think they can get away with immature
behaviors here in America because they are not in Japan ... And then, one middle-aged man
suddenly squeezed the waitress’s breasts! … We sincerely apologized ... In the mean time,
that man was proudly talking to the other guys about how good the breasts had felt. I was so
disgusted and embarrassed, but this is amae because they just expected … they could get
away with whatever they do here in the US (San Francisco Bay Area male).
Summary
The spontaneous responses to amae questions presented above by native Japa-
nese adults appear to characterize amae as an everyday phenomenon for Japanese
people today. Its uses vary considerably from the amae that Doi (1973) focused on.
It is worthy to note that Japanese living outside Japan seem to share experiences of
witnessing amae on the ground of ‘being Japanese’, as we have seen above. In fact,
it is possible some amae behaviors become more salient to Japanese living outside
Japan because they can compare it with the behaviors of non-Japanese in everyday
life. It is also interesting to note that nearly half of the interviewees in Japan first
discussed the way they would do amae to themselves, describing behaviors of ‘not
taking good care of myself, trying to carry over the project to the next day when it

22
Human Development 2004;47:1–27
Behrens
could be done in a few more hours’, ‘trying to run away from my own research’,
and so forth. Lebra [1976] referred to this amae as self-indulgence to imply
‘disapproval for not taking responsibility seriously’ (p. 55). We humans are some-
times known, however, to ‘treat ourselves as interactants’ [Fridlund, 1994, p. 160].
In this paper, I maintain my view of amae as a concept of relatedness and interac-
tion, thus this amae on oneself is not included in the chart.
Discussion
Amae has many functions and characteristics and should not be treated as a
global construct with a single meaning. All amae phenomena, however, do share
some implicit or explicit assumptions of being understood and accepted in a given
context. Nevertheless, for non-Japanese observers, amae phenomena in various
contexts may not be readily recognizable. Amae is undeniably part of Japanese cul-
ture, and Japanese researches should be paying more attention to it. In my view,
Doi’s [1973] and other theorists’ [Kumagai & Kumagai, 1986; Lebra, 1976; Ma-
ruta, 1992; Okonogi, 1992; Taketomo, 1986; Watanabe, 1992] claims about amae,
their critiques and proposals for a modification of Doi’s original definition are all
essentially correct, although they do not reach agreement between themselves. I
believe that Dale [1986] also has a valid point when stating that Doi and other Japa-
nese scholars should caution against using linguistic differences as an excuse to
gain a special permit to leave everything down to ‘myth’.
Like Doi, I do believe that amae-like phenomena, at least some of the amae
categories I introduced above, do exist in non-Japanese cultures, although there is
perhaps no single word, to my knowledge, equivalent to amae, which covers multi-
ple meanings and functions. A recent study [Lewis & Ozaki, 2002] compared amae
with the British slang term ‘mardy’ for its usage, which was thought to be very
similar, and concluded that mardy focuses only on the negative aspect of amae-like
phenomena, such as childish self-indulgence that should be discouraged. The new
approach that I propose here should help clarify the contextual meaning of a par-
ticular amae phenomenon by evaluating the relationship between the interactants,
the motive of the amae doer, and the likelihood of the amae being granted. Corre-
sponding terms in English, then, could perhaps be applied to each particular amae
phenomenon separately to promote better understanding. Furthermore, confusion
surrounding attachment and amae, as I stated earlier, should hopefully be settled
here by illuminating similarities and differences, as summarized below.
Similarities and Differences between Attachment and Amae
Similarities. As I have pointed out at the beginning, there are similarities be-
tween attachment and amae in rather objective aspects, e.g., their onset time (both
are emerging in the second half of the first year of life) and the interactive partners
(both phenomena are likely to be first observed in the mother-child relationship).
Behavioral strategies can also be similar between attachment and amae, as some
researchers have argued [e.g., Rothbaum et al., 2000; Mizuta, Zahn-Waxler, Cole,
& Hiruma, 1996]. In particular, certain attachment behaviors – insecure-ambivalent

Amae Reconsidered 23 Human Development 2004;47:1–27
behaviors – may indeed appear similar to amae behaviors, but are, in fact, only
similar to a particular category of amae, as I argued earlier.
Cassidy and Berlin’s [1994] intriguing discussion of a particular attachment
classification – insecure-ambivalent patterns of attachment – helps clarify how the
strategies work for both children and mothers to bring about such behaviors. For
example, the child’s strategy of emphasizing immaturity to regain the attention of
his mother upon a birth of a sibling was described by many Japanese mothers in our
sample in response to a question of the situations when their children want to do
amae. These preschool children come to their mothers to amaeru to compete with
their siblings for the mothers’ attention; I categorized this kind of amae as Amae
II – ‘Manipulative’ in Childhood. Cassidy and Berlin’s discussion of the strategy of
acting immaturely by adults to promote the feeling of competence in others by pro-
viding a chance to play a caregiver’s role was demonstrated by Taketomo’s [1986]
example of the Japanese woman in New York who consciously did amaeru to her
mother to make her mother feel needed. Furthermore, maternal strategies of inse-
cure-ambivalent children that Cassidy and Berlin discussed also appear to be simi-
lar to a description of some Japanese mothers. For example, they describe a mother
who ‘(consciously or nonconsciously) wants to be particularly assured of her im-
portance to the infant, of his dependence on her, and of his availability to meet her
own attachment needs’ (p.984). This mother resembles Iwao’s [1993] description
of a Japanese mother who is largely neglected by her husband and relates to her
male child as a ‘surrogate husband’, although this strategy appears to be that of
over-involvement rather than of being inconsistent, as Cassidy and Berlin claim.
Also, a mother who may find the child’s clinginess in some way satisfying and
comforting, as they describe, is similar to a mother amayakasu her child to fulfill
her egoistic needs, which, as Watanabe [1992] argues, can indeed be harmful.
Differences. The most important difference is perhaps that attachment is a be-
havioral system that is activated when there is a threat in the environment, the phe-
nomenon becomes most distinct when a child is under distress [Bowlby, 1982].
Amae behaviors, however, can be observed anytime when there is a desire to have
either purely affective or instrumental needs fulfilled. Due to society’s implicit
emphasis on status differentials, amae phenomena are often intricately linked to
status roles in Japanese socialization [DeVos, 1975], whereas attachment phenom-
ena depict the more basic human relatedness. In addition, attachment patterns rep-
resent relationship quality, quantity is never measured. By contrast, I believe that
amae can be assessed in terms of quantity or frequency to learn about a relation-
ship. Furthermore, although multiple attachment relationships are certainly recog-
nized, attachment classification to a specific individual is not expected to fluctuate
too easily [Bowlby, 1982]. On the contrary, one can engage in multiple categories
of amae with the same individual in a given day. For example, a child can approach
his mother to enjoy her warmth and affection, engaging in Amae I – ‘Affective’,
and the same child can also demand the mother’s attention, acting helplessly when
he gets tired later that day, engaging in Amae II – ‘Manipulative’.
Other differences may include that ‘attachment’ is not necessarily an everyday
word, whereas ‘amae’ can be for Japanese people. The word ‘attachment’ is used
primarily by the specialists in the field. Experts in the field tend to distinguish cas-
ual usage of the word ‘attachment’, such as ‘I’m really attached to this car’, from

24
Human Development 2004;47:1–27
Behrens
that in attachment theory. On the contrary, amae is a word that is frequently used
even among children and adults in Japan as well as experts in the field. Children
rarely say to their mothers, ‘I am going to amaeru now’, as was pointed out by
Okonogi [1992], but they may talk to a puppy or a younger child saying, ‘You are
amaeru-ing, aren’t you?’ in an endearing manner. Adults do often say that they are
going to engage in amae behavior, presuming upon the other’s kindness, it is one of
many formula-like expressions that exist in Japan. We have seen above how native
Japanese directly used the word amae in their narratives to describe certain behav-
iors in certain contexts.
Summary. The similarities between attachment and amae that I identified
above assert that there is an undeniable link between them. The concrete aspects of
attachment theory and the concept of amae, such as the onset time or the basic
module of the relationship, appear to reflect the nature of both concepts more gen-
erally. The behavioral strategies that reflect internal states of individuals of a par-
ticular attachment classification, however, suggest its link to only a particular cate-
gory of amae. The differences between attachment and amae that I summarized
above also confirm that attachment cannot be simply substituted by amae or vice
versa. Nor are attachment and amae mutually exclusive. I argue that securely as
well as insecurely attached children in Japan may engage in various kinds of amae
at different times. Likewise, adults with secure states of mind as well as adults with
insecure states of mind in Japan are likely to engage and experience amae in di-
verse contexts throughout their lifetime.
Future Amae Research
Lack of empirical amae research is perhaps partly to blame for confusions and
inconsistent interpretations surrounding the concept. Therefore there is an urgent
need to promote systematic investigations of amae. The conceptual framework that
I presented in this paper should serve as a guide to help select a particular aspect of
amae for investigation. Investigators should determine what population in what
context should be examined for amae behaviors. For example, children’s amae
behaviors should be observed either at home or at school. The children’s parents or
teachers should then be asked about their perception of amae behaviors, either
through interview or questionnaire. Involving both interactants in the investigation,
child and adult in this case, should enable the investigators to elucidate a compre-
hensive picture of the amae phenomena.
While it is essential that researchers continue their investigations on amae to
present more evidence and further conceptual clarifications, what is perhaps most
effective and immediately beneficial is a study that integrates amae research into
attachment research in Japan in order to investigate a direct link between them.
Japan has been considered a challenge for attachment researchers, partly because of
amae [e.g., van IJzendoorn & Sagi, 1999], and the validity of attachment theory has
been questioned because of amae [Rothbaum et al., 2000]. Japan has also been a
focus of controversy due to inconsistent findings from two attachment studies that
have been conducted in the past [Durrett, Otaki, & Richards, 1984; Takahashi,
1986]. Further replication studies of attachment in Japan are thus necessary. Inves-
tigators should examine whether distinct amae behaviors can be discriminated from
attachment behaviors during the procedure. An amae questionnaire should also be

Amae Reconsidered 25 Human Development 2004;47:1–27
given to the parents to examine their perception of the child’s amae. Do Japanese
children with different attachment classifications differ in their amae behaviors
toward the parents? Do Japanese parents’ perceptions of their child’s amae differ
for children with different attachment classifications? Furthermore, do Japanese
parents’ perceptions of their child’s amae differ according to their own attachment
status? Administering the Adult Attachment Interview [George, Kaplan, & Main,
1996] to parents would first allow us to investigate the concordance between child
and parent attachment classifications. It could also tell us how the parents’ own
attachment status might be related to their perception of their child’s amae.
Another amae questionnaire or interview should be administered to examine
adults’ more general amae experiences with other adults in intimate and noninti-
mate relationships and to further explore a link with attachment status. For exam-
ple, are Japanese adults with insecure states of mind likely to engage in or experi-
ence generally less favored amae (i.e., Amae II – ‘Manipulative’, Amae IV –
‘Obligatory’, or Amae V – ‘Presumptive’) more often than those who have secure
states of mind? Are Japanese adults with secure states of mind likely to engage in
or experience Amae I – ‘Affective’ more often than those with insecure states of
mind? Thus responses to the extensive amae questionnaire or amae interview, in-
corporated in attachment research, should not only illuminate Japanese people’s
thoughts and feelings about amae through their experiences, but also help advance
our understanding of its possible relation to attachment organizations. Hence, re-
search that combines investigations of attachment and amae will enhance cross-
cultural research of human relationships.
Acknowledgements
I thank all the participants who willingly shared their experiences. I am grateful to
Joseph J. Campos, George A. DeVos, and Susan D. Holloway for their valuable input on the
earlier version of the manuscript.
References
Ainsworth, M.D.S., Bell, S.M.V., & Slayon, D.J. (1971). Individual differences in strange situation
behaviour of one-year-olds. In H.R. Schaffer (Ed.), The origins of human social relations (pp. 17–
57). New York: Academic Press.
Balint, M. (1966). Primary love and psychoanalytic technique. Oxford, UK: Ernst Klett.
Bornstein, M.H., Tal, J., Rahn, C., Galperin, C.Z., Pechuex, M., Lamour, M., Toda, S., Azuma, H.,
Ogino, M., & Tamis-LeMonda, C. (1992). Functional analysis of the contents of maternal speech
to infants of 5 and 13 months in four cultures: Argentina, France, Japan, and the United States.
Developmental Psychology, 28, 593–603.
Bowlby, J. (1958). The nature of the child’s tie to his mother. International Journal of Psychoanalysis,
39, 350–373.
Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). New York: Basic Books.
Brazelton, T.B., Koslowski, B., & Main, M. (1974). The origins of reciprocity: The early mother-infant
interaction. In M. Lewis & L.A. Rosenblum (Eds.), The effect of the infant on its caregiver (pp.
49–76). New York: John Wiley & Sons.
Bretherton, I. (1992). The origin of attachment theory: John Bowlby and Mary Ainsworth. Developmen-
tal Psychology, 28, 759–775.
Cassidy, J., & Berlin, L.J. (1994). The insecure/ambivalent pattern of attachment: Theory and research.
Child Development, 65, 971–991.
Dale, P. (1986). The myth of Japanese uniqueness. New York: St. Martin’s.
DeVos, G.A. (1975). Apprenticeship and paternalism. In E. F. Vogel (Ed.), Modern Japanese organiza-
tion and decision-making (pp. 210–227). Berkeley, CA: University of California Press.

26
Human Development 2004;47:1–27
Behrens
Doi, T. (1973). The anatomy of dependence. New York: Kodansha.
Durrett, M.E., Otaki, M., & Richards, P. (1984). Attachment and the mother’s perception of support
from the father. International Journal of Behavioral Development, 7, 167–176.
Emde, R.N. (1992). Amae, intimacy, and the early moral self. Infant Mental Health Journal, 13, 34–42.
Fridlund, A. (1994). Human facial expression: An evolutionary view. New York: Academic Press.
George, C., Kaplan, N., & Main, M. (1996). Adult Attachment Interview Protocol (3rd ed.) Unpublished
manuscript, University of California at Berkeley.
Gopnik, A., & Wellman, H.M. (1992). Why the child’s theory of mind really is a theory. Mind and
Language, 7, 145–171.
Harris, P.L. (1992). From simulation to folk psychology: The case for development. Mind and Lan-
guage, 7, 120–144.
Holloway, S.D., & Behrens, K.Y. (2002). Parenting efficacy among Japanese mothers: Qualitative and
quantitative perspectives on its association with childhood memories of family relations. In J.
Bempechat & J. Elliot (Eds.), Learning in culture and context. New Directions for Child and
Adolescent Development, No. 96, 27–43. San Francisco: Jossey-Bass.
Iwao, S. (1993). The Japanese woman: Traditional image and changing reality. Cambridge, MA: Har-
vard University Press.
Johnson, F.A. (1993). Dependency and Japanese socialization: Psychoanalytic and anthropological
investigations into amae. New York: New York University Press.
Kashima, Y., & Callan, V.J. (1994). The Japanese work group. In H.C. Traindis, M.D. Dunnette, & M.
H. Leadtta (Eds.), Handbook of industrial and organizational psychology (Vol. 4, pp. 609–646).
Palo Alto, CA: Consulting Psychologists Press.
Kim, U., & Yamaguchi, S. (1995). Conceptual and empirical analysis of amae: Exploration into Japa-
nese psychological space. In Proceedings of the 43rd Annual Conference of the Japanese Group
Dynamics Association (pp. 158–159). Tokyo: Japanese Group Dynamics Association.
Kumagai, H.A., & Kumagai, A.K. (1986). The hidden ‘I’ in amae: ‘Passive love’ and Japanese social
perception. Ethos, 14, 305–320.
Lebra, T.S. (1976). Japanese patterns of behavior. Honolulu: University of Hawaii Press.
Lebra, T.S. (1994). Mother and child in Japanese socialization: A Japan-U.S. comparison. In P.M.
Greenfield & R.R. Cocking (Eds.), Cross-cultural roots of minority child development (pp. 259–
274). Hillsdale, NJ: Lawrence Erlbaum Associates.
Lewis, J.R., & Ozaki, R. (2002, July). Amae in the UK. Paper presented at the XIV International Asso-
ciation for Cross-Cultural Psychology Congress, Yogyakarta, Indonesia.
Lewis, M., & Ramsay, D. (1999). Intentions, consciousness, and pretend play. In P.D. Zelazo, J.W.
Astington, & D.R. Olson (Eds.), Developing theories of intention: Social understanding and self-
control (pp. 77–94). Mahwah, NJ: Lawrence Erlbaum Associates.
Mahler, M. (1968). On human symbiosis and the vicissitudes of individuation. New York: International
University Press.
Maruta, T. (1992). Does an American puppy amaeru? A comment on Dr. Doi’s paper. Infant Mental
Health Journal, 13, 12–17.
Mitchell, D.O. (1976). Amaeru: The expression of reciprocal dependency in Japanese politics and law.
Boulder, CO: Westview.
Mizuta, I., Zahn-Waxler, C., Cole, P. M., & Hiruma, N. (1996). A cross cultural study of preschoolers’
attachment: Security and sensitivity in Japanese and U.S. dyads. International Journal of Behav-
ioral Development, 19, 141–159.
Morikawa, H., Shand, N., & Kosawa, Y. (1988). Maternal speech to prelingual infants in Japan and the
United States: Relationships among functions, forms and referents. Journal of Child Language,
15, 237–256.
Morsbach, H., & Tyler, W.J. (1986). A Japanese emotion: Amae. In R. Harre (Ed.), The social con-
struction of emotion (pp. 289–307). Oxford: Basil Blackwell.
Osofsky, J. D. (1992). Editor’s note. Infant Mental Health Journal, 13, 4.
Ohsako, H., & Takahashi, S. (1994). Effects of ‘Amae’ on interpersonal emotions and conflict-solution
strategies in interpersonal conflict situations. The Japanese Journal of Experimental Social Psy-
chology, 34, 44–57.
Okonogi, K. (1992). Amae as seen in diverse interpersonal interactions. Infant Mental Health Journal,
13, 18–25.
Reddy, V., Hay, D., Murray, L., & Trevarthen, C. (1997). Communication in infancy: Mutual regulation
of affect and attention. In G. Bremner, A. Slater, & G. Butterworth (Eds.), Infant development:
Recent advances (pp. 247–273). East Sussex, UK: Psychology Press.
Rothbaum, F., Weisz, J., Pott, M., Miyake, K., & Morelli, G. (2000). Attachment and culture: Security
in the United States and Japan. American Psychologist, 55, 1093–1104.
Stern, D. (1977). The first relationship: Infant and mother. Cambridge, MA: Harvard University Press.
Stern, D. (1985). The interpersonal world of the infant: A view from psychoanalysis and developmental
psychology. New York: Basic Books.

Amae Reconsidered 27 Human Development 2004;47:1–27
Takahashi, K. (1986). Examining the strange-situation procedure with Japanese mothers and 12-month-
old infants. Developmental Psychology, 22, 265–270.
Taketomo, Y. (1986). Amae as metalanguage: A critique of Doi’s theory of amae. Journal of the Ameri-
can Academy of Psychoanalysis, 14, 525–544.
Toda, S., Fogel, A., & Kawai, M. (1990). Maternal speech to three-month-old infants in the United
States and Japan. Journal of Child Language, 27, 279–294.
Tomasello, M. (1999). Having intentions, understanding intentions, and understanding communicative
intentions. In P.D. Zelazo, J.W. Astington, & D.R. Olson (Eds.), Developing theories of intention:
Social understanding and self-control (pp. 63–75). Mahwah, NJ: Lawrence Erlbaum Associates.
van IJzendoorn, M.H., & Sagi, A. (1999). Cross-cultural patterns of attachment: Universal and contex-
tual dimensions. In J. Cassidy & P.R. Shaver (Eds.), Handbook of attachment: Theory, research,
and clinical applications (pp. 713–734). New York: Guilford Press.
Vereijken, C.J.J.L., Riksen-Walraven, J.M., & VanLieshout, C.F.M. (1997). Mother-infant relationships
in Japan: Attachment, dependency, and amae. Journal of Cross-Cultural Psychology, 28, 442–
462.
Watanabe, H. (1992). Difficulties in amae: A clinical perspective. Infant Mental Health Journal, 13,
26–33.
Yamaguchi, S., Ohashi, M., Zemba, Y., Murakami, F., Yamauchi, M., Watanabe, S., Maruyama, J., &
Yoshiwara, C. (1998, September). Nichijoogo to shiteno ‘Amae’ no tsukaikata [Usage of ‘Amae’
as everyday word]. Paper presented at the 46th
Congress of the Japanese Group Dynamics Asso-
ciation, Nagoya University, Japan.